From janssen at parc.com  Wed Mar  1 00:08:43 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 28 Feb 2006 15:08:43 PST
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: Your message of "Tue, 28 Feb 2006 13:07:57 PST."
	<C60DD163-5F85-480C-A757-F4061C49712B@mac.com> 
Message-ID: <06Feb28.150849pst."58633"@synergy1.parc.xerox.com>

Why don't we call this RC1?  I'll update the Web page next week, let's say.

Bill

> I've placed a DMG with an installer for a universal build of python  
> 2.4.2. This is an initial release of this installer, there are  
> probably issues with it.
> 
> This installer features:
> - a universal binary of Python.framework
> - readline 5.1
> - sleepycat db 4.4
> - IDLE with Mac keybindings
> - Python Documentation, both in IDLE and pydoc
> - pythonw and python are equivalent
> - the installer will update your shell profile if needed
> 
> Known issues:
> - "regrtest -uall" crashes somewhere in test_db3 (bus error)
> - I haven't tested this version on Intel or OSX 10.3 yet.
> - This version cannot build extensions on OSX 10.3, or any other  
> system without Xcode 2.2.1 and the 10.4u SDK.
> 
> The installer is here: http://homepage.mac.com/ronaldoussoren/.Public/ 
> Universal%20MacPython%202.4.2.dmg
> 
> Ronald
> 
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From janssen at parc.com  Wed Mar  1 00:12:14 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 28 Feb 2006 15:12:14 PST
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: Your message of "Tue, 28 Feb 2006 13:07:57 PST."
	<C60DD163-5F85-480C-A757-F4061C49712B@mac.com> 
Message-ID: <06Feb28.151214pst."58633"@synergy1.parc.xerox.com>

Ron,

For this release, we were going to change the name of the folder that
the IDLE appears in, from "MacPython-2.4" to "Python 2.4".  In this
installer, it seems to be "MacPython 2.4".

I think it would be a good idea to remove the old "MacPython-2.4"
folder, as well.

Bill

From ronaldoussoren at mac.com  Wed Mar  1 09:41:08 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 01 Mar 2006 09:41:08 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <06Feb28.151214pst.58633@synergy1.parc.xerox.com>
References: <06Feb28.151214pst.58633@synergy1.parc.xerox.com>
Message-ID: <10550985.1141202468600.JavaMail.ronaldoussoren@mac.com>

 
On Wednesday, March 01, 2006, at 00:13AM, Bill Janssen <janssen at parc.com> wrote:

>Ron,
^^^^^ Ronald :-)
>
>For this release, we were going to change the name of the folder that
>the IDLE appears in, from "MacPython-2.4" to "Python 2.4".  In this
>installer, it seems to be "MacPython 2.4".

I'd prefer to keep this 'MacPython 2.4'.

>
>I think it would be a good idea to remove the old "MacPython-2.4"
>folder, as well.

Good idea.  We should probably also provide something to uninstall
the software (either an uninstall package or a script like the one included
with Xcode).

Ronald


From ronaldoussoren at mac.com  Wed Mar  1 09:44:38 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 01 Mar 2006 09:44:38 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <06Feb28.150849pst.58633@synergy1.parc.xerox.com>
References: <06Feb28.150849pst.58633@synergy1.parc.xerox.com>
Message-ID: <7308263.1141202678218.JavaMail.ronaldoussoren@mac.com>

 
On Wednesday, March 01, 2006, at 00:10AM, Bill Janssen <janssen at parc.com> wrote:

>Why don't we call this RC1?  I'll update the Web page next week, let's say.

Because I don't want to attract newbies at the moment. Let's get this one
right before giving a lot of publicity to it.

Apart from the db4 issue I'm pretty sure that this version will work correctly
on OSX 10.4.  Support for 10.3 needs more testing, and some work if we want
to support building extensions on 10.3 as well (which we should).

Ronald

>
>Bill
>
>> I've placed a DMG with an installer for a universal build of python  
>> 2.4.2. This is an initial release of this installer, there are  
>> probably issues with it.
>> 
>> This installer features:
>> - a universal binary of Python.framework
>> - readline 5.1
>> - sleepycat db 4.4
>> - IDLE with Mac keybindings
>> - Python Documentation, both in IDLE and pydoc
>> - pythonw and python are equivalent
>> - the installer will update your shell profile if needed
>> 
>> Known issues:
>> - "regrtest -uall" crashes somewhere in test_db3 (bus error)
>> - I haven't tested this version on Intel or OSX 10.3 yet.
>> - This version cannot build extensions on OSX 10.3, or any other  
>> system without Xcode 2.2.1 and the 10.4u SDK.
>> 
>> The installer is here: http://homepage.mac.com/ronaldoussoren/.Public/ 
>> Universal%20MacPython%202.4.2.dmg
>> 
>> Ronald
>> 
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>
>

From bob at redivi.com  Wed Mar  1 11:05:49 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 1 Mar 2006 02:05:49 -0800
Subject: [Pythonmac-SIG] Fwd: framework on Intel
References: <6C30D77C-667C-4F4B-AAE5-B649FDA1F022@bx.psu.edu>
Message-ID: <61B827FB-4CB4-403A-823A-1235EDBCA1DA@redivi.com>



Begin forwarded message:

> From: James Taylor <james at bx.psu.edu>
> Date: February 28, 2006 2:40:35 PM PST
> To: sw at wordtech-software.com
> Cc: bob at redivi.com
> Subject: framework on Intel
>
>> ld: warning can't open dynamic library:
>> - -all_load/usr/lib/system/libmathCommon.A.dylib referenced from:
>> /usr/lib/gcc/powerpc-apple-darwin8/4.0.1/../../../libSystem.dylib
>> (checking for undefined symbols may be affected) (No such file or
>> directory, errno = 2)
>
> I believe this is the result of building without --enable- 
> universalsdk, which results in isysroot not being defined correctly  
> because of a little bug in the makefile (bash variable reference  
> used instead of make), one possible fix:
>
> Index: Makefile.pre.in
> ===================================================================
> --- Makefile.pre.in     (revision 23)
> +++ Makefile.pre.in     (working copy)
> @@ -379,7 +379,7 @@
>                 for X in `lipo -info $(LIBRARY) | cut -f 3 -s - 
> d :`; do \
>                         Z="$$Z $(LDLIBRARY).libtool.$$X" ;\
>                         $(CC) -o $(LDLIBRARY).libtool.$$X -arch $$X  
> -dynamiclib \
> -                               -isysroot ${UNIVERSALSDK:-"/"} \
> +                               -isysroot $(if $(UNIVERSALSDK),$ 
> (UNIVERSALSDK),"/") \
>                                 -all_load $(LIBRARY) -Wl,- 
> single_module \
>                                 -install_name $(DESTDIR)$ 
> (PYTHONFRAMEWORKINSTALLDIR)/Versions/$(VERSION)/Python \
>                                 -compatibility_version $(VERSION) \

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060301/5a02de98/attachment.html 

From janssen at parc.com  Wed Mar  1 17:27:54 2006
From: janssen at parc.com (Bill Janssen)
Date: Wed, 1 Mar 2006 08:27:54 PST
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: Your message of "Wed, 01 Mar 2006 00:44:38 PST."
	<7308263.1141202678218.JavaMail.ronaldoussoren@mac.com> 
Message-ID: <06Mar1.082755pst."58633"@synergy1.parc.xerox.com>

> >Why don't we call this RC1?  I'll update the Web page next week, let's say.
> 
> Because I don't want to attract newbies at the moment. Let's get this one
> right before giving a lot of publicity to it.

OK, fine with me.  But please change that folder name to "Python 2.4".

Thanks.

Bill

From zbir at urbanape.com  Wed Mar  1 17:32:11 2006
From: zbir at urbanape.com (Zachery Bir)
Date: Wed, 1 Mar 2006 11:32:11 -0500
Subject: [Pythonmac-SIG] NSUserDefaults causes crash on MOSXI
Message-ID: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>

Can someone help me read this Crash log?

I reduced a bug I was seeing down to a reproduceable minimum:

   <http://www.urbanape.com/downloads/TestUserDefaults.zip>

This has two PyObjC apps: TestUserDefaults, and TestUserDefaults- 
Working. The only difference between the two being the following  
lines in the former's init method:

   NSUserDefaults.resetStandardUserDefaults()
   self.sud = NSUserDefaults.standardUserDefaults()

At this point, the app crashes on Mac OS X Intel. It runs fine on my  
iMac G5 at home. Both are running Ronald's new Universal 2.4.2  
Python, and each is running SVN trunk PyObjC. I wasn't able to get  
PyObjC built using the AUTO_UNIVERSAL (it fails trying to find some  
temp directory in /private/tmp), so each is building apps native to  
its processor.

The crash log is located here:

   <http://www.urbanape.com/downloads/ 
TestUserDefaultsAppDelegate.crash.log>

Thanks,

Zac


From zbir at urbanape.com  Wed Mar  1 17:36:21 2006
From: zbir at urbanape.com (Zachery Bir)
Date: Wed, 1 Mar 2006 11:36:21 -0500
Subject: [Pythonmac-SIG] NSUserDefaults causes crash on MOSXI
In-Reply-To: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>
References: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>
Message-ID: <60AF98A3-3EB5-4905-9F4A-DF813C5C416A@urbanape.com>

On Mar 1, 2006, at 11:32 AM, Zachery Bir wrote:

> Can someone help me read this Crash log?
>
>    <http://www.urbanape.com/downloads/ 
> TestUserDefaultsAppDelegate.crash.log>

 From the crash log:

Exception:  EXC_BAD_INSTRUCTION (0x0002)
Code[0]:    0x0000000d
Code[1]:    0x00000000


Thread 0 Crashed:
0   com.apple.CoreFoundation       	0x90823daa _CFReadBytesFromFile + 76
1   com.apple.CoreFoundation       	0x90823793  
CFURLCreateDataAndPropertiesFromResource + 284
2   com.apple.CoreFoundation       	0x9082d67a _loadXMLDomainIfStale  
+ 299
3   com.apple.CoreFoundation       	0x90839978 fetchXMLValue + 38
4   com.apple.CoreFoundation       	0x908396ad  
_CFPreferencesDomainCreateValueForKey + 35
5   com.apple.Foundation           	0x926e0df8 -[NSUserDefaults  
initWithUser:] + 765
6   com.apple.Foundation           	0x926e096c +[NSUserDefaults  
standardUserDefaults] + 131

Reading bytes from file == endian issue?

Zac


From bob at redivi.com  Wed Mar  1 19:22:45 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 1 Mar 2006 10:22:45 -0800
Subject: [Pythonmac-SIG] NSUserDefaults causes crash on MOSXI
In-Reply-To: <60AF98A3-3EB5-4905-9F4A-DF813C5C416A@urbanape.com>
References: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>
	<60AF98A3-3EB5-4905-9F4A-DF813C5C416A@urbanape.com>
Message-ID: <2C9DD974-6B77-4EBD-B267-EB7E026F21A7@redivi.com>


On Mar 1, 2006, at 8:36 AM, Zachery Bir wrote:

> On Mar 1, 2006, at 11:32 AM, Zachery Bir wrote:
>
>> Can someone help me read this Crash log?
>>
>>    <http://www.urbanape.com/downloads/
>> TestUserDefaultsAppDelegate.crash.log>
>
>  From the crash log:
>
> Exception:  EXC_BAD_INSTRUCTION (0x0002)
> Code[0]:    0x0000000d
> Code[1]:    0x00000000
>
>
> Thread 0 Crashed:
> 0   com.apple.CoreFoundation       	0x90823daa _CFReadBytesFromFile  
> + 76
> 1   com.apple.CoreFoundation       	0x90823793
> CFURLCreateDataAndPropertiesFromResource + 284
> 2   com.apple.CoreFoundation       	0x9082d67a _loadXMLDomainIfStale
> + 299
> 3   com.apple.CoreFoundation       	0x90839978 fetchXMLValue + 38
> 4   com.apple.CoreFoundation       	0x908396ad
> _CFPreferencesDomainCreateValueForKey + 35
> 5   com.apple.Foundation           	0x926e0df8 -[NSUserDefaults
> initWithUser:] + 765
> 6   com.apple.Foundation           	0x926e096c +[NSUserDefaults
> standardUserDefaults] + 131
>
> Reading bytes from file == endian issue?

Maybe, but it says XML...

The first thing I would do is take gdb and figure out what file it's  
looking at... that's the second argument to  
CFURLCreateDataAndPropertiesFromResource.  I'm guessing it would live  
on the stack, but I haven't used gdb on intel enough lately to  
remember what the offset is.

Alternatively you could run it under ktrace to see which file it  
opens.  There should be a NAMI with the path it's opening.

-bob

From keith.ray at gmail.com  Wed Mar  1 22:06:23 2006
From: keith.ray at gmail.com (Keith Ray)
Date: Wed, 1 Mar 2006 13:06:23 -0800
Subject: [Pythonmac-SIG] [fitnesse] fitnesse PyFit on 2 Macs
In-Reply-To: <20060301191606.39782.qmail@web54514.mail.yahoo.com>
References: <207ec7f00603011014p509d4e61he39db9a9f7b7cf90@mail.gmail.com>
	<20060301191606.39782.qmail@web54514.mail.yahoo.com>
Message-ID: <207ec7f00603011306v46b8da71i8ed5a15b1f67dda0@mail.gmail.com>

so I do this (please excuse the different names than in previous msgs)

$ python
>>> import open_test_file

and get this:

Traceback (most recent call last):
  File "<stdin>", line 1 ,in ?
  File "open_test_file.py" ,line 10, in ?
    from appscript import *
  File "/Library/Python/2.3/site-packages/appscript/__init__.py line 15, in ?
    from findapp import ApplicationNotFoundError
  File "/Library/Python/2.3/site-packages/site-packages/appscript/findapp.py",
line 9, in ?
    from LaunchServices.Launch import LSFindApplicationForInfo
ImportError: No module named LaunchServices.Launch


On 3/1/06, Grig Gheorghiu <grig at gheorghiu.net> wrote:
> That may mean that there is a syntax or path-related error in your
> test_my_app.py file. Can you try running it through python at the
> command line? I've seen this before, when I had syntax errors in my
> fixtures, yet pyFIT was saying "fixture not found" instead of reporting
> the errors.
>

From smithsm at samuelsmith.org  Wed Mar  1 20:21:55 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Wed, 1 Mar 2006 12:21:55 -0700
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <7308263.1141202678218.JavaMail.ronaldoussoren@mac.com>
References: <06Feb28.150849pst.58633@synergy1.parc.xerox.com>
	<7308263.1141202678218.JavaMail.ronaldoussoren@mac.com>
Message-ID: <A002A373-2F29-4427-9FF5-71B622BCF3AE@samuelsmith.org>

It would be nice if the universal installer for 2.4.2 also installed  
the 2.4.2 version of the python documentation
in Mac Help format. I have the framework python 2.4.2 but the mac  
help docs are still 2.4.1. Somewhere sometime ago
there was an installer for the reference docs in mac help format. I  
haven't seen anything for 2.4.2


On 01 Mar, 2006, at 01:44, Ronald Oussoren wrote:

>
> On Wednesday, March 01, 2006, at 00:10AM, Bill Janssen  
> <janssen at parc.com> wrote:
>
>> Why don't we call this RC1?  I'll update the Web page next week,  
>> let's say.
>
> Because I don't want to attract newbies at the moment. Let's get  
> this one
> right before giving a lot of publicity to it.
>
> Apart from the db4 issue I'm pretty sure that this version will  
> work correctly
> on OSX 10.4.  Support for 10.3 needs more testing, and some work if  
> we want
> to support building extensions on 10.3 as well (which we should).
>
> Ronald
>
>>
>> Bill
>>
>>> I've placed a DMG with an installer for a universal build of python
>>> 2.4.2. This is an initial release of this installer, there are
>>> probably issues with it.
>>>
>>> This installer features:
>>> - a universal binary of Python.framework
>>> - readline 5.1
>>> - sleepycat db 4.4
>>> - IDLE with Mac keybindings
>>> - Python Documentation, both in IDLE and pydoc
>>> - pythonw and python are equivalent
>>> - the installer will update your shell profile if needed
>>>
>>> Known issues:
>>> - "regrtest -uall" crashes somewhere in test_db3 (bus error)
>>> - I haven't tested this version on Intel or OSX 10.3 yet.
>>> - This version cannot build extensions on OSX 10.3, or any other
>>> system without Xcode 2.2.1 and the 10.4u SDK.
>>>
>>> The installer is here: http://homepage.mac.com/ 
>>> ronaldoussoren/.Public/
>>> Universal%20MacPython%202.4.2.dmg
>>>
>>> Ronald
>>>
>>> _______________________________________________
>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>
>>
>>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig

**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From ronaldoussoren at mac.com  Wed Mar  1 22:50:28 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 1 Mar 2006 22:50:28 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <C60DD163-5F85-480C-A757-F4061C49712B@mac.com>
References: <C60DD163-5F85-480C-A757-F4061C49712B@mac.com>
Message-ID: <92C242EA-DFCE-4270-89BC-9E61ABD9475B@mac.com>


On 28-feb-2006, at 22:07, Ronald Oussoren wrote:
>
> Known issues:
> - "regrtest -uall" crashes somewhere in test_db3 (bus error)

I don't understand why yet, but the build seems to have picked up  
another version of libdb that I had expected. _bsddb.so is a PPC only  
binary and claims to be 4.3.29 instead of the 4.4.x version I wanted  
it to be.

Ronald


From bob at redivi.com  Wed Mar  1 23:07:34 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 1 Mar 2006 14:07:34 -0800
Subject: [Pythonmac-SIG] [fitnesse] fitnesse PyFit on 2 Macs
In-Reply-To: <207ec7f00603011306v46b8da71i8ed5a15b1f67dda0@mail.gmail.com>
References: <207ec7f00603011014p509d4e61he39db9a9f7b7cf90@mail.gmail.com>
	<20060301191606.39782.qmail@web54514.mail.yahoo.com>
	<207ec7f00603011306v46b8da71i8ed5a15b1f67dda0@mail.gmail.com>
Message-ID: <8D4CB0AC-ED13-46A0-B650-63A2E7B1648E@redivi.com>

You need to install Python23Compat to get the LaunchServices package.
http://pythonmac.org/packages/Python23Compat-0.0-py2.3-macosx10.3.zip


On Mar 1, 2006, at 1:06 PM, Keith Ray wrote:

> so I do this (please excuse the different names than in previous msgs)
>
> $ python
>>>> import open_test_file
>
> and get this:
>
> Traceback (most recent call last):
>   File "<stdin>", line 1 ,in ?
>   File "open_test_file.py" ,line 10, in ?
>     from appscript import *
>   File "/Library/Python/2.3/site-packages/appscript/__init__.py  
> line 15, in ?
>     from findapp import ApplicationNotFoundError
>   File "/Library/Python/2.3/site-packages/site-packages/appscript/ 
> findapp.py",
> line 9, in ?
>     from LaunchServices.Launch import LSFindApplicationForInfo
> ImportError: No module named LaunchServices.Launch
>
>
> On 3/1/06, Grig Gheorghiu <grig at gheorghiu.net> wrote:
>> That may mean that there is a syntax or path-related error in your
>> test_my_app.py file. Can you try running it through python at the
>> command line? I've seen this before, when I had syntax errors in my
>> fixtures, yet pyFIT was saying "fixture not found" instead of  
>> reporting
>> the errors.
>>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From skip at pobox.com  Thu Mar  2 00:56:44 2006
From: skip at pobox.com (skip at pobox.com)
Date: Wed, 1 Mar 2006 17:56:44 -0600
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <ACE7ECD6-983B-4350-BF72-83222535EF1B@mac.com>
References: <a47214750602250729g7dd17d90idd823ce36cc3f379@mail.gmail.com>
	<06Feb27.103207pst.58633@synergy1.parc.xerox.com>
	<17411.29468.745083.261535@montanaro.dyndns.org>
	<ACE7ECD6-983B-4350-BF72-83222535EF1B@mac.com>
Message-ID: <17414.13500.232643.205233@montanaro.dyndns.org>


    Ronald> The page looks fine, but it might be better to remove the icon
    Ronald> or replace it by the current MacPython icon (the 16 ton
    Ronald> weight). It's not that I don't like the snake, but some people
    Ronald> claim a lifelike snake will scare some (potential) users and
    Ronald> furthermore the snake-icon isn't actually used by MacPython
    Ronald> (yet?).

I believe Bill has svn access now, so I'll leave that to him to decide.

Skip

From keith.ray at gmail.com  Thu Mar  2 02:11:21 2006
From: keith.ray at gmail.com (Keith Ray)
Date: Wed, 1 Mar 2006 17:11:21 -0800
Subject: [Pythonmac-SIG] [fitnesse] fitnesse PyFit on 2 Macs
In-Reply-To: <8D4CB0AC-ED13-46A0-B650-63A2E7B1648E@redivi.com>
References: <207ec7f00603011014p509d4e61he39db9a9f7b7cf90@mail.gmail.com>
	<20060301191606.39782.qmail@web54514.mail.yahoo.com>
	<207ec7f00603011306v46b8da71i8ed5a15b1f67dda0@mail.gmail.com>
	<8D4CB0AC-ED13-46A0-B650-63A2E7B1648E@redivi.com>
Message-ID: <207ec7f00603011711k2809294xfb928f4b930cc958@mail.gmail.com>

Thanks. I discovered this. For some reason, I thought that installer
was only needed on MacOS X 10.3.x.

On 3/1/06, Bob Ippolito <bob at redivi.com> wrote:
> You need to install Python23Compat to get the LaunchServices package.
> http://pythonmac.org/packages/Python23Compat-0.0-py2.3-macosx10.3.zip
>
>
> On Mar 1, 2006, at 1:06 PM, Keith Ray wrote:
>
> > so I do this (please excuse the different names than in previous msgs)
> >
> > $ python
> >>>> import open_test_file
> >
> > and get this:
> >
> > Traceback (most recent call last):
> >   File "<stdin>", line 1 ,in ?
> >   File "open_test_file.py" ,line 10, in ?
> >     from appscript import *
> >   File "/Library/Python/2.3/site-packages/appscript/__init__.py
> > line 15, in ?
> >     from findapp import ApplicationNotFoundError
> >   File "/Library/Python/2.3/site-packages/site-packages/appscript/
> > findapp.py",
> > line 9, in ?
> >     from LaunchServices.Launch import LSFindApplicationForInfo
> > ImportError: No module named LaunchServices.Launch
> >
> >
> > On 3/1/06, Grig Gheorghiu <grig at gheorghiu.net> wrote:
> >> That may mean that there is a syntax or path-related error in your
> >> test_my_app.py file. Can you try running it through python at the
> >> command line? I've seen this before, when I had syntax errors in my
> >> fixtures, yet pyFIT was saying "fixture not found" instead of
> >> reporting
> >> the errors.
> >>
> > _______________________________________________
> > Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> > http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>


--

C. Keith Ray
<http://homepage.mac.com/keithray/blog/index.html>
<http://homepage.mac.com/keithray/xpminifaq.html>
<http://homepage.mac.com/keithray/resume2.html>

From bob at redivi.com  Thu Mar  2 02:16:44 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 1 Mar 2006 17:16:44 -0800
Subject: [Pythonmac-SIG] [fitnesse] fitnesse PyFit on 2 Macs
In-Reply-To: <207ec7f00603011711k2809294xfb928f4b930cc958@mail.gmail.com>
References: <207ec7f00603011014p509d4e61he39db9a9f7b7cf90@mail.gmail.com>
	<20060301191606.39782.qmail@web54514.mail.yahoo.com>
	<207ec7f00603011306v46b8da71i8ed5a15b1f67dda0@mail.gmail.com>
	<8D4CB0AC-ED13-46A0-B650-63A2E7B1648E@redivi.com>
	<207ec7f00603011711k2809294xfb928f4b930cc958@mail.gmail.com>
Message-ID: <8867F4E6-79C0-4FFB-91AF-5B971618912C@redivi.com>

No, that installer applies to all versions of Python 2.3.

There is an installer that is only needed on Mac OS X 10.4.x that  
allows you to use packages built for 10.3 with the vendor Python.

You actually need both of these in order to use appscript.

-bob

On Mar 1, 2006, at 5:11 PM, Keith Ray wrote:

> Thanks. I discovered this. For some reason, I thought that installer
> was only needed on MacOS X 10.3.x.
>
> On 3/1/06, Bob Ippolito <bob at redivi.com> wrote:
>> You need to install Python23Compat to get the LaunchServices package.
>> http://pythonmac.org/packages/Python23Compat-0.0-py2.3-macosx10.3.zip
>>
>>
>> On Mar 1, 2006, at 1:06 PM, Keith Ray wrote:
>>
>>> so I do this (please excuse the different names than in previous  
>>> msgs)
>>>
>>> $ python
>>>>>> import open_test_file
>>>
>>> and get this:
>>>
>>> Traceback (most recent call last):
>>>   File "<stdin>", line 1 ,in ?
>>>   File "open_test_file.py" ,line 10, in ?
>>>     from appscript import *
>>>   File "/Library/Python/2.3/site-packages/appscript/__init__.py
>>> line 15, in ?
>>>     from findapp import ApplicationNotFoundError
>>>   File "/Library/Python/2.3/site-packages/site-packages/appscript/
>>> findapp.py",
>>> line 9, in ?
>>>     from LaunchServices.Launch import LSFindApplicationForInfo
>>> ImportError: No module named LaunchServices.Launch
>>>
>>>
>>> On 3/1/06, Grig Gheorghiu <grig at gheorghiu.net> wrote:
>>>> That may mean that there is a syntax or path-related error in your
>>>> test_my_app.py file. Can you try running it through python at the
>>>> command line? I've seen this before, when I had syntax errors in my
>>>> fixtures, yet pyFIT was saying "fixture not found" instead of
>>>> reporting
>>>> the errors.
>>>>
>>> _______________________________________________
>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>
>>
>
>
> --
>
> C. Keith Ray
> <http://homepage.mac.com/keithray/blog/index.html>
> <http://homepage.mac.com/keithray/xpminifaq.html>
> <http://homepage.mac.com/keithray/resume2.html>


From wrw at ornl.gov  Thu Mar  2 13:40:13 2006
From: wrw at ornl.gov (W. R. Wing)
Date: Thu, 02 Mar 2006 07:40:13 -0500
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <17414.13500.232643.205233@montanaro.dyndns.org>
References: <a47214750602250729g7dd17d90idd823ce36cc3f379@mail.gmail.com>
	<06Feb27.103207pst.58633@synergy1.parc.xerox.com>
	<17411.29468.745083.261535@montanaro.dyndns.org>
	<ACE7ECD6-983B-4350-BF72-83222535EF1B@mac.com>
	<17414.13500.232643.205233@montanaro.dyndns.org>
Message-ID: <p06210270c02c94d7ba88@[160.91.212.99]>

At 5:56 PM -0600 3/1/06, skip at pobox.com wrote:
>     Ronald> The page looks fine, but it might be better to remove the icon
>     Ronald> or replace it by the current MacPython icon (the 16 ton
>     Ronald> weight). It's not that I don't like the snake, but some people
>     Ronald> claim a lifelike snake will scare some (potential) users and
>     Ronald> furthermore the snake-icon isn't actually used by MacPython
>     Ronald> (yet?).
>
>I believe Bill has svn access now, so I'll leave that to him to decide.
>
>Skip

I've been lurking on this list since early 2002 (and I can't tell you-all
how much I've picked up listening - it has been incredibly helpful), 
but I feel it is finally time to say something.

I LIKE the python squeezing the apple.  The 16 ton icon is not only 
amateurish, it is so indirect an "in" joke as to be inscrutable to 
anyone coming new to the party.  When I first saw it (back in OS 9 
days) I thought  it was supposed to represent Python's ability to do 
"heavy lifting" for the user.

Until, or unless  Guido himself officially signs off on an icon that 
_has_ to be used - I'd say stick with this one.  It at least looks 
professional.

			There, got it off my chest,
			Bill

PS:  If it REALLY is a problem for some people, offer them an alternative
      in the post-flight phase of the installer.


-- 
   _
  /~\ The ASCII       | William R. Wing, PhD.  wrw at ornl.gov 865-574-8839
  \ / Ribbon Campaign | Network Architect, Oak Ridge National Laboratory
   X  Against HTML    | Network Research Group, Computer Sci & Math Div.
  / \ Email!          |_________

From ronaldoussoren at mac.com  Thu Mar  2 14:17:39 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 02 Mar 2006 14:17:39 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <p06210270c02c94d7ba88@[160.91.212.99]>
References: <a47214750602250729g7dd17d90idd823ce36cc3f379@mail.gmail.com>
	<06Feb27.103207pst.58633@synergy1.parc.xerox.com>
	<17411.29468.745083.261535@montanaro.dyndns.org>
	<ACE7ECD6-983B-4350-BF72-83222535EF1B@mac.com>
	<17414.13500.232643.205233@montanaro.dyndns.org>
	<p06210270c02c94d7ba88@[160.91.212.99]>
Message-ID: <16491074.1141305460018.JavaMail.ronaldoussoren@mac.com>

 
On Thursday, March 02, 2006, at 01:40PM, W. R. Wing <wrw at ornl.gov> wrote:

>At 5:56 PM -0600 3/1/06, skip at pobox.com wrote:
>>     Ronald> The page looks fine, but it might be better to remove the icon
>>     Ronald> or replace it by the current MacPython icon (the 16 ton
>>     Ronald> weight). It's not that I don't like the snake, but some people
>>     Ronald> claim a lifelike snake will scare some (potential) users and
>>     Ronald> furthermore the snake-icon isn't actually used by MacPython
>>     Ronald> (yet?).
>>
>>I believe Bill has svn access now, so I'll leave that to him to decide.
>>
>>Skip
>
>I've been lurking on this list since early 2002 (and I can't tell you-all
>how much I've picked up listening - it has been incredibly helpful), 
>but I feel it is finally time to say something.
>
>I LIKE the python squeezing the apple.  The 16 ton icon is not only 
>amateurish, it is so indirect an "in" joke as to be inscrutable to 
>anyone coming new to the party.  When I first saw it (back in OS 9 
>days) I thought  it was supposed to represent Python's ability to do 
>"heavy lifting" for the user.
>
>Until, or unless  Guido himself officially signs off on an icon that 
>_has_ to be used - I'd say stick with this one.  It at least looks 
>professional.

The 16 ton icon stays until someone brings a better alternative. Please
remember that it is not just 1 icon, but a set of them: One each for
IDLE, BuildApplet and .py files. Ideally there would also be one for
py2app application bundles where the user hasn't provided his own
icon.

The reason I don't like having the snake on the website is that the icon
will not match the icons used by the various applications installed
by the macpython installer.

Ronald

From ronaldoussoren at mac.com  Thu Mar  2 14:57:14 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 02 Mar 2006 14:57:14 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <Pine.LNX.4.64.0603021342060.12820@gandalf.fraca7.homeunix.net>
References: <C60DD163-5F85-480C-A757-F4061C49712B@mac.com>
	<92C242EA-DFCE-4270-89BC-9E61ABD9475B@mac.com>
	<Pine.LNX.4.64.0603021342060.12820@gandalf.fraca7.homeunix.net>
Message-ID: <8402061.1141307835030.JavaMail.ronaldoussoren@mac.com>

 
On Thursday, March 02, 2006, at 02:20PM, Jerome Laheurte <jerome at gandalf.fraca7.homeunix.net> wrote:

>On Wed, 1 Mar 2006, Ronald Oussoren wrote:
>
>> On 28-feb-2006, at 22:07, Ronald Oussoren wrote:
>>>
>>> Known issues:
>>> - "regrtest -uall" crashes somewhere in test_db3 (bus error)
>>
>> I don't understand why yet, but the build seems to have picked up
>> another version of libdb that I had expected. _bsddb.so is a PPC only
>> binary and claims to be 4.3.29 instead of the 4.4.x version I wanted
>> it to be.
>
>Python 2.4.2 doesn't support 4.4; I submitted a patch here:
>
>http://sourceforge.net/tracker/index.php?func=detail&aid=1441660&group_id=5470&atid=305470
>

The tree at svn.pythonmac.org already contains that patch, maybe it is based on the 2.4.x branch of python instead of the 2.4.2 release. Bob?

test_bsddb3 seems to behave better after I patched setup.py to look for bsddb at the right location.

Ronald

>

From bob at redivi.com  Thu Mar  2 16:14:27 2006
From: bob at redivi.com (Bob Ippolito)
Date: Thu, 2 Mar 2006 07:14:27 -0800
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <8402061.1141307835030.JavaMail.ronaldoussoren@mac.com>
References: <C60DD163-5F85-480C-A757-F4061C49712B@mac.com>
	<92C242EA-DFCE-4270-89BC-9E61ABD9475B@mac.com>
	<Pine.LNX.4.64.0603021342060.12820@gandalf.fraca7.homeunix.net>
	<8402061.1141307835030.JavaMail.ronaldoussoren@mac.com>
Message-ID: <C43F93B2-11C3-4165-8199-860C0D89B96B@redivi.com>


On Mar 2, 2006, at 5:57 AM, Ronald Oussoren wrote:

>
> On Thursday, March 02, 2006, at 02:20PM, Jerome Laheurte  
> <jerome at gandalf.fraca7.homeunix.net> wrote:
>
>> On Wed, 1 Mar 2006, Ronald Oussoren wrote:
>>
>>> On 28-feb-2006, at 22:07, Ronald Oussoren wrote:
>>>>
>>>> Known issues:
>>>> - "regrtest -uall" crashes somewhere in test_db3 (bus error)
>>>
>>> I don't understand why yet, but the build seems to have picked up
>>> another version of libdb that I had expected. _bsddb.so is a PPC  
>>> only
>>> binary and claims to be 4.3.29 instead of the 4.4.x version I wanted
>>> it to be.
>>
>> Python 2.4.2 doesn't support 4.4; I submitted a patch here:
>>
>> http://sourceforge.net/tracker/index.php? 
>> func=detail&aid=1441660&group_id=5470&atid=305470
>>
>
> The tree at svn.pythonmac.org already contains that patch, maybe it  
> is based on the 2.4.x branch of python instead of the 2.4.2  
> release. Bob?

The tree at svn.pythonmac.org was made using release24-maint, not the  
2.4.2 tag.  I *think* that this is the last revision that it is  
synchronized with:

----------------------------------------------------------------------
r1050 (orig r42239):  barry.warsaw | 2006-02-04 15:45:12 -0800

Resolves SF bug #1423972.

-bob


From janssen at parc.com  Thu Mar  2 18:55:51 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 2 Mar 2006 09:55:51 PST
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: Your message of "Thu, 02 Mar 2006 05:17:39 PST."
	<16491074.1141305460018.JavaMail.ronaldoussoren@mac.com> 
Message-ID: <06Mar2.095553pst."58633"@synergy1.parc.xerox.com>

Ron Oussoren writes:
> The 16 ton icon stays until someone brings a better alternative.

W. R. Wing writes:
> I LIKE the python squeezing the apple.  The 16 ton icon is not only 
> amateurish, it is so indirect an "in" joke as to be inscrutable to 
> anyone coming new to the party.

I agree with Wing.  So there *is* a better alternative, and there's
now a MacPython.icns file at
http://bill.janssen.org/mac/MacPython.icns with the snake wrapped
around the Apple.  Feel free to use it where it fits.

But Ron objects:
> The page looks fine, but it might be better to remove the icon or
> replace it by the current MacPython icon (the 16 ton weight).

The 16-ton weight screams "done by amateurs for a toy computer".  The
sooner it's stamped out, the better.

And further explains:
> It's not that I don't like the snake, but some people claim a
> lifelike snake will scare some (potential) users and furthermore the
> snake-icon isn't actually used by MacPython (yet?).

As we know, some people will claim anything.  The snake/apple icon is
all over the MacPython pages, that's where it comes from.  The new
Python web site has finally switched to a snake-based logo, as well.

As for scaring users:  I feel sympathy for folks who have pathological
fears of almost anything, including drawings of snakes.  I think,
though, it might be kinder to repel those folks immediately, rather
than helping them overcome their incipient nausea at the very idea of
a language called "Python", and getting to use it.  Such use might
cause a nasty amount of subconscious angst, expressed in -- who knows
-- continuous nightmares?  Random acts of violence?

Surely better to act up front!  More snakes on the web page!

Bill

From dp at ulaluma.com  Thu Mar  2 19:08:03 2006
From: dp at ulaluma.com (Donovan Preston)
Date: Thu, 2 Mar 2006 10:08:03 -0800
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar2.095553pst."58633"@synergy1.parc.xerox.com>
References: <06Mar2.095553pst."58633"@synergy1.parc.xerox.com>
Message-ID: <433A1F5F-0A70-41F8-9577-328C79E8D4FC@ulaluma.com>


On Mar 2, 2006, at 9:55 AM, Bill Janssen wrote:

> Ron Oussoren writes:
>> The 16 ton icon stays until someone brings a better alternative.

Stop with the 16 ton weight/snakes/apples debate. Just use the new,  
official, python logo, which is a stylized icon that isn't going to  
offend anybody, is vector so will look good at any size, and looks  
very, very professional. I don't see any reason to indicate macness/ 
appleness in the logo. It's python. You are using it on a Mac.  
Everyone already knows Apple makes macs.

http://beta.python.org/images/python-logo.gif

dp


From berkowit at silcom.com  Thu Mar  2 19:25:31 2006
From: berkowit at silcom.com (Paul Berkowitz)
Date: Thu, 02 Mar 2006 10:25:31 -0800
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar2.095553pst."58633"@synergy1.parc.xerox.com>
Message-ID: <C02C789B.C4CEC%berkowit@silcom.com>

On 3/2/06 9:55 AM, "Bill Janssen" <janssen at parc.com> wrote:

> Ron Oussoren writes:
>> The 16 ton icon stays until someone brings a better alternative.
> 
> W. R. Wing writes:
>> I LIKE the python squeezing the apple.  The 16 ton icon is not only
>> amateurish, it is so indirect an "in" joke as to be inscrutable to
>> anyone coming new to the party.
> 
> I agree with Wing.  So there *is* a better alternative, and there's
> now a MacPython.icns file at
> http://bill.janssen.org/mac/MacPython.icns with the snake wrapped
> around the Apple.  Feel free to use it where it fits.

That leads to a web page that looks like this:

icns?^ics#H@?????????????????@?????????????????is32?
?
???????wwtx??????????|c?????????y|?B?}?rS?8HYo?{?;?s?cn{?t9|?{???L?z
}xxs{qt??WK???iy~{???y???e???xw????Q{?~_??????????-I???m??pG??????q8??
?/
????
Uc2E9{?Tfj?f??
?s2?.yq?
????????nutx??????????tX?????????tq?@?y?
pR>9FXjtn|8?n_aozj3k?s???y{srlvos??LE???ew?}5y???dk??]???wv????CdfiP????
?????r"=~??_??b=??p???^/m?x(
??x?
8Q(8.d?DRS?Q??j?\'?fZ?


(ad infinitum). Was it maybe supposed to have a different type of encoding,
or URL address? If there's another way to access it in Safari, I can't find
it. If I remove the "MacPython.icns" from the URL I get a Forbidden. If I
remove just the ".icns", I get a nice 404. If I remove "mac/MacPython.icns"
I get to your main page, but don't see a link.
> 
> But Ron objects:
>> The page looks fine, but it might be better to remove the icon or
>> replace it by the current MacPython icon (the 16 ton weight).
> 
> The 16-ton weight screams "done by amateurs for a toy computer".  The
> sooner it's stamped out, the better.
> 
> And further explains:
>> It's not that I don't like the snake, but some people claim a
>> lifelike snake will scare some (potential) users and furthermore the
>> snake-icon isn't actually used by MacPython (yet?).
> 
> As we know, some people will claim anything.  The snake/apple icon is
> all over the MacPython pages, that's where it comes from.  The new
> Python web site has finally switched to a snake-based logo, as well.
> 
> As for scaring users:  I feel sympathy for folks who have pathological
> fears of almost anything, including drawings of snakes.  I think,
> though, it might be kinder to repel those folks immediately, rather
> than helping them overcome their incipient nausea at the very idea of
> a language called "Python", and getting to use it.  Such use might
> cause a nasty amount of subconscious angst, expressed in -- who knows
> -- continuous nightmares?  Random acts of violence?
> 
> Surely better to act up front!  More snakes on the web page!

 The 16-ton icon looks horrible. I'm keen to see the snake, if you can
provide a link that works. Thanks.

But I think Donovan has the right idea. The new official Python icon looks
great. This is Python, use the Python logo.



-- 
Paul Berkowitz



From kevino at theolliviers.com  Thu Mar  2 19:57:53 2006
From: kevino at theolliviers.com (Kevin Ollivier)
Date: Thu, 2 Mar 2006 10:57:53 -0800
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <433A1F5F-0A70-41F8-9577-328C79E8D4FC@ulaluma.com>
References: <06Mar2.095553pst."58633"@synergy1.parc.xerox.com>
	<433A1F5F-0A70-41F8-9577-328C79E8D4FC@ulaluma.com>
Message-ID: <AF3FA733-6900-4F92-98A3-94EDC68D0823@theolliviers.com>

Hi Donovan,

On Mar 2, 2006, at 10:08 AM, Donovan Preston wrote:

>
> On Mar 2, 2006, at 9:55 AM, Bill Janssen wrote:
>
>> Ron Oussoren writes:
>>> The 16 ton icon stays until someone brings a better alternative.
>
> Stop with the 16 ton weight/snakes/apples debate. Just use the new,
> official, python logo, which is a stylized icon that isn't going to
> offend anybody, is vector so will look good at any size, and looks
> very, very professional. I don't see any reason to indicate macness/
> appleness in the logo. It's python. You are using it on a Mac.
> Everyone already knows Apple makes macs.
>
> http://beta.python.org/images/python-logo.gif

I'm actually in talks with a designer about creating an icon, and so  
far things are going well. I hope in a week or so to have something  
to show.

As for the beta Python icon, it really doesn't symbolize Python at  
all to me; Graphically, it's not horrible or anything, but what comes  
to mind when I look at that are "rigid/stiff", and "computation".  
Those snakes don't look very comfortable to me. ;-) It's worlds apart  
from, say, the Firefox icon. Fit that icon against the background of  
a nearly colorless page, the whole page says bland, boring, for the  
suits. It's one thing to be professional, another entirely to have a  
site that is not memorable nor makes a strong impression of any sort.  
(The best part of the page is the "Using Python for" and you have to  
scroll down to see that.) Well, I could go on, but anyway, neither  
the corporate-style logo, or the snake squeezing the Apple, seem to  
me to fit with Apple's HIG nor I think do they convey what Python is  
about. (And I have serious doubts that Apple, if they noticed the  
snake/apple icon, would allow it to get into the official distro  
anyway.)

Anyway, of course the issue is whether we can come up with some thing  
better for the icon. I'll let people know when I have something  
concrete to show.

Thanks,

Kevin

> dp
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Thu Mar  2 20:33:35 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 02 Mar 2006 11:33:35 -0800
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <433A1F5F-0A70-41F8-9577-328C79E8D4FC@ulaluma.com>
References: <06Mar2.095553pst.58633@synergy1.parc.xerox.com>
	<433A1F5F-0A70-41F8-9577-328C79E8D4FC@ulaluma.com>
Message-ID: <4407488F.6080304@noaa.gov>



Donovan Preston wrote:

> Stop with the 16 ton weight/snakes/apples debate. Just use the new,  
> official, python logo,

+1



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Thu Mar  2 21:44:36 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 2 Mar 2006 21:44:36 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar2.095553pst.58633@synergy1.parc.xerox.com>
References: <06Mar2.095553pst.58633@synergy1.parc.xerox.com>
Message-ID: <361B524F-11D8-479D-BB25-536059CA4201@mac.com>


On 2-mrt-2006, at 18:55, Bill Janssen wrote:

> Ron Oussoren writes:
>> The 16 ton icon stays until someone brings a better alternative.
>
> W. R. Wing writes:
>> I LIKE the python squeezing the apple.  The 16 ton icon is not only
>> amateurish, it is so indirect an "in" joke as to be inscrutable to
>> anyone coming new to the party.
>
> I agree with Wing.  So there *is* a better alternative, and there's
> now a MacPython.icns file at
> http://bill.janssen.org/mac/MacPython.icns with the snake wrapped
> around the Apple.  Feel free to use it where it fits.

That's just one icon, we need several icons. At least we now have a
consistent set of icons. An historical looking set of icons, but a *set*
of icons.

>
> Surely better to act up front!  More snakes on the web page!

But Python isn't named after snakes ;-) ;-)

Ronald

>
> Bill


From skip at pobox.com  Thu Mar  2 23:14:07 2006
From: skip at pobox.com (skip at pobox.com)
Date: Thu, 2 Mar 2006 16:14:07 -0600
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <361B524F-11D8-479D-BB25-536059CA4201@mac.com>
References: <06Mar2.095553pst.58633@synergy1.parc.xerox.com>
	<361B524F-11D8-479D-BB25-536059CA4201@mac.com>
Message-ID: <17415.28207.707096.648246@montanaro.dyndns.org>


    >> Surely better to act up front!  More snakes on the web page!

    Ronald> But Python isn't named after snakes ;-) ;-)

True enough, but I think even Guido has abandoned that argument.  Had it
been named "Monty Python", "Cleese" or "Cheeseshop" you might have a leg to
stand on, but it's just "Python".  Many people younger than 30 will not
understand the reference.  Most people outside North America and Europe
won't know what it means either.  I think you just have to accept that
snakes are going to be part of the package.

Skip

From janssen at parc.com  Thu Mar  2 23:28:24 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 2 Mar 2006 14:28:24 PST
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: Your message of "Thu, 02 Mar 2006 12:44:36 PST."
	<361B524F-11D8-479D-BB25-536059CA4201@mac.com> 
Message-ID: <06Mar2.142833pst."58633"@synergy1.parc.xerox.com>

> That's just one icon, we need several icons. At least we now have a
> consistent set of icons. An historical looking set of icons, but a *set*
> of icons.

Yes, I comprehend, several different icons for the various
applications are needed.  The base logo is just a starting place.

> > Surely better to act up front!  More snakes on the web page!
> 
> But Python isn't named after snakes ;-) ;-)

Doesn't matter.  The dominant connotation is snakish, and therefore
scary to susceptible folk.  I think they should be gently steered
towards "Lisp" (unfortunate connotation there) or perhaps "Ruby" (hard
to argue with that connotation).

Bill

From kevino at theolliviers.com  Fri Mar  3 00:27:21 2006
From: kevino at theolliviers.com (Kevin Ollivier)
Date: Thu, 2 Mar 2006 15:27:21 -0800
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar2.142833pst."58633"@synergy1.parc.xerox.com>
References: <06Mar2.142833pst."58633"@synergy1.parc.xerox.com>
Message-ID: <0C4B4A8A-98B4-409F-8C8B-4A251E7C4B7D@theolliviers.com>

Hi Bill,

On Mar 2, 2006, at 2:28 PM, Bill Janssen wrote:

>> That's just one icon, we need several icons. At least we now have a
>> consistent set of icons. An historical looking set of icons, but a  
>> *set*
>> of icons.
>
> Yes, I comprehend, several different icons for the various
> applications are needed.  The base logo is just a starting place.
>
>>> Surely better to act up front!  More snakes on the web page!
>>
>> But Python isn't named after snakes ;-) ;-)
>
> Doesn't matter.  The dominant connotation is snakish, and therefore
> scary to susceptible folk.  I think they should be gently steered
> towards "Lisp" (unfortunate connotation there) or perhaps "Ruby" (hard
> to argue with that connotation).

Okay, this is all getting a bit silly, but I'll respond just for the  
record. With all due respect, I think you're being condescending. No  
one's going to freak out by seeing the PyGame picture, or the  
mod_python Apache picture, or the beta.python.org logo, or the  
Windows icon, or need I go on? They don't scare or bother anyone, and  
I have yet to hear of anyone who is bothered by the very word "Python".

The issue is with that particular icon. It doesn't convey, shall we  
say, positive imagery. Scary, dangerous, and "the snake as temptor  
and devil" are not I think ideas which accurately represent the  
Python programming language, nor accurately reflect the outlook of  
its community. Do you?

Maybe you think it's a joke that some people would see the icon as  
creepy, or downright offensive, but whether or not the imagery in the  
icon was intended (which I doubt it was, except perhaps as a joke),  
it is there, and some people viewing it are certainly going to think  
it was intentional. If you want a new icon, and you all are simply  
unable to be patient and see what I'm able to come up with, do what  
Donovan said and go with the official Python icon. At least no one  
will be scared or offended by that.

Of course, if you do so and say the issue is settled, I'm going to be  
more than a little upset because I was asked to look into generating  
an icon, I have done so and even found someone who has already done  
some initial work on it, and I'm going to have to tell this person to  
scrap their work simply because you guys are impatient. We're not  
working on your dime, you know. You and Chris asked for a favor and I  
agreed, and in fact I put a fair amount of time into researching what  
was out there and finally found someone interested a few days ago,  
but then when I can't produce something for you at lightning speed  
you just want to say forget it? I mean, seriously, you can't wait a  
few weeks on changing an icon that has been used for what, almost a  
decade?

Kevin

> Bill
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From janssen at parc.com  Fri Mar  3 01:11:38 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 2 Mar 2006 16:11:38 PST
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: Your message of "Thu, 02 Mar 2006 15:27:21 PST."
	<0C4B4A8A-98B4-409F-8C8B-4A251E7C4B7D@theolliviers.com> 
Message-ID: <06Mar2.161138pst."58633"@synergy1.parc.xerox.com>

> Of course, if you do so and say the issue is settled, I'm going to be  
> more than a little upset because I was asked to look into generating  
> an icon, I have done so and even found someone who has already done  
> some initial work on it, and I'm going to have to tell this person to  
> scrap their work simply because you guys are impatient. We're not  
> working on your dime, you know. You and Chris asked for a favor and I  
> agreed, and in fact I put a fair amount of time into researching what  
> was out there and finally found someone interested a few days ago,  
> but then when I can't produce something for you at lightning speed  
> you just want to say forget it? I mean, seriously, you can't wait a  
> few weeks on changing an icon that has been used for what, almost a  
> decade?

Kevin,

I can't speak for Ron or W.R., but for my part, I'm perfectly willing
to wait and see what you come up with -- I think the ideas you floated
are interesting, I appreciate the work you are doing on this front,
and I've got no vested interest in the snake-around-the-apple logo.  I
just found it on the pythonmac.org wiki, and slapped it into the
download page.  Happy to change it to something else, too -- no great
amount of effort required.

I *am* interested in getting rid of the 16-ton weight, though.

Bill

From ronaldoussoren at mac.com  Fri Mar  3 16:27:06 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Fri, 03 Mar 2006 16:27:06 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar2.161138pst.58633@synergy1.parc.xerox.com>
References: <06Mar2.161138pst.58633@synergy1.parc.xerox.com>
Message-ID: <9147587.1141399626657.JavaMail.ronaldoussoren@mac.com>

 
On Friday, March 03, 2006, at 01:11AM, Bill Janssen <janssen at parc.com> wrote:

>> Of course, if you do so and say the issue is settled, I'm going to be  
>> more than a little upset because I was asked to look into generating  
>> an icon, I have done so and even found someone who has already done  
>> some initial work on it, and I'm going to have to tell this person to  
>> scrap their work simply because you guys are impatient. We're not  
>> working on your dime, you know. You and Chris asked for a favor and I  
>> agreed, and in fact I put a fair amount of time into researching what  
>> was out there and finally found someone interested a few days ago,  
>> but then when I can't produce something for you at lightning speed  
>> you just want to say forget it? I mean, seriously, you can't wait a  
>> few weeks on changing an icon that has been used for what, almost a  
>> decade?
>
>Kevin,
>
>I can't speak for Ron or W.R., but for my part, I'm perfectly willing
>to wait and see what you come up with -- I think the ideas you floated
>are interesting, I appreciate the work you are doing on this front,
>and I've got no vested interest in the snake-around-the-apple logo.  I
>just found it on the pythonmac.org wiki, and slapped it into the
>download page.  Happy to change it to something else, too -- no great
>amount of effort required.

BTW.  I do know about the American tendency to go for shortened names, but my name is Ronald. Please type the additional 4 characters. 

With that pet peeve out of the way, I'd be happy to replace the current set of  icons by something else. I'm just not happy about replacing just one icon and ignoring the rest, what some people seem to suggest. The new icons should also be an improvement, but that shouldn't be too hard when someone with actual grafical skills is involved :-). 

While I do like the Monty Python connotations of the current icon set, I wouldn't mind moving to something more snake related. For better or worse, the world associates Python with the snake instead of the comedians and that means the 16 ton weight won't be understood by the majority of Python users.

Ronald

>
>I *am* interested in getting rid of the 16-ton weight, though.
>
>Bill
>
>

From ronaldoussoren at mac.com  Fri Mar  3 18:24:36 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Fri, 3 Mar 2006 18:24:36 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <0C4B4A8A-98B4-409F-8C8B-4A251E7C4B7D@theolliviers.com>
References: <06Mar2.142833pst.58633@synergy1.parc.xerox.com>
	<0C4B4A8A-98B4-409F-8C8B-4A251E7C4B7D@theolliviers.com>
Message-ID: <B203AE89-4D25-4933-9C33-8CCC94CC761E@mac.com>


On 3-mrt-2006, at 0:27, Kevin Ollivier wrote:
>
> Of course, if you do so and say the issue is settled, I'm going to  
> be more than a little upset because I was asked to look into  
> generating an icon, I have done so and even found someone who has  
> already done some initial work on it, and I'm going to have to tell  
> this person to scrap their work simply because you guys are  
> impatient. We're not working on your dime, you know. You and Chris  
> asked for a favor and I agreed, and in fact I put a fair amount of  
> time into researching what was out there and finally found someone  
> interested a few days ago, but then when I can't produce something  
> for you at lightning speed you just want to say forget it? I mean,  
> seriously, you can't wait a few weeks on changing an icon that has  
> been used for what, almost a decade?

Please don't stop working on an icon.  The current icon-set has been  
ancient looking for years now, there's no need to rush this. I  
wouldn't mind if the new icons wouldn't be introduced until Python  
2.5, that should give us enough time to pick a new icon. Rushing this  
would only increase the chance that we'd pick an icon that we don't  
really like and have to change again soon.

And for the record: I'm not really thrilled by the icon on  
beta.python.org. I can live with it if it would get picked, but it  
isn't a very compelling image.

Ronald

>
> Kevin
>
>> Bill
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>


From leknarf at pacbell.net  Fri Mar  3 21:02:47 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Fri, 3 Mar 2006 12:02:47 -0800
Subject: [Pythonmac-SIG] PyObjC versions
Message-ID: <92F76BC6-F128-40C7-B6C6-7D831E3996F4@pacbell.net>


Exploring the brave new world of Python GUIs with PyObjC and I'm  
stymied at the download page.  This doesn't bode well ;)

Looking at http://pyobjc.sourceforge.net/software/, I've got download  
choices for a number of Python & OSX versions.  None match my setup:

	Python 2.4.1
	OSX 10.4.5

Which should I download?

Thanks in advance!
Scott






From sw at wordtech-software.com  Fri Mar  3 21:08:00 2006
From: sw at wordtech-software.com (Kevin Walzer)
Date: Fri, 03 Mar 2006 15:08:00 -0500
Subject: [Pythonmac-SIG] wxPython & "universal" Python
Message-ID: <4408A220.50906@wordtech-software.com>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

My installation of the "universal" Python installer went smoothly
(thanks Ronald and Bob!). However, it broke my existing installation of
wxPython 2.6.2. Should I just reinstall the current version of
wxPython-Mac from Sourceforge, or is there a "universal" build of
wxPython in the works?

- --
Kevin Walzer
iReveal: File Search Tool
http://www.wordtech-software.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFECKIfJmdQs+6YVcoRAp8MAJ9i0QPwQwOXPWBOKaeQ1jBwzi4AnwCffQJf
/bwff4E/wB7gkdubv7jcHT0=
=Ip5+
-----END PGP SIGNATURE-----

From bob at redivi.com  Fri Mar  3 21:15:22 2006
From: bob at redivi.com (Bob Ippolito)
Date: Fri, 3 Mar 2006 12:15:22 -0800
Subject: [Pythonmac-SIG] PyObjC versions
In-Reply-To: <92F76BC6-F128-40C7-B6C6-7D831E3996F4@pacbell.net>
References: <92F76BC6-F128-40C7-B6C6-7D831E3996F4@pacbell.net>
Message-ID: <59B25A03-2CFA-4C67-99AA-8C44DA1816A0@redivi.com>


On Mar 3, 2006, at 12:02 PM, Scott Frankel wrote:

>
> Exploring the brave new world of Python GUIs with PyObjC and I'm
> stymied at the download page.  This doesn't bode well ;)
>
> Looking at http://pyobjc.sourceforge.net/software/, I've got download
> choices for a number of Python & OSX versions.  None match my setup:
>
> 	Python 2.4.1
> 	OSX 10.4.5
>
> Which should I download?

Get the one for Python 2.4 for OS X 10.3.  For third party builds of  
Python, the OS version is the minimum compatible version.  It does  
not need to be an exact match!

-bob




From bob at redivi.com  Fri Mar  3 21:17:22 2006
From: bob at redivi.com (Bob Ippolito)
Date: Fri, 3 Mar 2006 12:17:22 -0800
Subject: [Pythonmac-SIG] wxPython & "universal" Python
In-Reply-To: <4408A220.50906@wordtech-software.com>
References: <4408A220.50906@wordtech-software.com>
Message-ID: <0C620D9D-626C-40DC-B923-1379F18648C7@redivi.com>


On Mar 3, 2006, at 12:08 PM, Kevin Walzer wrote:

> My installation of the "universal" Python installer went smoothly
> (thanks Ronald and Bob!). However, it broke my existing  
> installation of
> wxPython 2.6.2. Should I just reinstall the current version of
> wxPython-Mac from Sourceforge, or is there a "universal" build of
> wxPython in the works?

Reinstalling your extensions should work, but those extensions will  
be PPC only and will of course prevent you from building universal  
apps that depend on them.

-bob


From kevino at theolliviers.com  Fri Mar  3 21:21:36 2006
From: kevino at theolliviers.com (Kevin Ollivier)
Date: Fri, 3 Mar 2006 12:21:36 -0800
Subject: [Pythonmac-SIG] wxPython & "universal" Python
In-Reply-To: <4408A220.50906@wordtech-software.com>
References: <4408A220.50906@wordtech-software.com>
Message-ID: <6C5F8D91-9D16-4BA6-9CDD-DAB0740B5D78@theolliviers.com>

Hi Kevin,

On Mar 3, 2006, at 12:08 PM, Kevin Walzer wrote:

> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> My installation of the "universal" Python installer went smoothly
> (thanks Ronald and Bob!). However, it broke my existing  
> installation of
> wxPython 2.6.2. Should I just reinstall the current version of
> wxPython-Mac from Sourceforge, or is there a "universal" build of
> wxPython in the works?

wxPython 2.6.3, due out probably in a couple weeks, should support a  
Universal binary, although it will probably be Tiger only. Hopefully  
in the next few days I'll be able to iron out any issues. If you  
don't want to wait, you can checkout wxWidgets using the  
WX_2_6_BRANCH tag, then follow Robin's build instructions, except you  
also need to add --enable-universal_binary to the configure flags,  
and see if things work okay.

Thanks,

Kevin

> - --
> Kevin Walzer
> iReveal: File Search Tool
> http://www.wordtech-software.com
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.1 (Darwin)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFECKIfJmdQs+6YVcoRAp8MAJ9i0QPwQwOXPWBOKaeQ1jBwzi4AnwCffQJf
> /bwff4E/wB7gkdubv7jcHT0=
> =Ip5+
> -----END PGP SIGNATURE-----
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From leknarf at pacbell.net  Fri Mar  3 21:29:15 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Fri, 3 Mar 2006 12:29:15 -0800
Subject: [Pythonmac-SIG] PyObjC versions
In-Reply-To: <59B25A03-2CFA-4C67-99AA-8C44DA1816A0@redivi.com>
References: <92F76BC6-F128-40C7-B6C6-7D831E3996F4@pacbell.net>
	<59B25A03-2CFA-4C67-99AA-8C44DA1816A0@redivi.com>
Message-ID: <297C079D-9DFA-4D4E-8DD2-67B1A17AFFBF@pacbell.net>


Got it.  Thanks!


On Mar 3, 2006, at 12:15 PM, Bob Ippolito wrote:

>
> On Mar 3, 2006, at 12:02 PM, Scott Frankel wrote:
>
>>
>> Exploring the brave new world of Python GUIs with PyObjC and I'm
>> stymied at the download page.  This doesn't bode well ;)
>>
>> Looking at http://pyobjc.sourceforge.net/software/, I've got download
>> choices for a number of Python & OSX versions.  None match my setup:
>>
>> 	Python 2.4.1
>> 	OSX 10.4.5
>>
>> Which should I download?
>
> Get the one for Python 2.4 for OS X 10.3.  For third party builds  
> of Python, the OS version is the minimum compatible version.  It  
> does not need to be an exact match!
>
> -bob
>
>
>


From hengist.podd at virgin.net  Sat Mar  4 18:20:36 2006
From: hengist.podd at virgin.net (has)
Date: Sat, 4 Mar 2006 17:20:36 +0000
Subject: [Pythonmac-SIG] Intel/PyMac_GetOSType status?
Message-ID: <a05200f01c029fc30333c@[82.3.255.67]>

Hi folks,

Just wondering what progress (if any) there's been on fixing the endian bugs in PyMac_GetOSType and PyMac_BuildOSType?

https://sourceforge.net/tracker/?func=detail&atid=305470&aid=1218421&group_id=5470

Ta,

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From bob at redivi.com  Sat Mar  4 20:56:39 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sat, 4 Mar 2006 11:56:39 -0800
Subject: [Pythonmac-SIG] Intel/PyMac_GetOSType status?
In-Reply-To: <a05200f01c029fc30333c@[82.3.255.67]>
References: <a05200f01c029fc30333c@[82.3.255.67]>
Message-ID: <7922E229-36C1-49C4-8560-9952590D6E55@redivi.com>


On Mar 4, 2006, at 9:20 AM, has wrote:

> Hi folks,
>
> Just wondering what progress (if any) there's been on fixing the  
> endian bugs in PyMac_GetOSType and PyMac_BuildOSType?
>
> https://sourceforge.net/tracker/? 
> func=detail&atid=305470&aid=1218421&group_id=5470

That's been fixed in the universal branch for some time now.

-bob


From brendansimons at yahoo.ca  Sun Mar  5 01:36:01 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Sat, 4 Mar 2006 19:36:01 -0500
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page on
	www.python.org now updated)
In-Reply-To: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
Message-ID: <4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>

I spent about 5 hours today drawing up my own icon idea, but my poor  
illustrator skills thwarted my attempts (Corel, why hast thy forsaken  
the mac?)

Instead, I'm endorsing the official icon route.  I've made a first  
cut here:
Python Documents:  http://twototango.blogs.com/PythonDocument.icns
Python Launcher:  http://twototango.blogs.com/PythonLauncher.icns
Idle: http://twototango.blogs.com/Idle.icns

There's a bunch I'd like to do still:
  - Rebuild them with a high res copy of the official logo
  - Make low-res versions
  - Tint .py files different than .pyc and .pyo so they are easier to  
pick out of a folder
  - Add the code lines to the Python Documents icons
  - Find a better "floating slab" for the python launcher icon.  I  
know Apple uses one I can crib somewhere

I've run out of time now though.  I can follow through later if  
there's interest from this group.

    -Brendan


On 3-Mar-06, at 6:00 AM, pythonmac-sig-request at python.org wrote:

> Donovan Preston wrote:
>
>> Stop with the 16 ton weight/snakes/apples debate. Just use the  
>> new,  official, python logo,
>
> +1
>
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                     		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060304/c9f87ebd/attachment.htm 

From brendansimons at yahoo.ca  Sun Mar  5 01:43:21 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Sat, 4 Mar 2006 19:43:21 -0500
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page
	on www.python.org now updated)
In-Reply-To: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
Message-ID: <29944A53-67E7-4E06-9603-5D1F23ADB4A5@yahoo.ca>

Hmmm, Typepad is doing something funny with the mimetypes on those  
links.  If you have trouble opening them, here are 128x128 tiff  
versions:

http://twototango.blogs.com/PythonDocument.tif
http://twototango.blogs.com/Idle.tif
http://twototango.blogs.com/PythonLauncher.tif

  - Brendan

> I spent about 5 hours today drawing up my own icon idea, but my  
> poor illustrator skills thwarted my attempts (Corel, why hast thy  
> forsaken the mac?)
>
> Instead, I'm endorsing the official icon route.  I've made a first  
> cut here:
> Python Documents:  http://twototango.blogs.com/PythonDocument.icns
> Python Launcher:  http://twototango.blogs.com/PythonLauncher.icns
> Idle: http://twototango.blogs.com/Idle.icns
>
> There's a bunch I'd like to do still:
>  - Rebuild them with a high res copy of the official logo
>  - Make low-res versions
>  - Tint .py files different than .pyc and .pyo so they are easier  
> to pick out of a folder
>  - Add the code lines to the Python Documents icons
>  - Find a better "floating slab" for the python launcher icon.  I  
> know Apple uses one I can crib somewhere
>
> I've run out of time now though.  I can follow through later if  
> there's interest from this group.
>
>    -Brendan
>
>
> On 3-Mar-06, at 6:00 AM, pythonmac-sig-request at python.org wrote:
>
>> Donovan Preston wrote:
>>
>>> Stop with the 16 ton weight/snakes/apples debate. Just use the  
>>> new,  official, python logo,
>>
>> +1
>>
>>
>>
>> -- 
>> Christopher Barker, Ph.D.
>> Oceanographer
>>                                     		
>> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
>> 7600 Sand Point Way NE   (206) 526-6329   fax
>> Seattle, WA  98115       (206) 526-6317   main reception
>>
>> Chris.Barker at noaa.gov

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060304/edc32986/attachment.html 

From skip at pobox.com  Sun Mar  5 01:49:31 2006
From: skip at pobox.com (skip at pobox.com)
Date: Sat, 4 Mar 2006 18:49:31 -0600
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page
 on www.python.org now updated)
In-Reply-To: <4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
	<4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
Message-ID: <17418.13723.174976.832218@montanaro.dyndns.org>


    Brendan> Python Documents:  http://twototango.blogs.com/PythonDocument.icns
    Brendan> Python Launcher:  http://twototango.blogs.com/PythonLauncher.icns
    Brendan> Idle: http://twototango.blogs.com/Idle.icns

Dumb question probably, but what reads/writes those files?

Skip

From bob at redivi.com  Sun Mar  5 02:26:42 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sat, 4 Mar 2006 17:26:42 -0800
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page
	on www.python.org now updated)
In-Reply-To: <17418.13723.174976.832218@montanaro.dyndns.org>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
	<4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
	<17418.13723.174976.832218@montanaro.dyndns.org>
Message-ID: <309804F7-F6C5-425E-9AC5-A9EBB1CD03C7@redivi.com>


On Mar 4, 2006, at 4:49 PM, skip at pobox.com wrote:

>
>     Brendan> Python Documents:  http://twototango.blogs.com/ 
> PythonDocument.icns
>     Brendan> Python Launcher:  http://twototango.blogs.com/ 
> PythonLauncher.icns
>     Brendan> Idle: http://twototango.blogs.com/Idle.icns
>
> Dumb question probably, but what reads/writes those files?

/Developer/Applications/Utilities/Icon Composer.app writes them..  
just about anything Cocoa-based can read them.  PIL can read the  
highest resolution image out of them, too.

-bob


From brendansimons at yahoo.ca  Sun Mar  5 07:57:14 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Sun, 5 Mar 2006 01:57:14 -0500
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page
	on www.python.org now updated)
In-Reply-To: <17418.13723.174976.832218@montanaro.dyndns.org>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
	<4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
	<17418.13723.174976.832218@montanaro.dyndns.org>
Message-ID: <5A2CF681-864E-4415-939E-6C375C79C162@yahoo.ca>

Preview also.

  - Brendan

On 4-Mar-06, at 7:49 PM, skip at pobox.com wrote:

>
>     Brendan> Python Documents:  http://twototango.blogs.com/ 
> PythonDocument.icns
>     Brendan> Python Launcher:  http://twototango.blogs.com/ 
> PythonLauncher.icns
>     Brendan> Idle: http://twototango.blogs.com/Idle.icns
>
> Dumb question probably, but what reads/writes those files?
>
> Skip

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

From brendansimons at yahoo.ca  Sun Mar  5 08:04:29 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Sun, 5 Mar 2006 02:04:29 -0500
Subject: [Pythonmac-SIG] Icons for universal build (was: Download page
	on www.python.org now updated)
In-Reply-To: <17418.13723.174976.832218@montanaro.dyndns.org>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
	<4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
	<17418.13723.174976.832218@montanaro.dyndns.org>
Message-ID: <AEC72854-1710-448A-AE9C-5A832AF7FBEB@yahoo.ca>

For interest's sake, here's the one I tried to do.  (I'm not  
advocating it, but I put so much work into it ;)

http://twototango.blogs.com/PythonLauncherBlech.tiff

Brendan

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

From zpincus at stanford.edu  Sun Mar  5 08:46:51 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sat, 4 Mar 2006 23:46:51 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
Message-ID: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>

Hi folks,

I have a tricky issue with some python extension modules that need to  
share symbols between themselves.

The setup: I am creating python modules to wrap a large template- 
based C++ image processing library. The fundamental class that each  
of these modules needs to be able to create, pass around, and  
dynamic_cast from and to, is a templated image type.

This requires that at module load time, the C++ typeinfo objects  
describing the template instantiations of that class be loaded into  
the global symbol table and shared between all of the modules.  
Otherwise, each module will have its own typeinfo object for the  
class, causing dynamic_cast and other comparisons of object identity  
to fail (because GCC uses the address of the typeinfo object to make  
such comparisons).

I compiled the modules so that all the symbols are visible (nm -m  
tells me that the tyepinfo symbols are all declared as 'weak  
external', which is what they should be for sharing between modules).  
Before I load the modules in Python, I call sys.setdlopenflags(0xA),  
which is RTDL_NOW|RTDL_GLOBAL, as these constants are defined in  
dlfcn.h on OS X.

This is what I have been told is necessary and sufficient for this to  
work, yet somehow it is not working for me. When a collaborator  
compiles this all on Linux with GCC, things work for him (note that  
his dlopenflags are 0x102, which corresponds to RTDL_NOW|RTDL_GLOBAL  
on his system).

Has anyone else wrestled with these issues? Does python not respect  
sys.setdlopenflags() on OS X in some cases? Is there anything else I  
should try to do?

I'm on OS X 10.4.5, using python 2.4.1 compiled as a framework (built  
with GCC4, as are all of the modules I'm working on).

Any help or suggestions or thoughts would be greatly appreciated,


Zach Pincus

Program in Biomedical Informatics and Department of Biochemistry
Stanford University School of Medicine


From ronaldoussoren at mac.com  Sun Mar  5 11:05:03 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sun, 5 Mar 2006 11:05:03 +0100
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
Message-ID: <D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>


On 5-mrt-2006, at 8:46, Zachary Pincus wrote:

> Hi folks,
>
>
> Any help or suggestions or thoughts would be greatly appreciated,

Don't do that?

Python uses two-level namespaces on OSX, and for a purpose: to avoid
accidentally picking up a symbol from another extension. IIRC You should
build a shared library containing the shared code and link all  
extensions
to that.

Ronald

>
>
> Zach Pincus
>
> Program in Biomedical Informatics and Department of Biochemistry
> Stanford University School of Medicine
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From hengist.podd at virgin.net  Sun Mar  5 14:52:41 2006
From: hengist.podd at virgin.net (has)
Date: Sun, 5 Mar 2006 13:52:41 +0000
Subject: [Pythonmac-SIG] Intel/PyMac_GetOSType status?
In-Reply-To: <7922E229-36C1-49C4-8560-9952590D6E55@redivi.com>
References: <a05200f01c029fc30333c@[82.3.255.67]>
	<7922E229-36C1-49C4-8560-9952590D6E55@redivi.com>
Message-ID: <a05200f04c0309c9ec06d@[82.9.16.53]>

Bob wrote:

>>Just wondering what progress (if any) there's been on fixing the endian bugs in PyMac_GetOSType and PyMac_BuildOSType?
>>
>>https://sourceforge.net/tracker/?func=detail&atid=305470&aid=1218421&group_id=5470
>
>That's been fixed in the universal branch for some time now.

Cool. Mind if I pinch the relevant code and kludge it into the CarbonX AE and OSA extensions? Seems the simplest way to maintain compatibility with Apple-installed 2.3.5 on Intel Mac.

Thanks,

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From hraban at fiee.net  Sun Mar  5 19:49:31 2006
From: hraban at fiee.net (Henning Hraban Ramm)
Date: Sun, 5 Mar 2006 19:49:31 +0100
Subject: [Pythonmac-SIG] Icons for universal build
In-Reply-To: <AEC72854-1710-448A-AE9C-5A832AF7FBEB@yahoo.ca>
References: <mailman.8.1141383602.28723.pythonmac-sig@python.org>
	<4F8384F1-A33D-42C1-AA84-421677259021@yahoo.ca>
	<17418.13723.174976.832218@montanaro.dyndns.org>
	<AEC72854-1710-448A-AE9C-5A832AF7FBEB@yahoo.ca>
Message-ID: <71F7C093-6F5B-436B-ADC4-CAFE2A225AE3@fiee.net>

Am 2006-03-05 um 08:04 schrieb Brendan Simons:

> For interest's sake, here's the one I tried to do.  (I'm not
> advocating it, but I put so much work into it ;)
>
> http://twototango.blogs.com/PythonLauncherBlech.tiff

How about some Jack-in-the-box python icon for the Launcher?
(i mean: snake with a spiral spring tail coming out of a box)


Greetlings from Lake Constance!
Hraban
---
http://www.fiee.net
http://www.cacert.org (I'm an assurer)



From bob at redivi.com  Sun Mar  5 20:36:10 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sun, 5 Mar 2006 11:36:10 -0800
Subject: [Pythonmac-SIG] Intel/PyMac_GetOSType status?
In-Reply-To: <a05200f04c0309c9ec06d@[82.9.16.53]>
References: <a05200f01c029fc30333c@[82.3.255.67]>
	<7922E229-36C1-49C4-8560-9952590D6E55@redivi.com>
	<a05200f04c0309c9ec06d@[82.9.16.53]>
Message-ID: <BEF2C0F6-9953-4B3B-A096-D6D8F3CC5CF9@redivi.com>


On Mar 5, 2006, at 5:52 AM, has wrote:

> Bob wrote:
>
>>> Just wondering what progress (if any) there's been on fixing the  
>>> endian bugs in PyMac_GetOSType and PyMac_BuildOSType?
>>>
>>> https://sourceforge.net/tracker/? 
>>> func=detail&atid=305470&aid=1218421&group_id=5470
>>
>> That's been fixed in the universal branch for some time now.
>
> Cool. Mind if I pinch the relevant code and kludge it into the  
> CarbonX AE and OSA extensions? Seems the simplest way to maintain  
> compatibility with Apple-installed 2.3.5 on Intel Mac.

It's a really bad idea to even bother trying to support that.  What  
you should do is file bugs with Apple until they fix it, rather than  
trying to patch over the egregious bugs in all of the Mac extension  
modules.  Tell your users that in order to support Intel, they must  
use a third party build that isn't broken.

-bob


From zpincus at stanford.edu  Sun Mar  5 21:32:36 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sun, 5 Mar 2006 12:32:36 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
Message-ID: <D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>

>> [need to share symbols across dso boundaries snipped]
>> Any help or suggestions or thoughts would be greatly appreciated,
>
> Don't do that?
>
> Python uses two-level namespaces on OSX, and for a purpose: to avoid
> accidentally picking up a symbol from another extension. IIRC You  
> should
> build a shared library containing the shared code and link all  
> extensions
> to that.

Hmm, this is not exactly what I was hoping to hear.

About building python with a two-level namespace: isn't the best way  
to not pick up symbols from other extensions to not load the  
extensions with the RTLD_GLOBAL flag? Solving the problem with two- 
level namespaces seems a bit extreme... though perhaps there are  
other good reasons for this?

The python documentation for sys.setdlopenflags() specifically says  
that the sharing of symbols across modules can be enabled by the use  
of the RTLD_GLOBAL flag. It seems like Mac Python is at variance with  
this, and with the fact that modules can be set up on Linux Python to  
share symbols.

As to your first piece of advice: 'Don't do that,' unfortunately that  
is not an option. First, the modules need to support a plugin  
architecture where other modules created in the future will be able  
to pass these objects between them. Having to rebuild a mammoth  
central library each time someone wishes to add a plugin is not  
particularly attractive, and definitely not particularly dynamic.  
Second, the full set of core modules (~10 modules), if all linked  
together to form one central library, would be over 300 megabytes  
already, which totally chokes the loader.

The truth is that I really need to have multiple shared libraries,  
and they need to share the C++ typeinfo symbols between them. And  
*anyone* who wants to implement a plugin architecture with C++ Python  
extensions will need to do the same (or will need to not use C++  
dynamic_casts). Python on Linux supports this fine, and it would be a  
pity if Python on the Mac couldn't.

Does anyone have any additional insight as to (a) why the two-level  
namespace decision was made, (b) how it interacts with plugin  
architectures like I've described, and (c) if there's any way to load  
symbols globally even in the face of the two-level namespace?

Thanks,

Zach







From bob at redivi.com  Sun Mar  5 21:51:18 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sun, 5 Mar 2006 12:51:18 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
Message-ID: <5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>


On Mar 5, 2006, at 12:32 PM, Zachary Pincus wrote:

>>> [need to share symbols across dso boundaries snipped]
>>> Any help or suggestions or thoughts would be greatly appreciated,
>>
>> Don't do that?
>>
>> Python uses two-level namespaces on OSX, and for a purpose: to avoid
>> accidentally picking up a symbol from another extension. IIRC You
>> should
>> build a shared library containing the shared code and link all
>> extensions
>> to that.
>
> Hmm, this is not exactly what I was hoping to hear.
>
> About building python with a two-level namespace: isn't the best way
> to not pick up symbols from other extensions to not load the
> extensions with the RTLD_GLOBAL flag? Solving the problem with two-
> level namespaces seems a bit extreme... though perhaps there are
> other good reasons for this?

Actually it's because Python extensions are bundles.  Bundles are NOT  
shared libraries and can not do what you want.  Additionally, Python  
on Mac OS X doesn't use dl to load its extensions, it uses lower- 
level means... so the RTLD flags aren't going to do anything.

The "best" you can do is to build shared libraries that all of your  
extensions link to.  Phillip J Eby has done some work in order to  
allow setuptools to build shared libraries and to have them work with  
eggs.  There's probably some discussion about that in distutils-sig.

-bob


From martin at moellenbecks.de  Sun Mar  5 22:31:36 2006
From: martin at moellenbecks.de (=?ISO-8859-1?Q?Martin_M=F6llenbeck?=)
Date: Sun, 5 Mar 2006 22:31:36 +0100
Subject: [Pythonmac-SIG] Build ICE 3.0 under Mac OS X 10.4.5 with python
	2.4.1
Message-ID: <75184CDA-3195-406C-99E6-255A8370AD68@moellenbecks.de>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

has anybody tried to build the ICE Library for python under mac osx  
10.4.5.

I have configured the env with:

declare -x DYLD_LIBRARY_PATH=":/share/Projekte/network/Ice-3.0.1/lib"
declare -x ICE_HOME="/share/Projekte/network/Ice-3.0.1"
declare -x PYTHONPATH="/share/Projekte/network/Ice-3.0.1/python:"
declare -x PYTHON_HOME="/Library/Frameworks/Python.framework/Versions/ 
Current/"


After I start the build with 'make', the result was an error like:

iMac:/share/Projekte/network/IcePy-3.0.1 Martin$ make
making all in python
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/LocalException.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Communicator.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/CommunicatorF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Logger.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/LoggerF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/BuiltinSequences.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ObjectAdapter.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ObjectAdapterF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ServantLocator.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ServantLocatorF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Properties.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/PropertiesF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ObjectFactory.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ObjectFactoryF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Identity.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Current.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Router.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/RouterF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Plugin.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/PluginF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Locator.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/LocatorF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/StatsF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Stats.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Process.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ProcessF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/FacetMap.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Connection.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/ConnectionF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/SliceChecksumDict.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Ice_ --ice -- 
no-package -I../slice ../slice/Ice/Endpoint.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/RouterF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/Router.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/SessionF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/Session.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/PermissionsVerifierF.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix Glacier2_ -- 
checksum -I../slice ../slice/Glacier2/PermissionsVerifier.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceBox_ --ice  
- --checksum -I../slice ../slice/IceBox/IceBox.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceGrid_ -- 
ice --checksum -I../slice ../slice/IceGrid/Admin.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceGrid_ -- 
ice --checksum -I../slice ../slice/IceGrid/Descriptor.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceGrid_ -- 
ice --checksum -I../slice ../slice/IceGrid/Exception.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceGrid_ -- 
ice --checksum -I../slice ../slice/IceGrid/Observer.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceGrid_ -- 
ice --checksum -I../slice ../slice/IceGrid/Query.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IcePatch2_ -- 
ice --checksum -I../slice ../slice/IcePatch2/FileInfo.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IcePatch2_ -- 
ice --checksum -I../slice ../slice/IcePatch2/FileServer.ice
/share/Projekte/network/Ice-3.0.1/bin/slice2py --prefix IceStorm_ -- 
ice --checksum -I../slice ../slice/IceStorm/IceStorm.ice
making all in modules
making all in IcePy
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Communicator.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Connection.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Current.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Init.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Logger.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT ObjectAdapter.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT ObjectFactory.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Operation.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Properties.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Proxy.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Slice.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Types.cpp
c++ -c -I.  -I/share/Projekte/network/Ice-3.0.1/include -I/Library/ 
Frameworks/Python.framework/Versions/Current//include/python2.4 -g - 
ftemplate-depth-128 -Wall -D_REENTRANT Util.cpp
rm -f ../../python/IcePy.so.3.0.1
c++ -bundle -g -ftemplate-depth-128 -Wall -D_REENTRANT -L../../python  
- -o ../../python/IcePy.so.3.0.1 Communicator.o Connection.o Current.o  
Init.o Logger.o ObjectAdapter.o ObjectFactory.o Operation.o  
Properties.o Proxy.o Slice.o Types.o Util.o -L/share/Projekte/network/ 
Ice-3.0.1/lib -lIce -lIceUtil -lSlice -F/Library/Frameworks/ 
Python.framework/Versions/Current/ -framework Python
/usr/bin/ld: Undefined symbols:
typeinfo for Ice::LocalObject
typeinfo for IceUtil::NullHandleException
typeinfo for IceUtil::ThreadSyscallException
typeinfo for IceUtil::Thread
typeinfo for IceUtil::Exception
typeinfo for IceProxy::Ice::Object
typeinfo for IceProxy::Ice::Router
typeinfo for IceProxy::Ice::Locator
typeinfo for Ice::BlobjectAsync
typeinfo for Ice::OperationNotExistException
typeinfo for Ice::Object
typeinfo for IceInternal::OutgoingAsync
typeinfo for Ice::MarshalException
typeinfo for Ice::UnknownException
typeinfo for Ice::UnknownUserException
typeinfo for Ice::AMI_Object_ice_invoke
typeinfo for Ice::ObjectReader
typeinfo for Ice::ObjectWriter
typeinfo for Ice::NoObjectFactoryException
typeinfo for Ice::DNSException
typeinfo for Ice::UserException
typeinfo for Ice::SyscallException
typeinfo for Ice::NoEndpointException
typeinfo for Ice::ProxyParseException
typeinfo for Ice::TwowayOnlyException
typeinfo for Ice::UnknownLocalException
typeinfo for Ice::EndpointParseException
typeinfo for Ice::FacetNotExistException
typeinfo for Ice::IdentityParseException
typeinfo for Ice::NotRegisteredException
typeinfo for Ice::RequestFailedException
typeinfo for Ice::ObjectNotExistException
typeinfo for Ice::IllegalIdentityException
typeinfo for Ice::AlreadyRegisteredException
typeinfo for Ice::UnsupportedEncodingException
typeinfo for Ice::UnsupportedProtocolException
typeinfo for Ice::ObjectAdapterIdInUseException
typeinfo for Ice::PluginInitializationException
typeinfo for Ice::ObjectAdapterDeactivatedException
collect2: ld returned 1 exit status
make[2]: *** [../../python/IcePy.so.3.0.1] Error 1
make[1]: *** [all] Error 1
make: *** [all] Error 1

Can anybody help me?

Thanks
Martin
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)

iD8DBQFEC1i4iJE4wzxJLBwRAvLnAJ4we3MSHoltrJn8x3Bdu1cAe/QK8gCfT1ac
/bZ8uSXWbF9T8r6iciikdgQ=
=MYSw
-----END PGP SIGNATURE-----

From zpincus at stanford.edu  Mon Mar  6 01:16:35 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sun, 5 Mar 2006 16:16:35 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
Message-ID: <3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>

Thanks for the information, Bob. I appreciate your help.

> Actually it's because Python extensions are bundles.  Bundles are  
> NOT shared libraries and can not do what you want.  Additionally,  
> Python on Mac OS X doesn't use dl to load its extensions, it uses  
> lower-level means... so the RTLD flags aren't going to do anything.

If python isn't using dl to load extensions on the Mac, then I would  
suggest that sys.get/setdlopenflags() should not be made available on  
mac builds (as it is not in windows builds). This looks like a pretty  
easy patch to sysmodule.c.

> The "best" you can do is to build shared libraries that all of your  
> extensions link to.

Really? There's no way that different bundles can be loaded such that  
weak external references will be properly shared between the bundles?  
That's a shame, because it really unfits python on the mac for  
modules which use functionality from C++ libraries. It's a double  
shame, since Python on Linux handles this case just fine.

Are bundles in general unable to deal with shared global symbols,  
which are required by C++ for RTTI, or is it just how Python is  
dealing with the bundles?

Anyhow, I assume that there are good reasons for how this all works,  
so I will probably have to go the shared library route. Here's what I  
imagine I would need to do:
(1) Create shared libraries for each python module
(2) Write "stub" python modules that only serve to dlopen() the  
shared libraries (with the GLOBAL flag)
(3) Presumably all the stub would do is declare an initialization  
function that dlopens the shared lib and calls the shared lib's  
initialization function?

Basically, this looks like it would be replicating how Python loads  
modules on Linux. It's too bad that's how it would have to be done,  
unless there's a better way that I can't think of right now. Anyone  
have any thoughts/suggestions/guidance?

Zach


From bob at redivi.com  Mon Mar  6 01:44:02 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sun, 5 Mar 2006 16:44:02 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
Message-ID: <58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>


On Mar 5, 2006, at 4:16 PM, Zachary Pincus wrote:

> Thanks for the information, Bob. I appreciate your help.
>
>> Actually it's because Python extensions are bundles.  Bundles are
>> NOT shared libraries and can not do what you want.  Additionally,
>> Python on Mac OS X doesn't use dl to load its extensions, it uses
>> lower-level means... so the RTLD flags aren't going to do anything.
>
> If python isn't using dl to load extensions on the Mac, then I would
> suggest that sys.get/setdlopenflags() should not be made available on
> mac builds (as it is not in windows builds). This looks like a pretty
> easy patch to sysmodule.c.

Eh.  It'd be better to just switch to dl, but that would break 10.2  
support.

>> The "best" you can do is to build shared libraries that all of your
>> extensions link to.
>
> Really? There's no way that different bundles can be loaded such that
> weak external references will be properly shared between the bundles?
> That's a shame, because it really unfits python on the mac for
> modules which use functionality from C++ libraries. It's a double
> shame, since Python on Linux handles this case just fine.
>
> Are bundles in general unable to deal with shared global symbols,
> which are required by C++ for RTTI, or is it just how Python is
> dealing with the bundles?

MH_BUNDLE is not MH_DYLIB and neither are ELF.  Everything you think  
you know about shared libraries isn't quite the same on Mac OS X.  I  
suggest you read up on Mach-O, I don't really have time to answer all  
of your questions about it.

> Anyhow, I assume that there are good reasons for how this all works,
> so I will probably have to go the shared library route. Here's what I
> imagine I would need to do:
> (1) Create shared libraries for each python module
> (2) Write "stub" python modules that only serve to dlopen() the
> shared libraries (with the GLOBAL flag)
> (3) Presumably all the stub would do is declare an initialization
> function that dlopens the shared lib and calls the shared lib's
> initialization function?
>
> Basically, this looks like it would be replicating how Python loads
> modules on Linux. It's too bad that's how it would have to be done,
> unless there's a better way that I can't think of right now. Anyone
> have any thoughts/suggestions/guidance?

It'd probably factor it such that the shared stuff (C++ classes) is  
in shared libraries and the Python C API stuff is in the extensions.

-bob


From zpincus at stanford.edu  Mon Mar  6 02:56:08 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sun, 5 Mar 2006 17:56:08 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
Message-ID: <AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>

Thanks again Bob for the feedback.

> MH_BUNDLE is not MH_DYLIB and neither are ELF.  Everything you  
> think you know about shared libraries isn't quite the same on Mac  
> OS X.  I suggest you read up on Mach-O, I don't really have time to  
> answer all of your questions about it.

My admittedly limited knowledge of shared libraries is primarily mac- 
based -- I only brought up the ELF cases by way of comparison. At any  
rate, regardless of differences between the formats they *all* have a  
way to the support weak external references needed for C++ RTTI with  
templates.

The sections Apple's of documentation about coalesced sections and  
weak definitions discuss how Mach-O supports these features. I think  
it's not really that different from how they are supported on ELF.
http://developer.apple.com/documentation/DeveloperTools/Conceptual/ 
MachORuntime/Reference/reference.html

The problem is that on Macs, Python is not exposing the necessary  
mechanism for loading modules so that weak external references are  
resolved. On Linux and other platforms where Python respects  
sys.setdlopenflags(), this mechanism *is* exposed.

Specifically, the problem is here:
http://svn.python.org/view/python/trunk/Python/dynload_next.c? 
rev=36531&view=auto

Python is using NSCreateObjectFileImageFromFile and NSLinkImage to  
load and link the bundles. The flags passed to NSLinkImage are *hard  
coded* at compile time, so either all bundles are loaded globally or  
none are. Contrast this to the case where dlopen is used, and the  
dlopenflags are looked up at library load time via
'dlopenflags = PyThreadState_GET()->interp->dlopenflags':
http://svn.python.org/view/python/trunk/Python/dynload_shlib.c? 
rev=41910&view=auto

Now, perhaps it is really necessary that on the Mac, the flags be  
hard coded and that no run-time selection will work. If so, then I'll  
have to implement my own dlopen() based solution as I described earlier.

Otherwise, Mac Python's bundle loading code should have a way to  
expose the global symbol loading mechanism (necessary for C++ RTTI  
with templates) to user code. There are two ways that I imagine this  
could be added:

(1) Provide run-time support for setting the NSLinkImage flags. This  
could be as simple as consulting the dlopenflags set in the sys  
module (and just noting in the documentation that on Macs, the  
'dlopenflags' are really 'NSLinkImage flags'). Or there could be some  
mac-specific module where the NSLinkImage flags are defined, and  
where there is a SetNSLinkImageFlags command to call.

(2) Move to using dlopen() to load the libraries instead of  
NSLinkImage. This might need some minor changes to dynload_shlib.c,  
and it would need the configure script to be smart enough to choose  
dynload_next.c on 10.2 and below.

Neither of these seem too hard. I could try to work up a patch for  
one of these two fixes, if people agree that one or the other is a  
better approach. (I might need some guidance, but I would like to  
help get this issue resolved.)


>> Basically, this looks like it would be replicating how Python loads
>> modules on Linux. It's too bad that's how it would have to be done,
>> unless there's a better way that I can't think of right now. Anyone
>> have any thoughts/suggestions/guidance?
>
> It'd probably factor it such that the shared stuff (C++ classes) is  
> in shared libraries and the Python C API stuff is in the extensions.

Agreed. Unfortunately, parts of the first version of these libraries  
are SWIG-generated, and I'm not sure I can teach SWIG to do that  
refactoring. Would the stub/proxy approach I suggested work at all,  
or will I have to figure out a way to get swig to emit the C++ stuff  
in a different place?

Thanks for your suggestions and clarifications!

Zach




From bob at redivi.com  Mon Mar  6 03:24:49 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sun, 5 Mar 2006 18:24:49 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
Message-ID: <90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>


On Mar 5, 2006, at 5:56 PM, Zachary Pincus wrote:

> Otherwise, Mac Python's bundle loading code should have a way to  
> expose the global symbol loading mechanism (necessary for C++ RTTI  
> with templates) to user code. There are two ways that I imagine  
> this could be added:
>
> (1) Provide run-time support for setting the NSLinkImage flags.  
> This could be as simple as consulting the dlopenflags set in the  
> sys module (and just noting in the documentation that on Macs, the  
> 'dlopenflags' are really 'NSLinkImage flags'). Or there could be  
> some mac-specific module where the NSLinkImage flags are defined,  
> and where there is a SetNSLinkImageFlags command to call.
>
> (2) Move to using dlopen() to load the libraries instead of  
> NSLinkImage. This might need some minor changes to dynload_shlib.c,  
> and it would need the configure script to be smart enough to choose  
> dynload_next.c on 10.2 and below.
>
> Neither of these seem too hard. I could try to work up a patch for  
> one of these two fixes, if people agree that one or the other is a  
> better approach. (I might need some guidance, but I would like to  
> help get this issue resolved.)

If you come up with a patch for (2) against Python's svn trunk that  
is tested and works on 10.2 then I'll commit it.  It's not  
appropriate to make this sort of change against Python 2.4, though.   
*Maybe* for the universal branch that Ronald and I are doing, because  
that's also unlikely to be merged with the 2.4 branch.

-bob


From zpincus at stanford.edu  Mon Mar  6 09:20:14 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Mon, 6 Mar 2006 00:20:14 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
Message-ID: <B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>

>> (2) Move to using dlopen() to load the libraries instead of  
>> NSLinkImage. This might need some minor changes to  
>> dynload_shlib.c, and it would need the configure script to be  
>> smart enough to choose dynload_next.c on 10.2 and below.
>
> If you come up with a patch for (2) against Python's svn trunk that  
> is tested and works on 10.2 then I'll commit it.  It's not  
> appropriate to make this sort of change against Python 2.4,  
> though.  *Maybe* for the universal branch that Ronald and I are  
> doing, because that's also unlikely to be merged with the 2.4 branch.

Here's the patch, in all its glory.

--- configure.in        (revision 42867)
+++ configure.in        (working copy)
@@ -2105,7 +2105,8 @@
-       Darwin/*) DYNLOADFILE="dynload_next.o";;
+  # use dynload_next.o only on darwin versions without dlopen().
+       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;

(Also, autoconf must be re-run to generate a new configure file from  
the patched configure.in.)

I just built python with this patch and it passes all of the tests on  
10.4. Is there anything else I should test, or will the dynload  
functionality be tested plenty by the standard tests? (I assume  
that's the case.)

I don't have access to a 10.2 machine currently, so if someone is  
able to test this patch on 10.1, 10.2, and/or 10.3, I would be most  
grateful. I can try to find some 10.2 install disks (and hope that my  
computer can run 10.2) if nobody has easy access to such a machine.

One last thing -- given that OS X now has dlfcn.h and dlopen(), will  
the library dl module now work? Right now, setup.py won't build  
dlmodule.c on darwin, even if it can find dlfcn.h. Does anyone know  
the reason for this is still valid for 10.3 or 10.4?

Zach



From samrobertsmith at gmail.com  Mon Mar  6 11:12:02 2006
From: samrobertsmith at gmail.com (linda.s)
Date: Mon, 6 Mar 2006 02:12:02 -0800
Subject: [Pythonmac-SIG] building from source or installer
Message-ID: <1d987df30603060212o267e4651scfd8ddafad13563@mail.gmail.com>

what are the benefits of building python from source file?
any difference if using installer?
Linda

From hengist.podd at virgin.net  Mon Mar  6 13:14:50 2006
From: hengist.podd at virgin.net (has)
Date: Mon, 6 Mar 2006 12:14:50 +0000
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
Message-ID: <a05200f00c031d3330537@[82.3.255.67]>

Donovan Preston wrote:

>Stop with the 16 ton weight/snakes/apples debate. Just use the new,
>official, python logo

Absolutely agree, this is a solved problem already. It's not my most favourite logo design ever either (though still better than any others I've seen), but that's irrelevant. Building a strong, consistent brand identity matters FAR more than one's personal preferences.

The Mac Python community will improve its own recognition by self-identifying more closely with the larger Python community. Being Different simply for the sake of being different is not a sufficient rationale and just makes the place feel like a backwater.

has
http://freespace.virgin.net/hamish.sanderson/

From ronaldoussoren at mac.com  Mon Mar  6 14:49:54 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 06 Mar 2006 14:49:54 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <a05200f00c031d3330537@[82.3.255.67]>
References: <a05200f00c031d3330537@[82.3.255.67]>
Message-ID: <14586358.1141652994987.JavaMail.ronaldoussoren@mac.com>

 
On Monday, March 06, 2006, at 01:17PM, has <hengist.podd at virgin.net> wrote:

>Donovan Preston wrote:
>
>>Stop with the 16 ton weight/snakes/apples debate. Just use the new,
>>official, python logo
>
>Absolutely agree, this is a solved problem already. It's not my most favourite logo design ever either (though still better than any others I've seen), but that's irrelevant. Building a strong, consistent brand identity matters FAR more than one's personal preferences.

Fine by me, if someone actually creates the icons! Brendan Simons has created some icons that look good enough, although the launcher icon could be better.  I'm willing to use these icons when he says they are ready. However, I'd like to wait for Kevin Olliver, he's also working on new icons.

>
>The Mac Python community will improve its own recognition by self-identifying more closely with the larger Python community. Being Different simply for the sake of being different is not a sufficient rationale and just makes the place feel like a backwater.

Or is it forefront of design ;-) ;-) ;-)

Ronald


From ronaldoussoren at mac.com  Mon Mar  6 14:51:30 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 06 Mar 2006 14:51:30 +0100
Subject: [Pythonmac-SIG] building from source or installer
In-Reply-To: <1d987df30603060212o267e4651scfd8ddafad13563@mail.gmail.com>
References: <1d987df30603060212o267e4651scfd8ddafad13563@mail.gmail.com>
Message-ID: <7509842.1141653090260.JavaMail.ronaldoussoren@mac.com>

 
On Monday, March 06, 2006, at 11:12AM, linda.s <samrobertsmith at gmail.com> wrote:

>what are the benefits of building python from source file?
>any difference if using installer?

I'd use the installer, that way you don't have to worry about depedencies like readline and bsddb. 

Ronald

>Linda
>_______________________________________________
>Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>

From smithsm at samuelsmith.org  Mon Mar  6 17:12:12 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Mon, 6 Mar 2006 09:12:12 -0700
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
Message-ID: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>

I asked this before but no one responded.
Wouldn't it be a good idea to include the Python Reference docs in  
Mac Help format in the the
MacPython universal installer? or at least as another package


**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From brendansimons at yahoo.ca  Mon Mar  6 17:07:01 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Mon, 6 Mar 2006 11:07:01 -0500 (EST)
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <14586358.1141652994987.JavaMail.ronaldoussoren@mac.com>
Message-ID: <20060306160701.52977.qmail@web31113.mail.mud.yahoo.com>


--- Ronald Oussoren <ronaldoussoren at mac.com> wrote:
> Fine by me, if someone actually creates the icons!
> Brendan Simons has created some icons that look good
> enough, although the launcher icon could be better. 

I did some more digging in /system and /developer, and
found this icon: 
http://twototango.blogs.com/Developer_Applications_JavaTools_AppletLauncher.png

Better? (I would try to make the logo wrap around the
rocket better than the Java logo does)

> I'm willing to use these icons when he says they are
> ready. However, I'd like to wait for Kevin Olliver,
> he's also working on new icons.

Me too.  Please Kevin, don't be too upset.  I just
whipped up something quickly by badging the "official
logo" on the generic document & editor icons.  I can
do the same for any graphic.  I can also do folder and
disk image icons in the same way.

-Brendan



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

From ronaldoussoren at mac.com  Mon Mar  6 17:19:51 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 06 Mar 2006 17:19:51 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <20060306160701.52977.qmail@web31113.mail.mud.yahoo.com>
References: <20060306160701.52977.qmail@web31113.mail.mud.yahoo.com>
Message-ID: <9898295.1141661991770.JavaMail.ronaldoussoren@mac.com>

 
On Monday, March 06, 2006, at 05:07PM, Brendan Simons <brendansimons at yahoo.ca> wrote:

>
>--- Ronald Oussoren <ronaldoussoren at mac.com> wrote:
>> Fine by me, if someone actually creates the icons!
>> Brendan Simons has created some icons that look good
>> enough, although the launcher icon could be better. 
>
>I did some more digging in /system and /developer, and
>found this icon: 
>http://twototango.blogs.com/Developer_Applications_JavaTools_AppletLauncher.png
>
>Better? (I would try to make the logo wrap around the
>rocket better than the Java logo does)

That icon looks nice, but I wonder if reusing it would be legal.  

>
>> I'm willing to use these icons when he says they are
>> ready. However, I'd like to wait for Kevin Olliver,
>> he's also working on new icons.
>
>Me too.  Please Kevin, don't be too upset.  I just
>whipped up something quickly by badging the "official
>logo" on the generic document & editor icons.  I can
>do the same for any graphic.  I can also do folder and
>disk image icons in the same way.
>
>-Brendan
>
>
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around 
>http://mail.yahoo.com 
>
>

From ronaldoussoren at mac.com  Mon Mar  6 17:23:26 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 06 Mar 2006 17:23:26 +0100
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
In-Reply-To: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
Message-ID: <3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>

 
On Monday, March 06, 2006, at 05:13PM, Samuel M. Smith <smithsm at samuelsmith.org> wrote:

>I asked this before but no one responded.
>Wouldn't it be a good idea to include the Python Reference docs in  
>Mac Help format in the the
>MacPython universal installer? or at least as another package

The big question would be: why? The universal installer currently includes
the docs in HTML format, and whenever I republish the installer IDLE will
actually find those documents. 

If we'd include the documents in Help format I'd like to see a patch for
IDLE that will open that help book when the users selects 'Help/Python Docs'
in IDLE.

Ronald

>
>
>**********************************************************************
>Samuel M. Smith Ph.D.
>2966 Fort Hill Road
>Eagle Mountain, Utah 84043
>801-768-2768 voice
>801-768-2769 fax
>**********************************************************************
>"The greatest source of failure and unhappiness in the world is
>giving up what we want most for what we want at the moment"
>**********************************************************************
>
>_______________________________________________
>Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>

From brendansimons at yahoo.ca  Mon Mar  6 17:39:38 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Mon, 6 Mar 2006 11:39:38 -0500 (EST)
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <9898295.1141661991770.JavaMail.ronaldoussoren@mac.com>
Message-ID: <20060306163939.4505.qmail@web31104.mail.mud.yahoo.com>


--- Ronald Oussoren <ronaldoussoren at mac.com> wrote:
> On Monday, March 06, 2006, at 05:07PM, Brendan
> Simons <brendansimons at yahoo.ca> wrote:

> >I did some more digging in /system and /developer,
> and
> >found this icon: 
>
>http://twototango.blogs.com/Developer_Applications_JavaTools_AppletLauncher.png
> >
> >Better? (I would try to make the logo wrap around
> the
> >rocket better than the Java logo does)
> 
> That icon looks nice, but I wonder if reusing it
> would be legal.  

I'm not a lawyer, but it doesn't look like Apple is
enforcing their copyrights to stringently on this. 
The rocket icon is used for the Java launcher, and for

/Developer/Examples/Bluetooth/OBEXSampleSendVCard
So I would liken it to the folder, document, disk
image icons (ie, implied permission to use freely). 
That doesn't mean we shouldn't ask.

-Brendan

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

From bob at redivi.com  Mon Mar  6 18:30:39 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 6 Mar 2006 09:30:39 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
Message-ID: <88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>


On Mar 6, 2006, at 12:20 AM, Zachary Pincus wrote:

>>> (2) Move to using dlopen() to load the libraries instead of  
>>> NSLinkImage. This might need some minor changes to  
>>> dynload_shlib.c, and it would need the configure script to be  
>>> smart enough to choose dynload_next.c on 10.2 and below.
>>
>> If you come up with a patch for (2) against Python's svn trunk  
>> that is tested and works on 10.2 then I'll commit it.  It's not  
>> appropriate to make this sort of change against Python 2.4,  
>> though.  *Maybe* for the universal branch that Ronald and I are  
>> doing, because that's also unlikely to be merged with the 2.4 branch.
>
> Here's the patch, in all its glory.
>
> --- configure.in        (revision 42867)
> +++ configure.in        (working copy)
> @@ -2105,7 +2105,8 @@
> -       Darwin/*) DYNLOADFILE="dynload_next.o";;
> +  # use dynload_next.o only on darwin versions without dlopen().
> +       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;
>
> (Also, autoconf must be re-run to generate a new configure file  
> from the patched configure.in.)
>
> I just built python with this patch and it passes all of the tests  
> on 10.4. Is there anything else I should test, or will the dynload  
> functionality be tested plenty by the standard tests? (I assume  
> that's the case.)
>
> I don't have access to a 10.2 machine currently, so if someone is  
> able to test this patch on 10.1, 10.2, and/or 10.3, I would be most  
> grateful. I can try to find some 10.2 install disks (and hope that  
> my computer can run 10.2) if nobody has easy access to such a machine.

I think the Mac OS X box on the sourceforge compile farm may be 10.2.

> One last thing -- given that OS X now has dlfcn.h and dlopen(),  
> will the library dl module now work? Right now, setup.py won't  
> build dlmodule.c on darwin, even if it can find dlfcn.h. Does  
> anyone know the reason for this is still valid for 10.3 or 10.4?

That module should be available on 10.3+ (but isn't, obviously), but  
it's not all that useful on its own.  I'd just use ctypes instead,  
especially since it's slated for inclusion with Python 2.5.

-bob


From leknarf at pacbell.net  Mon Mar  6 18:33:59 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Mon, 6 Mar 2006 09:33:59 -0800
Subject: [Pythonmac-SIG] getting started with PyObjC
Message-ID: <50A553D6-65FF-40E2-B988-2FB3E2E47A75@pacbell.net>


Following the simple example app that Apple provides for PyObjC has  
lead me to a number of questions ... and a failed build.  The example  
demonstrates using Xcode & InterfaceBuilder to build a simple app,  
PyAverager.  My attempts at a build yield "Build failed for target  
"Development" errors in Xcode and nothing launchable from py2applet.   
There were no instructions on how to configure a target in the Apple  
doco.

Questions:

Does the target need to be more fully specified in Xcode (beyond what  
the PyObjC template provides), if so how?

Does the target need to be specified somehow in the py2app(let) step?

Is Xcode really necessary or advantageous for building PyObjC apps?

Can Interface Builder NIB files be used with py2app(let)?  If so, how  
are they specified on the cmd-line?  (i.e.: declared as the Foo.nib  
parent directory, or as each of the nib files contained in Foo.nib:  
classes.nib, info.nib, &c.)?

Feeling a little at sea; thanks in advance for a point in the right  
direction.
Scott


Xcode 2.1
Python 2.4.1
OSX 10.4.5





From delza at livingcode.org  Mon Mar  6 18:52:56 2006
From: delza at livingcode.org (Dethe Elza)
Date: Mon, 6 Mar 2006 09:52:56 -0800
Subject: [Pythonmac-SIG] getting started with PyObjC
In-Reply-To: <50A553D6-65FF-40E2-B988-2FB3E2E47A75@pacbell.net>
References: <50A553D6-65FF-40E2-B988-2FB3E2E47A75@pacbell.net>
Message-ID: <24d517dd0603060952u493be516xdb33eb8e4fe79cf3@mail.gmail.com>

Using XCode isn't really the best way to create PyObjC apps.  The
recommended way is to use InterfaceBuilder to create your UI,
implement it with PyObjC, and use py2app to build the application from
there.

Check out the tutorial on the PyObjC site, and the docs for py2app.

PyObjC tutorial: http://pyobjc.sourceforge.net/doc/tutorial.php
Py2app documentation: http://undefined.org/python/py2app.html

It sounds like you already have the nib and code, so that should work
for you.  If you have the nib, but no code yet, there is a script
which will generate a python files with stubs built from your nib to
get you started.  Using the nibclassbuilder script is demonstrated in
the tutorial.

I hope that helps to get you started.  Come back with any further
questions you have.

--Dethe

On 3/6/06, Scott Frankel <leknarf at pacbell.net> wrote:
>
> Following the simple example app that Apple provides for PyObjC has
> lead me to a number of questions ... and a failed build.  The example
> demonstrates using Xcode & InterfaceBuilder to build a simple app,
> PyAverager.  My attempts at a build yield "Build failed for target
> "Development" errors in Xcode and nothing launchable from py2applet.
> There were no instructions on how to configure a target in the Apple
> doco.
>
> Questions:
>
> Does the target need to be more fully specified in Xcode (beyond what
> the PyObjC template provides), if so how?
>
> Does the target need to be specified somehow in the py2app(let) step?
>
> Is Xcode really necessary or advantageous for building PyObjC apps?
>
> Can Interface Builder NIB files be used with py2app(let)?  If so, how
> are they specified on the cmd-line?  (i.e.: declared as the Foo.nib
> parent directory, or as each of the nib files contained in Foo.nib:
> classes.nib, info.nib, &c.)?
>
> Feeling a little at sea; thanks in advance for a point in the right
> direction.
> Scott
>
>
> Xcode 2.1
> Python 2.4.1
> OSX 10.4.5
>
>
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
>

From leknarf at pacbell.net  Mon Mar  6 19:05:44 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Mon, 6 Mar 2006 10:05:44 -0800
Subject: [Pythonmac-SIG] getting started with PyObjC
In-Reply-To: <24d517dd0603060952u493be516xdb33eb8e4fe79cf3@mail.gmail.com>
References: <50A553D6-65FF-40E2-B988-2FB3E2E47A75@pacbell.net>
	<24d517dd0603060952u493be516xdb33eb8e4fe79cf3@mail.gmail.com>
Message-ID: <8450B01A-950F-499C-BF0F-018EF744BC6A@pacbell.net>


Thanks for the tips!


On Mar 6, 2006, at 9:52 AM, Dethe Elza wrote:

> Using XCode isn't really the best way to create PyObjC apps.  The
> recommended way is to use InterfaceBuilder to create your UI,
> implement it with PyObjC, and use py2app to build the application from
> there.
>
> Check out the tutorial on the PyObjC site, and the docs for py2app.
>
> PyObjC tutorial: http://pyobjc.sourceforge.net/doc/tutorial.php
> Py2app documentation: http://undefined.org/python/py2app.html
>
> It sounds like you already have the nib and code, so that should work
> for you.  If you have the nib, but no code yet, there is a script
> which will generate a python files with stubs built from your nib to
> get you started.  Using the nibclassbuilder script is demonstrated in
> the tutorial.
>
> I hope that helps to get you started.  Come back with any further
> questions you have.
>
> --Dethe
>
> On 3/6/06, Scott Frankel <leknarf at pacbell.net> wrote:
>>
>> Following the simple example app that Apple provides for PyObjC has
>> lead me to a number of questions ... and a failed build.  The example
>> demonstrates using Xcode & InterfaceBuilder to build a simple app,
>> PyAverager.  My attempts at a build yield "Build failed for target
>> "Development" errors in Xcode and nothing launchable from py2applet.
>> There were no instructions on how to configure a target in the Apple
>> doco.
>>
>> Questions:
>>
>> Does the target need to be more fully specified in Xcode (beyond what
>> the PyObjC template provides), if so how?
>>
>> Does the target need to be specified somehow in the py2app(let) step?
>>
>> Is Xcode really necessary or advantageous for building PyObjC apps?
>>
>> Can Interface Builder NIB files be used with py2app(let)?  If so, how
>> are they specified on the cmd-line?  (i.e.: declared as the Foo.nib
>> parent directory, or as each of the nib files contained in Foo.nib:
>> classes.nib, info.nib, &c.)?
>>
>> Feeling a little at sea; thanks in advance for a point in the right
>> direction.
>> Scott
>>
>>
>> Xcode 2.1
>> Python 2.4.1
>> OSX 10.4.5
>>
>>
>>
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>


From gandreas at gandreas.com  Mon Mar  6 20:05:35 2006
From: gandreas at gandreas.com (gandreas at gandreas.com)
Date: Mon, 6 Mar 2006 13:05:35 -0600
Subject: [Pythonmac-SIG] ANN: ScrIDE 0.3
Message-ID: <08E06611-4636-4CBC-BAD8-46FEF7C2599F@gandreas.com>

So in an attempt to finally get rid of the potential morass that  
further work on PyOXIDE could be, I'm release a very early version of  
ScrIDE, a generic, extensible, scripting IDE.  There are several  
major differences between the two - first, ScrIDE is designed to  
support multiple scripting languages, and to be able to do that, the  
second different is that _everything_ is run in separate processes,  
which means that you can easily select from different versions of  
python (or even easily pick between python and pythonw).  Finally, it  
is highly extensible (all sorts of things controlled by various  
property lists which can be overridden and extended, including adding  
menu commands that pipe text to/from other processes).

Version 0.3 includes a debugger that supports breakpoints, and "rich"  
interactive shells (with things like code completion, call tips,  
syntax coloring), as well as a bunch of smaller features, bug fixes,  
internal "clean ups" and the like.



<http://projects.gandreas.com/scride/download.php>



Glenn Andreas                      gandreas at gandreas.com
  <http://www.gandreas.com/> wicked fun!
Dominogy | Just place all the dominos on the board...


From ronaldoussoren at mac.com  Mon Mar  6 20:14:53 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 6 Mar 2006 20:14:53 +0100
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <20060306163939.4505.qmail@web31104.mail.mud.yahoo.com>
References: <20060306163939.4505.qmail@web31104.mail.mud.yahoo.com>
Message-ID: <8BEF7927-C424-43E8-86F3-5F1361B74216@mac.com>


On 6-mrt-2006, at 17:39, Brendan Simons wrote:

>
> --- Ronald Oussoren <ronaldoussoren at mac.com> wrote:
>> On Monday, March 06, 2006, at 05:07PM, Brendan
>> Simons <brendansimons at yahoo.ca> wrote:
>
>>> I did some more digging in /system and /developer,
>> and
>>> found this icon:
>>
>> http://twototango.blogs.com/ 
>> Developer_Applications_JavaTools_AppletLauncher.png
>>>
>>> Better? (I would try to make the logo wrap around
>> the
>>> rocket better than the Java logo does)
>>
>> That icon looks nice, but I wonder if reusing it
>> would be legal.
>
> I'm not a lawyer, but it doesn't look like Apple is
> enforcing their copyrights to stringently on this.
> The rocket icon is used for the Java launcher, and for
>
> /Developer/Examples/Bluetooth/OBEXSampleSendVCard
> So I would liken it to the folder, document, disk
> image icons (ie, implied permission to use freely).
> That doesn't mean we shouldn't ask.

If we want to reuse that icon we should ask, I definitly
wouldn't want to add anything to the python repository
that has an unclear legal status.

But asking right now is not very useful as we don't know
yet what we'll end up with.

Ronald

>
> -Brendan
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com


From charles.hartman at conncoll.edu  Mon Mar  6 20:25:48 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Mon, 6 Mar 2006 14:25:48 -0500
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
In-Reply-To: <3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
	<3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>
Message-ID: <8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu>

But the Mac Help program has always been and still is a dog. I'd much  
rather have the HTML docs myself.

Charles



On Mar 6, 2006, at 11:23 AM, Ronald Oussoren wrote:

>
> On Monday, March 06, 2006, at 05:13PM, Samuel M. Smith  
> <smithsm at samuelsmith.org> wrote:
>
>> I asked this before but no one responded.
>> Wouldn't it be a good idea to include the Python Reference docs in
>> Mac Help format in the the
>> MacPython universal installer? or at least as another package
>
> The big question would be: why? The universal installer currently  
> includes
> the docs in HTML format, and whenever I republish the installer  
> IDLE will
> actually find those documents.
>
> If we'd include the documents in Help format I'd like to see a  
> patch for
> IDLE that will open that help book when the users selects 'Help/ 
> Python Docs'
> in IDLE.
>
> Ronald
>
>>
>>
>> ********************************************************************* 
>> *
>> Samuel M. Smith Ph.D.
>> 2966 Fort Hill Road
>> Eagle Mountain, Utah 84043
>> 801-768-2768 voice
>> 801-768-2769 fax
>> ********************************************************************* 
>> *
>> "The greatest source of failure and unhappiness in the world is
>> giving up what we want most for what we want at the moment"
>> ********************************************************************* 
>> *
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>
>>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Mon Mar  6 20:32:16 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 6 Mar 2006 20:32:16 +0100
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
In-Reply-To: <8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu>
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
	<3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>
	<8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu>
Message-ID: <8BB51A9B-AA03-455C-AD15-6D8E40C1D42B@mac.com>


On 6-mrt-2006, at 20:25, Charles Hartman wrote:

> But the Mac Help program has always been and still is a dog. I'd  
> much rather have the HTML docs myself.

One advantage of the Mac Help program is that it has a search  
function. IDLE's 'Help/Python Docs' just
opens a webbrowser. I am however happy enough with this that I don't  
volunteer to add Apple Help support :-)

Ronald

>
> Charles
>
>
>
> On Mar 6, 2006, at 11:23 AM, Ronald Oussoren wrote:
>
>>
>> On Monday, March 06, 2006, at 05:13PM, Samuel M. Smith  
>> <smithsm at samuelsmith.org> wrote:
>>
>>> I asked this before but no one responded.
>>> Wouldn't it be a good idea to include the Python Reference docs in
>>> Mac Help format in the the
>>> MacPython universal installer? or at least as another package
>>
>> The big question would be: why? The universal installer currently  
>> includes
>> the docs in HTML format, and whenever I republish the installer  
>> IDLE will
>> actually find those documents.
>>
>> If we'd include the documents in Help format I'd like to see a  
>> patch for
>> IDLE that will open that help book when the users selects 'Help/ 
>> Python Docs'
>> in IDLE.
>>
>> Ronald
>>
>>>
>>>
>>> ******************************************************************** 
>>> **
>>> Samuel M. Smith Ph.D.
>>> 2966 Fort Hill Road
>>> Eagle Mountain, Utah 84043
>>> 801-768-2768 voice
>>> 801-768-2769 fax
>>> ******************************************************************** 
>>> **
>>> "The greatest source of failure and unhappiness in the world is
>>> giving up what we want most for what we want at the moment"
>>> ******************************************************************** 
>>> **
>>>
>>> _______________________________________________
>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>>
>>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>


From smithsm at samuelsmith.org  Mon Mar  6 20:47:20 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Mon, 6 Mar 2006 12:47:20 -0700
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
In-Reply-To: <8BB51A9B-AA03-455C-AD15-6D8E40C1D42B@mac.com>
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
	<3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>
	<8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu>
	<8BB51A9B-AA03-455C-AD15-6D8E40C1D42B@mac.com>
Message-ID: <37E24660-1201-4B1C-A806-86D32164B16D@samuelsmith.org>

I find the search to be very helpful. Is it hard to convert from HTML  
to Apple Help format?

On 06 Mar, 2006, at 12:32, Ronald Oussoren wrote:

>
> On 6-mrt-2006, at 20:25, Charles Hartman wrote:
>
>> But the Mac Help program has always been and still is a dog. I'd  
>> much rather have the HTML docs myself.
>
> One advantage of the Mac Help program is that it has a search  
> function. IDLE's 'Help/Python Docs' just
> opens a webbrowser. I am however happy enough with this that I  
> don't volunteer to add Apple Help support :-)
>
> Ronald
>
>>
>> Charles
>>
>>
>>
>> On Mar 6, 2006, at 11:23 AM, Ronald Oussoren wrote:
>>
>>>
>>> On Monday, March 06, 2006, at 05:13PM, Samuel M. Smith  
>>> <smithsm at samuelsmith.org> wrote:
>>>
>>>> I asked this before but no one responded.
>>>> Wouldn't it be a good idea to include the Python Reference docs in
>>>> Mac Help format in the the
>>>> MacPython universal installer? or at least as another package
>>>
>>> The big question would be: why? The universal installer currently  
>>> includes
>>> the docs in HTML format, and whenever I republish the installer  
>>> IDLE will
>>> actually find those documents.
>>>
>>> If we'd include the documents in Help format I'd like to see a  
>>> patch for
>>> IDLE that will open that help book when the users selects 'Help/ 
>>> Python Docs'
>>> in IDLE.
>>>
>>> Ronald
>>>
>>>>
>>>>
>>>> ******************************************************************* 
>>>> ***
>>>> Samuel M. Smith Ph.D.
>>>> 2966 Fort Hill Road
>>>> Eagle Mountain, Utah 84043
>>>> 801-768-2768 voice
>>>> 801-768-2769 fax
>>>> ******************************************************************* 
>>>> ***
>>>> "The greatest source of failure and unhappiness in the world is
>>>> giving up what we want most for what we want at the moment"
>>>> ******************************************************************* 
>>>> ***
>>>>
>>>> _______________________________________________
>>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>>>
>>>>
>>> _______________________________________________
>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>
>

**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From ronaldoussoren at mac.com  Mon Mar  6 20:53:28 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 6 Mar 2006 20:53:28 +0100
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
In-Reply-To: <37E24660-1201-4B1C-A806-86D32164B16D@samuelsmith.org>
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org>
	<3518315.1141662206533.JavaMail.ronaldoussoren@mac.com>
	<8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu>
	<8BB51A9B-AA03-455C-AD15-6D8E40C1D42B@mac.com>
	<37E24660-1201-4B1C-A806-86D32164B16D@samuelsmith.org>
Message-ID: <6FC07863-6D6E-47A0-8D0A-EDDF855B230A@mac.com>


On 6-mrt-2006, at 20:47, Samuel M. Smith wrote:

> I find the search to be very helpful. Is it hard to convert from  
> HTML to Apple Help format?

No, there's a script in the python tree to do the transformation.  
There's
  another script that compiles the search index, but that doesn't  
reliably detect the end
of the indexing operation.

The hard part is probably to get IDLE to use the help book, although  
I haven't looked into this
yet.

Ronald

>
> On 06 Mar, 2006, at 12:32, Ronald Oussoren wrote:
>
>>
>> On 6-mrt-2006, at 20:25, Charles Hartman wrote:
>>
>>> But the Mac Help program has always been and still is a dog. I'd  
>>> much rather have the HTML docs myself.
>>
>> One advantage of the Mac Help program is that it has a search  
>> function. IDLE's 'Help/Python Docs' just
>> opens a webbrowser. I am however happy enough with this that I  
>> don't volunteer to add Apple Help support :-)
>>
>> Ronald
>>
>>>
>>> Charles
>>>
>>>
>>>
>>> On Mar 6, 2006, at 11:23 AM, Ronald Oussoren wrote:
>>>
>>>>
>>>> On Monday, March 06, 2006, at 05:13PM, Samuel M. Smith  
>>>> <smithsm at samuelsmith.org> wrote:
>>>>
>>>>> I asked this before but no one responded.
>>>>> Wouldn't it be a good idea to include the Python Reference docs in
>>>>> Mac Help format in the the
>>>>> MacPython universal installer? or at least as another package
>>>>
>>>> The big question would be: why? The universal installer  
>>>> currently includes
>>>> the docs in HTML format, and whenever I republish the installer  
>>>> IDLE will
>>>> actually find those documents.
>>>>
>>>> If we'd include the documents in Help format I'd like to see a  
>>>> patch for
>>>> IDLE that will open that help book when the users selects 'Help/ 
>>>> Python Docs'
>>>> in IDLE.
>>>>
>>>> Ronald
>>>>
>>>>>
>>>>>
>>>>> ****************************************************************** 
>>>>> ****
>>>>> Samuel M. Smith Ph.D.
>>>>> 2966 Fort Hill Road
>>>>> Eagle Mountain, Utah 84043
>>>>> 801-768-2768 voice
>>>>> 801-768-2769 fax
>>>>> ****************************************************************** 
>>>>> ****
>>>>> "The greatest source of failure and unhappiness in the world is
>>>>> giving up what we want most for what we want at the moment"
>>>>> ****************************************************************** 
>>>>> ****
>>>>>
>>>>> _______________________________________________
>>>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>>>>
>>>>>
>>>> _______________________________________________
>>>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>>>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>>
>>
>
> **********************************************************************
> Samuel M. Smith Ph.D.
> 2966 Fort Hill Road
> Eagle Mountain, Utah 84043
> 801-768-2768 voice
> 801-768-2769 fax
> **********************************************************************
> "The greatest source of failure and unhappiness in the world is
> giving up what we want most for what we want at the moment"
> **********************************************************************
>


From zpincus at stanford.edu  Tue Mar  7 02:04:50 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Mon, 6 Mar 2006 17:04:50 -0800
Subject: [Pythonmac-SIG] dl module on Mac python builds
Message-ID: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>

Hi folks,

Despite the fact that dlopen() is available on OS X for 10.3 and  
above, the python dl module has not been made available on 10.3 and  
10.4.

Specifically, some code in the python setup.py script explicitly  
prevents the dl module from being built on darwin even if python sees  
dlfcn.h:

     dl_inc = find_file('dlfcn.h', [], inc_dirs)
         if (dl_inc is not None) and (platform not in ['atheos',  
'darwin']):
             exts.append( Extension('dl', ['dlmodule.c']) )


Does anyone know the reason for this explicit check to keep the dl  
module from being built on darwin?

Now, currently dl is not too useful since mac python does not respect  
sys.setdlopenflags(), and since ctypes are coming in python 2.5.  
However, I am testing a patch to make python on the mac use dlopen()  
instead of the NeXT-derived extension loading mechanism, precisely so  
that sys.setdlopenflags() can be used meaningfully on may python.

In addition to providing functions to call dlopen(), the python dl  
module declares the values of the various dlopenflags that a user  
might want to pass to sys.setdlopenflags(). Thus, it would be  
worthwhile making the dl module work in the future.

So, is there any reason that one could not just remove the 'darwin'  
from the above platform check in setup.py? Presumably versions of OS  
X below 10.3 would not have dlfcn.h, and thus on those platforms, the  
dl module would not be built. But on 10.3 and up, it would be built.

Is this an acceptable solution? Or do there need to be explicit  
checks to keep dl from being build on 10.2 and below (say, because  
10.2 actually does have some inoperative dlfcn.h header somewhere?)

Zach Pincus


Program in Biomedical Informatics and Department of Biochemistry
Stanford University School of Medicine


From zpincus at stanford.edu  Tue Mar  7 02:05:38 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Mon, 6 Mar 2006 17:05:38 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
	<88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
Message-ID: <5AEC2D57-1B71-4860-B1D2-A2C4BBB900CE@stanford.edu>

Thanks for your feedback, Bob.

>> I don't have access to a 10.2 machine currently, so if someone is  
>> able to test this patch on 10.1, 10.2, and/or 10.3, I would be  
>> most grateful. I can try to find some 10.2 install disks (and hope  
>> that my computer can run 10.2) if nobody has easy access to such a  
>> machine.
>
> I think the Mac OS X box on the sourceforge compile farm may be 10.2.

Now, assuming the configure script works right, things aren't  
changing on the 10.2 case. So all that's needed to test there is that  
dynload_next and not dynload_shlib is picked up in the configure step.

This would be very easy to test for anyone who has a 10.2 box, since  
it doesn't actually involve compiling anything. I'll look into using  
the sourceforge one.

However, on 10.3, this patch actually changes the behavior to use  
dynload_shlib (same on 10.4). I've tested it on 10.4, but we ought to  
make sure that things work on 10.3 too. Does anyone have access to a  
10.3 box?


>> One last thing -- given that OS X now has dlfcn.h and dlopen(),  
>> will the library dl module now work? Right now, setup.py won't  
>> build dlmodule.c on darwin, even if it can find dlfcn.h. Does  
>> anyone know the reason for this is still valid for 10.3 or 10.4?
>
> That module should be available on 10.3+ (but isn't, obviously),  
> but it's not all that useful on its own.  I'd just use ctypes  
> instead, especially since it's slated for inclusion with Python 2.5.

I'm going to branch this discussion to a new email thread to see if  
anyone else has any thoughts. The main argument for having dl work on  
the mac (other than the fact that it is a part of the standard  
library!) is that if sys.setdlopenflags() is respected, there should  
be some in-python way to get the appropriate flags (which are  
different on darwin's dlfcn.h than on most other platforms). And  
these flags are defined in the dl module.

Zach



From zpincus at stanford.edu  Tue Mar  7 02:31:53 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Mon, 6 Mar 2006 17:31:53 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <5AEC2D57-1B71-4860-B1D2-A2C4BBB900CE@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
	<88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
	<5AEC2D57-1B71-4860-B1D2-A2C4BBB900CE@stanford.edu>
Message-ID: <6B27A861-7DDA-403A-9ED6-C24F7096D5C7@stanford.edu>

>> I think the Mac OS X box on the sourceforge compile farm may be 10.2.

Only if you're a developer on some SF project. Is there a python  
project someone can add me to so I can test this (and the dl module)  
patch? Or should I register some dummy project?

Zach


From charles.hartman at conncoll.edu  Tue Mar  7 04:51:16 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Mon, 6 Mar 2006 22:51:16 -0500
Subject: [Pythonmac-SIG] simple help?
Message-ID: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>

I'm getting really tired of not understanding this: Can someone point  
me to some single place that explains the whole path / PATH /  
PYTHONPATH arrangement for Python? I know there are two sorts on the  
Mac, framework and /usr, and I know some other bits and pieces, but I  
have no clear overall picture.

This came to my attention when I downloaded and installed IPython,  
which looks very nifty. I put it in my MacPython folder in / 
Applications -- but that's not really where it belongs, and I don't  
want to have to run it from the Terminal only from there . . . you  
see the kind of simple problem I'm not grasping.

Any help much appreciated as always.

(I'm using the 2.4.1 framework build.)

Charles


From ronaldoussoren at mac.com  Tue Mar  7 08:14:02 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 08:14:02 +0100
Subject: [Pythonmac-SIG] dl module on Mac python builds
In-Reply-To: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
Message-ID: <EA5ADC37-F142-4DC0-908B-3D6CC8E91B99@mac.com>


On 7-mrt-2006, at 2:04, Zachary Pincus wrote:

> Hi folks,
>
> Despite the fact that dlopen() is available on OS X for 10.3 and
> above, the python dl module has not been made available on 10.3 and
> 10.4.
>
> Specifically, some code in the python setup.py script explicitly
> prevents the dl module from being built on darwin even if python sees
> dlfcn.h:
>
>      dl_inc = find_file('dlfcn.h', [], inc_dirs)
>          if (dl_inc is not None) and (platform not in ['atheos',
> 'darwin']):
>              exts.append( Extension('dl', ['dlmodule.c']) )
>
>
> Does anyone know the reason for this explicit check to keep the dl
> module from being built on darwin?

I don't. Why don't you patch setup.py and see what breaks?

Ronald


From ronaldoussoren at mac.com  Tue Mar  7 08:20:31 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 08:20:31 +0100
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
Message-ID: <20586AAD-2E4F-4E35-A0C8-3A266877C772@mac.com>


On 6-mrt-2006, at 3:24, Bob Ippolito wrote:

>
> On Mar 5, 2006, at 5:56 PM, Zachary Pincus wrote:
>
>> Otherwise, Mac Python's bundle loading code should have a way to
>> expose the global symbol loading mechanism (necessary for C++ RTTI
>> with templates) to user code. There are two ways that I imagine
>> this could be added:
>>
>> (1) Provide run-time support for setting the NSLinkImage flags.
>> This could be as simple as consulting the dlopenflags set in the
>> sys module (and just noting in the documentation that on Macs, the
>> 'dlopenflags' are really 'NSLinkImage flags'). Or there could be
>> some mac-specific module where the NSLinkImage flags are defined,
>> and where there is a SetNSLinkImageFlags command to call.
>>
>> (2) Move to using dlopen() to load the libraries instead of
>> NSLinkImage. This might need some minor changes to dynload_shlib.c,
>> and it would need the configure script to be smart enough to choose
>> dynload_next.c on 10.2 and below.
>>
>> Neither of these seem too hard. I could try to work up a patch for
>> one of these two fixes, if people agree that one or the other is a
>> better approach. (I might need some guidance, but I would like to
>> help get this issue resolved.)
>
> If you come up with a patch for (2) against Python's svn trunk that
> is tested and works on 10.2 then I'll commit it.  It's not
> appropriate to make this sort of change against Python 2.4, though.
> *Maybe* for the universal branch that Ronald and I are doing, because
> that's also unlikely to be merged with the 2.4 branch.

What effect would this change have on the existing behaviour? Currently
modules cannot share symbols and IMO that is a good feature.

Zachary: how would you solve this on Windows? Are symbols shared between
modules there?

While we are on the subject of linking...  Bob: is '-undefined  
dynamic_lookup'
still necessary? I'd much rather link against the framework and  
without this
flag because this catches bugs earlier. IIRC we added this flag when  
Apple's
python and MacPython used the same directory for site-packages. Py2app
could (does?) rewrite all loader commands when building standalone  
applications.

Ronald

>
> -bob
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Tue Mar  7 08:39:04 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 06 Mar 2006 23:39:04 -0800
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
Message-ID: <440D3898.4070104@noaa.gov>

Charles Hartman wrote:
> I'm getting really tired of not understanding this: Can someone point  
> me to some single place that explains the whole path / PATH /  
> PYTHONPATH arrangement for Python? I know there are two sorts on the  
> Mac, framework and /usr, and I know some other bits and pieces, but I  
> have no clear overall picture.
> 
> This came to my attention when I downloaded and installed IPython,  
> which looks very nifty. I put it in my MacPython folder in / 
> Applications -- but that's not really where it belongs,

I think ipython is an executable script, so you want to put it where 
those go. For stuff not installed by Apple, that is usually:

/usr/local/bin

then you want to make sure /usr/local/bin is on your PATH. You do that 
by adding it to your /Users/YourName/.profile file, like this:

export PATH=/usr/local/bin:$PATH

A little googling should give you more info if you need it.

One more complication: the default dist-utils install directory for 
scripts is not /usr/local/bin. It's a bin directory buried in the depths 
of the Framework somewhere (I can't tell you where, as I'm at home on my 
Linux box right now). You may want to add that to your PATH too. The 
upcoming new installer is going to do that for you.

good luck,

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                      		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From stewart.midwinter at gmail.com  Tue Mar  7 09:11:42 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Tue, 7 Mar 2006 01:11:42 -0700
Subject: [Pythonmac-SIG] what is wxMacExecute Bad bundle?
Message-ID: <d4c0d4800603070011y669d8d33qa25a78fb91ddd03b@mail.gmail.com>

hi

I'm working on Firedrop2, a blogging client using wax and wxPython.  
I'm developing on OS X 10.4.4.   My problem occurs while starting the
app, long before a GUI is created.  I import several plugins.  The app
was running fine earlier, now suddenly it's choking when I try to
import the last of several plugins (and I haven't changed any code in
any plugins!).    The exact line causing the error is the one with
'exec':
name = OutBox
line = "import %s as mod" % name
exec line

Here's a debug exception statement that is generated:
[Debug] 00:38:53: wxMacExecute Bad bundle:
/Library/Frameworks/Python.framework/Versions/2.4/Resources/Python.app/Contents/MacOS/Python
[Debug] 00:38:53: pid=10825
[Debug] 00:38:53: Successfully added notification to the runloop
[Debug] 00:38:57: Process ended

So, what's a bad bundle?

thanks
S

From charles.hartman at conncoll.edu  Tue Mar  7 15:01:25 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 09:01:25 -0500
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <440D3898.4070104@noaa.gov>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
Message-ID: <6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>


On Mar 7, 2006, at 2:39 AM, Christopher Barker wrote:

> Charles Hartman wrote:
>> I'm getting really tired of not understanding this: Can someone  
>> point  me to some single place that explains the whole path /  
>> PATH /  PYTHONPATH arrangement for Python? I know there are two  
>> sorts on the  Mac, framework and /usr, and I know some other bits  
>> and pieces, but I  have no clear overall picture.
>> This came to my attention when I downloaded and installed  
>> IPython,  which looks very nifty. I put it in my MacPython folder  
>> in / Applications -- but that's not really where it belongs,
>
> I think ipython is an executable script, so you want to put it  
> where those go. For stuff not installed by Apple, that is usually:
>
> /usr/local/bin
>
> then you want to make sure /usr/local/bin is on your PATH. You do  
> that by adding it to your /Users/YourName/.profile file, like this:
>
> export PATH=/usr/local/bin:$PATH
>
> A little googling should give you more info if you need it.

Thank you. This much I knew, but it's useful to have it spelled out.

I think that part of what confuses a non-Unix person like me is the  
distinction PATH / (executable programs and scripts) vs. PYTHONPATH /  
modules. (This comes up with IPython because embedding its  
interpreter in another program requires importing a module  
IPShellEmbed.) I find that my environment contains no PYTHONPATH  
variable, presumably because I've just been using the modules  
supplied with Python and with wxPython, and the installers have put  
them in (some) default places. So I assume that the new installer,  
too, won't set a PYTHONPATH.

> One more complication: the default dist-utils install directory for  
> scripts is not /usr/local/bin. It's a bin directory buried in the  
> depths of the Framework somewhere (I can't tell you where, as I'm  
> at home on my Linux box right now). You may want to add that to  
> your PATH too. The upcoming new installer is going to do that for you.



From mcovill at mac.com  Tue Mar  7 15:24:56 2006
From: mcovill at mac.com (Mike Covill)
Date: Tue, 07 Mar 2006 09:24:56 -0500
Subject: [Pythonmac-SIG] 50-un_commentLines.py
In-Reply-To: <3092402.1141741309710.JavaMail.maico.a@mac.com>
References: <3092402.1141741309710.JavaMail.maico.a@mac.com>
Message-ID: <16379614.1141741496241.JavaMail.mcovill@mac.com>

Does anyone know what happened to or have a copy of
50-un_commentLines.py ('A Python version of the Apple-provided "Un-Comment" script to be Python-aware.')

referenced on:
http://pythonmac.org/wiki/XcodeIntegration ?

When I click on the link I just get the message:
"No attachments stored for XcodeIntegration"

Thanks, 
Mike Covill


From ronaldoussoren at mac.com  Tue Mar  7 15:56:33 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 15:56:33 +0100
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
Message-ID: <86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>


On 7-mrt-2006, at 15:01, Charles Hartman wrote:

>
> On Mar 7, 2006, at 2:39 AM, Christopher Barker wrote:
>
>> Charles Hartman wrote:
>>> I'm getting really tired of not understanding this: Can someone
>>> point  me to some single place that explains the whole path /
>>> PATH /  PYTHONPATH arrangement for Python? I know there are two
>>> sorts on the  Mac, framework and /usr, and I know some other bits
>>> and pieces, but I  have no clear overall picture.
>>> This came to my attention when I downloaded and installed
>>> IPython,  which looks very nifty. I put it in my MacPython folder
>>> in / Applications -- but that's not really where it belongs,
>>
>> I think ipython is an executable script, so you want to put it
>> where those go. For stuff not installed by Apple, that is usually:
>>
>> /usr/local/bin
>>
>> then you want to make sure /usr/local/bin is on your PATH. You do
>> that by adding it to your /Users/YourName/.profile file, like this:
>>
>> export PATH=/usr/local/bin:$PATH
>>
>> A little googling should give you more info if you need it.
>
> Thank you. This much I knew, but it's useful to have it spelled out.
>
> I think that part of what confuses a non-Unix person like me is the
> distinction PATH / (executable programs and scripts) vs. PYTHONPATH /
> modules. (This comes up with IPython because embedding its
> interpreter in another program requires importing a module
> IPShellEmbed.) I find that my environment contains no PYTHONPATH
> variable, presumably because I've just been using the modules
> supplied with Python and with wxPython, and the installers have put
> them in (some) default places. So I assume that the new installer,
> too, won't set a PYTHONPATH.

PYTHONPATH is a way to extend the search path for python modules,  
e.g. sys.path,
while PATH is used by the unix shell to look for programs.

The new installer will indeed not set PYTHONPATH, the default module  
search
path is good enough. It will update PATH, because the bin directory  
inside
the framework is not usually on PATH.

Ronald

From goedman at mac.com  Tue Mar  7 16:51:11 2006
From: goedman at mac.com (Rob J Goedman)
Date: Tue, 7 Mar 2006 07:51:11 -0800
Subject: [Pythonmac-SIG] wx.ART_NEW,
	'module' object has no attribute 'ART_NEW'
In-Reply-To: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
Message-ID: <87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>

Hi,

I'm new to Python. For a biometrix application I am investigating  
python or java (right
now its in VB).

Earlier this morning I installed (on OS 10.4.5) the latest version of  
MacPython-OSX-2.4.1,
the Mac fix (TigerPython24Fix-r2) and wxPython2.6-osx-unicode-2.6.2.1- 
macosx10.3-py2.4.

Most things work great (including the wxPython Demo.app), but double  
clicking demo.pyw
or pySketch.py (and I guess several others) produces below error  
message.

Any hints?

Thanks,
Rob



Traceback (most recent call last):
   File "/Users/rob/Projects/Python/Samples/samples/pySketch/ 
pySketch.py", line 2634, in ?
     main()
   File "/Users/rob/Projects/Python/Samples/samples/pySketch/ 
pySketch.py", line 2629, in main
     _app = SketchApp(0)
   File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/ 
Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3- 
mac-unicode/wx/_core.py", line 5301, in __init__
   File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/ 
Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3- 
mac-unicode/wx/_core.py", line 4980, in _BootstrapApp
   File "/Users/rob/Projects/Python/Samples/samples/pySketch/ 
pySketch.py", line 2599, in OnInit
     frame = DrawingFrame(None, -1, "Untitled")
   File "/Users/rob/Projects/Python/Samples/samples/pySketch/ 
pySketch.py", line 218, in __init__
     wx.ArtProvider.GetBitmap(wx.ART_NEW, wx.ART_TOOLBAR, tsize),
AttributeError: 'module' object has no attribute 'ART_NEW'


From kevino at theolliviers.com  Tue Mar  7 17:10:30 2006
From: kevino at theolliviers.com (Kevin Ollivier)
Date: Tue, 7 Mar 2006 08:10:30 -0800
Subject: [Pythonmac-SIG] wx.ART_NEW,
	'module' object has no attribute 'ART_NEW'
In-Reply-To: <87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
	<87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>
Message-ID: <13889738-AD68-48E0-8A00-6AFBF5B7C0A3@theolliviers.com>

Hi Rob,

On Mar 7, 2006, at 7:51 AM, Rob J Goedman wrote:

> Hi,
>
> I'm new to Python. For a biometrix application I am investigating
> python or java (right
> now its in VB).
>
> Earlier this morning I installed (on OS 10.4.5) the latest version of
> MacPython-OSX-2.4.1,
> the Mac fix (TigerPython24Fix-r2) and wxPython2.6-osx-unicode-2.6.2.1-
> macosx10.3-py2.4.
>
> Most things work great (including the wxPython Demo.app), but double
> clicking demo.pyw
> or pySketch.py (and I guess several others) produces below error
> message.
>
> Any hints?

I suspect what is happening is that Python Launcher is using the  
system Python 2.3 installation instead of the Python 2.4 one you  
installed. Do a Get Info on the demo.pyw file and see if it says  
PythonLauncher (2.3.5) as the default app to open that file. If it  
does, try changing it to PythonLauncher (PythonLauncher version 0.1)  
and see if that helps.

Kevin

> Thanks,
> Rob
>
>
>
> Traceback (most recent call last):
>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
> pySketch.py", line 2634, in ?
>      main()
>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
> pySketch.py", line 2629, in main
>      _app = SketchApp(0)
>    File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/
> Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3-
> mac-unicode/wx/_core.py", line 5301, in __init__
>    File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/
> Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3-
> mac-unicode/wx/_core.py", line 4980, in _BootstrapApp
>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
> pySketch.py", line 2599, in OnInit
>      frame = DrawingFrame(None, -1, "Untitled")
>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
> pySketch.py", line 218, in __init__
>      wx.ArtProvider.GetBitmap(wx.ART_NEW, wx.ART_TOOLBAR, tsize),
> AttributeError: 'module' object has no attribute 'ART_NEW'
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From vinogri at mcmaster.ca  Tue Mar  7 16:15:16 2006
From: vinogri at mcmaster.ca (Ivan Vinogradov)
Date: Tue, 7 Mar 2006 10:15:16 -0500
Subject: [Pythonmac-SIG]  universal binary installer, take 1
Message-ID: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>

 > I've placed a DMG with an installer for a universal build of
 > 2.4.2. This is an initial release of this installer, there are
 > probably issues with it.
 > ... <snip>

Guidance requested.

1) Does this installation still require TigerPython24Fix to be applied?

Hmm, the TigerPython24Fix-r2.mpkg refuses to install it anyways:
Under Select a Destination:
"You cannot install TigerPython24Fix (r2) on this volume. Make sure  
your system meet the requirements"
Other .mpkg packages didn't complain and seem to work fine.

2) As is having problems with building extensions:
Happy-Happy-Joy-Joy:~/Desktop/vector3-0.6 iv$ python2.4 setup.py install
running install
running build
running build_ext
building 'vector3' extension
Traceback (most recent call last):
   File "setup.py", line 46, in ?
     ext_modules = [
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/core.py", line 149, in setup
     dist.run_commands()
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/dist.py", line 946, in run_commands
     self.run_command(cmd)
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/dist.py", line 966, in run_command
     cmd_obj.run()
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/command/install.py", line 506, in run
     self.run_command('build')
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/cmd.py", line 333, in run_command
     self.distribution.run_command(command)
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/dist.py", line 966, in run_command
     cmd_obj.run()
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/command/build.py", line 112, in run
     self.run_command(cmd_name)
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/cmd.py", line 333, in run_command
     self.distribution.run_command(command)
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/dist.py", line 966, in run_command
     cmd_obj.run()
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/command/build_ext.py", line 279, in run
     self.build_extensions()
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/command/build_ext.py", line 405, in build_extensions
     self.build_extension(ext)
   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
python2.4/distutils/command/build_ext.py", line 442, in build_extension
     sources = self.swig_sources(sources, ext)
TypeError: swig_sources() takes exactly 2 arguments (3 given)

3) The included $PATH modifier doesn't work. Perhaps this is by  
design and is included just for testing, in this case apologies.

Currently, ~/.profile is modified to include:
| # Setting PATH for MacPython 2.4
| # The orginal version is saved in .cshrc.pysave
| PATH="/Library/Frameworks/Python.framework/Versions/Current/bin:$ 
{PATH}"

a) for *csh missing setenv to actually make the change outside the  
script
b) default shell since panther is bash, needs export
c) should this be a global setting anyways?
    ie. i-Installer latex installation add this to /etc/profile
| ## setloginpath added /usr/local/bin start at Mon Mar  6 16:18:17  
EST 2006
| ## Do not remove the previous line
| if [ `whoami` != "root" ]
|  then
|   PATH="$PATH:/usr/local/bin"
|   export PATH
| fi
| ## Do not remove the next line
| ## setloginpath added /usr/local/bin end at Mon Mar  6 16:18:17 EST  
2006


Regards, Ivan.


  

From ronaldoussoren at mac.com  Tue Mar  7 17:24:24 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 17:24:24 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
References: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
Message-ID: <BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>


Ivan,

First off, thanks for your feedback.

On 7-mrt-2006, at 16:15, Ivan Vinogradov wrote:

>> I've placed a DMG with an installer for a universal build of
>> 2.4.2. This is an initial release of this installer, there are
>> probably issues with it.
>> ... <snip>
>
> Guidance requested.
>
> 1) Does this installation still require TigerPython24Fix to be  
> applied?
>
> Hmm, the TigerPython24Fix-r2.mpkg refuses to install it anyways:
> Under Select a Destination:
> "You cannot install TigerPython24Fix (r2) on this volume. Make sure
> your system meet the requirements"
> Other .mpkg packages didn't complain and seem to work fine.

This package does not need other packages to function, the idea is that
this installer will work for anyone on Mac OS X 10.3 or later.

>
> 2) As is having problems with building extensions:
> Happy-Happy-Joy-Joy:~/Desktop/vector3-0.6 iv$ python2.4 setup.py  
> install
> running install
> running build
> running build_ext
> building 'vector3' extension
> Traceback (most recent call last):
>    File "setup.py", line 46, in ?
>      ext_modules = [
>    File "/Library/Frameworks/Python.framework/Versions/2.4//lib/
> python2.4/distutils/core.py", line 149, in setup
>      dist.run_commands()
>
[... strip most of stack trace ...]

>    File "/Library/Frameworks/Python.framework/Versions/2.4//lib/
> python2.4/distutils/command/build_ext.py", line 442, in  
> build_extension
>      sources = self.swig_sources(sources, ext)
> TypeError: swig_sources() takes exactly 2 arguments (3 given)


I don't use swig myself, but I do no this is not a part of the code I
have touched yet.  I seem to recall that setuptools contains a patch to
workaround a simular bug.

Is this by any change a open source software that I could play around  
with?

>
> 3) The included $PATH modifier doesn't work. Perhaps this is by
> design and is included just for testing, in this case apologies.

That is not intentional, I'll have a look.
>
> Currently, ~/.profile is modified to include:
> | # Setting PATH for MacPython 2.4
> | # The orginal version is saved in .cshrc.pysave
> | PATH="/Library/Frameworks/Python.framework/Versions/Current/bin:$
> {PATH}"
>
> a) for *csh missing setenv to actually make the change outside the
> script
> b) default shell since panther is bash, needs export
> c) should this be a global setting anyways?
>     ie. i-Installer latex installation add this to /etc/profile
> | ## setloginpath added /usr/local/bin start at Mon Mar  6 16:18:17
> EST 2006
> | ## Do not remove the previous line
> | if [ `whoami` != "root" ]
> |  then
> |   PATH="$PATH:/usr/local/bin"
> |   export PATH
> | fi
> | ## Do not remove the next line
> | ## setloginpath added /usr/local/bin end at Mon Mar  6 16:18:17 EST
> 2006
>

I don't want to update system files, you never know if some update
from apple will overwrite these changes. That seems unlikely with
/etc/profile, but I have run into this multiple times with /etc/ 
rc.common
(enabling network logging for syslogd).

Ronald

>
> Regards, Ivan.
>
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From goedman at mac.com  Tue Mar  7 17:29:25 2006
From: goedman at mac.com (Rob J Goedman)
Date: Tue, 7 Mar 2006 08:29:25 -0800
Subject: [Pythonmac-SIG] wx.ART_NEW,
	'module' object has no attribute 'ART_NEW'
In-Reply-To: <13889738-AD68-48E0-8A00-6AFBF5B7C0A3@theolliviers.com>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
	<87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>
	<13889738-AD68-48E0-8A00-6AFBF5B7C0A3@theolliviers.com>
Message-ID: <E9DEB986-CF13-4E7B-8803-9DFDAE94080A@mac.com>

Thanks Kevin,

That does the trick!

Thanks again.

Rob


On Mar 7, 2006, at 8:10 AM, Kevin Ollivier wrote:

> Hi Rob,
>
> On Mar 7, 2006, at 7:51 AM, Rob J Goedman wrote:
>
>> Hi,
>>
>> I'm new to Python. For a biometrix application I am investigating
>> python or java (right
>> now its in VB).
>>
>> Earlier this morning I installed (on OS 10.4.5) the latest version of
>> MacPython-OSX-2.4.1,
>> the Mac fix (TigerPython24Fix-r2) and wxPython2.6-osx- 
>> unicode-2.6.2.1-
>> macosx10.3-py2.4.
>>
>> Most things work great (including the wxPython Demo.app), but double
>> clicking demo.pyw
>> or pySketch.py (and I guess several others) produces below error
>> message.
>>
>> Any hints?
>
> I suspect what is happening is that Python Launcher is using the  
> system Python 2.3 installation instead of the Python 2.4 one you  
> installed. Do a Get Info on the demo.pyw file and see if it says  
> PythonLauncher (2.3.5) as the default app to open that file. If it  
> does, try changing it to PythonLauncher (PythonLauncher version  
> 0.1) and see if that helps.
>
> Kevin
>
>> Thanks,
>> Rob
>>
>>
>>
>> Traceback (most recent call last):
>>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
>> pySketch.py", line 2634, in ?
>>      main()
>>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
>> pySketch.py", line 2629, in main
>>      _app = SketchApp(0)
>>    File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/
>> Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3-
>> mac-unicode/wx/_core.py", line 5301, in __init__
>>    File "/BinaryCache/wxWidgets/wxWidgets-2.root~174/System/Library/
>> Frameworks/Python.framework/Versions/2.3/Extras/lib/python/wx-2.5.3-
>> mac-unicode/wx/_core.py", line 4980, in _BootstrapApp
>>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
>> pySketch.py", line 2599, in OnInit
>>      frame = DrawingFrame(None, -1, "Untitled")
>>    File "/Users/rob/Projects/Python/Samples/samples/pySketch/
>> pySketch.py", line 218, in __init__
>>      wx.ArtProvider.GetBitmap(wx.ART_NEW, wx.ART_TOOLBAR, tsize),
>> AttributeError: 'module' object has no attribute 'ART_NEW'
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>


From smithsm at samuelsmith.org  Tue Mar  7 17:49:13 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Tue, 7 Mar 2006 09:49:13 -0700
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
Message-ID: <016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>

I use PYTHONPATH to add my own python development directories so that
modules I create myself or don't want to install in the default:
/library/frameworks/python.framework/versions/2.4/lib/python2.4/site- 
packages/

Here is I how set mine up in .bashrc

PYTHONPATH=~/Data/Code/python:~/Data/Adept/Code/python:~/Data/ 
ProSapien/Code/python:~/Sites/ht/lib
export PYTHONPATH

To see what this produces open python

import sys
 >>> sys.path
['', '/Users/smithsm/Data/Code/python',
'/Users/smithsm/Data/Adept/Code/python',
'/Users/smithsm/Data/ProSapien/Code/python',
'/Users/smithsm/Sites/ht/lib',
'/Library/Frameworks/Python.framework/Versions/2.4/lib/python24.zip',  
'/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4', '/ 
Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/plat- 
darwin', '/Library/Frameworks/Python.framework/Versions/2.4/lib/ 
python2.4/plat-mac', '/Library/Frameworks/Python.framework/Versions/ 
2.4/lib/python2.4/plat-mac/lib-scriptpackages', '/Library/Frameworks/ 
Python.framework/Versions/2.4/lib/python2.4/lib-tk', '/Library/ 
Frameworks/Python.framework/Versions/2.4/lib/python2.4/lib-dynload',  
'/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site- 
packages', '/Library/Frameworks/Python.framework/Versions/2.4/lib/ 
python2.4/site-packages/Numeric', '/Library/Frameworks/ 
Python.framework/Versions/2.4/lib/python2.4/site-packages/PIL', '/ 
Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site- 
packages/PyObjC', '/Library/Frameworks/Python.framework/Versions/2.4/ 
lib/python2.4/site-packages/setuptools-0.6a10dev_r42195-py2.4.egg', '/ 
Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site- 
packages/numpy-0.9.5.2006-py2.4-macosx-10.4-ppc.egg', '/Library/ 
Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-packages/ 
numarray-1.5.1-py2.4-macosx-10.4-ppc.egg', '/Library/Frameworks/ 
Python.framework/Versions/2.4/lib/python2.4/site-packages/py2app', '/ 
Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site- 
packages/wx-2.6-mac-ansi']

So the stuff you put in PYTHONPATH get prepended to the list of  
places python looks into for modules when you
do an import.

I don't know how the other stuff (besides what is in PYTHONPATH) gets  
assigned to the sys.path variable. But
apparently different package installers know how to assign stuff to  
sys.path.

For comparison

Here is my PATH
echo $PATH
/usr/local/bin:/opt/local/bin:/usr/local/teTeX/bin/powerpc-apple- 
darwin-current:/bin:/sbin:/usr/bin:/usr/sbin:/Developer/Tools:/ 
Library/Frameworks/Python.framework/Versions/Current/bin/

Then to complicate things even more there are the .pth files that can  
further modify the path
python uses to lookup modules


On 07 Mar, 2006, at 07:56, Ronald Oussoren wrote:

>
> On 7-mrt-2006, at 15:01, Charles Hartman wrote:
>
>>
>> On Mar 7, 2006, at 2:39 AM, Christopher Barker wrote:
>>
>>> Charles Hartman wrote:
>>>> I'm getting really tired of not understanding this: Can someone
>>>> point  me to some single place that explains the whole path /
>>>> PATH /  PYTHONPATH arrangement for Python? I know there are two
>>>> sorts on the  Mac, framework and /usr, and I know some other bits
>>>> and pieces, but I  have no clear overall picture.
>>>> This came to my attention when I downloaded and installed
>>>> IPython,  which looks very nifty. I put it in my MacPython folder
>>>> in / Applications -- but that's not really where it belongs,
>>>
>>> I think ipython is an executable script, so you want to put it
>>> where those go. For stuff not installed by Apple, that is usually:
>>>
>>> /usr/local/bin
>>>
>>> then you want to make sure /usr/local/bin is on your PATH. You do
>>> that by adding it to your /Users/YourName/.profile file, like this:
>>>
>>> export PATH=/usr/local/bin:$PATH
>>>
>>> A little googling should give you more info if you need it.
>>
>> Thank you. This much I knew, but it's useful to have it spelled out.
>>
>> I think that part of what confuses a non-Unix person like me is the
>> distinction PATH / (executable programs and scripts) vs. PYTHONPATH /
>> modules. (This comes up with IPython because embedding its
>> interpreter in another program requires importing a module
>> IPShellEmbed.) I find that my environment contains no PYTHONPATH
>> variable, presumably because I've just been using the modules
>> supplied with Python and with wxPython, and the installers have put
>> them in (some) default places. So I assume that the new installer,
>> too, won't set a PYTHONPATH.
>
> PYTHONPATH is a way to extend the search path for python modules,
> e.g. sys.path,
> while PATH is used by the unix shell to look for programs.
>
> The new installer will indeed not set PYTHONPATH, the default module
> search
> path is good enough. It will update PATH, because the bin directory
> inside
> the framework is not usually on PATH.
>
> Ronald
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig

**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From ronaldoussoren at mac.com  Tue Mar  7 18:00:48 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 18:00:48 +0100
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
Message-ID: <8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>


On 7-mrt-2006, at 17:49, Samuel M. Smith wrote:

> I use PYTHONPATH to add my own python development directories so that
> modules I create myself or don't want to install in the default:
> /library/frameworks/python.framework/versions/2.4/lib/python2.4/site-
> packages/

The time machine strikes again :-). You can also install your own
packages in ~/Library/Python/2.4/site-packages. That directory is
a proper site-package, .pth files work in there.

[...]
>
> So the stuff you put in PYTHONPATH get prepended to the list of
> places python looks into for modules when you
> do an import.
>
> I don't know how the other stuff (besides what is in PYTHONPATH) gets
> assigned to the sys.path variable. But
> apparently different package installers know how to assign stuff to
> sys.path.

It is initialised to the directory containing the script followed by  
the python library.
PYTHONPATH is then prepended to sys.path. site.py will then add the  
site packages
directories to sys.path. The contents of .pth files in the site  
packages directories
is then added sys.path (again by site.py).

That's how package installers update your sys.path: they install .pth  
files.

Ronald



From daniel at brightfire.com  Tue Mar  7 17:01:33 2006
From: daniel at brightfire.com (Daniel Lord)
Date: Tue, 7 Mar 2006 08:01:33 -0800
Subject: [Pythonmac-SIG] re Icon discussion
In-Reply-To: <mailman.9612.1141674831.27774.pythonmac-sig@python.org>
References: <mailman.9612.1141674831.27774.pythonmac-sig@python.org>
Message-ID: <010EB49D-1C48-452B-9633-9292185F5495@brightfire.com>

(deleted thread as it was getting too long)

About this icon discussion, why don't we use a solution that  
maintains, reinforces, and capitalizes on the overall identity of  
cross-platform Python while distinguishing the OS X Mac-Python port  
by just taking the new Blue-Yellow Python logo and 'aquafying it' by  
beveling and 'glassing' it? As long as the Python logo caretakers  
don't think that dlitues the brand or trademark significantly that  
is. HAS has expressed concerns about regarding our deviating too much  
from the Python brand and I agree with him on that. Perhaps we could  
then get Active State, which has ported Komodo to OS X (which I use  
and fidn pretty good as cross platform IDE for Perl and Python) to  
maybe put it on their page. Yes, I know they do their own version but  
I also know their version doesn't include a lot this one does. But  
they are 'employee-owned' now and might agree that helping teh  
MacPython brand was in the community interest. Worth a try, but first  
you need to agree on a logo.

(side note: Active State Python omits not only Apple stuff--they  
don't include SSL incredibly. I discovered this while creating a  
Gmail mailing script. Gmail requires TLS, but it failed to find SSL  
on my WIndows box because I was using Active State's Python--it was  
found once I removed it and installed the Python.org Windows port.  
Shame.)



From vinogri at mcmaster.ca  Tue Mar  7 18:13:44 2006
From: vinogri at mcmaster.ca (Ivan Vinogradov)
Date: Tue, 7 Mar 2006 12:13:44 -0500
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
References: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
	<BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
Message-ID: <2E9A8FA9-53DE-4CC2-981D-33EBB5075DAE@mcmaster.ca>


On 7-Mar-06, at 11:24 AM, Ronald Oussoren wrote:

>
> Ivan,
>
> First off, thanks for your feedback.

Mutual well-being :)

> ...<snip>
>
> I don't use swig myself, but I do no this is not a part of the code I
> have touched yet.  I seem to recall that setuptools contains a  
> patch to
> workaround a simular bug.
>
> Is this by any change a open source software that I could play  
> around with?
>

3D vector operations module using Pyrex. Here's the link:
http://arbutus.physics.mcmaster.ca/dmc/software/vector3.html

> ... <snap>
> I don't want to update system files, you never know if some update
> from apple will overwrite these changes. That seems unlikely with
> /etc/profile, but I have run into this multiple times with /etc/ 
> rc.common
> (enabling network logging for syslogd).

Then perhaps a SetPath.sh script distributed in MacPython 2.4 folder  
for users other than the original installer?

Cheers, Ivan.



From Chris.Barker at noaa.gov  Tue Mar  7 18:25:10 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 07 Mar 2006 09:25:10 -0800
Subject: [Pythonmac-SIG] wx.ART_NEW,
	'module' object has no attribute 'ART_NEW'
In-Reply-To: <E9DEB986-CF13-4E7B-8803-9DFDAE94080A@mac.com>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
	<87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>
	<13889738-AD68-48E0-8A00-6AFBF5B7C0A3@theolliviers.com>
	<E9DEB986-CF13-4E7B-8803-9DFDAE94080A@mac.com>
Message-ID: <440DC1F6.7050303@noaa.gov>

> On Mar 7, 2006, at 8:10 AM, Kevin Ollivier wrote:

>> I suspect what is happening is that Python Launcher is using the  
>> system Python 2.3 installation instead of the Python 2.4 one you  
>> installed. Do a Get Info on the demo.pyw file and see if it says  
>> PythonLauncher (2.3.5) as the default app to open that file. If it  
>> does, try changing it to PythonLauncher (PythonLauncher version  
>> 0.1) and see if that helps.

as I understand it, PythonLauncher respects the #! line of start up 
scripts. That means that you could change the #! line in the relevant 
scripts to something like:

#!/usr/bin/env pythonw2.4

I like to put a version in my #! lines, as it document what version I've 
tested with, and it lets me have scripts that use different version on 
the same system.

Will changing the default app for a given *.pyw script change it for all 
*.pyw scripts?

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Chris.Barker at noaa.gov  Tue Mar  7 18:25:39 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 07 Mar 2006 09:25:39 -0800
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
Message-ID: <440DC213.7080301@noaa.gov>

Charles Hartman wrote:
> I think that part of what confuses a non-Unix person like me is the 
> distinction PATH / (executable programs and scripts) vs. PYTHONPATH / 
> modules. (This comes up with IPython because embedding its interpreter 
> in another program requires importing a module IPShellEmbed.)

Usually, they really are totally unrelated, but I guess it can be 
confusing when a package contains a program and also modules for use 
with other programs.

> I find 
> that my environment contains no PYTHONPATH variable, presumably because 
> I've just been using the modules supplied with Python and with wxPython, 
> and the installers have put them in (some) default places. So I assume 
> that the new installer, too, won't set a PYTHONPATH.

I have never needed a PYTHONPATH. Almost everything gets installed to 
the default location (site-packages). That's probably the case for 
almost all people that are the primary user of their machine.

The main exception is if you have a bunch of modules you want to use 
that you don't want to install in the system python. I've never had a 
need for that, but some do.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From vinogri at mcmaster.ca  Tue Mar  7 19:06:57 2006
From: vinogri at mcmaster.ca (Ivan Vinogradov)
Date: Tue, 7 Mar 2006 13:06:57 -0500
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
References: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
	<BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
Message-ID: <3D40BB44-5F65-461A-B565-BACFE4A1A487@mcmaster.ca>

Follow-up. (Yes, most answers were staring me in the face. I blame  
the cold medicine.)

> <snip>
>>    File "/Library/Frameworks/Python.framework/Versions/2.4//lib/
>> python2.4/distutils/command/build_ext.py", line 442, in  
>> build_extension
>>      sources = self.swig_sources(sources, ext)
>> TypeError: swig_sources() takes exactly 2 arguments (3 given)
>
>
> I don't use swig myself, but I do no this is not a part of the code I
> have touched yet.  I seem to recall that setuptools contains a  
> patch to
> workaround a simular bug.

This also can be resolved by using Pyrex-0.9.3.1 setup.py install.
The package at pythonmac.org is Pyrex-0.9.3 .
The bdist_mpkg is scary smooth in resolving that.

>> 3) The included $PATH modifier doesn't work.

Scratch this. Now it works fine. I'm just not sure how.

From charles.hartman at conncoll.edu  Tue Mar  7 19:26:07 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 13:26:07 -0500
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <440DC213.7080301@noaa.gov>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<440DC213.7080301@noaa.gov>
Message-ID: <DBE23CF2-91B4-4632-83FB-9D4736629C7A@conncoll.edu>

It turns out there's a rather helpful page at

http://geosci.uchicago.edu/~tobis/pylab.html

which (since I'm not using matplotlib) I might never have found, but  
which is referred to in the manual for IPython.

Some of this stuff could be gathered together for the wiki . . . I'll  
look at doing that, but everybody else will have to correct it.

Charles


From ronaldoussoren at mac.com  Tue Mar  7 19:54:57 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 19:54:57 +0100
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <3D40BB44-5F65-461A-B565-BACFE4A1A487@mcmaster.ca>
References: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
	<BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
	<3D40BB44-5F65-461A-B565-BACFE4A1A487@mcmaster.ca>
Message-ID: <25E98CE9-7D10-4BC0-9819-34758166DE88@mac.com>


On 7-mrt-2006, at 19:06, Ivan Vinogradov wrote:

> Follow-up. (Yes, most answers were staring me in the face. I blame
> the cold medicine.)
>
>> <snip>
>>>    File "/Library/Frameworks/Python.framework/Versions/2.4//lib/
>>> python2.4/distutils/command/build_ext.py", line 442, in
>>> build_extension
>>>      sources = self.swig_sources(sources, ext)
>>> TypeError: swig_sources() takes exactly 2 arguments (3 given)
>>
>>
>> I don't use swig myself, but I do no this is not a part of the code I
>> have touched yet.  I seem to recall that setuptools contains a
>> patch to
>> workaround a simular bug.
>
> This also can be resolved by using Pyrex-0.9.3.1 setup.py install.
> The package at pythonmac.org is Pyrex-0.9.3 .
> The bdist_mpkg is scary smooth in resolving that.

I don't understand why this helps. What does pyrex have to do
with swig?

>
>>> 3) The included $PATH modifier doesn't work.
>
> Scratch this. Now it works fine. I'm just not sure how.

Just after I had updated the script :-). I've also added
"Update Shell Profile.command" to the "/Application/MacPython 2.4"
folder.

I haven't updated the package on my iDisk yet, I'll do that
later.

Ronald

> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From vinogri at mcmaster.ca  Tue Mar  7 20:30:08 2006
From: vinogri at mcmaster.ca (Ivan Vinogradov)
Date: Tue, 7 Mar 2006 14:30:08 -0500
Subject: [Pythonmac-SIG] universal binary installer, take 1
In-Reply-To: <25E98CE9-7D10-4BC0-9819-34758166DE88@mac.com>
References: <2C44316F-64C7-47B0-813A-466EAB979D85@mcmaster.ca>
	<BDD902F2-F9AA-4DDB-AB29-77E9BEF94C83@mac.com>
	<3D40BB44-5F65-461A-B565-BACFE4A1A487@mcmaster.ca>
	<25E98CE9-7D10-4BC0-9819-34758166DE88@mac.com>
Message-ID: <6942B0C7-51AA-41E5-936F-76F8D827A532@mcmaster.ca>


On 7-Mar-06, at 1:54 PM, Ronald Oussoren wrote:
>>> <snip>
>>>>    File "/Library/Frameworks/Python.framework/Versions/2.4//lib/
>>>> python2.4/distutils/command/build_ext.py", line 442, in
>>>> build_extension
>>>>      sources = self.swig_sources(sources, ext)
>>>> TypeError: swig_sources() takes exactly 2 arguments (3 given)
>>>
>>>
>>> I don't use swig myself, but I do no this is not a part of the  
>>> code I
>>> have touched yet.  I seem to recall that setuptools contains a
>>> patch to
>>> workaround a simular bug.
>>
>> This also can be resolved by using Pyrex-0.9.3.1 setup.py install.
>> The package at pythonmac.org is Pyrex-0.9.3 .
>> The bdist_mpkg is scary smooth in resolving that.
>
> I don't understand why this helps. What does pyrex have to do
> with swig?

Uneducated guess: it has probably more to do with distutil's way of  
building extensions.
distutil's build_ext.py seems to only know about swig extensions 
(cursory review, cold medicine).

In this project's setup.py(select pieces):
...
try:
     from Pyrex.Distutils import build_ext
     vector3_source = 'vector3.pyx'
     setup_kw = {'cmdclass' : {'build_ext' : build_ext}}
except ImportError:
     vector3_source = 'vector3.c'
     setup_kw = {}
...
setup(
   ...
     ext_modules = [
         Extension('vector3', [vector3_source,]),
         ],
**setup_kw)

The above should somehow educate distutils on how to build Pyrex  
extensions; probably something has gone out of sync between distutils  
and Pyrex after 0.9.3.
It seems to be a well known bug indeed.




From charles.hartman at conncoll.edu  Tue Mar  7 21:54:37 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 15:54:37 -0500
Subject: [Pythonmac-SIG] a bit for the FAQ on IPython
Message-ID: <3821F535-70E0-4586-8137-78AF79DA2B42@conncoll.edu>


(I don't know where this goes. I've phrased it for Terminal beginners  
-- now *there's* a phrase.)

========================

IPython is a replacement for the default Python interpreter. It  
includes better history and tab-completion features, provides more  
elaborate ways of inspecting objects, and can be embedded in your own  
applications to help you debug them. See http://ipython.scipy.org/  
for a full description.

(If you also want to install matplotlib, follow the instructions at  
http://geosci.uchicago.edu/~tobis/pylab.html instead of these. The  
steps given there include the steps given here.)

The scipy.org site gives instructions for installing IPython, but  
they are not quite correct for the Mac. Assuming that you have  
installed Python 2.4, here are the correct steps:

1. Download the IPython archive. As of 7 March 2006 it is at
	http://ipython.scipy.org/dist/ipython-0.7.1.fix1.tar.gz
Move the downloaded file to any convenient place, such as the  
MacPython folder in your Applications folder. Double-click  
the .tar.gz file to unpack the archive. Then double-click the  
resulting folder (currently ipython-0.7.1.fix1) to open it.

2. Start the Terminal. Type "cd " (with a trailing space) at the  
command prompt, but don't press Return. Click once on the folder icon  
in the title bar of the folder you just opened. Then click and drag  
the icon into the Terminal window. The name of the folder will be  
added to your "cd" ("change directory") command. Press Return.

4. At the command prompt type
	sudo pythonw setup.py install --install-scripts=/usr/local/bin
and press Return.

 From the Terminal command prompt, from any directory, you can now  
enter the command "ipython" to use the enhanced interpreter. These  
steps also install the IPythonShell module where it can be found by  
an "import" statement, which is what you do to embed the interpreter  
in your own program.

==================================================

Charles


From epalmore at pixar.com  Tue Mar  7 20:07:50 2006
From: epalmore at pixar.com (Elizabeth Palmore)
Date: Tue, 7 Mar 2006 11:07:50 -0800
Subject: [Pythonmac-SIG] Software Development Engineer in Test Position
	Available @ Pixar!
Message-ID: <223FBFD8-0C3B-4F82-B8B5-1E350820CD02@pixar.com>

We are seeking an experienced QA Engineer with a passion for quality  
to participate in the software quality assurance efforts for our  
proprietary in house toolset.

This position requires close collaboration with the engineering staff  
to define, develop, execute, and automate API level test plans and  
test cases. Identify and communicate a strategy for API and other  
testing methodologies, review and report on test progress, status,  
and coverage, and meet test completion and delivery milestones that  
you help define. Work closely with development, project management  
and documentation to coordinate testing responsibilities.

Responsibilities:

? Develop and execute an API testing strategy, using API testing  
methodology and programming language knowledge
? Develop and implement API tests
? Contribute to setting and evaluating milestone criteria such that  
product is released on schedule with high quality
? Design and implement quality processes for a small team of senior  
software developers
? Work closely with the core development teams during all phases of  
the product life cycle
? Evaluate completeness and effectiveness of developer's unit tests

Qualifications:

? 5+ years of experience in QA or software development
? Bachelor's degree in Computer Science or equivalent
? 3+ years of experience testing APIs and complex data structures
? Excellent design and coding skills ( C/C++ or  and a scripting  
language such as Python or Perl)
? Knowledge of QA methodology, processes, and tools
? Direct experience testing software under Unix (e.g., Linux or OSX)
? Cross platform development experience is preferred (Macintosh, Linux)
? Experience with 3D graphics applications is a plus
? Experience with Objective-C/Cocoa is a plus

Pixar Animation Studios
Emeryville, CA
Full - Time
(Sorry, new visa sponsorship is not available.  Must be currently  
authorized to work in the US)

About Pixar Animation Studios:

Pixar Animation Studios combines creative and technical artistry to  
create original stories in the medium of computer animation.  Pixar  
has created six of the most successful and beloved animated films of  
all time: Toy Story, A Bug's Life, Toy Story 2, Monsters, Inc.,  
Finding Nemo and The Incredibles.  Pixar has won 18 Academy Awards?  
and its six films have grossed more than $3.2 billion at the  
worldwide box office to date.  The Northern California studio will  
release its next film, Cars, on June 9, 2006.  JOIN OUR TEAM!


To apply please go to www.pixar.com. Click on "Company Info" then  
"Jobs" the "Sofware Development Engineer in Test." Or send resume AND  
cover letter to epalmore at pixar.com.

From Jack.Jansen at cwi.nl  Tue Mar  7 22:18:05 2006
From: Jack.Jansen at cwi.nl (Jack Jansen)
Date: Tue, 7 Mar 2006 22:18:05 +0100
Subject: [Pythonmac-SIG] Job postings - allow or not?
Message-ID: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>

Folks,
I've just let a job posting through to the list, after looking at it  
for a long time and unsure whether or not we want these.

If you have a strong opinion (or a weak opinion, or anything  
resembling an opinion:-) on the matter: please let me know, then I'll  
use the collective feedback as a guidance in the future.
--
Jack Jansen, <Jack.Jansen at cwi.nl>, http://www.cwi.nl/~jack
If I can't dance I don't want to be part of your revolution -- Emma  
Goldman



From kquirk at solidworks.com  Tue Mar  7 22:23:18 2006
From: kquirk at solidworks.com (Kent Quirk)
Date: Tue, 7 Mar 2006 16:23:18 -0500
Subject: [Pythonmac-SIG] Job postings - allow or not?
Message-ID: <AD85EF0CD3800548B3B22C832F08788E028A6601@CORP-MAIL6.solidworks.swk>

I think the occasional Python-related job posting is Just Fine. I'd be
bugged if there were too many of them, but I wouldn't kick them to
another list unless and until there were so many that they formed a
measurable fraction of the list traffic. At which point I'd rejoice at
Python's success.

	Kent


-----Original Message-----
From: pythonmac-sig-bounces+kquirk=solidworks.com at python.org
[mailto:pythonmac-sig-bounces+kquirk=solidworks.com at python.org] On
Behalf Of Jack Jansen
Sent: Tuesday, March 07, 2006 4:18 PM
To: pythonmac-sig at python.org
Subject: [Pythonmac-SIG] Job postings - allow or not?

Folks,
I've just let a job posting through to the list, after looking at it  
for a long time and unsure whether or not we want these.

If you have a strong opinion (or a weak opinion, or anything  
resembling an opinion:-) on the matter: please let me know, then I'll  
use the collective feedback as a guidance in the future.

From charles.hartman at conncoll.edu  Tue Mar  7 22:25:22 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 16:25:22 -0500
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
References: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
Message-ID: <CAE2774E-8C7E-4652-94B2-2CE6883E4E32@conncoll.edu>

Quick reaction: As a list-member with no possible personal use for  
job postings here, I say, Fine. Interesting to see. If there get to  
be a lot, they could be broken out into a sublist perhaps. And if  
they look bogus, we get to get together & pelt the senders with eggs.  
(Courtesy of cheeseshop I hope.)

Charles

On Mar 7, 2006, at 4:18 PM, Jack Jansen wrote:

> Folks,
> I've just let a job posting through to the list, after looking at it
> for a long time and unsure whether or not we want these.
>
> If you have a strong opinion (or a weak opinion, or anything
> resembling an opinion:-) on the matter: please let me know, then I'll
> use the collective feedback as a guidance in the future.
> --
> Jack Jansen, <Jack.Jansen at cwi.nl>, http://www.cwi.nl/~jack
> If I can't dance I don't want to be part of your revolution -- Emma
> Goldman
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar  7 22:32:00 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 7 Mar 2006 22:32:00 +0100
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <AD85EF0CD3800548B3B22C832F08788E028A6601@CORP-MAIL6.solidworks.swk>
References: <AD85EF0CD3800548B3B22C832F08788E028A6601@CORP-MAIL6.solidworks.swk>
Message-ID: <61889D2F-A576-401B-8C18-4F333B0E2742@mac.com>


On 7-mrt-2006, at 22:23, Kent Quirk wrote:

> I think the occasional Python-related job posting is Just Fine. I'd be
> bugged if there were too many of them, but I wouldn't kick them to
> another list unless and until there were so many that they formed a
> measurable fraction of the list traffic. At which point I'd rejoice at
> Python's success.
>

I agree. Unless there'd be a lot of job postings I don't mind.

Ronald

> 	Kent
>
>
> -----Original Message-----
> From: pythonmac-sig-bounces+kquirk=solidworks.com at python.org
> [mailto:pythonmac-sig-bounces+kquirk=solidworks.com at python.org] On
> Behalf Of Jack Jansen
> Sent: Tuesday, March 07, 2006 4:18 PM
> To: pythonmac-sig at python.org
> Subject: [Pythonmac-SIG] Job postings - allow or not?
>
> Folks,
> I've just let a job posting through to the list, after looking at it
> for a long time and unsure whether or not we want these.
>
> If you have a strong opinion (or a weak opinion, or anything
> resembling an opinion:-) on the matter: please let me know, then I'll
> use the collective feedback as a guidance in the future.
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From delza at livingcode.org  Tue Mar  7 22:48:42 2006
From: delza at livingcode.org (Dethe Elza)
Date: Tue, 7 Mar 2006 13:48:42 -0800
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
References: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
Message-ID: <24d517dd0603071348i422a282br268335b629192f5c@mail.gmail.com>

I'm OK with job postings on the list.  Its interesting to watch Python
and OS X taking off.

--Dethe

From bob at redivi.com  Tue Mar  7 22:57:51 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 7 Mar 2006 13:57:51 -0800
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <61889D2F-A576-401B-8C18-4F333B0E2742@mac.com>
References: <AD85EF0CD3800548B3B22C832F08788E028A6601@CORP-MAIL6.solidworks.swk>
	<61889D2F-A576-401B-8C18-4F333B0E2742@mac.com>
Message-ID: <7D654075-0D98-4EB0-8CD1-1F69335967F7@redivi.com>


On Mar 7, 2006, at 1:32 PM, Ronald Oussoren wrote:

>
> On 7-mrt-2006, at 22:23, Kent Quirk wrote:
>
>> I think the occasional Python-related job posting is Just Fine.  
>> I'd be
>> bugged if there were too many of them, but I wouldn't kick them to
>> another list unless and until there were so many that they formed a
>> measurable fraction of the list traffic. At which point I'd  
>> rejoice at
>> Python's success.
>>
>
> I agree. Unless there'd be a lot of job postings I don't mind.

+1

-bob


From ed at leafe.com  Tue Mar  7 23:22:51 2006
From: ed at leafe.com (Ed Leafe)
Date: Tue, 7 Mar 2006 17:22:51 -0500
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <61889D2F-A576-401B-8C18-4F333B0E2742@mac.com>
References: <AD85EF0CD3800548B3B22C832F08788E028A6601@CORP-MAIL6.solidworks.swk>
	<61889D2F-A576-401B-8C18-4F333B0E2742@mac.com>
Message-ID: <159E6601-0BAF-41B4-98B4-C78C654211CC@leafe.com>

On Mar 7, 2006, at 4:32 PM, Ronald Oussoren wrote:

> I agree. Unless there'd be a lot of job postings I don't mind.

	If they were real jobs, listed by real companies, then I don't mind.  
What I would object to would be recruiters and HR people using the  
list as a resum?-gathering device.


-- Ed Leafe
-- http://leafe.com
-- http://dabodev.com




From goedman at mac.com  Wed Mar  8 01:31:28 2006
From: goedman at mac.com (Rob J Goedman)
Date: Tue, 7 Mar 2006 16:31:28 -0800
Subject: [Pythonmac-SIG] wx.ART_NEW,
	'module' object has no attribute 'ART_NEW'
In-Reply-To: <440DC1F6.7050303@noaa.gov>
References: <BCE6D68F-E430-40F7-B11D-C5DAC7A94019@stanford.edu>
	<87C23A93-0565-4373-B9FD-3401756ACB81@mac.com>
	<13889738-AD68-48E0-8A00-6AFBF5B7C0A3@theolliviers.com>
	<E9DEB986-CF13-4E7B-8803-9DFDAE94080A@mac.com>
	<440DC1F6.7050303@noaa.gov>
Message-ID: <B919ADA1-675C-44D5-AB86-9E41D280B3B6@mac.com>

Chris,

Thanks. Kevin's suggestion works fine and I prefer not to edit  
individual files.

I have noticed that changing the default app refuses to pick up the  
new PythonLauncher
0.1 for some reason, even if you tell it in 'Get Info' to change it  
for all .py and .pyw files.
Weird.

Rob



On Mar 7, 2006, at 9:25 AM, Christopher Barker wrote:

>> On Mar 7, 2006, at 8:10 AM, Kevin Ollivier wrote:
>
>>> I suspect what is happening is that Python Launcher is using the   
>>> system Python 2.3 installation instead of the Python 2.4 one you   
>>> installed. Do a Get Info on the demo.pyw file and see if it says   
>>> PythonLauncher (2.3.5) as the default app to open that file. If  
>>> it  does, try changing it to PythonLauncher (PythonLauncher  
>>> version  0.1) and see if that helps.
>
> as I understand it, PythonLauncher respects the #! line of start up  
> scripts. That means that you could change the #! line in the  
> relevant scripts to something like:
>
> #!/usr/bin/env pythonw2.4
>
> I like to put a version in my #! lines, as it document what version  
> I've tested with, and it lets me have scripts that use different  
> version on the same system.
>
> Will changing the default app for a given *.pyw script change it  
> for all *.pyw scripts?
>
> -Chris
>
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                     		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov


From charles.hartman at conncoll.edu  Wed Mar  8 02:05:36 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 20:05:36 -0500
Subject: [Pythonmac-SIG] Unix-amateur question
Message-ID: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>

(Sorry! How simple can they get? But I don't know of a better place  
to ask -- does anyone?)

How do I execute a Mac application from the Terminal command line?

Specifically, I'm trying to specify BBEdit in the EDITOR environment  
variable which is consulted by IPython. EDITOR= what? Not / 
Applications/BBEdit, or /Applications/BBEdit.app. Probably something  
that continues with /Contents/ . . . but then I get lost. And I don't  
quite know how to experiment because I'm not sure how to attempt to  
run it from the bash prompt . . .

Charles


From ed at leafe.com  Wed Mar  8 02:23:12 2006
From: ed at leafe.com (Ed Leafe)
Date: Tue, 7 Mar 2006 20:23:12 -0500
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
Message-ID: <44565C8D-7549-48E4-BC80-28CED5E29894@leafe.com>

On Mar 7, 2006, at 8:05 PM, Charles Hartman wrote:

> How do I execute a Mac application from the Terminal command line?

	For most apps, simply type 'open <name of app>', including any  
necessary pathing. Keep in mind that Cocoa apps are actually bundles,  
which means that the file name usually has .app at the end. For  
example, to open Safari from the Terminal, type: open /Applications/ 
Safari.app/

> Specifically, I'm trying to specify BBEdit in the EDITOR environment
> variable which is consulted by IPython. EDITOR= what? Not /
> Applications/BBEdit, or /Applications/BBEdit.app. Probably something
> that continues with /Contents/ . . . but then I get lost. And I don't
> quite know how to experiment because I'm not sure how to attempt to
> run it from the bash prompt . . .

	This is even easier! BBEdit has a command-line tool that you can  
install from the 'Tools' section of the preferences. The tool's name  
is 'bbedit' (lower case), so from the Terminal you can simply type  
'bbedit myfile', and 'myfile' will be opened up in BBEdit. For your  
environmental variable, just set: EDITOR=bbedit

-- Ed Leafe
-- http://leafe.com
-- http://dabodev.com




From leknarf at pacbell.net  Wed Mar  8 02:53:27 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Tue, 7 Mar 2006 17:53:27 -0800
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <44565C8D-7549-48E4-BC80-28CED5E29894@leafe.com>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
	<44565C8D-7549-48E4-BC80-28CED5E29894@leafe.com>
Message-ID: <756E8520-9579-4B82-8846-96596C2BD0BE@pacbell.net>


It's also possible to launch the executable directly by typing this  
is a terminal window:

	/Applications/Safari.app/Contents/MacOS/Safari

For applications that want to talk back to you by printing info in  
the shell window, this can be very useful.  To avoid having to type &  
retype those long commands each time, you can save them in an  
"aliases" file using syntax like this:

	alias safari 'open -a Safari'
or
	alias safari '/Applications/Safari.app/Contents/MacOS/Safari'


Scott


On Mar 7, 2006, at 5:23 PM, Ed Leafe wrote:

> On Mar 7, 2006, at 8:05 PM, Charles Hartman wrote:
>
>> How do I execute a Mac application from the Terminal command line?
>
> 	For most apps, simply type 'open <name of app>', including any
> necessary pathing. Keep in mind that Cocoa apps are actually bundles,
> which means that the file name usually has .app at the end. For
> example, to open Safari from the Terminal, type: open /Applications/
> Safari.app/
>
>> Specifically, I'm trying to specify BBEdit in the EDITOR environment
>> variable which is consulted by IPython. EDITOR= what? Not /
>> Applications/BBEdit, or /Applications/BBEdit.app. Probably something
>> that continues with /Contents/ . . . but then I get lost. And I don't
>> quite know how to experiment because I'm not sure how to attempt to
>> run it from the bash prompt . . .
>
> 	This is even easier! BBEdit has a command-line tool that you can
> install from the 'Tools' section of the preferences. The tool's name
> is 'bbedit' (lower case), so from the Terminal you can simply type
> 'bbedit myfile', and 'myfile' will be opened up in BBEdit. For your
> environmental variable, just set: EDITOR=bbedit
>
> -- Ed Leafe
> -- http://leafe.com
> -- http://dabodev.com
>
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From wrw at ornl.gov  Wed Mar  8 02:17:03 2006
From: wrw at ornl.gov (W. R. Wing)
Date: Tue, 07 Mar 2006 20:17:03 -0500
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
Message-ID: <p0602040fc033e04e7eb7@[70.219.182.191]>

At 8:05 PM -0500 3/7/06, Charles Hartman wrote:
>(Sorry! How simple can they get? But I don't know of a better place 
>to ask -- does anyone?)
>
>How do I execute a Mac application from the Terminal command line?
>

Use the open command - as in:

$open /Applications/GoogleEarth.app

			Bill

-- 
   _
  /~\ The ASCII       | William R. Wing, PhD.  wrw at ornl.gov 865-574-8839
  \ / Ribbon Campaign | Network Architect, Oak Ridge National Laboratory
   X  Against HTML    | Network Research Group, Computer Sci & Math Div.
  / \ Email!          |_________

From daniel at brightfire.com  Wed Mar  8 03:05:58 2006
From: daniel at brightfire.com (Daniel Lord)
Date: Tue, 7 Mar 2006 18:05:58 -0800
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 17
In-Reply-To: <mailman.9821.1141767131.27774.pythonmac-sig@python.org>
References: <mailman.9821.1141767131.27774.pythonmac-sig@python.org>
Message-ID: <A73C8197-374B-4BEB-80A5-0FFDAC265428@brightfire.com>

I agree with the main sentiments here. In addition, it depends not  
just on the volume being reasonable (a subjective word most of us  
intuitively understand), but also upon the quality of the firm and  
legitimacy of the offer. I personally find it interesting to see  
Pixar using Python and more interesting to guess what for.
If not for the posting, I never would have known Pixar used Python.

Daniel

On Mar 7, 2006, at 1:32 PM, pythonmac-sig-request at python.org wrote:

> From: Ronald Oussoren <ronaldoussoren at mac.com>
> Date: March 7, 2006 1:32:00 PM PST
> To: Kent Quirk <kquirk at solidworks.com>
> Cc: pythonmac-sig at python.org, Jack Jansen <Jack.Jansen at cwi.nl>
> Subject: Re: [Pythonmac-SIG] Job postings - allow or not?
>
>
>
> On 7-mrt-2006, at 22:23, Kent Quirk wrote:
>
>> I think the occasional Python-related job posting is Just Fine.  
>> I'd be
>> bugged if there were too many of them, but I wouldn't kick them to
>> another list unless and until there were so many that they formed a
>> measurable fraction of the list traffic. At which point I'd  
>> rejoice at
>> Python's success.
>>
>
> I agree. Unless there'd be a lot of job postings I don't mind.
>
> Ronald
>
>> 	Kent
>>
>>
>> -----Original Message-----
>> From: pythonmac-sig-bounces+kquirk=solidworks.com at python.org
>> [mailto:pythonmac-sig-bounces+kquirk=solidworks.com at python.org] On
>> Behalf Of Jack Jansen
>> Sent: Tuesday, March 07, 2006 4:18 PM
>> To: pythonmac-sig at python.org
>> Subject: [Pythonmac-SIG] Job postings - allow or not?
>>
>> Folks,
>> I've just let a job posting through to the list, after looking at it
>> for a long time and unsure whether or not we want these.
>>
>> If you have a strong opinion (or a weak opinion, or anything
>> resembling an opinion:-) on the matter: please let me know, then I'll
>> use the collective feedback as a guidance in the future.
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060307/0e9d91d0/attachment-0001.htm 

From robert.kern at gmail.com  Wed Mar  8 03:10:46 2006
From: robert.kern at gmail.com (Robert Kern)
Date: Tue, 07 Mar 2006 20:10:46 -0600
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <p0602040fc033e04e7eb7@[70.219.182.191]>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
	<p0602040fc033e04e7eb7@[70.219.182.191]>
Message-ID: <440E3D26.90300@gmail.com>

W. R. Wing wrote:
> At 8:05 PM -0500 3/7/06, Charles Hartman wrote:
> 
>>(Sorry! How simple can they get? But I don't know of a better place 
>>to ask -- does anyone?)
>>
>>How do I execute a Mac application from the Terminal command line?
> 
> Use the open command - as in:
> 
> $open /Applications/GoogleEarth.app

In this particular case, IPython executes

  $EDITOR temp_file

and waits for the process to exit. open starts the application and exits
immediately, so IPython thinks the user has finished editing the file. As Ed
Leafe mentioned, BBEdit comes with a convenient command-line executable for just
this use. Since this is a common idiom for programs that start external editors,
this executable ought to have the desired behavior (possibly available through a
command line option). One should use this executable instead.

-- 
Robert Kern
robert.kern at gmail.com

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco

From ed at leafe.com  Wed Mar  8 04:05:03 2006
From: ed at leafe.com (Ed Leafe)
Date: Tue, 7 Mar 2006 22:05:03 -0500
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <440E3D26.90300@gmail.com>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
	<p0602040fc033e04e7eb7@[70.219.182.191]> <440E3D26.90300@gmail.com>
Message-ID: <73DBE288-B2D3-4DE9-89DF-D08EFEDFEF9A@leafe.com>

On Mar 7, 2006, at 9:10 PM, Robert Kern wrote:

> In this particular case, IPython executes
>
>   $EDITOR temp_file
>
> and waits for the process to exit. open starts the application and  
> exits
> immediately, so IPython thinks the user has finished editing the  
> file. As Ed
> Leafe mentioned, BBEdit comes with a convenient command-line  
> executable for just
> this use. Since this is a common idiom for programs that start  
> external editors,
> this executable ought to have the desired behavior (possibly  
> available through a
> command line option). One should use this executable instead.

	It does. I use BBEdit for my Subversion editor, so I have:

SVN_EDITOR='bbedit -w'

in my environmental settings. The '-w' option means 'wait for the  
file to close'. What's good about this is that you don't have to quit  
BBEdit; simply closing the file ends the 'wait'.

-- Ed Leafe
-- http://leafe.com
-- http://dabodev.com




From charles.hartman at conncoll.edu  Wed Mar  8 04:37:26 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 7 Mar 2006 22:37:26 -0500
Subject: [Pythonmac-SIG] Unix-amateur question
In-Reply-To: <73DBE288-B2D3-4DE9-89DF-D08EFEDFEF9A@leafe.com>
References: <567E4287-AD1C-4B35-AE2E-F53B73E4D547@conncoll.edu>
	<p0602040fc033e04e7eb7@[70.219.182.191]> <440E3D26.90300@gmail.com>
	<73DBE288-B2D3-4DE9-89DF-D08EFEDFEF9A@leafe.com>
Message-ID: <739C2DC5-A551-41BD-BB8A-34F05A157FDC@conncoll.edu>

Many thanks to all who have enlightened me on this! Little by  
little . . .

Charles


From smithsm at samuelsmith.org  Wed Mar  8 06:11:16 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Tue, 7 Mar 2006 22:11:16 -0700
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
Message-ID: <845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>

>
>> I use PYTHONPATH to add my own python development directories so that
>> modules I create myself or don't want to install in the default:
>> /library/frameworks/python.framework/versions/2.4/lib/python2.4/site-
>> packages/
>
> The time machine strikes again :-). You can also install your own
> packages in ~/Library/Python/2.4/site-packages. That directory is
> a proper site-package, .pth files work in there.
>
> [...]

Well I missed that one. Why doesn't ~/Library/Python/2.4/site- 
packages show up in sys.path?
In what order is it searched relative to sys.path?


**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From ronaldoussoren at mac.com  Wed Mar  8 07:51:19 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 8 Mar 2006 07:51:19 +0100
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
Message-ID: <EB045138-B924-408F-AFB3-31F17BC04057@mac.com>


On 8-mrt-2006, at 6:11, Samuel M. Smith wrote:

>>
>>> I use PYTHONPATH to add my own python development directories so  
>>> that
>>> modules I create myself or don't want to install in the default:
>>> /library/frameworks/python.framework/versions/2.4/lib/python2.4/ 
>>> site-
>>> packages/
>>
>> The time machine strikes again :-). You can also install your own
>> packages in ~/Library/Python/2.4/site-packages. That directory is
>> a proper site-package, .pth files work in there.
>>
>> [...]
>
> Well I missed that one. Why doesn't ~/Library/Python/2.4/site- 
> packages show up in sys.path?
> In what order is it searched relative to sys.path?

It should be just after the system-wide site packages directory. It  
is added in site.py
when 'Python.framework' is in sys.prefix, and the environment  
variable HOME is set.

Ronald

>
>
> **********************************************************************
> Samuel M. Smith Ph.D.
> 2966 Fort Hill Road
> Eagle Mountain, Utah 84043
> 801-768-2768 voice
> 801-768-2769 fax
> **********************************************************************
> "The greatest source of failure and unhappiness in the world is
> giving up what we want most for what we want at the moment"
> **********************************************************************
>


From clajo04 at mac.com  Wed Mar  8 13:32:53 2006
From: clajo04 at mac.com (John Clark)
Date: Wed, 08 Mar 2006 07:32:53 -0500
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <159E6601-0BAF-41B4-98B4-C78C654211CC@leafe.com>
Message-ID: <C0343925.93AE9%clajo04@mac.com>

I would also say that I would appreciate it if the postings allowed were
significantly Python oriented rather than ones that just say:

    ...and a scripting language such as Python or Perl

(Not that I am not intrigued at the idea of working at Pixar, but I don't
need another place to look for C/C++ programming jobs).

Perhaps this is asking for too much of Jack's time - but if we can't
moderate the job postings, I would vote for not having them on the list.
--

-jdc 

On 3/7/06 5:22 PM, "Ed Leafe" <ed at leafe.com> wrote:

> On Mar 7, 2006, at 4:32 PM, Ronald Oussoren wrote:
> 
>> I agree. Unless there'd be a lot of job postings I don't mind.
> 
> If they were real jobs, listed by real companies, then I don't mind.
> What I would object to would be recruiters and HR people using the
> list as a resum?-gathering device.
> 
> 
> -- Ed Leafe
> -- http://leafe.com Ed Leafe <ed at leafe.com>
> -- http://dabodev.com
> 
> 
> 
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig



From bob at redivi.com  Wed Mar  8 18:19:49 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 8 Mar 2006 09:19:49 -0800
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <C0343925.93AE9%clajo04@mac.com>
References: <C0343925.93AE9%clajo04@mac.com>
Message-ID: <3137BF0F-FD56-4715-B036-29B1269D49A2@redivi.com>

On Mar 8, 2006, at 4:32 AM, John Clark wrote:

> On 3/7/06 5:22 PM, "Ed Leafe" <ed at leafe.com> wrote:
>
>> On Mar 7, 2006, at 4:32 PM, Ronald Oussoren wrote:
>>
>>> I agree. Unless there'd be a lot of job postings I don't mind.
>>
>> If they were real jobs, listed by real companies, then I don't mind.
>> What I would object to would be recruiters and HR people using the
>> list as a resum?-gathering device.
>>
>
> I would also say that I would appreciate it if the postings allowed  
> were
> significantly Python oriented rather than ones that just say:
>
>     ...and a scripting language such as Python or Perl
>
> (Not that I am not intrigued at the idea of working at Pixar, but I  
> don't
> need another place to look for C/C++ programming jobs).
>
> Perhaps this is asking for too much of Jack's time - but if we can't
> moderate the job postings, I would vote for not having them on the  
> list.

Pixar does actually use Python (and TCL and probably Perl I guess).   
This particular posting also lives on the python.org job board, so  
it's passed that level of scrutiny.

Personally I'm far more annoyed by top posters than job postings...   
Especially if every other reply in the thread was bottom posted!

-bob


From bobsavage at mac.com  Wed Mar  8 20:05:55 2006
From: bobsavage at mac.com (Bob Savage)
Date: Wed, 8 Mar 2006 13:05:55 -0600
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <C0343925.93AE9%clajo04@mac.com>
References: <C0343925.93AE9%clajo04@mac.com>
Message-ID: <F4DF5B6D-9C5A-44B2-83E2-7D3F1F185C4A@mac.com>


On Mar 8, 2006, at 6:32 AM, John Clark wrote:

> I would also say that I would appreciate it if the postings allowed  
> were
> significantly Python oriented

Speaking as a person who is actively looking for work, I agree with  
just about everything said so far. I especially agree that the job  
should include Python skills, else it is not worth anyone's  
bandwidth. I also agree that a decision to let some posts that look  
appropriate through should not be taken as a binding requirement to  
always do so in the future.

best wishes to all,

Bob Savage

From goodger at python.org  Wed Mar  8 21:20:46 2006
From: goodger at python.org (David Goodger)
Date: Wed, 08 Mar 2006 15:20:46 -0500
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
References: <9D623F05-5F9B-4983-A736-88EAE792558C@cwi.nl>
Message-ID: <440F3C9E.3060500@python.org>

[Jack Jansen]
> I've just let a job posting through to the list, after looking at it  
> for a long time and unsure whether or not we want these.

The Python Job Board is made for this very purpose:
http://www.python.org/community/jobs/
New jobs are also picked up by Planet Python and its RSS feed
(the job board itself may have a feed, but I can't find it).

Perhaps redirect posters to the job board?

-- 
David Goodger <http://python.net/~goodger>

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 249 bytes
Desc: OpenPGP digital signature
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060308/39053b14/attachment.pgp 

From ronaldoussoren at mac.com  Wed Mar  8 22:43:26 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 8 Mar 2006 22:43:26 +0100
Subject: [Pythonmac-SIG] updated universal python installer (rc1?)
Message-ID: <FBDA89FE-AF2D-4BCE-9D99-AC56C1B49608@mac.com>

I've spent some time on an 10.3 box and updated the installer as a  
result of that.

The universal python installer should now be complete, except for  
remaining bugs obviously. I have however not tested it on an Intel  
box yet.

* The installer works on 10.3 and 10.4
* regrtest -uall passes on 10.3/ppc, 10.4/ppc
* compiling extensions works on 10.3
* the existing (non-universal) wxPython installer installs fine and  
results in a working wxPython installation (on PPC only of course) on  
10.3, which indicates that existing C++ extensions should work just  
fine with the universal build
* IDLE works, including the local documentation
* Profile updater works

New from the previous release:
* Locally installed documentation can be accessed from IDLE
* Minor changes to the profile updater, and it is now also installed  
in /Application/MacPython 2.4
* Building extensions works on 10.3
* Various minor bugfixes

There is one minor issue: the curses module doesn't work on OSX 10.3,  
the extension is linked to a version of libncurses that isn't  
available on 10.3.

Please test this version and let me know of the results (both  
negative and positive). Unless major issues turn up we can do a  
golden release this weekend. That release will NOT have new icons, we  
can do a new release with updated icons later on.

The installer is in the Public folder of my iDisk: http:// 
homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython% 
202.4.2.dmg

I'm switching to this version for my daily work, and will be working  
on a universal build for PyObjC and friends.

Ronald

From Jack.Jansen at cwi.nl  Wed Mar  8 23:54:32 2006
From: Jack.Jansen at cwi.nl (Jack Jansen)
Date: Wed, 8 Mar 2006 23:54:32 +0100
Subject: [Pythonmac-SIG] Job postings - allow or not?
In-Reply-To: <F4DF5B6D-9C5A-44B2-83E2-7D3F1F185C4A@mac.com>
References: <C0343925.93AE9%clajo04@mac.com>
	<F4DF5B6D-9C5A-44B2-83E2-7D3F1F185C4A@mac.com>
Message-ID: <D90EBCD8-BEDD-483C-B290-B329844BA682@cwi.nl>

The majority of the people seem to think the occasional job posting  
is fine, with this policy to be reviewed when we're sick and tired of  
the number of job postings.

But as we've so far had only one in the 10+ years this list exists I  
think we're pretty safe for the time being:-)
--
Jack Jansen, <Jack.Jansen at cwi.nl>, http://www.cwi.nl/~jack
If I can't dance I don't want to be part of your revolution -- Emma  
Goldman



From janssen at parc.com  Thu Mar  9 01:10:37 2006
From: janssen at parc.com (Bill Janssen)
Date: Wed, 8 Mar 2006 16:10:37 PST
Subject: [Pythonmac-SIG] updated universal python installer (rc1?)
In-Reply-To: Your message of "Wed, 08 Mar 2006 13:43:26 PST."
	<FBDA89FE-AF2D-4BCE-9D99-AC56C1B49608@mac.com> 
Message-ID: <06Mar8.161044pst."58633"@synergy1.parc.xerox.com>

Let me know when you're happy, and I'll update the downloads Web page.

Bill

From smithsm at samuelsmith.org  Thu Mar  9 01:28:25 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Wed, 8 Mar 2006 17:28:25 -0700
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
Message-ID: <3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>

>>>
>>> The time machine strikes again :-). You can also install your own
>>> packages in ~/Library/Python/2.4/site-packages. That directory is
>>> a proper site-package, .pth files work in there.
>>>
>>> [...]
>>
>> Well I missed that one. Why doesn't ~/Library/Python/2.4/site- 
>> packages show up in sys.path?
>> In what order is it searched relative to sys.path?
>
> It should be just after the system-wide site packages directory. It  
> is added in site.py
> when 'Python.framework' is in sys.prefix, and the environment  
> variable HOME is set.
>

Since my sys.path did not have /Users/smithsm/Library/Python/2.4/site- 
packages

I checked
  echo $HOME
/Users/smithsm

 >>sys.prefix
'/Library/Frameworks/Python.framework/Versions/2.4'

Then I checked ~/library/python and there was no  /2.4/site-packages
So I created /Users/smithsm/Library/Python/2.4/site-packages
  and then re ran python and this time it was in sys.path at the very  
end.

So does the new Python Installer create ~/Library/Python/2.4/site- 
packages? Or must one create it manually?


From bob at redivi.com  Thu Mar  9 01:33:00 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 8 Mar 2006 16:33:00 -0800
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
	<3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
Message-ID: <43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>


On Mar 8, 2006, at 4:28 PM, Samuel M. Smith wrote:

>>>>
>>>> The time machine strikes again :-). You can also install your own
>>>> packages in ~/Library/Python/2.4/site-packages. That directory is
>>>> a proper site-package, .pth files work in there.
>>>>
>>>> [...]
>>>
>>> Well I missed that one. Why doesn't ~/Library/Python/2.4/site-
>>> packages show up in sys.path?
>>> In what order is it searched relative to sys.path?
>>
>> It should be just after the system-wide site packages directory. It
>> is added in site.py
>> when 'Python.framework' is in sys.prefix, and the environment
>> variable HOME is set.
>>
>
> Since my sys.path did not have /Users/smithsm/Library/Python/2.4/site-
> packages
>
> I checked
>   echo $HOME
> /Users/smithsm
>
>>> sys.prefix
> '/Library/Frameworks/Python.framework/Versions/2.4'
>
> Then I checked ~/library/python and there was no  /2.4/site-packages
> So I created /Users/smithsm/Library/Python/2.4/site-packages
>   and then re ran python and this time it was in sys.path at the very
> end.
>
> So does the new Python Installer create ~/Library/Python/2.4/site-
> packages? Or must one create it manually?

One must create it manually.

-bob


From leehinde at gmail.com  Thu Mar  9 06:25:28 2006
From: leehinde at gmail.com (Lee Hinde)
Date: Wed, 8 Mar 2006 21:25:28 -0800
Subject: [Pythonmac-SIG] Postgresql
Message-ID: <d959f9b30603082125x3e6e8b33p37a1124d4c056f5e@mail.gmail.com>

Hi all;

I'm having a heck of a time getting Python to talk to Postgres. My
near-term goal is to play with Pylons, but I would like to get a
simple 'Insert "Hello World"' going first.

Before I regale you with my travails, I'd be grateful for pointers to
FAQs, tips, etc. I.e., I really want to do my homework first.

(OS X 10.4.x, fwiw.)

Thanks!

 - Lee

From ronaldoussoren at mac.com  Thu Mar  9 08:22:50 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 9 Mar 2006 08:22:50 +0100
Subject: [Pythonmac-SIG] updated universal python installer (rc1?)
In-Reply-To: <06Mar8.161044pst.58633@synergy1.parc.xerox.com>
References: <06Mar8.161044pst.58633@synergy1.parc.xerox.com>
Message-ID: <68A63B81-4F02-4782-B280-E7CBF3292033@mac.com>


On 9-mrt-2006, at 1:10, Bill Janssen wrote:

> Let me know when you're happy, and I'll update the downloads Web page.

Let me know if it works for you, and I'll know if I'm happy :-). I'm  
going to use this build for a while before declaring it finished.

The only thing that bothers me (a little) right now is the issue with  
the curses extension.

Ronald
>
> Bill


From leknarf at pacbell.net  Thu Mar  9 18:36:47 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Thu, 9 Mar 2006 09:36:47 -0800
Subject: [Pythonmac-SIG] Postgresql
In-Reply-To: <d959f9b30603082125x3e6e8b33p37a1124d4c056f5e@mail.gmail.com>
References: <d959f9b30603082125x3e6e8b33p37a1124d4c056f5e@mail.gmail.com>
Message-ID: <E90E2890-2EB9-4DD1-9AA1-B835CF6DA52B@pacbell.net>


I use psycopg to talk to postgres via python.  Was relatively  
painless to setup.  Take a look at

	http://initd.org/projects/psycopg1

and make sure to follow the link the latest release, psycopg2.

Good luck!
Scott




On Mar 8, 2006, at 9:25 PM, Lee Hinde wrote:

> Hi all;
>
> I'm having a heck of a time getting Python to talk to Postgres. My
> near-term goal is to play with Pylons, but I would like to get a
> simple 'Insert "Hello World"' going first.
>
> Before I regale you with my travails, I'd be grateful for pointers to
> FAQs, tips, etc. I.e., I really want to do my homework first.
>
> (OS X 10.4.x, fwiw.)
>
> Thanks!
>
>  - Lee
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Thu Mar  9 18:38:35 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 09 Mar 2006 09:38:35 -0800
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
	<3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
	<43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>
Message-ID: <4410681B.6010702@noaa.gov>

Bob Ippolito wrote:
> On Mar 8, 2006, at 4:28 PM, Samuel M. Smith wrote:
>>So does the new Python Installer create ~/Library/Python/2.4/site-
>>packages? Or must one create it manually?
> 
> One must create it manually.

I think it should create it, it will make things a tiny bit easier for 
some folks, and what's the harm?

-CHB


-- 
Christopher Barker, Ph.D.
Oceanographer

NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

From ronaldoussoren at mac.com  Thu Mar  9 20:31:14 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 9 Mar 2006 20:31:14 +0100
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <4410681B.6010702@noaa.gov>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
	<3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
	<43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>
	<4410681B.6010702@noaa.gov>
Message-ID: <A967A27A-2C1B-4E02-AE5B-B57A82D9EAB8@mac.com>


On 9-mrt-2006, at 18:38, Christopher Barker wrote:

> Bob Ippolito wrote:
>> On Mar 8, 2006, at 4:28 PM, Samuel M. Smith wrote:
>>> So does the new Python Installer create ~/Library/Python/2.4/site-
>>> packages? Or must one create it manually?
>>
>> One must create it manually.
>
> I think it should create it, it will make things a tiny bit easier for
> some folks, and what's the harm?

Nothing will automaticly install in ~/Library/Python/2.4/site-packages,
which means automaticly creating the directory is not very useful. I'm
sure I wouldn't notice the directory if it were created, and I'm sure
I wouldn't be the only one.

This doesn't mean we shouldn't improve the documentation :-).

Ronald
>
> -CHB
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From rowen at cesmail.net  Thu Mar  9 22:40:56 2006
From: rowen at cesmail.net (Russell E. Owen)
Date: Thu, 09 Mar 2006 13:40:56 -0800
Subject: [Pythonmac-SIG] Postgresql
References: <d959f9b30603082125x3e6e8b33p37a1124d4c056f5e@mail.gmail.com>
Message-ID: <rowen-C14A58.13405609032006@sea.gmane.org>

In article 
<d959f9b30603082125x3e6e8b33p37a1124d4c056f5e at mail.gmail.com>,
 "Lee Hinde" <leehinde at gmail.com> wrote:

> Hi all;
> 
> I'm having a heck of a time getting Python to talk to Postgres. My
> near-term goal is to play with Pylons, but I would like to get a
> simple 'Insert "Hello World"' going first.
> 
> Before I regale you with my travails, I'd be grateful for pointers to
> FAQs, tips, etc. I.e., I really want to do my homework first.

I use pypgsql (which I installed from source). It seems to work fine. 
All DB-API-2 clients work about the same for the basics, but the connect 
method varies.

-- Russell

P.S. you might find the routines in RO.SQLUtil handy (for your own use 
or as an example). See 
<http://www.astro.washington.edu/rowen/ROPython.html>.


From smithsm at samuelsmith.org  Thu Mar  9 22:49:27 2006
From: smithsm at samuelsmith.org (Samuel M. Smith)
Date: Thu, 9 Mar 2006 14:49:27 -0700
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <A967A27A-2C1B-4E02-AE5B-B57A82D9EAB8@mac.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
	<3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
	<43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>
	<4410681B.6010702@noaa.gov>
	<A967A27A-2C1B-4E02-AE5B-B57A82D9EAB8@mac.com>
Message-ID: <2EBF1FE7-D757-466E-98CE-7A9B8D957D7B@samuelsmith.org>

I would guess that the PATH, PYTHONPATH, pth, and site-packages  
issues are close to the number one
FAQ for mac python newbies (and not really fully understood by many  
oldies like me)

I would be happy to take a shot at writing up some documentation  
probably should go under
setting up your user environment.

I remember that there are some obscure details associated with the   
pythonpath for cgi scripts in apache

The python library ref for cgi states (or stated some time ago).
"In particular, don't count on ...the Python module search path  
(PYTHONPATH) to be set to anything interesting.

One way to fix it is to use the setEnv directive in apache

<IfModule mod_env.c>
       SetEnv PYTHONPATH "/Users/smithsm/Development/python:/Users/ 
smithsm/Development/python/ht:/sw/lib/python2.2:/sw/lib/python2.2/ 
plat-darwin:/sw/lib/python2.2/lib-tk:/sw/lib/python2.2/lib-dynload:/ 
sw/lib/python2.2/site-packages:/sw/lib/python2.2/site-packages/ 
Numeric:/sw/lib/python2.2/site-packages/PIL"
</IfModule>
#quotes optional

The final path that python sees as sys.path in the cgi script is  
composed of a list in order of

the directory wherein resides the cgi script
the apache setenv PYTHONPATH
the default sys.path

Another suggestion was to put a .pth file in the cgi-bin directory  
holding the python cgi script. I never tested that.

Any changes to this behavior in 2.4.2 or with mod_python?



On 09 Mar, 2006, at 12:31, Ronald Oussoren wrote:

>
> On 9-mrt-2006, at 18:38, Christopher Barker wrote:
>
>> Bob Ippolito wrote:
>>> On Mar 8, 2006, at 4:28 PM, Samuel M. Smith wrote:
>>>> So does the new Python Installer create ~/Library/Python/2.4/site-
>>>> packages? Or must one create it manually?
>>>
>>> One must create it manually.
>>
>> I think it should create it, it will make things a tiny bit easier  
>> for
>> some folks, and what's the harm?
>
> Nothing will automaticly install in ~/Library/Python/2.4/site- 
> packages,
> which means automaticly creating the directory is not very useful. I'm
> sure I wouldn't notice the directory if it were created, and I'm sure
> I wouldn't be the only one.
>
> This doesn't mean we shouldn't improve the documentation :-).
>
> Ronald
>>
>> -CHB
>>
>>
>> -- 
>> Christopher Barker, Ph.D.
>> Oceanographer
>>
>> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
>> 7600 Sand Point Way NE   (206) 526-6329   fax
>> Seattle, WA  98115       (206) 526-6317   main reception
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig

**********************************************************************
Samuel M. Smith Ph.D.
2966 Fort Hill Road
Eagle Mountain, Utah 84043
801-768-2768 voice
801-768-2769 fax
**********************************************************************
"The greatest source of failure and unhappiness in the world is
giving up what we want most for what we want at the moment"
**********************************************************************


From Chris.Barker at noaa.gov  Thu Mar  9 23:26:02 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 09 Mar 2006 14:26:02 -0800
Subject: [Pythonmac-SIG] simple help?
In-Reply-To: <A967A27A-2C1B-4E02-AE5B-B57A82D9EAB8@mac.com>
References: <D4239254-E30C-4225-B91C-15134D9A38F7@conncoll.edu>
	<440D3898.4070104@noaa.gov>
	<6A86A70D-BF35-4B01-BC16-15BF0FFD9913@conncoll.edu>
	<86CC5BA9-8959-4F45-8472-23132DDD39AD@mac.com>
	<016544B1-2C85-48EB-BDE5-E9755378C87F@samuelsmith.org>
	<8B379EC2-4D42-4F32-B06A-06262188F80E@mac.com>
	<845FE05B-575F-437E-A513-4D1524B483E1@samuelsmith.org>
	<EB045138-B924-408F-AFB3-31F17BC04057@mac.com>
	<3FC0755F-9949-42F9-A442-9BE3EDC874D8@samuelsmith.org>
	<43E75E5E-DCED-45F1-8D66-AB859F19C0A8@redivi.com>
	<4410681B.6010702@noaa.gov>
	<A967A27A-2C1B-4E02-AE5B-B57A82D9EAB8@mac.com>
Message-ID: <4410AB7A.6060209@noaa.gov>


Ronald Oussoren wrote:
> Nothing will automaticly install in ~/Library/Python/2.4/site-packages,
> which means automaticly creating the directory is not very useful. I'm
> sure I wouldn't notice the directory if it were created, and I'm sure
> I wouldn't be the only one.
> 
> This doesn't mean we shouldn't improve the documentation :-).

My thought is that if we have docs that say something like:

If you have modules that you want accessible by a particular user, 
rather than the whole system, put them in:

Users/TheUserName/Library/Python/2.4/site-packages


It would be nice if that directory existed.

On the other hand, when the installer runs, do we create one only in the 
user's directory that's installing, or all users directories, or???

Maybe it's just as easy to just tell people to create it if they need it.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From hengist.podd at virgin.net  Fri Mar 10 00:12:07 2006
From: hengist.podd at virgin.net (has)
Date: Thu, 9 Mar 2006 23:12:07 +0000
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
Message-ID: <a05200f10c036649a652f@[82.12.228.35]>

Hey,

I notice the MacPython wiki page on python.org <http://wiki.python.org/moin/MacPython> is showing outdated information - might be an idea to redirect it to the new download page or copy over some of the new content.

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From bear42 at code-bear.com  Fri Mar 10 05:33:42 2006
From: bear42 at code-bear.com (bear)
Date: Thu, 09 Mar 2006 23:33:42 -0500
Subject: [Pythonmac-SIG] updated universal python installer (rc1?)
In-Reply-To: <FBDA89FE-AF2D-4BCE-9D99-AC56C1B49608@mac.com>
References: <FBDA89FE-AF2D-4BCE-9D99-AC56C1B49608@mac.com>
Message-ID: <441101A6.7000901@code-bear.com>

I installed the dmg you linked to on an Intel iMac just now - it went 
smoothly.

I ran IDLE and a bunch of the included samples.  I also ran some small 
utilities of my own and did not see any issues.

Anything particular you want me to try?



Ronald Oussoren wrote:
> I've spent some time on an 10.3 box and updated the installer as a  
> result of that.
>
> The universal python installer should now be complete, except for  
> remaining bugs obviously. I have however not tested it on an Intel  
> box yet.
>
> * The installer works on 10.3 and 10.4
> * regrtest -uall passes on 10.3/ppc, 10.4/ppc
> * compiling extensions works on 10.3
> * the existing (non-universal) wxPython installer installs fine and  
> results in a working wxPython installation (on PPC only of course) on  
> 10.3, which indicates that existing C++ extensions should work just  
> fine with the universal build
> * IDLE works, including the local documentation
> * Profile updater works
>
> New from the previous release:
> * Locally installed documentation can be accessed from IDLE
> * Minor changes to the profile updater, and it is now also installed  
> in /Application/MacPython 2.4
> * Building extensions works on 10.3
> * Various minor bugfixes
>
> There is one minor issue: the curses module doesn't work on OSX 10.3,  
> the extension is linked to a version of libncurses that isn't  
> available on 10.3.
>
> Please test this version and let me know of the results (both  
> negative and positive). Unless major issues turn up we can do a  
> golden release this weekend. That release will NOT have new icons, we  
> can do a new release with updated icons later on.
>
> The installer is in the Public folder of my iDisk: http:// 
> homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython% 
> 202.4.2.dmg
>
> I'm switching to this version for my daily work, and will be working  
> on a universal build for PyObjC and friends.
>
> Ronald
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>   

From rogerb at rogerbinns.com  Fri Mar 10 08:31:24 2006
From: rogerb at rogerbinns.com (Roger Binns)
Date: Thu, 9 Mar 2006 23:31:24 -0800
Subject: [Pythonmac-SIG] Python Ref Docs in OSX help format
References: <823CDC42-331B-4ADF-ADC5-44CF46145C93@samuelsmith.org><3518315.1141662206533.JavaMail.ronaldoussoren@mac.com><8A7BB37E-FD4C-435B-BF52-739277E27828@conncoll.edu><8BB51A9B-AA03-455C-AD15-6D8E40C1D42B@mac.com><37E24660-1201-4B1C-A806-86D32164B16D@samuelsmith.org>
	<6FC07863-6D6E-47A0-8D0A-EDDF855B230A@mac.com>
Message-ID: <001601c64414$a2df3690$3501a8c0@rogersqyvr14d3>

> The hard part is probably to get IDLE to use the help book, although  
> I haven't looked into this yet.

If you ever get his working, please let us know.  I have been able
to associate a help book with a wxPython application using the
APIs in Carbon.AH and everything looked like it worked (no API
errors, help shown).  However typing into the search box
never gave any results.  I had built the search index using the
gui tool from Apple.  There is probably some magic you have to
get right and I'm hoping you'll find it :-)

Roger

From altern2 at gmail.com  Fri Mar 10 11:31:52 2006
From: altern2 at gmail.com (altern)
Date: Fri, 10 Mar 2006 11:31:52 +0100
Subject: [Pythonmac-SIG] indentation problem
Message-ID: <44115598.9000700@gmail.com>

hi all

i have been developing some tutorials with PyOxide and now i find that 
when i open them on Smultron the indentation is wrong. Students are vey 
likely to be using Smultron (no need to insatall pyObjC).

It tried to fix the indentation in one but there is something wrong 
under the hood because then it complains at end of lines, it says the 
syntax is wrong, and i have to remove the line end, hit return again and 
save. This for each line in blocks that had indentation wrong.

I was wondering if there is some way to automatically fix this or if i 
have to go file by file fixing the indentation and line breaks.

The funny thing is that when i open them in Kate under debian they look ok.

thanks!

--
enrike

From lists at mostrom.pp.se  Fri Mar 10 14:11:26 2006
From: lists at mostrom.pp.se (=?UTF-8?Q?Jan_Erik_Mostr=C3=B6?= =?UTF-8?Q?m?=)
Date: Fri, 10 Mar 2006 14:11:26 +0100
Subject: [Pythonmac-SIG] indentation problem
In-Reply-To: <44115598.9000700@gmail.com>
Message-ID: <A0202000f-1045-F612E7B819D14D9BAB3C2256B5D174CB@klient105.wireless.umu.se>

altern <altern2 at gmail.com> 2006-03-10 11:31:

> It tried to fix the indentation in one but there is something wrong 
> under the hood because then it complains at end of lines, it says the 
> syntax is wrong, and i have to remove the line end, hit return again and 
> save. This for each line in blocks that had indentation wrong.

Without knowing anything about Smultron or PyOxide I would say that one of
them is saving documents using Windows new line conventions and the other
Unix (or Mac/Unix).

You can easily write a Python script that fixes this and I think there are
number of small apps on versiontracker that handles stuff like this (or
perhaps Smulton/PyOxide can be scripted to do it)

                    jem
-- 
Jan Erik Mostr?m, www.mostrom.pp.se

From gandreas at gandreas.com  Fri Mar 10 16:28:06 2006
From: gandreas at gandreas.com (gandreas at gandreas.com)
Date: Fri, 10 Mar 2006 09:28:06 -0600
Subject: [Pythonmac-SIG] indentation problem
In-Reply-To: <44115598.9000700@gmail.com>
References: <44115598.9000700@gmail.com>
Message-ID: <A5F37B8C-6390-4256-BE4F-F6DB09853154@gandreas.com>


On Mar 10, 2006, at 4:31 AM, altern wrote:

> hi all
>
> i have been developing some tutorials with PyOxide and now i find that
> when i open them on Smultron the indentation is wrong. Students are  
> vey
> likely to be using Smultron (no need to insatall pyObjC).
>


I would strongly recommend switching from PyOXIDE to ScrIDE <http:// 
projects.gandreas.com/scride>:

1) ScrIDE doesn't require PyObjC to be installed
2) ScrIDE does pretty much everything that PyOXIDE did (including  
full source level debugging)
3) ScrIDE has features that PyOXIDE doesn't
4) ScrIDE is more stable
5) ScrIDE is easier to extend with custom scripts (which can do more  
things)
6) ScrIDE has a much brighter future (bugs in ScrIDE will be fixed,  
features added, while PyOXIDE is pretty much "end of life")



Glenn Andreas                      gandreas at gandreas.com
  <http://www.gandreas.com/> wicked fun!
Mad, Bad, and Dangerous to Know


From altern2 at gmail.com  Fri Mar 10 16:52:34 2006
From: altern2 at gmail.com (altern)
Date: Fri, 10 Mar 2006 16:52:34 +0100
Subject: [Pythonmac-SIG] indentation problem
In-Reply-To: <A0202000f-1045-F612E7B819D14D9BAB3C2256B5D174CB@klient105.wireless.umu.se>
References: <A0202000f-1045-F612E7B819D14D9BAB3C2256B5D174CB@klient105.wireless.umu.se>
Message-ID: <4411A0C2.5050902@gmail.com>

Jan Erik Mostr?m(e)k dio:
> altern <altern2 at gmail.com> 2006-03-10 11:31:
> 
>> It tried to fix the indentation in one but there is something wrong 
>> under the hood because then it complains at end of lines, it says the 
>> syntax is wrong, and i have to remove the line end, hit return again and 
>> save. This for each line in blocks that had indentation wrong.
> 
> Without knowing anything about Smultron or PyOxide I would say that one of
> them is saving documents using Windows new line conventions and the other
> Unix (or Mac/Unix).
> 
> You can easily write a Python script that fixes this and I think there are
> number of small apps on versiontracker that handles stuff like this (or
> perhaps Smulton/PyOxide can be scripted to do it)

hi jan

After reading your email i have been checking the preferences of both 
Smultron and PyOxide trying to find some configuration option.
I also read some old posts on the list about tabs vs spaces regarding 
python indentation problems.

Basically looks like the problem lies on that PyOxide does the 
indentation with tabs by default. I changed this to spaces 
(Preferences/editing/tabs&font), saved the files and now they get opened 
ok in Smultron.

but now it is working, i just have to save all files again.

thanks for your tips!


From altern2 at gmail.com  Fri Mar 10 17:25:34 2006
From: altern2 at gmail.com (altern)
Date: Fri, 10 Mar 2006 17:25:34 +0100
Subject: [Pythonmac-SIG] indentation problem
In-Reply-To: <A5F37B8C-6390-4256-BE4F-F6DB09853154@gandreas.com>
References: <44115598.9000700@gmail.com>
	<A5F37B8C-6390-4256-BE4F-F6DB09853154@gandreas.com>
Message-ID: <4411A87E.4000101@gmail.com>

gandreas at gandreas.com(e)k dio:
> 
> On Mar 10, 2006, at 4:31 AM, altern wrote:
> 
>> hi all
>>
>> i have been developing some tutorials with PyOxide and now i find that
>> when i open them on Smultron the indentation is wrong. Students are vey
>> likely to be using Smultron (no need to insatall pyObjC).
>>
> 
> 
> I would strongly recommend switching from PyOXIDE to ScrIDE 
> <http://projects.gandreas.com/scride>:
> 
> 1) ScrIDE doesn't require PyObjC to be installed
> 2) ScrIDE does pretty much everything that PyOXIDE did (including full 
> source level debugging)
> 3) ScrIDE has features that PyOXIDE doesn't
> 4) ScrIDE is more stable
> 5) ScrIDE is easier to extend with custom scripts (which can do more 
> things)
> 6) ScrIDE has a much brighter future (bugs in ScrIDE will be fixed, 
> features added, while PyOXIDE is pretty much "end of life")

thanks for pointing this Glenn. i didnt know about this app yet. Looks 
nice. Good to know that you keep developing the editor further.


> Glenn Andreas                      gandreas at gandreas.com
>  <http://www.gandreas.com/> wicked fun!
> Mad, Bad, and Dangerous to Know
> 
> 


From janssen at parc.com  Fri Mar 10 18:45:54 2006
From: janssen at parc.com (Bill Janssen)
Date: Fri, 10 Mar 2006 09:45:54 PST
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: Your message of "Thu, 09 Mar 2006 15:12:07 PST."
	<a05200f10c036649a652f@[82.12.228.35]> 
Message-ID: <06Mar10.094556pst."58633"@synergy1.parc.xerox.com>

> Hey,
> 
> I notice the MacPython wiki page on python.org <http://wiki.python.org/moin/MacPython> is showing outdated information - might be an idea to redirect it to the new download page or copy over some of the new content.
> 
> has

The nifty thing about Wikis is that anyone can edit it...  (hint, hint).

Bill

From stewart.midwinter at gmail.com  Fri Mar 10 19:17:42 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Fri, 10 Mar 2006 11:17:42 -0700
Subject: [Pythonmac-SIG] indentation problem
Message-ID: <d4c0d4800603101017mb94a7dfm767daf057e964d3d@mail.gmail.com>

> From: altern <altern2 at gmail.com>
> Subject: [Pythonmac-SIG] indentation problem
> hi all
>
> i have been developing some tutorials with PyOxide and now i find that
> when i open them on Smultron the indentation is wrong.

I use Smultron as well.  What do you mean, "the indentation is wrong"?
 Too much indentation, not enough?

> It tried to fix the indentation in one but there is something wrong
> under the hood because then it complains at end of lines, it says the
> syntax is wrong,

What complains, Smultron?  What's the exact error message?

> I was wondering if there is some way to automatically fix this or if i
> have to go file by file fixing the indentation and line breaks.

The most common source of problems with indentation is mixed use of
tabs and space.  You should do one or other, preferably spaces.

In Preferences > Advanced, check "Indent with spaces not tabs"
In View, select Show Invisible Characters.

This should help you track down the problem

If you do have to change the indentation of a file, you don't need to
edit lines individually. Select all lines in a block, then use
Option-L to indent by 4 chars, or Option-K to outdent one char at a
time.

> The funny thing is that when i open them in Kate under debian they look ok.

Kate is not a programming editor, so that proves nothing.

cheers
Stewart

From altern2 at gmail.com  Fri Mar 10 19:31:37 2006
From: altern2 at gmail.com (altern)
Date: Fri, 10 Mar 2006 19:31:37 +0100
Subject: [Pythonmac-SIG] indentation problem
In-Reply-To: <d4c0d4800603101017mb94a7dfm767daf057e964d3d@mail.gmail.com>
References: <d4c0d4800603101017mb94a7dfm767daf057e964d3d@mail.gmail.com>
Message-ID: <4411C609.20709@gmail.com>

Stewart Midwinter(e)k dio:
>> From: altern <altern2 at gmail.com>
>> Subject: [Pythonmac-SIG] indentation problem
>> hi all
>>
>> i have been developing some tutorials with PyOxide and now i find that
>> when i open them on Smultron the indentation is wrong.
> 
> I use Smultron as well.  What do you mean, "the indentation is wrong"?
>  Too much indentation, not enough?
> 
>> It tried to fix the indentation in one but there is something wrong
>> under the hood because then it complains at end of lines, it says the
>> syntax is wrong,
> 
> What complains, Smultron?  What's the exact error message?

sorry. the exact problem was that something like this in pyoxide

def a():
	if x:
		pass

would look like this in smultron

def a():
	if x:
	pass

if i indented the pass line, then when running the code i got an error 
saying there was wrong syntax at the line return after "if x:"

But i solved this by changing the default option to save tabs as spaces 
in PyOxide, now Smultron opens them ok.

thanks for the tips

>> I was wondering if there is some way to automatically fix this or if i
>> have to go file by file fixing the indentation and line breaks.
> 
> The most common source of problems with indentation is mixed use of
> tabs and space.  You should do one or other, preferably spaces.
> 
> In Preferences > Advanced, check "Indent with spaces not tabs"
> In View, select Show Invisible Characters.
> 
> This should help you track down the problem
> 
> If you do have to change the indentation of a file, you don't need to
> edit lines individually. Select all lines in a block, then use
> Option-L to indent by 4 chars, or Option-K to outdent one char at a
> time.
> 
>> The funny thing is that when i open them in Kate under debian they look ok.
> 
> Kate is not a programming editor, so that proves nothing.
> 
> cheers
> Stewart
> 


From hengist.podd at virgin.net  Fri Mar 10 20:14:43 2006
From: hengist.podd at virgin.net (has)
Date: Fri, 10 Mar 2006 19:14:43 +0000
Subject: [Pythonmac-SIG] Download page on www.python.org now updated
In-Reply-To: <06Mar10.094556pst."58633"@synergy1.parc.xerox.com>
References: <06Mar10.094556pst."58633"@synergy1.parc.xerox.com>
Message-ID: <a05200f01c0377ee8ab21@[82.2.90.111]>

Hey,

> > I notice the MacPython wiki page on python.org <http://wiki.python.org/moin/MacPython> is showing outdated information - might be an idea to redirect it to the new download page or copy over some of the new content.
> >
>> has
>
>The nifty thing about Wikis is that anyone can edit it...  (hint, hint).

I know, I know. Just preoccupied with other things right now; hence the heads-up in case anyone feels like sorting it sooner. (People grow old and die in less time than it takes me to get around to doing stuff.:p)

Cheers,

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From zpincus at stanford.edu  Fri Mar 10 23:25:44 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Fri, 10 Mar 2006 14:25:44 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
	<88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
Message-ID: <E592E4EB-8329-4ACF-AB4F-2B5C07367CFC@stanford.edu>

> If you come up with a patch for (2) against Python's svn trunk that  
> is tested and works on 10.2 then I'll commit it.  It's not  
> appropriate to make this sort of change against Python 2.4,  
> though.  *Maybe* for the universal branch that Ronald and I are  
> doing, because that's also unlikely to be merged with the 2.4 branch.

Well, after wasting the better part of a day tracking down a computer  
capable of running 10.2, installing 10.2 on it, getting the latest  
updates and the newest 10.2 developer tools, I find that neither  
python 2.5 from svn (today's snapshot or a checkout from the 8th)  
*NOR* the svn snapshot of python 2.4 will build on 10.2.8.

Is python 2.5 (or even 2.4) supposed to be officially supported for  
10.2? (It certainly doesn't seem to be based on what's on the python  
web pages.) If not, I rather resent the goose chase I've been sent on  
just trying to submit a patch to get python to work the way the  
documentation says it does.

Anyhow, I was able to get python on 10.2 configure properly, which is  
all that the patch changes. So if python 2.5 ever were to work on  
10.2, my patch wouldn't prevent that.

Here's the patch. This includes a change to setup.py to build the dl  
module if dlfcn.h is available (which it is not on 10.2, so no  
incompatibility worries there either). I think this latter change is  
important because the 'standard idiom' for setting dlopenflags looks  
like this:
import dl
sys.setdlopenflags(dl.RTLD_GLOBAL|dl.RTLD_NOW)
(as noted in the standard docs: http://docs.python.org/lib/module- 
sys.html )

If sys.setdlopenflags() is going to be respected, dl should be  
available to allow the flags to be looked up. (I know the ctypes  
module provides some of these flags, but given that the idiom on  
other platforms has been to use the dl module, not including dl when  
it could be seems like arbitrary breakage.)

Zach


Patch follows:

Index: configure
===================================================================
--- configure   (revision 42924)
+++ configure   (working copy)
@@ -1,5 +1,5 @@
#! /bin/sh
-# From configure.in Revision: 42437 .
+# From configure.in Revision: 42563 .
# Guess values for system-dependent variables and create Makefiles.
# Generated by GNU Autoconf 2.59 for python 2.5.
#
@@ -13974,7 +13974,8 @@
         ;;
         BeOS*) DYNLOADFILE="dynload_beos.o";;
         hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
-       Darwin/*) DYNLOADFILE="dynload_next.o";;
+       # Use dynload_next.c only on 10.2 and below, which don't have  
native dlopen()
+       Darwin/[0123456].*) DYNLOADFILE="dynload_next.o";;
         atheos*) DYNLOADFILE="dynload_atheos.o";;
         *)
         # use dynload_shlib.c and dlopen() if we have it; otherwise  
stub
Index: configure.in
===================================================================
--- configure.in        (revision 42924)
+++ configure.in        (working copy)
@@ -2105,7 +2105,8 @@
         ;;
         BeOS*) DYNLOADFILE="dynload_beos.o";;
         hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
-       Darwin/*) DYNLOADFILE="dynload_next.o";;
+       # Use dynload_next.c only on 10.2 and below, which don't have  
native dlopen()
+       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;
         atheos*) DYNLOADFILE="dynload_atheos.o";;
         *)
         # use dynload_shlib.c and dlopen() if we have it; otherwise  
stub
Index: setup.py
===================================================================
--- setup.py    (revision 42924)
+++ setup.py    (working copy)
@@ -885,7 +885,7 @@
          if sys.maxint == 0x7fffffff:
              # This requires sizeof(int) == sizeof(long) == sizeof 
(char*)
              dl_inc = find_file('dlfcn.h', [], inc_dirs)
-            if (dl_inc is not None) and (platform not in ['atheos',  
'darwin']):
+            if (dl_inc is not None) and (platform not in ['atheos']):
                  exts.append( Extension('dl', ['dlmodule.c']) )
          # Thomas Heller's _ctypes module



From zbir at urbanape.com  Sat Mar 11 04:22:09 2006
From: zbir at urbanape.com (Zachery Bir)
Date: Fri, 10 Mar 2006 22:22:09 -0500
Subject: [Pythonmac-SIG] NSUserDefaults causes crash on MOSXI
In-Reply-To: <2C9DD974-6B77-4EBD-B267-EB7E026F21A7@redivi.com>
References: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>
	<60AF98A3-3EB5-4905-9F4A-DF813C5C416A@urbanape.com>
	<2C9DD974-6B77-4EBD-B267-EB7E026F21A7@redivi.com>
Message-ID: <78C2A3B9-95C4-42D7-AE63-AE7E54EC2971@urbanape.com>

[background: So, I tried this again with Ronald's latest Universal  
build of 2.4.2 and PyObjC from the head of SVN trunk (not using Ben's  
work). Same crash during NSUerDefaults.standardUserDefaults()]

On Mar 1, 2006, at 1:22 PM, Bob Ippolito wrote:

> The first thing I would do is take gdb and figure out what file  
> it's looking at... that's the second argument to  
> CFURLCreateDataAndPropertiesFromResource.  I'm guessing it would  
> live on the stack, but I haven't used gdb on intel enough lately to  
> remember what the offset is.

Not in any way I can make sense of :^\ Here follows a mini  
transcript, complete with help text running on my MacBook Pro:

(gdb) bt
#0  0x90823daa in _CFReadBytesFromFile ()
#1  0x90823793 in CFURLCreateDataAndPropertiesFromResource ()
#2  0x9082d67a in _loadXMLDomainIfStale ()
.
.
.
(gdb) frame 0
#0  0x90823daa in _CFReadBytesFromFile ()
(gdb) info frame
Stack level 0, frame at 0xbfffd1f8:
eip = 0x90823daa in _CFReadBytesFromFile; saved eip 0x90823793
called by frame at 0xbfffd2c8
Arglist at 0xbfffd1f0, args:
Locals at 0xbfffd1f0, Previous frame's sp is 0xbfffd1f8
Saved registers:
   ebx at 0xbfffd1e4, ebp at 0xbfffd1f0, esi at 0xbfffd1e8, edi at  
0xbfffd1ec, eip at 0xbfffd1f4
(gdb) frame 1
#1  0x90823793 in CFURLCreateDataAndPropertiesFromResource ()
(gdb) info frame
Stack level 1, frame at 0xbfffd2c8:
eip = 0x90823793 in CFURLCreateDataAndPropertiesFromResource; saved  
eip 0x9082d67a
called by frame at 0xbfffd338, caller of frame at 0xbfffd1f8
Arglist at 0xbfffd2c0, args:
Locals at 0xbfffd2c0, Previous frame's sp is 0xbfffd2c8
Saved registers:
   ebx at 0xbfffd2b4, ebp at 0xbfffd2c0, esi at 0xbfffd2b8, edi at  
0xbfffd2bc, eip at 0xbfffd2c4
(gdb) frame 2
#2  0x9082d67a in _loadXMLDomainIfStale ()
(gdb) info frame
Stack level 2, frame at 0xbfffd338:
eip = 0x9082d67a in _loadXMLDomainIfStale; saved eip 0x90839978
called by frame at 0xbfffd358, caller of frame at 0xbfffd2c8
Arglist at 0xbfffd330, args:
Locals at 0xbfffd330, Previous frame's sp is 0xbfffd338
Saved registers:
   ebx at 0xbfffd324, ebp at 0xbfffd330, esi at 0xbfffd328, edi at  
0xbfffd32c, eip at 0xbfffd334

You can see that in none of the frames are any arguments present. I'm  
a complete n00b when it comes to gdb in general and, if as you  
suspect it performs significantly differently on intel, well, that's  
just a whole other sport :^)

On a lark (uneducated, but trying to poke around), I e(x)amined some  
of the memory registers listed - specifically the arglist address:

(gdb) x 0xbfffd2c0
0xbfffd2c0:	0xbfffd330
(gdb) x 0xbfffd330
0xbfffd330:	0xbfffd350
(gdb) x 0xbfffd350
0xbfffd350:	0xbfffd370
(gdb) x 0xbfffd370
0xbfffd370:	0xbfffd3e0

I'm a fish hopelessly out of water here, but I'm willing to poke  
around some more with some guidance.

> Alternatively you could run it under ktrace to see which file it  
> opens.  There should be a NAMI with the path it's opening.

Boy does that make some kinda output ;^) I'm more than willing to  
post this to my site if it'll be helpful.

Zac


From spe.stani.be at gmail.com  Sat Mar 11 06:13:43 2006
From: spe.stani.be at gmail.com (SPE Stani's Python Editor)
Date: Sat, 11 Mar 2006 06:13:43 +0100
Subject: [Pythonmac-SIG] ANN: The Python multiple IDE Collaboration calls
	for coders
Message-ID: <2078a7ad0603102113nff2d8a3hfa8bfe0ee3d4fde3@mail.gmail.com>

(cross posting from comp.lang.python)

Dear Pythoneers,

Looking at IDE's I can have three observations:

1. For some reasons numerous users prefer to use an open source IDE.

2. For some reasons numerous python programmers like to develop an
open source IDE.

3. For some reasons the open source python IDE developers are not
collaborating at all.

The reasons for 1 or 2 are obvious, at least to me. Recently I have
been wondering about the reason for 3. (Probably a lot of python
programmers have wondered about this already for ages, but ok I might
be slow ;-) I came to the conclusion that there was *NO* reason.

As this was so clear, I started to invite all the authors of IDE's
personally to collaborate all together. I hope that I didn't forget
any, because there are so many. What is really nice, is that we feel
the same: we should work together and share as much as possible. We
don't want to waste our (often spare) time on reinventing wheels.
Almost all IDE's (except of two) are participating no matter if they
use Tkinter, wxPython, pyQT, Cocoa, pyGTK, ... (So this could open
doors for an ajax python editor, who knows. Any python web framework
like Django, Turbogears, ... interested in that?)

These projects are participating: NewEdit, scrIDE, Eric3, Leo IDE,
ActiveGrid, PIDA, drPython, pyDev, PyCrust, IPython, WinPdb debugger,
Extended Python Debugger, PyLint, Gaphor, Envisage, Dabo, SilverCity &
SPE.

It is not about unification, but about a little bit more
collaboration. There are always libraries to share, more as we might
think.

In order to give the project shape I started building a (wiki based)
website in plone which together with a mailing list should give a good
platform for collaboration.  (You need to login to edit wiki's.) All
the developers are already invited, but everyone willing to code or
contribute (documentation, translation, artwork, plone website, ..) is
welcome.  If you work on open source project which might be of
interest (parsing, uml, framework, ...) please join or invite the
projects which you think should participate as well.

We will probably work in smaller teams on the various aspects of IDE's
and tools. If this project succeeds it could be a major win for the
Python community.

These are some useful links:

- homepage: http://pyxides.stani.be
- mail list: http://pyxides.stani.be/polls/mailing
- starting mail: http://pyxides.stani.be/wiki/StartingEmail
- developers reaction: http://pyxides.stani.be/wiki/AuthorsOfIDEsTools
- poll: http://pyxides.stani.be/polls/20060310-firstfocus/PlonePopoll_results2

Stani

PS IDLE is the only one which didn't answer my invitation yet, but
we'd love them to be in the team as well. (Kurt?)

--
http://pythonide.stani.be

From spe.stani.be at gmail.com  Sat Mar 11 06:13:43 2006
From: spe.stani.be at gmail.com (SPE Stani's Python Editor)
Date: Sat, 11 Mar 2006 06:13:43 +0100
Subject: [Pythonmac-SIG] [wxPython-users] ANN: The Python multiple IDE
	Collaboration calls for coders
Message-ID: <2078a7ad0603102113nff2d8a3hfa8bfe0ee3d4fde3@mail.gmail.com>

(cross posting from comp.lang.python)

Dear Pythoneers,

Looking at IDE's I can have three observations:

1. For some reasons numerous users prefer to use an open source IDE.

2. For some reasons numerous python programmers like to develop an
open source IDE.

3. For some reasons the open source python IDE developers are not
collaborating at all.

The reasons for 1 or 2 are obvious, at least to me. Recently I have
been wondering about the reason for 3. (Probably a lot of python
programmers have wondered about this already for ages, but ok I might
be slow ;-) I came to the conclusion that there was *NO* reason.

As this was so clear, I started to invite all the authors of IDE's
personally to collaborate all together. I hope that I didn't forget
any, because there are so many. What is really nice, is that we feel
the same: we should work together and share as much as possible. We
don't want to waste our (often spare) time on reinventing wheels.
Almost all IDE's (except of two) are participating no matter if they
use Tkinter, wxPython, pyQT, Cocoa, pyGTK, ... (So this could open
doors for an ajax python editor, who knows. Any python web framework
like Django, Turbogears, ... interested in that?)

These projects are participating: NewEdit, scrIDE, Eric3, Leo IDE,
ActiveGrid, PIDA, drPython, pyDev, PyCrust, IPython, WinPdb debugger,
Extended Python Debugger, PyLint, Gaphor, Envisage, Dabo, SilverCity &
SPE.

It is not about unification, but about a little bit more
collaboration. There are always libraries to share, more as we might
think.

In order to give the project shape I started building a (wiki based)
website in plone which together with a mailing list should give a good
platform for collaboration.  (You need to login to edit wiki's.) All
the developers are already invited, but everyone willing to code or
contribute (documentation, translation, artwork, plone website, ..) is
welcome.  If you work on open source project which might be of
interest (parsing, uml, framework, ...) please join or invite the
projects which you think should participate as well.

We will probably work in smaller teams on the various aspects of IDE's
and tools. If this project succeeds it could be a major win for the
Python community.

These are some useful links:

- homepage: http://pyxides.stani.be
- mail list: http://pyxides.stani.be/polls/mailing
- starting mail: http://pyxides.stani.be/wiki/StartingEmail
- developers reaction: http://pyxides.stani.be/wiki/AuthorsOfIDEsTools
- poll: http://pyxides.stani.be/polls/20060310-firstfocus/PlonePopoll_results2

Stani

PS IDLE is the only one which didn't answer my invitation yet, but
we'd love them to be in the team as well. (Kurt?)

--
http://pythonide.stani.be

---------------------------------------------------------------------
To unsubscribe, e-mail: wxPython-users-unsubscribe at lists.wxwidgets.org
For additional commands, e-mail: wxPython-users-help at lists.wxwidgets.org

From ronaldoussoren at mac.com  Sat Mar 11 17:04:20 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sat, 11 Mar 2006 17:04:20 +0100
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <E592E4EB-8329-4ACF-AB4F-2B5C07367CFC@stanford.edu>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
	<88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
	<E592E4EB-8329-4ACF-AB4F-2B5C07367CFC@stanford.edu>
Message-ID: <99568F6E-E9BE-45A1-81B8-9F6E66EF0D01@mac.com>


On 10-mrt-2006, at 23:25, Zachary Pincus wrote:

>> If you come up with a patch for (2) against Python's svn trunk that
>> is tested and works on 10.2 then I'll commit it.  It's not
>> appropriate to make this sort of change against Python 2.4,
>> though.  *Maybe* for the universal branch that Ronald and I are
>> doing, because that's also unlikely to be merged with the 2.4 branch.
>
> Well, after wasting the better part of a day tracking down a computer
> capable of running 10.2, installing 10.2 on it, getting the latest
> updates and the newest 10.2 developer tools, I find that neither
> python 2.5 from svn (today's snapshot or a checkout from the 8th)
> *NOR* the svn snapshot of python 2.4 will build on 10.2.8.
>
> Is python 2.5 (or even 2.4) supposed to be officially supported for
> 10.2? (It certainly doesn't seem to be based on what's on the python
> web pages.) If not, I rather resent the goose chase I've been sent on
> just trying to submit a patch to get python to work the way the
> documentation says it does.

Do you know why it doesn't build? A patch to fix build problems would
of course be appricated :-). I know of nobody that actively
supports Python 2.4 or 2.5 on Jaguar, which means stuff will break.


Ronald


From zpincus at stanford.edu  Sat Mar 11 21:12:11 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sat, 11 Mar 2006 12:12:11 -0800
Subject: [Pythonmac-SIG] dlopenflags and sharing symbols across python
	extension modules
In-Reply-To: <99568F6E-E9BE-45A1-81B8-9F6E66EF0D01@mac.com>
References: <8221A3B4-41EA-4077-AB9A-053EDE08B3DE@stanford.edu>
	<D6F655CA-9154-4E76-B09D-F1391DC36178@mac.com>
	<D0678D51-3FAC-4F30-B040-8F00DFE8D5FF@stanford.edu>
	<5C4EDD8B-FA3B-46E1-927A-A23ED86DCED2@redivi.com>
	<3B430DF9-BFE3-498B-933F-40946D648016@stanford.edu>
	<58DFF96A-207F-4E52-9482-AFD8DCACDE21@redivi.com>
	<AC3D7954-7262-4CF2-8D45-F6ACC2486260@stanford.edu>
	<90C67763-5C18-4452-972A-59CCD2B5D1DD@redivi.com>
	<B4871DF2-F0BD-4E45-9A97-D10226FCD09F@stanford.edu>
	<88870F03-2EF3-407E-84DD-B695FCDE3C65@redivi.com>
	<E592E4EB-8329-4ACF-AB4F-2B5C07367CFC@stanford.edu>
	<99568F6E-E9BE-45A1-81B8-9F6E66EF0D01@mac.com>
Message-ID: <4A2824A0-8A1A-4039-AD73-AD93E8ECE2AC@stanford.edu>

>> Well, after wasting the better part of a day tracking down a computer
>> capable of running 10.2, installing 10.2 on it, getting the latest
>> updates and the newest 10.2 developer tools, I find that neither
>> python 2.5 from svn (today's snapshot or a checkout from the 8th)
>> *NOR* the svn snapshot of python 2.4 will build on 10.2.8.
>
> Do you know why it doesn't build? A patch to fix build problems would
> of course be appricated :-). I know of nobody that actively
> supports Python 2.4 or 2.5 on Jaguar, which means stuff will break.

The 2.4 problems stemmed from lack of a 'uint32' type in the system  
headers, I think (which surprised me) -- maybe python is including  
the wrong headers for 10.2 or something...

The 2.5 problems were more complex link errors with printf -- I'd  
seen that type of error before, but not for a while.

I don't really have time right now to track these errors down -- I've  
been spending enough time trying to get the dlopen patch tested and  
verfified, while still doing my actual work. Maybe once the dlopen  
stuff gets put to bed, but even then I can't guarantee anything.

Zach




From zbir at urbanape.com  Sun Mar 12 00:05:46 2006
From: zbir at urbanape.com (Zachery Bir)
Date: Sat, 11 Mar 2006 18:05:46 -0500
Subject: [Pythonmac-SIG] NSUserDefaults causes crash on MOSXI
In-Reply-To: <78C2A3B9-95C4-42D7-AE63-AE7E54EC2971@urbanape.com>
References: <2F100638-92A8-4932-81D1-753E2A1961F8@urbanape.com>
	<60AF98A3-3EB5-4905-9F4A-DF813C5C416A@urbanape.com>
	<2C9DD974-6B77-4EBD-B267-EB7E026F21A7@redivi.com>
	<78C2A3B9-95C4-42D7-AE63-AE7E54EC2971@urbanape.com>
Message-ID: <4BF92D85-E6EA-49D8-BAF8-1601C8265A16@urbanape.com>

On Mar 10, 2006, at 10:22 PM, Zachery Bir wrote:

> On Mar 1, 2006, at 1:22 PM, Bob Ippolito wrote:
>
>> Alternatively you could run it under ktrace to see which file it
>> opens.  There should be a NAMI with the path it's opening.
>
> Boy does that make some kinda output ;^) I'm more than willing to
> post this to my site if it'll be helpful.

I've uploaded the results of:

   ~/Developer/Thumbscrew/branches/1.0-branch
   zbir at Silverback $ kdump -f ktrace.out > /tmp/kdump.out

to my site:

   <http://www.urbanape.com/downloads/kdump.out>

I'd appreciate any help making sense of that as well :^)

Zac


From vivacarlie at gmail.com  Mon Mar 13 06:43:44 2006
From: vivacarlie at gmail.com (Nehemiah Dacres)
Date: Sun, 12 Mar 2006 23:43:44 -0600
Subject: [Pythonmac-SIG] let me get the "colaboration of opensource python
	IDEs" idea right...
Message-ID: <65fadfc30603122143x37f49cbev9946a4146e318ab3@mail.gmail.com>

so the idea is

get all the people working on open-source IDEs for python to
collaborate on the effort so they dont keep redoing the same job.

this will also improve the end product because it will take the best
parts of all the others.

and finally
it will be more modular in essence by making the core framework
separate from the gui or front end

that separation should make more opportunity for portable code, like
for example the main project in c++, the GUI parts in Object oriented
c and other operating system specific code to make it pretty and most
functional to each operating. and to make it the most accessible, some
of the more sensitive code should be in python. Also it allows for
some kind of uniformity in their behaviors between operating systems.
(this way we don't have such finger slowing differences like there
would be between versions of a video game but you don't have ctrl-s
saving on apples. i just hope this, as an IDE is giving all the
functionality as expected, like a Ctrl-r/Command-r for running the
code with the associated framework of python. and syntax coloring (
including the word 'self' BareBones)
 these are just my suggestions though.

From glassfordm at hotmail.com  Mon Mar 13 19:51:38 2006
From: glassfordm at hotmail.com (Michael Glassford)
Date: Mon, 13 Mar 2006 13:51:38 -0500
Subject: [Pythonmac-SIG] PyObjC:
 dragImageForRowsWithIndexes_tableColumns_event_offset_() causes
 "PyObjCPointer created" error
Message-ID: <dv4evq$or8$1@sea.gmane.org>

I added a dragImageForRowsWithIndexes_tableColumns_event_offset_() 
method to an NSTableView subclass that I have created. It's working OK, 
but every time it is called, a message like this is printed to the error 
log:

"PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
dragImageForRows_event_dragImageOffset_()"

I had a similar problem with 
tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which 
caused a crash on some machines. It was fixed by changing the signature 
of that function in AppKit/protocols.py. I looked for a way to change 
the signature of dragImageForRowsWithIndexes_tableColumns_event_offset_, 
but couldn't find it anywhere.

Suggestions?

Mike


From rodneys at io.com  Tue Mar 14 14:42:46 2006
From: rodneys at io.com (Rodney Somerstein)
Date: Tue, 14 Mar 2006 08:42:46 -0500
Subject: [Pythonmac-SIG] Bad download link on python.org
Message-ID: <p06230900c03c7822dff6@[10.71.0.200]>

I didn't notice anyone mentioning this on previous discussions, but 
currently, if you click the download link on the left side of the 
main python.org site, it takes you to a page that states this about 
the Mac version of Python:

Python 2.3 OS X 10.2 installer (requires admin privileges -- see 
MacPython download page for details). Note that as of the 2.4 Python 
release, the Mac OS X installer is still at version 2.3.

The actual link in that paragraph to the MacPython download page 
directs people to a non-existent page at Jack Jansen's old site. So, 
not only do people think that the latest version of MacPython is 2.3, 
but they can't even download that.

If this has already been pointed out here on the list, then nevermind.

-Rodney

From rodneys at io.com  Tue Mar 14 14:44:45 2006
From: rodneys at io.com (Rodney Somerstein)
Date: Tue, 14 Mar 2006 08:44:45 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
Message-ID: <p06230901c03c79492502@[192.168.2.234]>

As the title of this message asks, which Python should I use? And 
why? From following discussions here, I know why I don't want to use 
the built in Apple Python.

So, that leaves 2 choices that both seem good, the macpython 
framework build and ActiveState. The main reason that I would want to 
use the "official" framework build is that it is endorsed by this 
community. It is, however, a volunteer effort and actually states 
that it is usually one micro-version behind the most recent. For 
ActiveState, it seems to be well supported and more up to date. 
Though I understand that there could be issues that people here might 
have trouble helping me with because I'm using a different Python.

So, why would I really want to use one vs. the other? Will I have a 
harder time finding cross-platform packages that work with the 
ActiveState build? Am I going to have to compile many of the packages 
that I want anyway, regardless of the Python I use? Other than 
possible commercial distribution issues are there any real 
disadvantages of ActiveState's distribution? How about the "official" 
framework build? What is one or the other missing that is present in 
the other one?

I know this has been asked before, but I couldn't find it in a quick 
scan of the archives. The official page and Wiki don't seem to give 
me these answers and neither does ActiveState's page.

If it influences the answers, I am looking to build a cross-platform 
application that I eventually want to be able to package for easy 
installation by non-Python savvy users.

Thanks for your help and suggestions with this one,

-Rodney

From bob at redivi.com  Tue Mar 14 15:21:38 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 14 Mar 2006 06:21:38 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <p06230901c03c79492502@[192.168.2.234]>
References: <p06230901c03c79492502@[192.168.2.234]>
Message-ID: <D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>


On Mar 14, 2006, at 5:44 AM, Rodney Somerstein wrote:

> As the title of this message asks, which Python should I use? And
> why? From following discussions here, I know why I don't want to use
> the built in Apple Python.
>
> So, that leaves 2 choices that both seem good, the macpython
> framework build and ActiveState. The main reason that I would want to
> use the "official" framework build is that it is endorsed by this
> community. It is, however, a volunteer effort and actually states
> that it is usually one micro-version behind the most recent. For
> ActiveState, it seems to be well supported and more up to date.
> Though I understand that there could be issues that people here might
> have trouble helping me with because I'm using a different Python.
>
> So, why would I really want to use one vs. the other? Will I have a
> harder time finding cross-platform packages that work with the
> ActiveState build? Am I going to have to compile many of the packages
> that I want anyway, regardless of the Python I use? Other than
> possible commercial distribution issues are there any real
> disadvantages of ActiveState's distribution? How about the "official"
> framework build? What is one or the other missing that is present in
> the other one?

There is very little difference between ActiveState Python 2.4.2 and  
the 2.4.1 release.  The 2.4.1 release ships with a working readline,  
and ActiveState's doesn't need the header fix for distutils on 10.4.

At this point your best bet is probably to use the universal branch  
of 2.4.2.  Ronald has recently posted binaries to this list.  You  
will have some trouble making redistributable app bundles with this  
one for now, but that should be resolved sometime soon.

-bob


From ronaldoussoren at mac.com  Tue Mar 14 15:48:46 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 15:48:46 +0100
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
Message-ID: <FE6889B7-365D-46E8-B0A8-8497BE27838C@mac.com>


On 14-mrt-2006, at 15:21, Bob Ippolito wrote:

>
> On Mar 14, 2006, at 5:44 AM, Rodney Somerstein wrote:
>
>> As the title of this message asks, which Python should I use? And
>> why? From following discussions here, I know why I don't want to use
>> the built in Apple Python.
>>
>> So, that leaves 2 choices that both seem good, the macpython
>> framework build and ActiveState. The main reason that I would want to
>> use the "official" framework build is that it is endorsed by this
>> community. It is, however, a volunteer effort and actually states
>> that it is usually one micro-version behind the most recent. For
>> ActiveState, it seems to be well supported and more up to date.
>> Though I understand that there could be issues that people here might
>> have trouble helping me with because I'm using a different Python.
>>
>> So, why would I really want to use one vs. the other? Will I have a
>> harder time finding cross-platform packages that work with the
>> ActiveState build? Am I going to have to compile many of the packages
>> that I want anyway, regardless of the Python I use? Other than
>> possible commercial distribution issues are there any real
>> disadvantages of ActiveState's distribution? How about the "official"
>> framework build? What is one or the other missing that is present in
>> the other one?
>
> There is very little difference between ActiveState Python 2.4.2 and
> the 2.4.1 release.  The 2.4.1 release ships with a working readline,
> and ActiveState's doesn't need the header fix for distutils on 10.4.
>
> At this point your best bet is probably to use the universal branch
> of 2.4.2.  Ronald has recently posted binaries to this list.  You
> will have some trouble making redistributable app bundles with this
> one for now, but that should be resolved sometime soon.

The universal build is in the public folder of my iDisk (http:// 
homepage.mac.com/ronaldoussoren). I'm happy with this version and  
consider it stable.

BTW. I intent to publish a 2.4.3 version of the universal build  
whenever 2.4.3 is released.

Ronald

>
> -bob
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From bob at redivi.com  Tue Mar 14 16:30:45 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 14 Mar 2006 07:30:45 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <p06230903c03c88cbc797@[192.168.2.234]>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
	<p06230903c03c88cbc797@[192.168.2.234]>
Message-ID: <2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>


On Mar 14, 2006, at 7:00 AM, Rodney Somerstein wrote:

> Thanks for the answer, Bob, and thanks for the work on the  
> Universal build, Ronald. If someone could answer my other questions  
> as well, I would really appreciate it.
>
> As a beginner, what does having a working readline actually mean to  
> me? If I'm not building command line apps do I need that for user  
> input? And why wouldn't ActiveState have one? Given that  
> ActiveState seems to put forth the effort to make a release of  
> Python that is compatible across multiple platforms, including a  
> Universal Mac build, why does the MacPython community maintain a  
> separate framework build? (No criticism intended here, I want to  
> understand this)

readline provides for command history (the up arrow, etc.).   
ActiveState does not provide it because it is GPL.  It's possible to  
install this after the fact, but it's only easy because I've made it  
available as a separate package.. normally it's a pain in the ass to  
do because readline is part of Python's source tree and doesn't  
readily build externally (not a big deal, but most people aren't up  
to writing setup.py files for code they didn't write).

ActiveState's efforts on the various platforms really don't have  
anything to do with each other.  There isn't really anything  
technical that differentiates ActiveState Python from the mainline  
distro, at least on OS X.

> Again, is there really any reason that I would want to use one  
> release over the other? Is it simply a matter of readline, whatever  
> that buys me (I'm obviously a beginner to Python even though I've  
> read a bit about it over the years) or is there some other major  
> reason? Such as, will I have problems creating redistributable app  
> bundles with ActiveState since Bob seems to be working mostly with  
> the MacPython build? How about other add-on libraries I might want  
> to use?

The PPC builds are indistinguishable technically beyond the changelog  
from 2.4.1 to 2.4.2 (and the absence of readline -- maybe also bsddb  
support? -- in ActiveState's distro).  These questions are  
irrelevant.  The universal build is what you want to be using,  
though.  However, caveat emptor, because py2app doesn't *yet* work  
universal.  It will soon, and that's the only scenario I'm really  
going to be testing future versions of py2app with.

> If I go with the MacPython framework build, how likely is it to  
> catch up to the current release of Python? I notice that it has  
> been 6 months since the 2.4.2 release and it isn't easy for a new  
> user to find links to a "official" Mac build of this version. I do  
> note that Ronald has stated he will put out a 2.4.3 build when  
> 2.4.3 is release, but I can't even find links to 2.4.2 on  
> python.org. Is this likely to change?

First off, the biggest reason that 2.4.2 hasn't been released in  
binary form by us is that I didn't want to bother with another  
release until we handled the i386 situation, and secondly it doesn't  
have all that many relevant changes.  It simply wasn't worth the  
effort of doing a micro release when we knew we were going to be  
obsoleting the PPC-only builds altogether ASAP.

The universal build is considered to be of release candidate  
quality.  It will be the build on python.org when we've decided that  
it's final.  At this point, we don't expect to make any meaningful  
changes to what Ronald has in his disk image.  The universal build  
was built against the Python 2.4 maintenance branch as of a few weeks  
ago and includes patches that aren't yet in 2.4.2 and many patches  
relevant to i386.  It's likely to have bugfixes that ActiveState's  
distro does not have and will not have until the release of 2.4.3,  
especially for i386, unless they decide to switch to using the same  
universal branch.

-bob


From rodneys at io.com  Tue Mar 14 16:30:38 2006
From: rodneys at io.com (Rodney Somerstein)
Date: Tue, 14 Mar 2006 10:30:38 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
Message-ID: <p06230904c03c9216f510@[192.168.2.234]>

Thanks for the answer, Bob, and thanks for the work on the Universal 
build, Ronald. If someone could answer my other questions as well, I 
would really appreciate it.

As a beginner, what does having a working readline actually mean to 
me? If I'm not building command line apps do I need that for user 
input? And why wouldn't ActiveState have one? Given that ActiveState 
seems to put forth the effort to make a release of Python that is 
compatible across multiple platforms, including a Universal Mac 
build, why does the MacPython community maintain a separate framework 
build? (No criticism intended here, I want to understand this)

Again, is there really any reason that I would want to use one 
release over the other? Is it simply a matter of readline, whatever 
that buys me (I'm obviously a beginner to Python even though I've 
read a bit about it over the years) or is there some other major 
reason? Such as, will I have problems creating redistributable app 
bundles with ActiveState since Bob seems to be working mostly with 
the MacPython build? How about other add-on libraries I might want to 
use?

If I go with the MacPython framework build, how likely is it to catch 
up to the current release of Python? I notice that it has been 6 
months since the 2.4.2 release and it isn't easy for a new user to 
find links to a "official" Mac build of this version. I do note that 
Ronald has stated he will put out a 2.4.3 build when 2.4.3 is 
release, but I can't even find links to 2.4.2 on python.org. Is this 
likely to change?

Thanks,

-Rodney

From rodneys at io.com  Tue Mar 14 16:39:40 2006
From: rodneys at io.com (Rodney Somerstein)
Date: Tue, 14 Mar 2006 10:39:40 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
Message-ID: <p06230907c03c933237be@[192.168.2.234]>

Thanks for your extensive answer, Bob. I will definitely go with the 
universal build given what you have said, just to facilitate getting 
help with any problems I run into. As for py2app not working with 
that yet, I'm still a long way from having anything ready to package 
anyway, so that isn't an issue for me.

-Rodney

From ronaldoussoren at mac.com  Tue Mar 14 16:47:05 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 16:47:05 +0100
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <p06230904c03c9216f510@[192.168.2.234]>
References: <p06230904c03c9216f510@[192.168.2.234]>
Message-ID: <D60E310B-CC3F-44DF-84EE-CFFAF083EC2C@mac.com>


On 14-mrt-2006, at 16:30, Rodney Somerstein wrote:

> Thanks for the answer, Bob, and thanks for the work on the Universal
> build, Ronald. If someone could answer my other questions as well, I
> would really appreciate it.
>
> As a beginner, what does having a working readline actually mean to
> me?

Command-line history in the python shell. Using the array keys you
can bring back previous statements and edit and reexecute them. This
makes the python shell much more convenient.

> If I'm not building command line apps do I need that for user
> input? And why wouldn't ActiveState have one? Given that ActiveState
> seems to put forth the effort to make a release of Python that is
> compatible across multiple platforms, including a Universal Mac
> build, why does the MacPython community maintain a separate framework
> build? (No criticism intended here, I want to understand this)

Our framework build predates ActivePython support for OSX. We (Jack,
Bob, me and several others) are also the ones doing the actual OSX
support such as creating support for framework builds, Apple specific
extensions and support for univeral binaries.

Note that I have never used ActivePython and therefore cannot comment
on the quality of their work, or even on what they do or do not include.

As one of the maintainers for the universal builds I obviously prefer
out build :-)

> Again, is there really any reason that I would want to use one
> release over the other? Is it simply a matter of readline, whatever
> that buys me (I'm obviously a beginner to Python even though I've
> read a bit about it over the years) or is there some other major
> reason? Such as, will I have problems creating redistributable app
> bundles with ActiveState since Bob seems to be working mostly with
> the MacPython build? How about other add-on libraries I might want to
> use?

Both should work in both builds. However, as you noted Bob seems to be
using the MacPython build hence actual support for that will be better.

Bob has provided binary installers for lots of extension packages in
the past, and I'm sure simular installers will emerge for the new
universal build in the near future (if he doesn't create new installers
I will).

>
> If I go with the MacPython framework build, how likely is it to catch
> up to the current release of Python? I notice that it has been 6
> months since the 2.4.2 release and it isn't easy for a new user to
> find links to a "official" Mac build of this version. I do note that
> Ronald has stated he will put out a 2.4.3 build when 2.4.3 is
> release, but I can't even find links to 2.4.2 on python.org. Is this
> likely to change?

Yes, Bill Janssen has edit-rights on the python.org site and will
update the link sometime in the future (Bill: now would be fine :-))

The mac world is a bit in-between maintainers at the moment, the
official maintainer (Jack Jansen) is mostly absent (very busy with
other work), and nobody has picked up the slack. Things seem to
have improved a little recently, but we'll have to see how that
works out.

Ronald


From Hubert.Holin at lmd.polytechnique.fr  Tue Mar 14 17:04:27 2006
From: Hubert.Holin at lmd.polytechnique.fr (Hubert Holin)
Date: Tue, 14 Mar 2006 17:04:27 +0100
Subject: [Pythonmac-SIG] Python (2.3, 2.4...) + XCode (2.1,
	2.2) + Extensions...
Message-ID: <80498F04-735E-4B36-96DD-E73F6F648129@lmd.polytechnique.fr>

Paris (U.E.), le 14/03/2006

	Bonjour

		Back when XCode 2.1 was current (no *so* long ago), I used it to  
build extensions for Python (first Apple's 2.3, then Bob's "official  
unofficial'" 2.4). Finding the right combination of flags to make the  
thing work was not funny.

		Then XCode 2.2 came along, which was fine for some C++ work.

		I have come to revisit some older XCode projects, which included  
extensions to Python, and, lo and behold, Apple has played "mix the  
flags" again! This means I can't build new projects which includes  
custom extensions to Python using only *documented* flags, unless I  
once again go into a hair-tearing flag hunting session... Unless  
somebody else has done that and is willing to share the result of his  
labor.

		For the record, some of the projects I was revisiting were for  
getting my feet wet with MacTels during an Apple Workshop (Universal  
Binaries) in Paris, Friday. Guess I'll have some interesting  
questions to ask them...

	Merci

Hubert Holin


From janssen at parc.com  Tue Mar 14 17:34:23 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 14 Mar 2006 08:34:23 PST
Subject: [Pythonmac-SIG] Bad download link on python.org
In-Reply-To: Your message of "Tue, 14 Mar 2006 05:42:46 PST."
	<p06230900c03c7822dff6@[10.71.0.200]> 
Message-ID: <06Mar14.083426pst."58633"@synergy1.parc.xerox.com>

Yes, I know about it, thanks.

If I can figure out how to change it, I'll do so.

Bill

> I didn't notice anyone mentioning this on previous discussions, but 
> currently, if you click the download link on the left side of the 
> main python.org site, it takes you to a page that states this about 
> the Mac version of Python:
> 
> Python 2.3 OS X 10.2 installer (requires admin privileges -- see 
> MacPython download page for details). Note that as of the 2.4 Python 
> release, the Mac OS X installer is still at version 2.3.
> 
> The actual link in that paragraph to the MacPython download page 
> directs people to a non-existent page at Jack Jansen's old site. So, 
> not only do people think that the latest version of MacPython is 2.3, 
> but they can't even download that.
> 
> If this has already been pointed out here on the list, then nevermind.
> 
> -Rodney
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From janssen at parc.com  Tue Mar 14 17:36:40 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 14 Mar 2006 08:36:40 PST
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: Your message of "Tue, 14 Mar 2006 05:44:45 PST."
	<p06230901c03c79492502@[192.168.2.234]> 
Message-ID: <06Mar14.083641pst."58633"@synergy1.parc.xerox.com>

> If it influences the answers, I am looking to build a cross-platform 
> application that I eventually want to be able to package for easy 
> installation by non-Python savvy users.

I don't see a good alternative to using Apple's version, then.

Or, at least, I don't know how to build a Mac installer that installs
a complete Python along with the other code it needs.  If there's a
cookbook way of doing that, I'd be interested, too.

Bill

From charles.hartman at conncoll.edu  Tue Mar 14 17:50:39 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 14 Mar 2006 11:50:39 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
	<p06230903c03c88cbc797@[192.168.2.234]>
	<2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>
Message-ID: <E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>

I'm still not *quite* clear on the interactions here.

1. wxPython still isn't Universal, though there are rumblings in the  
distance. Until that happens, is it possible to use the emerging  
Universal build of Python with the PPC build of wxPython? I guess I  
mean, possible to make command-line Python scripts that call (PPC)  
wx, on an Intel machine? In general, I guess: does Python care how  
its libraries are built, as long as they're built to run *somehow* on  
the platform that Python is running on? (Is that even the right  
question??)

2. Until Universal py2app (thanks Bob!) is ready, it's not possible  
to build Mac standalone apps with Universal Python -- is that correct?

3. (the combination) When Universal py2app is ready -- long before  
Universal wxPython, I'll bet -- will it be possible to build  
standalone apps for Mac (PPC and Intel) using (PPC) wxPython?

Sorry to be so foggy-headed about the innards of this. It isn't  
urgent, since *everything* seems to work fine in the PPC versions on  
an Intel Mac.

Charles



From charles.hartman at conncoll.edu  Tue Mar 14 17:54:57 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 14 Mar 2006 11:54:57 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar14.083641pst."58633"@synergy1.parc.xerox.com>
References: <06Mar14.083641pst."58633"@synergy1.parc.xerox.com>
Message-ID: <169AC96E-3C6A-4E41-83F7-1CEB285DE608@conncoll.edu>


On Mar 14, 2006, at 11:36 AM, Bill Janssen wrote:

>> If it influences the answers, I am looking to build a cross-platform
>> application that I eventually want to be able to package for easy
>> installation by non-Python savvy users.
>
> I don't see a good alternative to using Apple's version, then.
>
> Or, at least, I don't know how to build a Mac installer that installs
> a complete Python along with the other code it needs.  If there's a
> cookbook way of doing that, I'd be interested, too.

I don't think I understand -- did I miss part of the context of the  
question? For a long time I've been building apps with MacPython and  
wxPython that run just fine on Mac and Windows, using py2app and  
py2exe to build the executables, and packaging the Win version with  
InnoSetup and for Mac just making a double-clickable .dmg -- both  
dead simple for users (whom I expect to know even less than I do).

Charles


From Chris.Barker at noaa.gov  Tue Mar 14 18:01:56 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 14 Mar 2006 09:01:56 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar14.083641pst.58633@synergy1.parc.xerox.com>
References: <06Mar14.083641pst.58633@synergy1.parc.xerox.com>
Message-ID: <4416F704.9030600@noaa.gov>

Bill Janssen wrote:
> Or, at least, I don't know how to build a Mac installer that installs
> a complete Python along with the other code it needs.  If there's a
> cookbook way of doing that, I'd be interested, too.

Bill,

Do you really not know about Py2App? It occasionally has some tricky 
bits, and doesn't yet work with the new universal build, but it is still 
better than anything else out there for any platform. In fact, with my 
limited testing, it has always "just worked" which I can not say for py2exe.

As a note, a big part of the discussion about what version of python the 
macpython community should endorse centered around the advantages of 
using a non-Apple build with py2app.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From bob at redivi.com  Tue Mar 14 18:46:51 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 14 Mar 2006 09:46:51 -0800
Subject: [Pythonmac-SIG] Python (2.3, 2.4...) + XCode (2.1,
	2.2) + Extensions...
In-Reply-To: <80498F04-735E-4B36-96DD-E73F6F648129@lmd.polytechnique.fr>
References: <80498F04-735E-4B36-96DD-E73F6F648129@lmd.polytechnique.fr>
Message-ID: <C081D19B-D743-4AAC-9091-8EC8BB087494@redivi.com>


On Mar 14, 2006, at 8:04 AM, Hubert Holin wrote:

> 		Back when XCode 2.1 was current (no *so* long ago), I used it to
> build extensions for Python (first Apple's 2.3, then Bob's "official
> unofficial'" 2.4). Finding the right combination of flags to make the
> thing work was not funny.
>
> 		Then XCode 2.2 came along, which was fine for some C++ work.
>
> 		I have come to revisit some older XCode projects, which included
> extensions to Python, and, lo and behold, Apple has played "mix the
> flags" again! This means I can't build new projects which includes
> custom extensions to Python using only *documented* flags, unless I
> once again go into a hair-tearing flag hunting session... Unless
> somebody else has done that and is willing to share the result of his
> labor.

The right answer is *use distutils*.  Then you don't have to touch  
the flags because you know that all the code will be compiled the  
same way Python was.

-bob


From nicholas.cole at gmail.com  Tue Mar 14 19:00:12 2006
From: nicholas.cole at gmail.com (Nicholas Cole)
Date: Tue, 14 Mar 2006 18:00:12 +0000
Subject: [Pythonmac-SIG] resizing terminal and curses
Message-ID: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>

I couldn't find anything in the archives related to this.

Using both Apple's python and a build of python2.4.2, there seems to
be a bug when using ncurses.  curses.newwin() is supposed to give a
new window filling the pysical terminal, but if the terminal has been
resized this does not seem to happen.  Instead, the window seems to be
the size that the terminal was when curses was initialised.  This is
not the case on other platforms, where the window created reflects the
new size of the terminal.  Is this a known bug?

(it seems to affect both xterm and terminal.app)

Best wishes,

N.

From janssen at parc.com  Tue Mar 14 19:00:35 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 14 Mar 2006 10:00:35 PST
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: Your message of "Tue, 14 Mar 2006 09:01:56 PST."
	<4416F704.9030600@noaa.gov> 
Message-ID: <06Mar14.100041pst."58633"@synergy1.parc.xerox.com>

> Do you really not know about Py2App? It occasionally has some tricky 
> bits, and doesn't yet work with the new universal build, but it is still 
> better than anything else out there for any platform. In fact, with my 
> limited testing, it has always "just worked" which I can not say for py2exe.

Chris,

I haven't paid much attention to py2app, but that's because I guessed
it didn't fit my needs.  I'm not writing apps in Python because I
don't write Mac-specific code, and therefore need a portable GUI
toolkit, and none of the available-for-Python-on-the-Mac portable GUI
systems are good enough (I have hopes for GTK+ on Cairo, with the
Imendio-sponsored port at
http://developer.imendio.com/wiki/Gtk_Mac_OS_X).  My system consists
of a Python daemon which uses PIL, ReportLab, Medusa, and PyLucene,
along with 4 Python command-line programs, along with 2 Mac apps
written in Java as bundled jar files (for Swing).

When building an installer, I build the Python extensions from source
(using the system Python) and install them under /usr/local/mysys/...
The post-install script in the Mac installer then inserts a .pth file
for my system's extensions in the system Python's site-packages
directory.  This allows all the command-line Python scripts to use
them, along with the daemon.

Can you suggest a better approach?

Bill

From rogerb at rogerbinns.com  Tue Mar 14 19:16:52 2006
From: rogerb at rogerbinns.com (Roger Binns)
Date: Tue, 14 Mar 2006 10:16:52 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
References: <p06230904c03c9216f510@[192.168.2.234]>
	<D60E310B-CC3F-44DF-84EE-CFFAF083EC2C@mac.com>
Message-ID: <001f01c64793$77d91d80$3501a8c0@rogersqyvr14d3>

> Note that I have never used ActivePython and therefore cannot comment
> on the quality of their work, or even on what they do or do not include.

Also don't forget that the license for ActivePython is different than
for "vanilla" Python:

http://www.activestate.com/Products/ActivePython/license_agreement.plex

In the past it didn't allow for use with py2app/py2exe style 
redistribution - see section 2.  Section 4 now does allow
it.

Roger

From bob at redivi.com  Tue Mar 14 19:17:43 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 14 Mar 2006 10:17:43 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
	<p06230903c03c88cbc797@[192.168.2.234]>
	<2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>
	<E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
Message-ID: <D584D4AB-F2D6-4811-A51E-43AB023F4ACF@redivi.com>


On Mar 14, 2006, at 8:50 AM, Charles Hartman wrote:

> I'm still not *quite* clear on the interactions here.
>
> 1. wxPython still isn't Universal, though there are rumblings in  
> the distance. Until that happens, is it possible to use the  
> emerging Universal build of Python with the PPC build of wxPython?  
> I guess I mean, possible to make command-line Python scripts that  
> call (PPC) wx, on an Intel machine? In general, I guess: does  
> Python care how its libraries are built, as long as they're built  
> to run *somehow* on the platform that Python is running on? (Is  
> that even the right question??)

I'd expect that universal wxPython should hop along pretty quickly  
after we make the universal build official.. they'll have to, more or  
less.

You can not use plugins cross-arch, so no.  You can coerce rosette  
into running Python and doing the whole Python process in PPC emulation.

> 2. Until Universal py2app (thanks Bob!) is ready, it's not possible  
> to build Mac standalone apps with Universal Python -- is that correct?

Correct.  Haven't I said this like 5 times today?

> 3. (the combination) When Universal py2app is ready -- long before  
> Universal wxPython, I'll bet -- will it be possible to build  
> standalone apps for Mac (PPC and Intel) using (PPC) wxPython?
>
> Sorry to be so foggy-headed about the innards of this. It isn't  
> urgent, since *everything* seems to work fine in the PPC versions  
> on an Intel Mac.

Yes, but the standalone apps will have to run under Rosetta on  
Intel.  There's an Info.plist key that you can set to say that the  
bundle needs to start up for PPC.

-bob


From rodneys at io.com  Tue Mar 14 19:47:04 2006
From: rodneys at io.com (Rodney Somerstein)
Date: Tue, 14 Mar 2006 13:47:04 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
	<p06230903c03c88cbc797@[192.168.2.234]>
	<2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>
	<E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
Message-ID: <p0623090ac03cbe0841ce@[192.168.2.234]>

(Lots of stuff about non-Universal builds of stuff and other comments 
from various people removed...)

Luckily for me, I'm still a good ways away from needing to actually 
create a standalone app. That is just my ultimate goal. If there 
isn't a universal build of py2app by the time that I'm ready for it, 
we can probably assume that there never will be. As far as I can 
tell, Bob is working MUCH faster than my current time scale since I'm 
just in the learning stages.

The same thing would hold true for wxPython, which I will likely use. 
I still have to finish learning Python before really diving into 
that. And in the meantime, if I start before there is a universal 
build, people who have Intel Macs (not me at the moment) will just 
have to run under Rosetta. I expect they will probably never notice 
the difference for the kinds of things I'm likely to be doing.

As for the comment about not knowing how to build a standalone 
application that installs Python... Well, unless I'm horribly 
mistaken, this is exactly what py2app on the Mac and py2exe does 
under Windows. I'm assuming there is a Unix equivalent as well. Based 
on the discussions I've seen on this list, I can't come up with any 
reason I would want to use the Apple supplied version of Python 
rather than having my application include it. There is always 
application size, I suppose, but I don't think that will be a major 
issue for me.

-Rodney

From charles.hartman at conncoll.edu  Tue Mar 14 20:08:25 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Tue, 14 Mar 2006 14:08:25 -0500
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar14.100041pst."58633"@synergy1.parc.xerox.com>
References: <06Mar14.100041pst."58633"@synergy1.parc.xerox.com>
Message-ID: <36271C61-5BFF-4DF0-842A-55FE960FB20B@conncoll.edu>

I know I'm ignorant, but this seems to me really bizarre. Why would  
you go through all that? What doesn't wxPython (for example) do, that  
you need a GUI to do?

Charles


On Mar 14, 2006, at 1:00 PM, Bill Janssen wrote:

>> Do you really not know about Py2App? It occasionally has some tricky
>> bits, and doesn't yet work with the new universal build, but it is  
>> still
>> better than anything else out there for any platform. In fact,  
>> with my
>> limited testing, it has always "just worked" which I can not say  
>> for py2exe.
>
> Chris,
>
> I haven't paid much attention to py2app, but that's because I guessed
> it didn't fit my needs.  I'm not writing apps in Python because I
> don't write Mac-specific code, and therefore need a portable GUI
> toolkit, and none of the available-for-Python-on-the-Mac portable GUI
> systems are good enough (I have hopes for GTK+ on Cairo, with the
> Imendio-sponsored port at
> http://developer.imendio.com/wiki/Gtk_Mac_OS_X).  My system consists
> of a Python daemon which uses PIL, ReportLab, Medusa, and PyLucene,
> along with 4 Python command-line programs, along with 2 Mac apps
> written in Java as bundled jar files (for Swing).
>
> When building an installer, I build the Python extensions from source
> (using the system Python) and install them under /usr/local/mysys/...
> The post-install script in the Mac installer then inserts a .pth file
> for my system's extensions in the system Python's site-packages
> directory.  This allows all the command-line Python scripts to use
> them, along with the daemon.
>
> Can you suggest a better approach?
>
> Bill
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Tue Mar 14 20:28:35 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 14 Mar 2006 11:28:35 -0800
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar14.100041pst.58633@synergy1.parc.xerox.com>
References: <06Mar14.100041pst.58633@synergy1.parc.xerox.com>
Message-ID: <44171963.5090305@noaa.gov>

Bill,

You have unique needs that are far beyond what most people mean by 
"distribute my app"!

Bill Janssen wrote:
> none of the available-for-Python-on-the-Mac portable GUI
> systems are good enough 

wxPython is really the only portable option right now -- what is not 
good enough about it for your needs? I'm curious.

(I have hopes for GTK+ on Cairo, with the
> Imendio-sponsored port at
> http://developer.imendio.com/wiki/Gtk_Mac_OS_X).

That would be nice, but I'm not all that optimistic that it will be 
ready that soon, or look&feel very Mac-ish. People complain about that 
enough with wxPython.

> along with 2 Mac apps
> written in Java as bundled jar files (for Swing).

What does SWING give you that wxPython doesn't -- again, I'm curios.

> When building an installer, I build the Python extensions from source
> (using the system Python) and install them under /usr/local/mysys/...
> The post-install script in the Mac installer then inserts a .pth file
> for my system's extensions in the system Python's site-packages
> directory.  This allows all the command-line Python scripts to use
> them, along with the daemon.
> 
> Can you suggest a better approach?

I don't know that it's better, but once you're doing all that, it really 
wouldn't be a bid deal to distribute the new FrameWork installer as 
well, and then not be dependent on Apple's Python. What you have now 
will most likely break if the user upgrades OS-X, as Apple is likely to 
upgrade their python, and then your .pth file could disappear and, more 
fatally, your compiled extensions may not work with the new one. The 
rest of the process would be the same.

I'd like to be able to build a "Python Runtime" that included all of the 
Framework build, plus the full set of modules that I need for my apps. 
then I'd like to be able to run command line apps against it, and use 
Py2Applet (no, it doesn't exist), to build Application bundles that 
relied on that runtime. This seems to me to be  a good solution for 
folks like us that are distributing not one app, but a bunch of apps 
that all depend on the same stuff. In my case, I have a set of small GUI 
utilities, and some command line scripts -- it seems a bit crazy to 
bundle up the entire python and wxPython with a 200 line script! Or more 
to the point, the same pile with a number of small scripts.

I think all this could be accomplished with a small amount of hacking on 
Py2App, but I have no idea if I'll ever get around to that.

On the other hand, I wonder if the "runtime" could be built by simply 
installing the Framework Build, installing all the extensions you want, 
then making a big tarball out of:

/Library/Frameworks/Python.Framework

After all, I think one of the points of Frameworks is that they are 
self-contained.

You could then even give it a different name (MySpecialRunTime) and put 
a few links into usr/local/bin to whatever you need in:

/Library/FrameWorks.MySpecialRuntime.framework/Version/2.4/bin/

Would this work?

I may just try this some day.

-Chris













-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Tue Mar 14 21:26:42 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:26:42 +0100
Subject: [Pythonmac-SIG] PyObjC:
	dragImageForRowsWithIndexes_tableColumns_event_offset_()
	causes "PyObjCPointer created" error
In-Reply-To: <dv4evq$or8$1@sea.gmane.org>
References: <dv4evq$or8$1@sea.gmane.org>
Message-ID: <3FA4207C-9311-4714-B528-63974FFBC6FA@mac.com>


On 13-mrt-2006, at 19:51, Michael Glassford wrote:

> I added a dragImageForRowsWithIndexes_tableColumns_event_offset_()
> method to an NSTableView subclass that I have created. It's working  
> OK,
> but every time it is called, a message like this is printed to the  
> error
> log:
>
> "PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
> dragImageForRows_event_dragImageOffset_()"

That means I have to spent some quality time with PyObjC :-). One of the
arguments of this method is an input/output argument and I haven't added
the required annotations to PyObjC yet.

>
> I had a similar problem with
> tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which
> caused a crash on some machines. It was fixed by changing the  
> signature
> of that function in AppKit/protocols.py. I looked for a way to change
> the signature of  
> dragImageForRowsWithIndexes_tableColumns_event_offset_,
> but couldn't find it anywhere.

You can add a simular defintion for  
dragImageForRows_event_dragImageOffset_
in the same file.

Ronald

>
> Suggestions?
>
> Mike
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 14 21:30:20 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:30:20 +0100
Subject: [Pythonmac-SIG] Python (2.3, 2.4...) + XCode (2.1,
	2.2) + Extensions...
In-Reply-To: <80498F04-735E-4B36-96DD-E73F6F648129@lmd.polytechnique.fr>
References: <80498F04-735E-4B36-96DD-E73F6F648129@lmd.polytechnique.fr>
Message-ID: <BA32D2FF-BDDF-4BC7-8073-73E66F73E402@mac.com>


On 14-mrt-2006, at 17:04, Hubert Holin wrote:

> Paris (U.E.), le 14/03/2006
>
> 	Bonjour
>
> 		Back when XCode 2.1 was current (no *so* long ago), I used it to
> build extensions for Python (first Apple's 2.3, then Bob's "official
> unofficial'" 2.4). Finding the right combination of flags to make the
> thing work was not funny.
>
> 		Then XCode 2.2 came along, which was fine for some C++ work.
>
> 		I have come to revisit some older XCode projects, which included
> extensions to Python, and, lo and behold, Apple has played "mix the
> flags" again! This means I can't build new projects which includes
> custom extensions to Python using only *documented* flags, unless I
> once again go into a hair-tearing flag hunting session... Unless
> somebody else has done that and is willing to share the result of his
> labor.

I use distutils for building python extensions, which means Python  
will provide
the right compiler flags for me ;-). PyObjC includes Xcode templates  
for Python
based applications, you could use this as a starting point for using  
distutils to
build python extensions from an Xcode project.

Ronald

>
> 		For the record, some of the projects I was revisiting were for
> getting my feet wet with MacTels during an Apple Workshop (Universal
> Binaries) in Paris, Friday. Guess I'll have some interesting
> questions to ask them...
>
> 	Merci
>
> Hubert Holin
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 14 21:34:07 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:34:07 +0100
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar14.083641pst.58633@synergy1.parc.xerox.com>
References: <06Mar14.083641pst.58633@synergy1.parc.xerox.com>
Message-ID: <3946CA2C-0E36-45EE-AF3A-96C66AD54052@mac.com>


On 14-mrt-2006, at 17:36, Bill Janssen wrote:

>> If it influences the answers, I am looking to build a cross-platform
>> application that I eventually want to be able to package for easy
>> installation by non-Python savvy users.
>
> I don't see a good alternative to using Apple's version, then.
>
> Or, at least, I don't know how to build a Mac installer that installs
> a complete Python along with the other code it needs.  If there's a
> cookbook way of doing that, I'd be interested, too.

The depends on what the application is. If the application is a GUI
application you can use py2app to build an application bundle, that
will include the python framework inside the application (unless you
use Apple's python).

Building a .pkg is also quite easy, see Mac/OSX/Dist in the python24-fat
tree for the code that is used to build the univeral binary installer.

Ronald

From ronaldoussoren at mac.com  Tue Mar 14 21:38:51 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:38:51 +0100
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
References: <p06230901c03c79492502@[192.168.2.234]>
	<D0BA657B-C02B-41BA-983A-A70353FCCD70@redivi.com>
	<p06230903c03c88cbc797@[192.168.2.234]>
	<2C0098B5-D254-499D-83DE-F2C8A72169F3@redivi.com>
	<E77D4FF1-98F3-492C-A388-3D05D99E80A2@conncoll.edu>
Message-ID: <D879CAEA-1E30-42AE-8886-D45DD7BF1F48@mac.com>


On 14-mrt-2006, at 17:50, Charles Hartman wrote:

> I'm still not *quite* clear on the interactions here.
>
> 1. wxPython still isn't Universal, though there are rumblings in the
> distance. Until that happens, is it possible to use the emerging
> Universal build of Python with the PPC build of wxPython? I guess I
> mean, possible to make command-line Python scripts that call (PPC)
> wx, on an Intel machine? In general, I guess: does Python care how
> its libraries are built, as long as they're built to run *somehow* on
> the platform that Python is running on? (Is that even the right
> question??)

You can use native extensions with the universal build, but you cannot
use PPC extensions on Intel. To answer you're question: you're going to
need a universal or Intel build of wxPython to create wx gui's on an
intel mac.

Rosetta is quite good, but it cannot be used to emulate just part of
an application.

>
> 2. Until Universal py2app (thanks Bob!) is ready, it's not possible
> to build Mac standalone apps with Universal Python -- is that correct?
Right. You might be able to get away with rebuilding the application
launcher in py2app, but there's no guarantee that this will work.

>
> 3. (the combination) When Universal py2app is ready -- long before
> Universal wxPython, I'll bet -- will it be possible to build
> standalone apps for Mac (PPC and Intel) using (PPC) wxPython?
No, see answer 1.

Ronald

>
> Sorry to be so foggy-headed about the innards of this. It isn't
> urgent, since *everything* seems to work fine in the PPC versions on
> an Intel Mac.
>
> Charles
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 14 21:41:24 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:41:24 +0100
Subject: [Pythonmac-SIG] resizing terminal and curses
In-Reply-To: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>
References: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>
Message-ID: <C9FD21B1-D825-4323-AAA3-D8D5C14936BB@mac.com>


On 14-mrt-2006, at 19:00, Nicholas Cole wrote:

> I couldn't find anything in the archives related to this.
>
> Using both Apple's python and a build of python2.4.2, there seems to
> be a bug when using ncurses.  curses.newwin() is supposed to give a
> new window filling the pysical terminal, but if the terminal has been
> resized this does not seem to happen.  Instead, the window seems to be
> the size that the terminal was when curses was initialised.  This is
> not the case on other platforms, where the window created reflects the
> new size of the terminal.  Is this a known bug?

I dont known, I don't use curses. Could you provide a small example  
program?

Ronald
>
> (it seems to affect both xterm and terminal.app)
>
> Best wishes,
>
> N.
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 14 21:52:50 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 14 Mar 2006 21:52:50 +0100
Subject: [Pythonmac-SIG] PyObjC:
	dragImageForRowsWithIndexes_tableColumns_event_offset_()
	causes "PyObjCPointer created" error
In-Reply-To: <dv4evq$or8$1@sea.gmane.org>
References: <dv4evq$or8$1@sea.gmane.org>
Message-ID: <9875F2D6-E758-44A6-B870-ED806AC16732@mac.com>


On 13-mrt-2006, at 19:51, Michael Glassford wrote:

> I added a dragImageForRowsWithIndexes_tableColumns_event_offset_()
> method to an NSTableView subclass that I have created. It's working  
> OK,
> but every time it is called, a message like this is printed to the  
> error
> log:
>
> "PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
> dragImageForRows_event_dragImageOffset_()"
>
> I had a similar problem with
> tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which
> caused a crash on some machines. It was fixed by changing the  
> signature
> of that function in AppKit/protocols.py. I looked for a way to change
> the signature of  
> dragImageForRowsWithIndexes_tableColumns_event_offset_,
> but couldn't find it anywhere.

Add this call to AppKit/_AppKitSignatures.py:

setSignatureForSelector(
	"NSTableView",
	"dragImageForRows:event:dragImageOffset:",
	"@@:@@N^{_NSPoint=ff}")

I haven't tested this addition yet, but It Should Work(TM).

Ronald

From glassfordm at hotmail.com  Tue Mar 14 21:48:40 2006
From: glassfordm at hotmail.com (Michael Glassford)
Date: Tue, 14 Mar 2006 15:48:40 -0500
Subject: [Pythonmac-SIG] PyObjC:
 dragImageForRowsWithIndexes_tableColumns_event_offset_() causes
 "PyObjCPointer created" error
In-Reply-To: <3FA4207C-9311-4714-B528-63974FFBC6FA@mac.com>
References: <dv4evq$or8$1@sea.gmane.org>
	<3FA4207C-9311-4714-B528-63974FFBC6FA@mac.com>
Message-ID: <44172C28.6070605@hotmail.com>

Ronald Oussoren wrote:
> On 13-mrt-2006, at 19:51, Michael Glassford wrote:
> 
>> I added a dragImageForRowsWithIndexes_tableColumns_event_offset_()
>> method to an NSTableView subclass that I have created. It's working  
>> OK,
>> but every time it is called, a message like this is printed to the  
>> error
>> log:
>>
>> "PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
>> dragImageForRows_event_dragImageOffset_()"
> 
> That means I have to spent some quality time with PyObjC :-). One of the
> arguments of this method is an input/output argument and I haven't added
> the required annotations to PyObjC yet.
> 
>> I had a similar problem with
>> tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which
>> caused a crash on some machines. It was fixed by changing the  
>> signature
>> of that function in AppKit/protocols.py. I looked for a way to change
>> the signature of  
>> dragImageForRowsWithIndexes_tableColumns_event_offset_,
>> but couldn't find it anywhere.
> 
> You can add a simular defintion for  
> dragImageForRows_event_dragImageOffset_
> in the same file.

I did: it has the same problem, and I couldn't find a way to fix that 
either. Did I miss something?

Mike

> Ronald
> 
>> Suggestions?
>>
>> Mike
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
> 
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
> 


From glassfordm at hotmail.com  Tue Mar 14 23:30:33 2006
From: glassfordm at hotmail.com (Michael Glassford)
Date: Tue, 14 Mar 2006 17:30:33 -0500
Subject: [Pythonmac-SIG] PyObjC:
 dragImageForRowsWithIndexes_tableColumns_event_offset_() causes
 "PyObjCPointer created" error
In-Reply-To: <9875F2D6-E758-44A6-B870-ED806AC16732@mac.com>
References: <dv4evq$or8$1@sea.gmane.org>
	<9875F2D6-E758-44A6-B870-ED806AC16732@mac.com>
Message-ID: <dv7g6a$4eh$1@sea.gmane.org>

Ronald Oussoren wrote:
> On 13-mrt-2006, at 19:51, Michael Glassford wrote:
> 
>> I added a dragImageForRowsWithIndexes_tableColumns_event_offset_()
>> method to an NSTableView subclass that I have created. It's working  
>> OK,
>> but every time it is called, a message like this is printed to the  
>> error
>> log:
>>
>> "PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
>> dragImageForRows_event_dragImageOffset_()"
>>
>> I had a similar problem with
>> tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which
>> caused a crash on some machines. It was fixed by changing the  
>> signature
>> of that function in AppKit/protocols.py. I looked for a way to change
>> the signature of  
>> dragImageForRowsWithIndexes_tableColumns_event_offset_,
>> but couldn't find it anywhere.
> 
> Add this call to AppKit/_AppKitSignatures.py:
> 
> setSignatureForSelector(
> 	"NSTableView",
> 	"dragImageForRows:event:dragImageOffset:",
> 	"@@:@@N^{_NSPoint=ff}")
> 
> I haven't tested this addition yet, but It Should Work(TM).

Thanks, it did work. I also added:

     _objc.setSignatureForSelector(
         "NSTableView",
         "dragImageForRowsWithIndexes:tableColumns:event:offset:",
         "@@:@@@N^{_NSPoint=ff}"
         )

Mike

> Ronald
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
> 


From ronaldoussoren at mac.com  Wed Mar 15 07:41:48 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 15 Mar 2006 07:41:48 +0100
Subject: [Pythonmac-SIG] PyObjC:
	dragImageForRowsWithIndexes_tableColumns_event_offset_()
	causes "PyObjCPointer created" error
In-Reply-To: <dv7g6a$4eh$1@sea.gmane.org>
References: <dv4evq$or8$1@sea.gmane.org>
	<9875F2D6-E758-44A6-B870-ED806AC16732@mac.com>
	<dv7g6a$4eh$1@sea.gmane.org>
Message-ID: <6F04EBE6-9A79-4DE2-B675-2CB768F2806B@mac.com>


On 14-mrt-2006, at 23:30, Michael Glassford wrote:

>
>>
>> Add this call to AppKit/_AppKitSignatures.py:
>>
>> setSignatureForSelector(
>> 	"NSTableView",
>> 	"dragImageForRows:event:dragImageOffset:",
>> 	"@@:@@N^{_NSPoint=ff}")
>>
>> I haven't tested this addition yet, but It Should Work(TM).
>
> Thanks, it did work. I also added:
>
>      _objc.setSignatureForSelector(
>          "NSTableView",
>          "dragImageForRowsWithIndexes:tableColumns:event:offset:",
>          "@@:@@@N^{_NSPoint=ff}"
>          )


Thanks for the confirmation. I've added these to the repository
and will be checking for other methods that need annotations in the
coming time.

Ronald

>
> Mike
>
>> Ronald
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From janssen at parc.com  Wed Mar 15 12:15:28 2006
From: janssen at parc.com (Bill Janssen)
Date: Wed, 15 Mar 2006 03:15:28 PST
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: Your message of "Tue, 14 Mar 2006 11:28:35 PST."
	<44171963.5090305@noaa.gov> 
Message-ID: <06Mar15.031537pst."58633"@synergy1.parc.xerox.com>

> I'm not all that optimistic that it will be ready that soon, or
> look&feel very Mac-ish.

I don't care about that -- I don't use the built-in widgets anyway.  I
use Java for UI largely because (1) widgets built from Swing will run in
applets as well as in apps, and (2) because the graphics model is
cairo-rich.  The last time I looked at wxWindows, it didn't satisfy
either of these criteria.  Perhaps for a future project.

See http://www.parc.com/janssen/pubs/TR-05-3.pdf for more than you
want to know about my application.

> What you have now 
> will most likely break if the user upgrades OS-X, as Apple is likely to 
> upgrade their python

I don't care.  That's a rare enough occurrence, and understood by
users to be a big enough occurrence, that no one blinks an eye if I
tell them to re-install if they do that.

> I'd like to be able to build a "Python Runtime" that included all of the 
> Framework build, plus the full set of modules that I need for my apps. 
> then I'd like to be able to run command line apps against it, and use 
> Py2Applet (no, it doesn't exist), to build Application bundles that 
> relied on that runtime. This seems to me to be  a good solution for 
> folks like us that are distributing not one app, but a bunch of apps 
> that all depend on the same stuff.

Yes, I agree, that's a good idea.  I'd probably still rely on the
system Python for my particular application, because it don't need the
latest and greatest version of Python, but I suppose there are
applications that might.

Bill


From janssen at parc.com  Wed Mar 15 12:20:12 2006
From: janssen at parc.com (Bill Janssen)
Date: Wed, 15 Mar 2006 03:20:12 PST
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: Your message of "Tue, 14 Mar 2006 12:34:07 PST."
	<3946CA2C-0E36-45EE-AF3A-96C66AD54052@mac.com> 
Message-ID: <06Mar15.032022pst."58633"@synergy1.parc.xerox.com>

> The depends on what the application is. If the application is a GUI
> application you can use py2app to build an application bundle, that
> will include the python framework inside the application (unless you
> use Apple's python).

What's the "framework"?  If that's the entire Python interpreter and
library, installed in a way that can't interfere with whatever the
user is using for whatever *other* Python needs they have on their
machine -- that's what I'd need if I were to use something other than
the system Python.

> Building a .pkg is also quite easy, see Mac/OSX/Dist in the python24-fat
> tree for the code that is used to build the univeral binary installer.

Yes, I've got my own python program that builds .pkg installers.

Bill

From ronaldoussoren at mac.com  Wed Mar 15 15:45:21 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 15 Mar 2006 15:45:21 +0100
Subject: [Pythonmac-SIG] Recurring question - which python should I use?
In-Reply-To: <06Mar15.032022pst.58633@synergy1.parc.xerox.com>
References: <06Mar15.032022pst.58633@synergy1.parc.xerox.com>
Message-ID: <16327217.1142433921595.JavaMail.ronaldoussoren@mac.com>

 
On Wednesday, March 15, 2006, at 12:20PM, Bill Janssen <janssen at parc.com> wrote:

>> The depends on what the application is. If the application is a GUI
>> application you can use py2app to build an application bundle, that
>> will include the python framework inside the application (unless you
>> use Apple's python).
>
>What's the "framework"?  If that's the entire Python interpreter and
>library, installed in a way that can't interfere with whatever the
>user is using for whatever *other* Python needs they have on their
>machine -- that's what I'd need if I were to use something other than
>the system Python.

And that is what py2app does. It will add the subsection of Python.framework
that is needed for your application (based on the contents of setup.py and
static analysis of the application) into the app bundle. Loader commands of
binaries in the app bundle are rewriten to make sure they refer to the framework
(and other libraries) inside the bundle instead of versions outside of the bundle.

Please check out py2app, it's documentation might be lacking on some points,
but not w.r.t. its technical capabilities.

>
>> Building a .pkg is also quite easy, see Mac/OSX/Dist in the python24-fat
>> tree for the code that is used to build the univeral binary installer.
>
>Yes, I've got my own python program that builds .pkg installers.

Ronald
>
>Bill
>
>

From glassfordm at gmail.com  Tue Mar 14 21:48:40 2006
From: glassfordm at gmail.com (Michael Glassford)
Date: Tue, 14 Mar 2006 15:48:40 -0500
Subject: [Pythonmac-SIG] PyObjC:
 dragImageForRowsWithIndexes_tableColumns_event_offset_() causes
 "PyObjCPointer created" error
In-Reply-To: <3FA4207C-9311-4714-B528-63974FFBC6FA@mac.com>
References: <dv4evq$or8$1@sea.gmane.org>
	<3FA4207C-9311-4714-B528-63974FFBC6FA@mac.com>
Message-ID: <44172C28.6070605@hotmail.com>

Ronald Oussoren wrote:
> On 13-mrt-2006, at 19:51, Michael Glassford wrote:
> 
>> I added a dragImageForRowsWithIndexes_tableColumns_event_offset_()
>> method to an NSTableView subclass that I have created. It's working  
>> OK,
>> but every time it is called, a message like this is printed to the  
>> error
>> log:
>>
>> "PyObjCPointer created: at 0xbfffcb50 of type {_NSPoint=ff}16
>> dragImageForRows_event_dragImageOffset_()"
> 
> That means I have to spent some quality time with PyObjC :-). One of the
> arguments of this method is an input/output argument and I haven't added
> the required annotations to PyObjC yet.
> 
>> I had a similar problem with
>> tableView_toolTipForCell_rect_tableColumn_row_mouseLocation_(), which
>> caused a crash on some machines. It was fixed by changing the  
>> signature
>> of that function in AppKit/protocols.py. I looked for a way to change
>> the signature of  
>> dragImageForRowsWithIndexes_tableColumns_event_offset_,
>> but couldn't find it anywhere.
> 
> You can add a simular defintion for  
> dragImageForRows_event_dragImageOffset_
> in the same file.

I did: it has the same problem, and I couldn't find a way to fix that 
either. Did I miss something?

Mike

> Ronald
> 
>> Suggestions?
>>
>> Mike
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
> 
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
> 


From danielculver at mac.com  Wed Mar 15 02:40:18 2006
From: danielculver at mac.com (Daniel Culver)
Date: Tue, 14 Mar 2006 17:40:18 -0800
Subject: [Pythonmac-SIG] (no subject)
Message-ID: <8721DEF8-E743-4345-B84E-081AB37A6EAB@mac.com>

Hi.
I am using Inkscape on OS 10.4 w/ X11 and many of the effects rely on  
Python, but I am having trouble arranging the framework. I am getting  
back an error; cannot import name boolean. There is a file in the  
directory called 'boolean.c' is this correct and if so why won't it  
run. If not correct how and where can I get the right one.
Thank you for any help.
Daniel Culver
danielculver at mac.com




From gic at cns.montana.edu  Thu Mar 16 03:01:27 2006
From: gic at cns.montana.edu (Graham Cummins)
Date: Wed, 15 Mar 2006 19:01:27 -0700
Subject: [Pythonmac-SIG] Universal binary installer
Message-ID: <1D312B2E-92E0-4D80-AE44-08C6AC0A4EC3@cns.montana.edu>

I recently downloaded Ronald Doussoren's universal binary installer  
for MacPython. This installed fine on my Macbook Pro, and the  
resulting python version was able to build most of my favorite  
extensions (except PyOpenGL, which I can't get to build on any Mac  
recently - I'll post a separate issue about that.).

I also compiled the source from the universal svn tree (revision 41).  
This required that I edit
Lib/distutils/unixccompiler.py as follows:

--- python/Lib/distutils/unixccompiler.py       (revision 41)
+++ python/Lib/distutils/unixccompiler.py       (working copy)
@@ -42,8 +42,11 @@
#     should just happily stuff them into the preprocessor/compiler/ 
linker
#     options and carry on.
+
def _darwin_compiler(compiler_so, cc_args):
      compiler_so = list(compiler_so)
+    stripArch=0
+    stripSysroot=0
      if os.uname()[2] < '8.':
          stripArch = stripSysroot = 1

This just clears up a bug where some variables can end up undefined  
if an if condition comes up false. After this modification, the  
source complied fine with --enable-framework. I didn't use --enable- 
universal-sdk, so I guess I compiled an Intel-only version of the  
framework. I then built some extensions for this version also.

My reason to comment here has to do with the relevant performance of  
the Universal vs locally compiled pythons. In particular, I make  
heavy use of numarray, so I have a standard benchmark that tests many  
of the most computation intensive numarray routines with a variety of  
different data types. According to this benchmark, I'm getting much  
(>3X) better performance out of the local version than out of the  
Universal one.

For both python frameworks, I built numarray 1.5.1 using the basic  
"python setup.py install" (starting with clean source). The  
benchmarks I got were (in seconds to completion) about 24 seconds for  
the Universal, and only 7.2 seconds for the locally compiled python.  
For comparison, the older PPC only MacPython 2.4.1, with numarray  
installed via the included package manager took 32.6 seconds.

The native code on the MacBook compares very well to other machines.  
Native code on my dual G5 takes 8.4 seconds on this task. The only  
machine I've seen that's as fast as this MacBook was an SGI Altix 330  
(Itanium 2), and even it wasn't any faster. This makes me pretty  
happy about the Intel Core Duo, but somewhat worried about Universal  
binaries (in general, but for python in particular) since the binary  
seems closer in performance to rosetta than to native code.



From bob at redivi.com  Thu Mar 16 04:43:00 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 15 Mar 2006 19:43:00 -0800
Subject: [Pythonmac-SIG] Universal binary installer
In-Reply-To: <1D312B2E-92E0-4D80-AE44-08C6AC0A4EC3@cns.montana.edu>
References: <1D312B2E-92E0-4D80-AE44-08C6AC0A4EC3@cns.montana.edu>
Message-ID: <58B04051-82FF-41DD-A718-65D64D1ED958@redivi.com>


On Mar 15, 2006, at 6:01 PM, Graham Cummins wrote:

> I recently downloaded Ronald Doussoren's universal binary installer
> for MacPython. This installed fine on my Macbook Pro, and the
> resulting python version was able to build most of my favorite
> extensions (except PyOpenGL, which I can't get to build on any Mac
> recently - I'll post a separate issue about that.).
>
> I also compiled the source from the universal svn tree (revision 41).
> This required that I edit
> Lib/distutils/unixccompiler.py as follows:
>
> --- python/Lib/distutils/unixccompiler.py       (revision 41)
> +++ python/Lib/distutils/unixccompiler.py       (working copy)
> @@ -42,8 +42,11 @@
> #     should just happily stuff them into the preprocessor/compiler/
> linker
> #     options and carry on.
> +
> def _darwin_compiler(compiler_so, cc_args):
>       compiler_so = list(compiler_so)
> +    stripArch=0
> +    stripSysroot=0
>       if os.uname()[2] < '8.':
>           stripArch = stripSysroot = 1
>
> This just clears up a bug where some variables can end up undefined
> if an if condition comes up false. After this modification, the
> source complied fine with --enable-framework. I didn't use --enable-
> universal-sdk, so I guess I compiled an Intel-only version of the
> framework. I then built some extensions for this version also.
>
> My reason to comment here has to do with the relevant performance of
> the Universal vs locally compiled pythons. In particular, I make
> heavy use of numarray, so I have a standard benchmark that tests many
> of the most computation intensive numarray routines with a variety of
> different data types. According to this benchmark, I'm getting much
> (>3X) better performance out of the local version than out of the
> Universal one.

I don't believe that.  3X performance different can only be explained  
by Rosetta, not universal vs. i386-only.  I think you were probably  
running the benchmark on 2.4.1 or something.

> For both python frameworks, I built numarray 1.5.1 using the basic
> "python setup.py install" (starting with clean source). The
> benchmarks I got were (in seconds to completion) about 24 seconds for
> the Universal, and only 7.2 seconds for the locally compiled python.
> For comparison, the older PPC only MacPython 2.4.1, with numarray
> installed via the included package manager took 32.6 seconds.
>
> The native code on the MacBook compares very well to other machines.
> Native code on my dual G5 takes 8.4 seconds on this task. The only
> machine I've seen that's as fast as this MacBook was an SGI Altix 330
> (Itanium 2), and even it wasn't any faster. This makes me pretty
> happy about the Intel Core Duo, but somewhat worried about Universal
> binaries (in general, but for python in particular) since the binary
> seems closer in performance to rosetta than to native code.

I blame the tests.  Make sure to check platform.mac_ver()[2] to make  
sure that it's actually running the version that's compiled for  
i386.  Perhaps you had a lingering python executable that was  
compiled PPC only that was executing instead of the i386 version.  It  
could've still linked against the universal framework, but would've  
started up under Rosetta.

$ /usr/local/bin/python -c "import platform; print platform.mac_ver() 
[2]; import test.pystone; test.pystone.main(50000)"
i386
Pystone(1.1) time for 50000 passes = 1.93
This machine benchmarks at 25906.7 pystones/second

$ /usr/libexec/oah/oah750 /usr/local/bin/python -c "import platform;  
print platform.mac_ver()[2]; import test.pystone; test.pystone.main 
(50000)"
PowerPC
Pystone(1.1) time for 50000 passes = 5.27
This machine benchmarks at 9487.67 pystones/second

The difference in pystones between PPC and i386 is ~2.73, which  
roughly proves that you were benchmarking the universal build under  
PPC emulation.

-bob


From M.Laloux at mrw.wallonie.be  Thu Mar 16 11:43:59 2006
From: M.Laloux at mrw.wallonie.be (LALOUX Martin)
Date: Thu, 16 Mar 2006 11:43:59 +0100
Subject: [Pythonmac-SIG] Veusz, Pyqt
Message-ID: <6.1.2.0.0.20060316113428.02d33b78@pop.promibra.intra.mrw.wallonie.be>

I discover the scientific package written in python, Veusz  in macupdate
http://www.macupdate.com/info.php/id/21029
The package is very interressant in oneself  but, moreover, it installs 
a  fully operational version of Pyqt in MacPython 2.4.1
"The installer installs Veusz 0.9 and the required PyQt 3.1.5 (incl. SIP 
4.3.1) - replacing at least in parts an already existing PyQt installation!
REQUIREMENTS
Mac OS X 10.3 or later, MacPython 2.4.x, NumArray (python module) AND 
Trolltech?s Qt/Mac 3.3.x framework on your computer! (download bookmarks 
are provided with the installer)"
Martin Laloux


From ronaldoussoren at mac.com  Thu Mar 16 12:43:44 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 16 Mar 2006 12:43:44 +0100
Subject: [Pythonmac-SIG] Veusz, Pyqt
In-Reply-To: <6.1.2.0.0.20060316113428.02d33b78@pop.promibra.intra.mrw.wallonie.be>
References: <6.1.2.0.0.20060316113428.02d33b78@pop.promibra.intra.mrw.wallonie.be>
Message-ID: <397B1D63-21FF-4EF3-87C5-2C4CFA94172E@mac.com>

Shees, how lame can one get.  It is not that hard to use py2ap or  
even macho_standalone
to create an application bundle that doesn't need to place files all  
over the place.

Don't install this package, it will replace files in /Library/ 
Frameworks/Python.framework.

Ronald

On 16-mrt-2006, at 11:43, LALOUX Martin wrote:

> I discover the scientific package written in python, Veusz  in  
> macupdate
> http://www.macupdate.com/info.php/id/21029
> The package is very interressant in oneself  but, moreover, it  
> installs
> a  fully operational version of Pyqt in MacPython 2.4.1
> "The installer installs Veusz 0.9 and the required PyQt 3.1.5  
> (incl. SIP
> 4.3.1) - replacing at least in parts an already existing PyQt  
> installation!
> REQUIREMENTS
> Mac OS X 10.3 or later, MacPython 2.4.x, NumArray (python module) AND
> Trolltech?s Qt/Mac 3.3.x framework on your computer! (download  
> bookmarks
> are provided with the installer)"
> Martin Laloux
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Thu Mar 16 17:04:04 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 16 Mar 2006 17:04:04 +0100
Subject: [Pythonmac-SIG] Veusz, Pyqt
In-Reply-To: <397B1D63-21FF-4EF3-87C5-2C4CFA94172E@mac.com>
References: <6.1.2.0.0.20060316113428.02d33b78@pop.promibra.intra.mrw.wallonie.be>
	<397B1D63-21FF-4EF3-87C5-2C4CFA94172E@mac.com>
Message-ID: <7B437D5E-BA1D-4844-98C4-2C6CA771671D@mac.com>


On 16-mrt-2006, at 12:43, Ronald Oussoren wrote:

> Shees, how lame can one get.  It is not that hard to use py2ap or
> even macho_standalone
> to create an application bundle that doesn't need to place files all
> over the place.
>
> Don't install this package, it will replace files in /Library/
> Frameworks/Python.framework.

Never mind, I'm an idiot. It will replace files in /Library/ 
Frameworks/Python.framework,
but only those related to PyQt and Veusz itself.

Ronald

>
> Ronald
>
> On 16-mrt-2006, at 11:43, LALOUX Martin wrote:
>
>> I discover the scientific package written in python, Veusz  in
>> macupdate
>> http://www.macupdate.com/info.php/id/21029
>> The package is very interressant in oneself  but, moreover, it
>> installs
>> a  fully operational version of Pyqt in MacPython 2.4.1
>> "The installer installs Veusz 0.9 and the required PyQt 3.1.5
>> (incl. SIP
>> 4.3.1) - replacing at least in parts an already existing PyQt
>> installation!
>> REQUIREMENTS
>> Mac OS X 10.3 or later, MacPython 2.4.x, NumArray (python module) AND
>> Trolltech?s Qt/Mac 3.3.x framework on your computer! (download
>> bookmarks
>> are provided with the installer)"
>> Martin Laloux
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From stewart.midwinter at gmail.com  Thu Mar 16 17:44:29 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Thu, 16 Mar 2006 09:44:29 -0700
Subject: [Pythonmac-SIG] py2app question
Message-ID: <d4c0d4800603160844o3199769fgcfd4c544d0a45fad@mail.gmail.com>

I have an application I'd like to have located on a USB stick, and be
able to run in a self-contained manner on any PC it's connected to,
whether running OS X or Windows.   Can I use py2app to help create
this stand-alone distribution?

Or, could I count on Python always being installed on any OS X -
equipped PC, and thus not need to install a separate copy on my USB
stick in order to be able to run my app?

thanks
S

From delza at livingcode.org  Thu Mar 16 17:58:21 2006
From: delza at livingcode.org (Dethe Elza)
Date: Thu, 16 Mar 2006 08:58:21 -0800
Subject: [Pythonmac-SIG] Fwd:  py2app question
In-Reply-To: <24d517dd0603160856h72351cebla3a3a35a99a4a3dd@mail.gmail.com>
References: <d4c0d4800603160844o3199769fgcfd4c544d0a45fad@mail.gmail.com>
	<24d517dd0603160856h72351cebla3a3a35a99a4a3dd@mail.gmail.com>
Message-ID: <24d517dd0603160858y54914de4j58db24bd27e1ebc7@mail.gmail.com>

Forgot to respond to list in my reply:

On 3/16/06, Stewart Midwinter <stewart.midwinter at gmail.com> wrote:
> I have an application I'd like to have located on a USB stick, and be
> able to run in a self-contained manner on any PC it's connected to,
> whether running OS X or Windows.   Can I use py2app to help create
> this stand-alone distribution?

There's no way to create an entirely self-contained application that
runs on either Windows or OS X, there are just too many differences in
how they are built and bootstrapped.  But py2app will let you build a
self-contained application which will run off of a memory stick.  You
might be able to use filesystem tricks to hide the OS X application on
Windows and the Windows application on OS X, to give the illusion that
there is only one application (some CD ROMs do this, I believe).

> Or, could I count on Python always being installed on any OS X -
> equipped PC, and thus not need to install a separate copy on my USB
> stick in order to be able to run my app?

You cannot count on the version of Python that will be installed, or
any external libraries you depend on.  The best thing is to bundle it
all into your application using py2app.

--Dethe

>
> thanks
> S
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig
>

From gic at cns.montana.edu  Thu Mar 16 17:53:27 2006
From: gic at cns.montana.edu (Graham Cummins)
Date: Thu, 16 Mar 2006 09:53:27 -0700
Subject: [Pythonmac-SIG] PyOpenGL Mac
Message-ID: <FCC14F8C-038F-4BB5-BBBF-DD2D69EA980A@cns.montana.edu>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I've been having a beast of a time trying to build PyOpenGL on Mac  
recently. My last successful build was 2.0.1.09, and currently I  
can't even repeat that.

If I remember correctly, I had to switch to gcc-3.3 to get that build  
to work. Currently, with gcc-4.0 on version 2.0.2.01 I get lots of  
errors of the form:

from src/interface/GL.3DFX._multisample.c:6:
/System/Library/Frameworks/Kernel.framework/Headers/sys/stat.h:225:  
error: field 'st_atimespec' has incomplete type

I'm trying to go back and repeat my build with gcc-3.3 on my powerpc  
mac, but even if I rediscover that secret, it won't help me on my new  
Intel mac, since gcc-3.3 doesn't target the Intel platform.

Anyway, if anyone knows how to build this package on a recent (OS  
10.4.5) Mac, especially an Intel Mac, please let me know. Thanks.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)

iD8DBQFEGZgPm4P5KE2ufQoRAqQbAKCAvM94rS7PKr4PEVekeauNZ+j3kACgrf4d
+DhvOSOSsBx7mduJYFMUV1I=
=Jv2G
-----END PGP SIGNATURE-----

From bob at redivi.com  Thu Mar 16 19:41:18 2006
From: bob at redivi.com (Bob Ippolito)
Date: Thu, 16 Mar 2006 10:41:18 -0800
Subject: [Pythonmac-SIG] Fwd:  py2app question
In-Reply-To: <24d517dd0603160858y54914de4j58db24bd27e1ebc7@mail.gmail.com>
References: <d4c0d4800603160844o3199769fgcfd4c544d0a45fad@mail.gmail.com>
	<24d517dd0603160856h72351cebla3a3a35a99a4a3dd@mail.gmail.com>
	<24d517dd0603160858y54914de4j58db24bd27e1ebc7@mail.gmail.com>
Message-ID: <F7505727-9274-4F70-AA02-B309B72ED1C4@redivi.com>


On Mar 16, 2006, at 8:58 AM, Dethe Elza wrote:

> Forgot to respond to list in my reply:
>
> On 3/16/06, Stewart Midwinter <stewart.midwinter at gmail.com> wrote:
>> I have an application I'd like to have located on a USB stick, and be
>> able to run in a self-contained manner on any PC it's connected to,
>> whether running OS X or Windows.   Can I use py2app to help create
>> this stand-alone distribution?
>
> There's no way to create an entirely self-contained application that
> runs on either Windows or OS X, there are just too many differences in
> how they are built and bootstrapped.  But py2app will let you build a
> self-contained application which will run off of a memory stick.  You
> might be able to use filesystem tricks to hide the OS X application on
> Windows and the Windows application on OS X, to give the illusion that
> there is only one application (some CD ROMs do this, I believe).

I don't think you can ISO format a memory stick.. at least not in a  
way that Windows would recognize...

>> Or, could I count on Python always being installed on any OS X -
>> equipped PC, and thus not need to install a separate copy on my USB
>> stick in order to be able to run my app?
>
> You cannot count on the version of Python that will be installed, or
> any external libraries you depend on.  The best thing is to bundle it
> all into your application using py2app.

py2exe does something similar to py2app, but for Windows.  Just do it  
on both systems, then drop them both on the memory stick.  They won't  
be sharing any files, but the only thing that they would be able to  
share is python bytecode, and that's insignificant compared to the  
size of Python and the compiled extensions.

-bob


From gic at cns.montana.edu  Fri Mar 17 02:19:01 2006
From: gic at cns.montana.edu (Graham Cummins)
Date: Thu, 16 Mar 2006 18:19:01 -0700
Subject: [Pythonmac-SIG] PyOpenGL Mac
In-Reply-To: <FCC14F8C-038F-4BB5-BBBF-DD2D69EA980A@cns.montana.edu>
References: <FCC14F8C-038F-4BB5-BBBF-DD2D69EA980A@cns.montana.edu>
Message-ID: <2FCDA336-FD47-42BE-975F-982CA35B1879@cns.montana.edu>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

- -----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On Mar 16, 2006, at 9:53 AM, Graham Cummins wrote:
>
> I've been having a beast of a time trying to build PyOpenGL on Mac
> recently. My last successful build was 2.0.1.09, and currently I
> can't even repeat that.
>
> from src/interface/GL.3DFX._multisample.c:6:
> /System/Library/Frameworks/Kernel.framework/Headers/sys/stat.h:225:
> error: field 'st_atimespec' has incomplete type

Never mind, I've answered my own question. This error is caused by  
having /System/Library/Frameworks/Kernel.frameworks/Headers in the  
"include_dirs" defined in config/darwin.cfg. It seems that this  
causes Python.h to include the wrong stat.h. Removing that line fixes  
that issue.

  I got some other errors from files using #include <GL/glext.h>, but  
I was able to solve those with a (very) nasty hack of symlinking / 
System/Library/Frameworks/OpenGL.framework/Headers into itself under  
the alternate name "GL".

I was still not able to compile 2.1.2.01, but I with these two  
modifications I could compile 2.0.1.09 for intel macintosh using  
gcc-4.0, which is good enough for me for now.


- -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)

iD8DBQFEGg5Dm4P5KE2ufQoRAnK7AJ9KfcCmJswfuild01nzsIRSLEpvpgCeO2an
VmGG0Zu4Q0td7+mPKJ24UNM=
=aDYp
- -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)

iD8DBQFEGg6Gm4P5KE2ufQoRApY6AKCCnKhqrsRrYbgcyOabKA3d0oy4PwCeI7BT
45GzWqMiRwSNeIYjnoMWVz0=
=x7kB
-----END PGP SIGNATURE-----

From jreese99 at verizon.net  Fri Mar 17 03:43:08 2006
From: jreese99 at verizon.net (James Reese)
Date: Thu, 16 Mar 2006 21:43:08 -0500
Subject: [Pythonmac-SIG] Gdbm with Python 2.4.1?
Message-ID: <F8A70207-EE6E-484B-9DAF-13284DD9C294@verizon.net>

Can anyone tell me how to get Python 2.4.1 to support gdbm short of  
building it from source? Any hints will be greatly appreciated.

Jim

From bob at redivi.com  Fri Mar 17 03:51:48 2006
From: bob at redivi.com (Bob Ippolito)
Date: Thu, 16 Mar 2006 18:51:48 -0800
Subject: [Pythonmac-SIG] Gdbm with Python 2.4.1?
In-Reply-To: <F8A70207-EE6E-484B-9DAF-13284DD9C294@verizon.net>
References: <F8A70207-EE6E-484B-9DAF-13284DD9C294@verizon.net>
Message-ID: <CCDE6FC2-AAB6-49DD-BC21-332AADE3653E@redivi.com>


On Mar 16, 2006, at 6:43 PM, James Reese wrote:

> Can anyone tell me how to get Python 2.4.1 to support gdbm short of
> building it from source? Any hints will be greatly appreciated.

If you want gdbm you will have to build from source (either extract  
the gdbmmodule.c, write a setup.py for it, and build that separately,  
or build all of Python).  There is no binary for gdbm available.

-bob


From mforget at mtotelecom.com  Fri Mar 17 21:22:32 2006
From: mforget at mtotelecom.com (Martin Forget)
Date: Fri, 17 Mar 2006 15:22:32 -0500
Subject: [Pythonmac-SIG] building pyobjc/pypoppengl/pygame on intel
Message-ID: <8C2B1711-D131-44CA-AD0F-532D383320F5@mtotelecom.com>


hello everyone,

i have been trying to build different python modules for macosx-intel  
(min-core-solo) with no luck.


my closest try was with ronald's universal binary of python 2.4.2 from:
http://homepage.mac.com/ronaldoussoren/

i then tried to build an SVN version of PyObjC
running:
"... /python2.4 build.py build"  works fine, the build finishes with  
no critical errors,

"python build.py test"   fails with "Illegal Error"

i tried a "build.py install" anyway which succeeds,

but when i run the following code:
import AppKit
from AppKit import *
app = NSApplication.sharedApplication()

i have an illegal error.

also, the pygame.init() also gives an illegal error... but i figure  
they are related.

(i was able to compile cvs/svn versions of pygame and pyopengl with  
"minor" guesswork to the code...)

my target is a pygame/pyobjc/pyopengl kitchen sink setup for a game  
like application in python.

note: my application works  fine on my powerbook-g4
and even in a ppc build of python and the packages from pythonmac.org/ 
packages/ using rosetta on the intel-mini.  (just very very slowly)

thanks in advance, you help is appreciated.

-mforget



From sw at wordtech-software.com  Sat Mar 18 19:03:17 2006
From: sw at wordtech-software.com (Kevin Walzer)
Date: Sat, 18 Mar 2006 13:03:17 -0500
Subject: [Pythonmac-SIG] Would bundlebuilder work "universally?"
Message-ID: <441C4B65.4070008@wordtech-software.com>

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I know py2app doesn't support universal builds right now...would
bundlebuilder?

- --
Kevin Walzer
iReveal: File Search Tool
http://www.wordtech-software.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.1 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFEHEtlJmdQs+6YVcoRAlogAJ96mrJT653PzSOvbFUiT66TVWNQOACgi7it
CTo2s81Z0rF2k86oo30971Q=
=JJXd
-----END PGP SIGNATURE-----

From just at letterror.com  Sat Mar 18 19:24:38 2006
From: just at letterror.com (Just van Rossum)
Date: Sat, 18 Mar 2006 19:24:38 +0100
Subject: [Pythonmac-SIG] Would bundlebuilder work "universally?"
In-Reply-To: <441C4B65.4070008@wordtech-software.com>
Message-ID: <r01050400-1039-77234297B6AC11DA951F001124365170@[10.0.0.24]>

Kevin Walzer wrote:

> I know py2app doesn't support universal builds right now...would
> bundlebuilder?

I don't think bundlebuilder supports anything at all currently...
bundlebuilder is dead, we should probably just remove it from svn.

Just

From stewart.midwinter at gmail.com  Sun Mar 19 01:22:14 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Sat, 18 Mar 2006 17:22:14 -0700
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )
Message-ID: <d4c0d4800603181622k2b5c398ar31c04a97100f75c3@mail.gmail.com>

I have a question on use of the tilde symbol (~) to access the current
user's home directory.

If you are in a bash shell, you can "cd ~" and be in the default
user's home directory.

I want my python app to be able to switch to the user's directory. 
But I can't use os.chdir('~') since Python doesn't understand the
tilde.   Nor can I do this:
file = os.path.join('~', filename)

What options do I have?

thanks
S

From kquirk at solidworks.com  Sun Mar 19 01:35:01 2006
From: kquirk at solidworks.com (Kent Quirk)
Date: Sat, 18 Mar 2006 19:35:01 -0500
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )
Message-ID: <AD85EF0CD3800548B3B22C832F08788E02944A28@CORP-MAIL6.solidworks.swk>

Try this:

import os.path
os.path.expanduser('~')

You might also find a use for
os.path.expandvars()

	Kent


-----Original Message-----
From: pythonmac-sig-bounces at python.org
[mailto:pythonmac-sig-bounces at python.org] On Behalf Of Stewart Midwinter
Sent: Saturday, March 18, 2006 7:22 PM
To: pythonmac-sig at python.org
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )

I have a question on use of the tilde symbol (~) to access the current
user's home directory.

If you are in a bash shell, you can "cd ~" and be in the default
user's home directory.

I want my python app to be able to switch to the user's directory. 
But I can't use os.chdir('~') since Python doesn't understand the
tilde.   Nor can I do this:
file = os.path.join('~', filename)

What options do I have?

thanks
S
_______________________________________________
Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
http://mail.python.org/mailman/listinfo/pythonmac-sig

From stewart.midwinter at gmail.com  Sun Mar 19 01:40:06 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Sat, 18 Mar 2006 17:40:06 -0700
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )
In-Reply-To: <AD85EF0CD3800548B3B22C832F08788E02944A28@CORP-MAIL6.solidworks.swk>
References: <AD85EF0CD3800548B3B22C832F08788E02944A28@CORP-MAIL6.solidworks.swk>
Message-ID: <d4c0d4800603181640m5e16c143yde3fe43143e78800@mail.gmail.com>

On 3/18/06, Kent Quirk <kquirk at solidworks.com> wrote:
> Try this:
>
> import os.path
> os.path.expanduser('~')

you da man!     That works perfectly.     I mostly use Python on
WinXP, so I haven't been familiar with that function up until now.

thanks again
S

From nicholas.cole at gmail.com  Sun Mar 19 08:20:51 2006
From: nicholas.cole at gmail.com (Nicholas Cole)
Date: Sun, 19 Mar 2006 07:20:51 +0000
Subject: [Pythonmac-SIG] os.environ
Message-ID: <a5c3d8350603182320k388baedehc5847a9e7330b484@mail.gmail.com>

I hope that this is not an FAQ, but I can't find an answer on it.

I know that Terminal.app is setting the environment variables LINES
and COLUMNS (I can see them with the bash "set" command), but
os.environ['LINES'] and os.environ['COLUMNS'] don't exist.  Why is
that?

Best,

N.

From njriley at uiuc.edu  Sun Mar 19 09:01:11 2006
From: njriley at uiuc.edu (Nicholas Riley)
Date: Sun, 19 Mar 2006 02:01:11 -0600
Subject: [Pythonmac-SIG] os.environ
In-Reply-To: <a5c3d8350603182320k388baedehc5847a9e7330b484@mail.gmail.com>
References: <a5c3d8350603182320k388baedehc5847a9e7330b484@mail.gmail.com>
Message-ID: <20060319080111.GA25543@uiuc.edu>

On Sun, Mar 19, 2006 at 07:20:51AM +0000, Nicholas Cole wrote:
> I hope that this is not an FAQ, but I can't find an answer on it.
> 
> I know that Terminal.app is setting the environment variables LINES
> and COLUMNS (I can see them with the bash "set" command), but
> os.environ['LINES'] and os.environ['COLUMNS'] don't exist.  Why is
> that?

LINES and COLUMNS are not typically exported by the shell (passed to
child processes).  You can use 'export LINES' or 'export COLUMNS' to
do so if you wish.

-- 
Nicholas Riley <njriley at uiuc.edu> | <http://www.uiuc.edu/ph/www/njriley>

From nicholas.cole at gmail.com  Sun Mar 19 14:55:20 2006
From: nicholas.cole at gmail.com (Nicholas Cole)
Date: Sun, 19 Mar 2006 13:55:20 +0000
Subject: [Pythonmac-SIG] resizing terminal and curses
In-Reply-To: <a5c3d8350603181437w48204a32lf7bc5f6edefb9621@mail.gmail.com>
References: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>
	<C9FD21B1-D825-4323-AAA3-D8D5C14936BB@mac.com>
	<a5c3d8350603150411j54055f95vab068e3673e5647c@mail.gmail.com>
	<12424448.1142434313465.JavaMail.ronaldoussoren@mac.com>
	<a5c3d8350603150822o4283f3afh2d73ca1cbeead4be@mail.gmail.com>
	<a5c3d8350603181437w48204a32lf7bc5f6edefb9621@mail.gmail.com>
Message-ID: <a5c3d8350603190555u4f8343a7ndf692e58c3f035e5@mail.gmail.com>

I'm still trying to investigate this problem with resizing terminals. 
I've found a method which can reliably determine the size of the
terminal on OS X (not completely tested), which is to do the
following:

struct.unpack('hh', fcntl.ioctl(sys.stdout.fileno(),
termios.TIOCGWINSZ, 'xxxx'))

However, there still seems to be a problem in the underlying curses
code.  The following python script will crash on OS X 10.4 in X11
(xterm) or terminal.app if the terminal is made larger than its
original size - curses refuses to believe it is possible to create a
window larger than the size of the terminal when curses was
initialised.

Terminal.app behaives particularly strangely - crashing when the
window is enlarged for the second time (xterm crashes immediately),
though not the first;  both terminals have problems displaying the
window as it is made smaller, even though they do not crash.

Best wishes,

N.

import curses
import os
import sys
import termios
import fcntl
import struct

def main(screen):
        while 1:
                y, x = struct.unpack('hh',
fcntl.ioctl(sys.stderr.fileno(), termios.TIOCGWINSZ, 'xxxx'))
                win = curses.newwin(y-1, x-2, 0,0)
                #y, x = win.getmaxyx()

                win.addstr("Terminal Size = %s x %s" % (y, x))
                win.refresh()
                curses.napms(100)


if __name__ == '__main__':
        curses.wrapper(main)

From daniel at brightfire.com  Sun Mar 19 21:15:48 2006
From: daniel at brightfire.com (Daniel Lord)
Date: Sun, 19 Mar 2006 12:15:48 -0800
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 36
In-Reply-To: <mailman.10.1142766002.11336.pythonmac-sig@python.org>
References: <mailman.10.1142766002.11336.pythonmac-sig@python.org>
Message-ID: <680AAB1C-A520-447D-BF7C-738497DE7123@brightfire.com>

I prefer:

import os
os.environ['HOME']

but I have no strong argument for that bias other than I sometimes  
use different environment variables and I standardized on the call.

Daniel

On Mar 19, 2006, at 3:00 AM, pythonmac-sig-request at python.org wrote:

> From: "Stewart Midwinter" <stewart.midwinter at gmail.com>
> Date: March 18, 2006 4:40:06 PM PST
> To: "Kent Quirk" <kquirk at solidworks.com>
> Cc: pythonmac-sig at python.org
> Subject: Re: [Pythonmac-SIG] Question on user's directory ( ~ )
>
>
> On 3/18/06, Kent Quirk <kquirk at solidworks.com> wrote:
>> Try this:
>>
>> import os.path
>> os.path.expanduser('~')
>
> you da man!     That works perfectly.     I mostly use Python on
> WinXP, so I haven't been familiar with that function up until now.
>
> thanks again
> S

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060319/133db34b/attachment.html 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060319/133db34b/attachment.pgp 

From zpincus at stanford.edu  Mon Mar 20 04:13:52 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Sun, 19 Mar 2006 19:13:52 -0800
Subject: [Pythonmac-SIG] [PATCH]: Python should use dlopen() on 10.3+
Message-ID: <8C9BE417-AA5B-42A3-952D-F06F5E17600A@stanford.edu>

Hello folks,

Here is a patch to make Python on OS X 10.3 and above use dlopen()  
(via dynload_shlib.c) to load extension modules, and to make the dl  
module avaliable. It ought to be incorporated into the SVN for python  
2.5, if not for further development on 2.4.

If this is an inappropriate channel for this patch, please let me  
know to where I should direct it. Bob Ippolito initially suggested  
this patch be sent hither.

Zach Pincus
Program in Biomedical Informatics and Department of Biochemistry
Stanford University School of Medicine


RATIONALE
---------
On most other unix-like operating systems, Python uses the dlopen()  
call to load extension modules. In addition, the way that these  
modules are opened can be modified via 'sys.setdlopenflags(...)'.

Modifications of how extensions are loaded are useful for several  
reasons (enough so that the standard Python docs (e.g. http:// 
docs.python.org/lib/module-sys.html ). In particular, if multiple  
modules need to share symbols, a call to sys.setdlopenflags is  
necessary.

Symbol sharing is especially important for interoperability of  
modules which wrap C++ classes, because GCC creates classes that  
resolve their run-time type identification by symbol identity. Thus,  
symbols must be shared globally for many C++ features to work  
properly for objects passed between modules.

On OS X / Darwin, Python uses some NeXT-derived APIs to load modules.  
Though these APIs provide analogues to the dlopenflags used to  
control how dlopen() loads modules, this interface is *not* exposed  
to the Python interpreter. Worse, sys.setdlopenflags remains  
available on Darwin, though calls to it are never heeded.

Fortunately, on OS X 10.3 and above, Apple has included dlopen as a  
standard function. In 10.3, this call is provided by a compatibility  
API; in 10.4, the dlopen() call is written to interface directly with  
the library loading mechanism and is now the recommended method for  
libraries to be opened for non Carbon/Cocoa tools.

IMPLEMENTATION
--------------
This (trivial) patch instructs the Python build process to use  
dynload_shlib.c (which uses dlopen) instead of dynload_next.c (which  
uses the NeXT-derived APIs). It also allows for the dl module to be  
built in order to provide access to the proper values for the various  
dlopen flags.

TESTING
-------
This patch can be configured and built into executables that build  
and test correctly on 10.3 and 10.4. The patch is nominally targeted  
at the SVN head as of today, but can be applied (and I have done so  
with success) to Python 2.4.2.

Because Python 2.5 and 2.4 do not currently compile properly on OS X  
10.2, I have not built or tested this patch on that OS version.  
However, the configure and compile process does select the  
appropriate dynload_next.c file to use, and compiles that correctly  
before breaking elsewhere. Thus, if the other errors are fixed for  
10.2, these patches will work fine. (This is because they only change  
Python's behavior for 10.3 and up.)

PATCHES
-------
Here are three sets of patches, as unified diffs.

FIRST: Patch the configure.in file to use dynload_shlib.c when it  
can. Also regenerate the configure script.

Index: configure.in
===================================================================
--- configure.in        (revision 43150)
+++ configure.in        (working copy)
@@ -2105,7 +2105,8 @@
         ;;
         BeOS*) DYNLOADFILE="dynload_beos.o";;
         hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
-       Darwin/*) DYNLOADFILE="dynload_next.o";;
+       # Use dynload_next.c only on 10.2 and below, which don't have  
native dlopen()
+       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;
         atheos*) DYNLOADFILE="dynload_atheos.o";;
         *)
         # use dynload_shlib.c and dlopen() if we have it; otherwise  
stub
Index: configure
===================================================================
--- configure   (revision 43150)
+++ configure   (working copy)
@@ -1,5 +1,5 @@
#! /bin/sh
-# From configure.in Revision: 42437 .
+# From configure.in Revision: 42563 .
# Guess values for system-dependent variables and create Makefiles.
# Generated by GNU Autoconf 2.59 for python 2.5.
#
@@ -13974,7 +13974,8 @@
         ;;
         BeOS*) DYNLOADFILE="dynload_beos.o";;
         hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
-       Darwin/*) DYNLOADFILE="dynload_next.o";;
+       # Use dynload_next.c only on 10.2 and below, which don't have  
native dlopen()
+       Darwin/[0123456].*) DYNLOADFILE="dynload_next.o";;
         atheos*) DYNLOADFILE="dynload_atheos.o";;
         *)
         # use dynload_shlib.c and dlopen() if we have it; otherwise  
stub



SECOND: Allow the dl module to be built and tested


Index: setup.py
===================================================================
--- setup.py    (revision 43150)
+++ setup.py    (working copy)
@@ -883,7 +883,7 @@
          if sys.maxint == 0x7fffffff:
              # This requires sizeof(int) == sizeof(long) == sizeof 
(char*)
              dl_inc = find_file('dlfcn.h', [], inc_dirs)
-            if (dl_inc is not None) and (platform not in ['atheos',  
'darwin']):
+            if (dl_inc is not None) and (platform not in ['atheos']):
                  exts.append( Extension('dl', ['dlmodule.c']) )
          # Thomas Heller's _ctypes module
Index: Lib/test/test_dl.py
===================================================================
--- Lib/test/test_dl.py (revision 43150)
+++ Lib/test/test_dl.py (working copy)
@@ -10,6 +10,7 @@
      ('/usr/lib/libc.so', 'getpid'),
      ('/lib/libc.so.6', 'getpid'),
      ('/usr/bin/cygwin1.dll', 'getpid'),
+    ('/usr/lib/libc.dylib', 'getpid'),
      ]
for s, func in sharedlibs:



THIRD: (optional) Tell Python that the dl test is not expected to be  
skipped anymore. This is optional because if Python is ever built on  
10.2, the test script will expect dl to work, when it only works on  
10.3 and above. However, if Python on 10.2 is officially not  
supported, then this change should be made to properly test the dl  
functionality.


Index: Lib/test/regrtest.py
===================================================================
--- Lib/test/regrtest.py        (revision 43150)
+++ Lib/test/regrtest.py        (working copy)
@@ -930,7 +930,6 @@
          test_cd
          test_cl
          test_curses
-        test_dl
          test_gdbm
          test_gl
          test_imgfile


From Benjamin.Schollnick at xerox.com  Mon Mar 20 14:32:40 2006
From: Benjamin.Schollnick at xerox.com (Schollnick, Benjamin)
Date: Mon, 20 Mar 2006 08:32:40 -0500
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )
Message-ID: <266589E1B9392B4C9195CC25A07C73B90183D318@usa0300ms04.na.xerox.net>

> > import os.path
> > os.path.expanduser('~')
> 
> you da man!     That works perfectly.     I mostly use Python on
> WinXP, so I haven't been familiar with that function up until now.

It's a common problem...  I ran into that once when I was moving an
PC based Python application over to MOSX...  I knew there was a way
expand it, but forgot that it wasn't automagically performed...

	- Ben

From leknarf at pacbell.net  Mon Mar 20 17:27:15 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Mon, 20 Mar 2006 08:27:15 -0800
Subject: [Pythonmac-SIG] Question on user's directory ( ~ )
In-Reply-To: <d4c0d4800603181622k2b5c398ar31c04a97100f75c3@mail.gmail.com>
References: <d4c0d4800603181622k2b5c398ar31c04a97100f75c3@mail.gmail.com>
Message-ID: <5C1B28D9-62F3-4DD4-B3E5-0FFF0DCE8936@pacbell.net>


The tilde is just UNIX shorthand for the environment variable,  
"home."  You can access environment variables in python like this:

	os.getenv("HOME")

Scott


On Mar 18, 2006, at 4:22 PM, Stewart Midwinter wrote:

> I have a question on use of the tilde symbol (~) to access the current
> user's home directory.
>
> If you are in a bash shell, you can "cd ~" and be in the default
> user's home directory.
>
> I want my python app to be able to switch to the user's directory.
> But I can't use os.chdir('~') since Python doesn't understand the
> tilde.   Nor can I do this:
> file = os.path.join('~', filename)
>
> What options do I have?
>
> thanks
> S
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Mon Mar 20 10:05:43 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 20 Mar 2006 10:05:43 +0100
Subject: [Pythonmac-SIG] [PATCH]: Python should use dlopen() on 10.3+
In-Reply-To: <8C9BE417-AA5B-42A3-952D-F06F5E17600A@stanford.edu>
References: <8C9BE417-AA5B-42A3-952D-F06F5E17600A@stanford.edu>
Message-ID: <F519F3ED-94C3-46C3-A6E5-1FDFCEED107A@mac.com>


On 20-mrt-2006, at 4:13, Zachary Pincus wrote:

> Hello folks,
>
> Here is a patch to make Python on OS X 10.3 and above use dlopen()
> (via dynload_shlib.c) to load extension modules, and to make the dl
> module avaliable. It ought to be incorporated into the SVN for python
> 2.5, if not for further development on 2.4.
>
> If this is an inappropriate channel for this patch, please let me
> know to where I should direct it. Bob Ippolito initially suggested
> this patch be sent hither.


Please file this patch on sourceforge.net (www.sf.net/projects/python)

Ronald

>
> Zach Pincus
> Program in Biomedical Informatics and Department of Biochemistry
> Stanford University School of Medicine
>
>
> RATIONALE
> ---------
> On most other unix-like operating systems, Python uses the dlopen()
> call to load extension modules. In addition, the way that these
> modules are opened can be modified via 'sys.setdlopenflags(...)'.
>
> Modifications of how extensions are loaded are useful for several
> reasons (enough so that the standard Python docs (e.g. http://
> docs.python.org/lib/module-sys.html ). In particular, if multiple
> modules need to share symbols, a call to sys.setdlopenflags is
> necessary.
>
> Symbol sharing is especially important for interoperability of
> modules which wrap C++ classes, because GCC creates classes that
> resolve their run-time type identification by symbol identity. Thus,
> symbols must be shared globally for many C++ features to work
> properly for objects passed between modules.
>
> On OS X / Darwin, Python uses some NeXT-derived APIs to load modules.
> Though these APIs provide analogues to the dlopenflags used to
> control how dlopen() loads modules, this interface is *not* exposed
> to the Python interpreter. Worse, sys.setdlopenflags remains
> available on Darwin, though calls to it are never heeded.
>
> Fortunately, on OS X 10.3 and above, Apple has included dlopen as a
> standard function. In 10.3, this call is provided by a compatibility
> API; in 10.4, the dlopen() call is written to interface directly with
> the library loading mechanism and is now the recommended method for
> libraries to be opened for non Carbon/Cocoa tools.
>
> IMPLEMENTATION
> --------------
> This (trivial) patch instructs the Python build process to use
> dynload_shlib.c (which uses dlopen) instead of dynload_next.c (which
> uses the NeXT-derived APIs). It also allows for the dl module to be
> built in order to provide access to the proper values for the various
> dlopen flags.
>
> TESTING
> -------
> This patch can be configured and built into executables that build
> and test correctly on 10.3 and 10.4. The patch is nominally targeted
> at the SVN head as of today, but can be applied (and I have done so
> with success) to Python 2.4.2.
>
> Because Python 2.5 and 2.4 do not currently compile properly on OS X
> 10.2, I have not built or tested this patch on that OS version.
> However, the configure and compile process does select the
> appropriate dynload_next.c file to use, and compiles that correctly
> before breaking elsewhere. Thus, if the other errors are fixed for
> 10.2, these patches will work fine. (This is because they only change
> Python's behavior for 10.3 and up.)
>
> PATCHES
> -------
> Here are three sets of patches, as unified diffs.
>
> FIRST: Patch the configure.in file to use dynload_shlib.c when it
> can. Also regenerate the configure script.
>
> Index: configure.in
> ===================================================================
> --- configure.in        (revision 43150)
> +++ configure.in        (working copy)
> @@ -2105,7 +2105,8 @@
>          ;;
>          BeOS*) DYNLOADFILE="dynload_beos.o";;
>          hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
> -       Darwin/*) DYNLOADFILE="dynload_next.o";;
> +       # Use dynload_next.c only on 10.2 and below, which don't have
> native dlopen()
> +       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;
>          atheos*) DYNLOADFILE="dynload_atheos.o";;
>          *)
>          # use dynload_shlib.c and dlopen() if we have it; otherwise
> stub
> Index: configure
> ===================================================================
> --- configure   (revision 43150)
> +++ configure   (working copy)
> @@ -1,5 +1,5 @@
> #! /bin/sh
> -# From configure.in Revision: 42437 .
> +# From configure.in Revision: 42563 .
> # Guess values for system-dependent variables and create Makefiles.
> # Generated by GNU Autoconf 2.59 for python 2.5.
> #
> @@ -13974,7 +13974,8 @@
>          ;;
>          BeOS*) DYNLOADFILE="dynload_beos.o";;
>          hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
> -       Darwin/*) DYNLOADFILE="dynload_next.o";;
> +       # Use dynload_next.c only on 10.2 and below, which don't have
> native dlopen()
> +       Darwin/[0123456].*) DYNLOADFILE="dynload_next.o";;
>          atheos*) DYNLOADFILE="dynload_atheos.o";;
>          *)
>          # use dynload_shlib.c and dlopen() if we have it; otherwise
> stub
>
>
>
> SECOND: Allow the dl module to be built and tested
>
>
> Index: setup.py
> ===================================================================
> --- setup.py    (revision 43150)
> +++ setup.py    (working copy)
> @@ -883,7 +883,7 @@
>           if sys.maxint == 0x7fffffff:
>               # This requires sizeof(int) == sizeof(long) == sizeof
> (char*)
>               dl_inc = find_file('dlfcn.h', [], inc_dirs)
> -            if (dl_inc is not None) and (platform not in ['atheos',
> 'darwin']):
> +            if (dl_inc is not None) and (platform not in ['atheos']):
>                   exts.append( Extension('dl', ['dlmodule.c']) )
>           # Thomas Heller's _ctypes module
> Index: Lib/test/test_dl.py
> ===================================================================
> --- Lib/test/test_dl.py (revision 43150)
> +++ Lib/test/test_dl.py (working copy)
> @@ -10,6 +10,7 @@
>       ('/usr/lib/libc.so', 'getpid'),
>       ('/lib/libc.so.6', 'getpid'),
>       ('/usr/bin/cygwin1.dll', 'getpid'),
> +    ('/usr/lib/libc.dylib', 'getpid'),
>       ]
> for s, func in sharedlibs:
>
>
>
> THIRD: (optional) Tell Python that the dl test is not expected to be
> skipped anymore. This is optional because if Python is ever built on
> 10.2, the test script will expect dl to work, when it only works on
> 10.3 and above. However, if Python on 10.2 is officially not
> supported, then this change should be made to properly test the dl
> functionality.
>
>
> Index: Lib/test/regrtest.py
> ===================================================================
> --- Lib/test/regrtest.py        (revision 43150)
> +++ Lib/test/regrtest.py        (working copy)
> @@ -930,7 +930,6 @@
>           test_cd
>           test_cl
>           test_curses
> -        test_dl
>           test_gdbm
>           test_gl
>           test_imgfile
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Mon Mar 20 21:44:24 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 20 Mar 2006 12:44:24 -0800
Subject: [Pythonmac-SIG] New Python icons.
Message-ID: <441F1428.6000301@noaa.gov>

Hi all,

I just noticed on c.l.p that someone has whipped up a set of icons based 
on the new one on the website.

I know Kevin O. is working to come up with something better, but I 
figured it wouldn't hurt to take a look at these:

"""

I've taken the opportunity to knock up some icons using it, finally
banishing the poor old standard-VGA-palette snake from my desktop. If
you like, you can grab them from:

   http://www.doxdesk.com/img/software/py/icons.zip

in .ICO format for Windows - containing all resolutions/depths up to
and including Windows Vista's crazy new enormo-icons. Also contains the
vector graphics source file in Xara format. You can also see a preview
here:

   http://www.doxdesk.com/img/software/py/icons.png

-- And Clover
mailto:and at doxdesk.com
http://www.doxdesk.com/
"""



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From abarnert at yahoo.com  Tue Mar 21 19:48:57 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Tue, 21 Mar 2006 10:48:57 -0800 (PST)
Subject: [Pythonmac-SIG] universal binary installer, take 1
Message-ID: <20060321184857.81415.qmail@web82305.mail.mud.yahoo.com>

Hi there. I just joined the list, and I noticed you're
looking for feedback on the new universal binary. I
haven't read all of the archives, but if there's
anything specific you want tested, let me know.

Meanwhile, here's what I've done:

I installed the universal 2.4 on an Intel iMac, built
Psyco with it, and ran a Psyco-ized version of a
curses game (http://cartagena.sf.net). It works; no
apparent bugs beyond those already in the game. It
even successfully detects whether your terminal is
80x24 or 80x25. I haven't verified whether Psyco is
actually working, but it's not reporting any errors.

Then I wrote a script to find all the modules I built
for Darwinports, pull the source, and rebuild and
install for the /usr/local build. It's about two
thirds of the way through, with no failures yet, but I
haven't tested any of these. If I find any problems,
I'll try to build them manually (and fetch more
up-to-date versions) before reporting anything.

The .so files look like universal binaries, but I
haven't tested in Rosetta or my PPC (which is running
pre-10.3.9 Panther, so most universals don't run on
it, and the only gcc4 it has is darwinports, so it
can't build universals).


From rroberts at adobe.com  Tue Mar 21 18:31:03 2006
From: rroberts at adobe.com (Read Roberts)
Date: Tue, 21 Mar 2006 09:31:03 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
Message-ID: <C0457857.2BA%rroberts@adobe.com>

I can't figure out why py2app is building a only semi-standalone bundle app.
Any advice appreciated, including where to look for why the decision is made
to build semi-standalone.


I conclude that the result is semi-standalone because:
- it fails to run if I move the system python framework out of the paths in
sys.path, but otherwise works fine
- there is nothing under <bundle app name>.app/Content/Frameworks


Context:
- Brand new (well old, but recently crashed and now with a new hard-drive)
Mac OSX system with OSX 10.3 with all updates
- Apple developer tools from a disk image 'Panther_7B85_DeveloperCD.dmg"
- Using the Apple-installed python 2.3 in /System/Library/Frameworks, no
other Python installed
Have installed three third-party Python packages;
   - fontTools, by Just von Rossum,
   - numpy
   - py2app from the zip file py2app-0.2-py2.3-macosx10.3.zip

My setup.py file is invoked with the command "python FDKsetup.py py2app",
and generates no error messages. The setup file contains following:
*************************************************************************
from distutils.core import setup
import py2app

import os
import sys

programsDir = os.path.dirname(os.path.abspath(__file__))
programsDir = os.path.dirname(programsDir)
focusDir = os.path.join(programsDir, "focus","focusdll","exe","osx")
acDir = os.path.join(programsDir, "AC","exe","osx")

sys.path.append(focusDir)
sys.path.append(acDir)

py2app_options = dict(
    packages=["fontTools", "numpy"],
     includes = ["PyAC", "PyACDebug",
                "focusdll", "focusdllDebug",
                "array",
                "cStringIO",
                "glob",
                "pdb",
                "re",
                "shutil",
                "socket",
                "string",
                "StringIO",
                "struct",
                "smtplib",
                "tempfile",
                "time",
                "types"
            ]

)

setup(
    app=['FDK.py'],
    options=dict(
        # Each command is allowed to have its own
        # options, so we must specify that these
        # options are py2app specific.
        py2app=py2app_options,
    )
)


From bob at redivi.com  Tue Mar 21 20:01:24 2006
From: bob at redivi.com (Bob Ippolito)
Date: Tue, 21 Mar 2006 11:01:24 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <C0457857.2BA%rroberts@adobe.com>
References: <C0457857.2BA%rroberts@adobe.com>
Message-ID: <328EB11C-5271-4DEF-903E-ACC0D168E72F@redivi.com>


On Mar 21, 2006, at 9:31 AM, Read Roberts wrote:

> I can't figure out why py2app is building a only semi-standalone  
> bundle app.
> Any advice appreciated, including where to look for why the  
> decision is made
> to build semi-standalone.
>
>
> I conclude that the result is semi-standalone because:
> - it fails to run if I move the system python framework out of the  
> paths in
> sys.path, but otherwise works fine
> - there is nothing under <bundle app name>.app/Content/Frameworks
>
>
> Context:
> - Brand new (well old, but recently crashed and now with a new hard- 
> drive)
> Mac OSX system with OSX 10.3 with all updates
> - Apple developer tools from a disk image  
> 'Panther_7B85_DeveloperCD.dmg"
> - Using the Apple-installed python 2.3 in /System/Library/ 
> Frameworks, no
> other Python installed

This is why.  It refuses to build a standalone bundle out of the  
system framework.

-bob


From zpincus at stanford.edu  Tue Mar 21 20:26:28 2006
From: zpincus at stanford.edu (Zachary Pincus)
Date: Tue, 21 Mar 2006 11:26:28 -0800
Subject: [Pythonmac-SIG] [PATCH]: Python should use dlopen() on 10.3+
In-Reply-To: <F519F3ED-94C3-46C3-A6E5-1FDFCEED107A@mac.com>
References: <8C9BE417-AA5B-42A3-952D-F06F5E17600A@stanford.edu>
	<F519F3ED-94C3-46C3-A6E5-1FDFCEED107A@mac.com>
Message-ID: <08C20550-555F-4D8F-98F4-6E1079B80AD4@stanford.edu>

I have added the patch to sourceforge's patch tracker:
http://sourceforge.net/tracker/index.php? 
func=detail&aid=1454844&group_id=5470&atid=305470

Since this patch is going to be of primary interest to mac people, I  
would appreciate it if anyone interested in making mac python (a)  
load extensions like other pythons, and (b) load extensions as per  
the python documentation would look at, comment on, and/or test the  
patch.

This patch has already been tested by myself and one other on 10.2,  
10.3, and 10.4, so I'm confident that it works as advertised, and  
that there's no reason it should not be applied to the python codebase.

Thanks,
Zach Pincus

Program in Biomedical Informatics and Department of Biochemistry
Stanford University School of Medicine



On Mar 20, 2006, at 1:05 AM, Ronald Oussoren wrote:

>
> On 20-mrt-2006, at 4:13, Zachary Pincus wrote:
>
>> Hello folks,
>>
>> Here is a patch to make Python on OS X 10.3 and above use dlopen()
>> (via dynload_shlib.c) to load extension modules, and to make the dl
>> module avaliable. It ought to be incorporated into the SVN for python
>> 2.5, if not for further development on 2.4.
>>
>> If this is an inappropriate channel for this patch, please let me
>> know to where I should direct it. Bob Ippolito initially suggested
>> this patch be sent hither.
>
>
> Please file this patch on sourceforge.net (www.sf.net/projects/python)
>
> Ronald
>
>>
>> Zach Pincus
>> Program in Biomedical Informatics and Department of Biochemistry
>> Stanford University School of Medicine
>>
>>
>> RATIONALE
>> ---------
>> On most other unix-like operating systems, Python uses the dlopen()
>> call to load extension modules. In addition, the way that these
>> modules are opened can be modified via 'sys.setdlopenflags(...)'.
>>
>> Modifications of how extensions are loaded are useful for several
>> reasons (enough so that the standard Python docs (e.g. http://
>> docs.python.org/lib/module-sys.html ). In particular, if multiple
>> modules need to share symbols, a call to sys.setdlopenflags is
>> necessary.
>>
>> Symbol sharing is especially important for interoperability of
>> modules which wrap C++ classes, because GCC creates classes that
>> resolve their run-time type identification by symbol identity. Thus,
>> symbols must be shared globally for many C++ features to work
>> properly for objects passed between modules.
>>
>> On OS X / Darwin, Python uses some NeXT-derived APIs to load modules.
>> Though these APIs provide analogues to the dlopenflags used to
>> control how dlopen() loads modules, this interface is *not* exposed
>> to the Python interpreter. Worse, sys.setdlopenflags remains
>> available on Darwin, though calls to it are never heeded.
>>
>> Fortunately, on OS X 10.3 and above, Apple has included dlopen as a
>> standard function. In 10.3, this call is provided by a compatibility
>> API; in 10.4, the dlopen() call is written to interface directly with
>> the library loading mechanism and is now the recommended method for
>> libraries to be opened for non Carbon/Cocoa tools.
>>
>> IMPLEMENTATION
>> --------------
>> This (trivial) patch instructs the Python build process to use
>> dynload_shlib.c (which uses dlopen) instead of dynload_next.c (which
>> uses the NeXT-derived APIs). It also allows for the dl module to be
>> built in order to provide access to the proper values for the various
>> dlopen flags.
>>
>> TESTING
>> -------
>> This patch can be configured and built into executables that build
>> and test correctly on 10.3 and 10.4. The patch is nominally targeted
>> at the SVN head as of today, but can be applied (and I have done so
>> with success) to Python 2.4.2.
>>
>> Because Python 2.5 and 2.4 do not currently compile properly on OS X
>> 10.2, I have not built or tested this patch on that OS version.
>> However, the configure and compile process does select the
>> appropriate dynload_next.c file to use, and compiles that correctly
>> before breaking elsewhere. Thus, if the other errors are fixed for
>> 10.2, these patches will work fine. (This is because they only change
>> Python's behavior for 10.3 and up.)
>>
>> PATCHES
>> -------
>> Here are three sets of patches, as unified diffs.
>>
>> FIRST: Patch the configure.in file to use dynload_shlib.c when it
>> can. Also regenerate the configure script.
>>
>> Index: configure.in
>> ===================================================================
>> --- configure.in        (revision 43150)
>> +++ configure.in        (working copy)
>> @@ -2105,7 +2105,8 @@
>>          ;;
>>          BeOS*) DYNLOADFILE="dynload_beos.o";;
>>          hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
>> -       Darwin/*) DYNLOADFILE="dynload_next.o";;
>> +       # Use dynload_next.c only on 10.2 and below, which don't have
>> native dlopen()
>> +       Darwin/@<:@0123456@:>@.*) DYNLOADFILE="dynload_next.o";;
>>          atheos*) DYNLOADFILE="dynload_atheos.o";;
>>          *)
>>          # use dynload_shlib.c and dlopen() if we have it; otherwise
>> stub
>> Index: configure
>> ===================================================================
>> --- configure   (revision 43150)
>> +++ configure   (working copy)
>> @@ -1,5 +1,5 @@
>> #! /bin/sh
>> -# From configure.in Revision: 42437 .
>> +# From configure.in Revision: 42563 .
>> # Guess values for system-dependent variables and create Makefiles.
>> # Generated by GNU Autoconf 2.59 for python 2.5.
>> #
>> @@ -13974,7 +13974,8 @@
>>          ;;
>>          BeOS*) DYNLOADFILE="dynload_beos.o";;
>>          hp*|HP*) DYNLOADFILE="dynload_hpux.o";;
>> -       Darwin/*) DYNLOADFILE="dynload_next.o";;
>> +       # Use dynload_next.c only on 10.2 and below, which don't have
>> native dlopen()
>> +       Darwin/[0123456].*) DYNLOADFILE="dynload_next.o";;
>>          atheos*) DYNLOADFILE="dynload_atheos.o";;
>>          *)
>>          # use dynload_shlib.c and dlopen() if we have it; otherwise
>> stub
>>
>>
>>
>> SECOND: Allow the dl module to be built and tested
>>
>>
>> Index: setup.py
>> ===================================================================
>> --- setup.py    (revision 43150)
>> +++ setup.py    (working copy)
>> @@ -883,7 +883,7 @@
>>           if sys.maxint == 0x7fffffff:
>>               # This requires sizeof(int) == sizeof(long) == sizeof
>> (char*)
>>               dl_inc = find_file('dlfcn.h', [], inc_dirs)
>> -            if (dl_inc is not None) and (platform not in ['atheos',
>> 'darwin']):
>> +            if (dl_inc is not None) and (platform not in  
>> ['atheos']):
>>                   exts.append( Extension('dl', ['dlmodule.c']) )
>>           # Thomas Heller's _ctypes module
>> Index: Lib/test/test_dl.py
>> ===================================================================
>> --- Lib/test/test_dl.py (revision 43150)
>> +++ Lib/test/test_dl.py (working copy)
>> @@ -10,6 +10,7 @@
>>       ('/usr/lib/libc.so', 'getpid'),
>>       ('/lib/libc.so.6', 'getpid'),
>>       ('/usr/bin/cygwin1.dll', 'getpid'),
>> +    ('/usr/lib/libc.dylib', 'getpid'),
>>       ]
>> for s, func in sharedlibs:
>>
>>
>>
>> THIRD: (optional) Tell Python that the dl test is not expected to be
>> skipped anymore. This is optional because if Python is ever built on
>> 10.2, the test script will expect dl to work, when it only works on
>> 10.3 and above. However, if Python on 10.2 is officially not
>> supported, then this change should be made to properly test the dl
>> functionality.
>>
>>
>> Index: Lib/test/regrtest.py
>> ===================================================================
>> --- Lib/test/regrtest.py        (revision 43150)
>> +++ Lib/test/regrtest.py        (working copy)
>> @@ -930,7 +930,6 @@
>>           test_cd
>>           test_cl
>>           test_curses
>> -        test_dl
>>           test_gdbm
>>           test_gl
>>           test_imgfile
>>
>> _______________________________________________
>> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>> http://mail.python.org/mailman/listinfo/pythonmac-sig
>


From rroberts at adobe.com  Wed Mar 22 02:43:59 2006
From: rroberts at adobe.com (Read Roberts)
Date: Tue, 21 Mar 2006 17:43:59 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <328EB11C-5271-4DEF-903E-ACC0D168E72F@redivi.com>
Message-ID: <C045EBDF.2D6%rroberts@adobe.com>

Dear Bob;

Thank you, I am now a happy camper with a real stqnd-alone py2app.

For others:

To get a separate installation of Python 2.3.x under Mac OSX 10.3, I
downloaded the installer for MacPython-OSX 2.3.5 from:
http://homepages.cwi.nl/~jack/macpython/beta.html
This installs under /Library/Frameworks, and when used to run py2app, builds
a stand-alone installer.

Two wrinkles to be aware of:
1) location of command-line 'python'
In a plain Mac OS X 10.3 installation, the command 'python' works because
the system environment variable "PATH" contains a reference to "/usr/bin",
and this directory contains a symbolic link "python" that points to the
python program under /System/Library/Frameworks/Python.framework. However,
the Python 2.3.5 installer makes the equivalent symbolic link to the python
program under '/Library/Frameworks/Python.framework", but puts the symbolic
link under "/user/local/bin", which is not in the default PATH environment
variable. I added a symbolic "python2.3.5"  in the directory "/usr/bin,
pointing to the python program under 'Library/Frameworks/Python.framework",
so I can invoke this python with the command "python2.3.5".

2) Location of py2app.
The installer in " py2app-0.2-py2.3-macosx10.3.zip" puts the py2app package
under /Library/Python/2.3. This works for the system Python, because its
site-packages directory is actually a symbolic link pointing to the
directory '/Library/Python/2.3". However, the Python2.3.5 installer creates
a Python whose site-packages directory is a real directory, so it will never
see the py2app package. I chose to fix this by deleting the real directory,
and making a symbolic link with the same name to '/Library/Python/2.3"

- Read Roberts

On 3/21/06 11:01 AM, "Bob Ippolito" <bob at redivi.com> wrote:

> 
> On Mar 21, 2006, at 9:31 AM, Read Roberts wrote:
> 
>> I can't figure out why py2app is building a only semi-standalone
>> bundle app.
>> Any advice appreciated, including where to look for why the
>> decision is made
>> to build semi-standalone.
>> 
>> 
>> I conclude that the result is semi-standalone because:
>> - it fails to run if I move the system python framework out of the
>> paths in
>> sys.path, but otherwise works fine
>> - there is nothing under <bundle app name>.app/Content/Frameworks
>> 
>> 
>> Context:
>> - Brand new (well old, but recently crashed and now with a new hard-
>> drive)
>> Mac OSX system with OSX 10.3 with all updates
>> - Apple developer tools from a disk image
>> 'Panther_7B85_DeveloperCD.dmg"
>> - Using the Apple-installed python 2.3 in /System/Library/
>> Frameworks, no
>> other Python installed
> 
> This is why.  It refuses to build a standalone bundle out of the
> system framework.
> 
> -bob
> 


From abarnert at yahoo.com  Wed Mar 22 04:04:10 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Tue, 21 Mar 2006 19:04:10 -0800 (PST)
Subject: [Pythonmac-SIG] resizing terminal and curses
Message-ID: <20060322030411.80489.qmail@web82304.mail.mud.yahoo.com>

> Terminal.app behaives particularly strangely -
crashing when the
> window is enlarged for the second time (xterm
crashes immediately),
> though not the first

For me, it crashes in Terminal.app the first time I
try to enlarge the window, not the second. I see no
difference between Terminal and xterm. This is on a
10.4.5 Intel iMac, with all versions of Python from
Apple's 2.3 to a custom post-2.4 build.


From brendansimons at yahoo.ca  Wed Mar 22 04:52:08 2006
From: brendansimons at yahoo.ca (Brendan Simons)
Date: Tue, 21 Mar 2006 22:52:08 -0500
Subject: [Pythonmac-SIG] New Python icons.
In-Reply-To: <mailman.6.1142938802.27684.pythonmac-sig@python.org>
References: <mailman.6.1142938802.27684.pythonmac-sig@python.org>
Message-ID: <4CD2004C-D610-467E-A801-7490C6AF5846@yahoo.ca>

Chris, And  These look great!  Which one is for what?

I asked Tim Parkin to send me a transparent version of the logo, so  
you can experiment without the white border.   Find it attached.

The only other thing I'd request would be proper document icon type  
badging per http://developer.apple.com/documentation/UserExperience/ 
Conceptual/OSXHIGuidelines/XHIGIcons/chapter_14_section_2.html#// 
apple_ref/doc/uid/20000967-TPXREF117 (holy that's a long url).  I was  
going to tint the .py document icons as well so that they stand out  
in a folder full of .pyc and .pyo

   -Brendan
--
Brendan Simons
?On 21-Mar-06, at 6:00 AM, pythonmac-sig-request at python.org wrote:

> From: Christopher Barker <Chris.Barker at noaa.gov>
> Date: March 20, 2006 3:44:24 PM EST (CA)
> To: MacPython <pythonmac-sig at python.org>
> Subject: [Pythonmac-SIG] New Python icons.
>
>
> Hi all,
>
> I just noticed on c.l.p that someone has whipped up a set of icons  
> based on the new one on the website.
>
> I know Kevin O. is working to come up with something better, but I  
> figured it wouldn't hurt to take a look at these:
>
> """
>
> I've taken the opportunity to knock up some icons using it, finally
> banishing the poor old standard-VGA-palette snake from my desktop. If
> you like, you can grab them from:
>
>   http://www.doxdesk.com/img/software/py/icons.zip
>
> in .ICO format for Windows - containing all resolutions/depths up to
> and including Windows Vista's crazy new enormo-icons. Also contains  
> the
> vector graphics source file in Xara format. You can also see a preview
> here:
>
>   http://www.doxdesk.com/img/software/py/icons.png
>
> -- And Clover
> mailto:and at doxdesk.com
> http://www.doxdesk.com/
> """
>
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                     		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060321/9f67b59a/attachment-0001.htm 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: python-logo-transparent.png
Type: image/png
Size: 11446 bytes
Desc: not available
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060321/9f67b59a/attachment-0001.png 

From abarnert at yahoo.com  Wed Mar 22 09:45:37 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Wed, 22 Mar 2006 00:45:37 -0800 (PST)
Subject: [Pythonmac-SIG] Questions about various bits of Intel status
Message-ID: <20060322084537.51313.qmail@web82310.mail.mud.yahoo.com>

I'm guessing these answers are all out there
somewhere, but I can't find them. So I apologize if
I'm being stupid, but:

First, what's up with ctypes and other things that
need libffi? They complain that it hasn't been ported
to darwin x86 and punt. The latest information I could
find was a blog at bob.pythonmac.org that says it's
been ported, but has no link or information. It looks
like PyObjC has a port, but it doesn't work. Should I
be looking on the PyObjC and/or ctypes lists instead
of here? (A quick search didn't turn up any obvious
answers, but I didn't look in-depth.)

Next, have the python24-fat changes been ported to the
2.5 trunk? If not, should it be relatively easy to do
myself, or should I just wait for the gurus to check
it in? If it should be doable, at what version and tag
was python24-fat forked from the tree (so I can pull
the stock equivalent and make diffs to start from)?

How exactly are you supposed to properly install a
framework build? I assume that if you sudo make
installframeworks, you can't install any modules via
.pkg files. (At least the docs say not to do this, and
it seems the most likely reason.) But if you just make
installframeworks, it fails on /usr/local/{bin,lib}. I
ended up sudo'ing it and chgrp/chmod -R on the
frameworks, but there must be a better way. Maybe
change the makefiles to install -m775/664 instead of
755/644, and -gadmin for the framework bits only? But
that may not be what you want when building a
package/darwinport/etc.? Without digging through the
makefile structure, I'm not sure....

Are trunk builds supposed to mostly work on an Intel
Mac? By "mostly work" I mean that if something works
on a linux box off the trunk, and on it's in the 2.4
official unofficial release candidate universal
installer and works on my Intel box, it'll also work
on building off the trunk on my Intel Mac. I'm not
expecting any better.... But there are a bunch of
things that seem to work in trunk/linux and
2.4.2/universal but not trunk/Intel mac:
* test_curses is skipped, even though curses appears
to work
* all tests that require networking are skipped, even
though at least
   some (maybe all) of the relevant modules work
* test_re fails
* test_{unicode,unicodedata,codecs} fail
* test_{aepack,applesingle,macostools} fail

I checked one of the tests, and the Unicode stuff does
seem to be wrong (u'\u2000'.isspace() returns False on
trunk, True on everything else). I haven't checked the
rest. No configure settings seem to make a difference
(unless I disable the relevant modules or totally
break the build).

Finally: Is there some known place to look for this
kind of info besides this list, pythonmac.org's wiki,
the current trunk code, the bug tracker, and random
googling? Is there any way to follow threads from this
list's archive without stopping and rooting around at
each month boundary? Is there a way to search the
archives other than googling with
site:http://mail.python.org/pipermail/pythonmac-sig?

Thanks.


From ronaldoussoren at mac.com  Wed Mar 22 11:45:46 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 22 Mar 2006 11:45:46 +0100
Subject: [Pythonmac-SIG] Questions about various bits of Intel status
In-Reply-To: <20060322084537.51313.qmail@web82310.mail.mud.yahoo.com>
References: <20060322084537.51313.qmail@web82310.mail.mud.yahoo.com>
Message-ID: <4876C93F-50A0-419B-A1A1-EEE5EB265A91@mac.com>


On 22-mrt-2006, at 9:45, Andrew Barnert wrote:

> I'm guessing these answers are all out there
> somewhere, but I can't find them. So I apologize if
> I'm being stupid, but:
>
> First, what's up with ctypes and other things that
> need libffi? They complain that it hasn't been ported
> to darwin x86 and punt. The latest information I could
> find was a blog at bob.pythonmac.org that says it's
> been ported, but has no link or information. It looks
> like PyObjC has a port, but it doesn't work. Should I
> be looking on the PyObjC and/or ctypes lists instead
> of here? (A quick search didn't turn up any obvious
> answers, but I didn't look in-depth.)

PyObjC contains a port of libffi to darwin/x86, but sadly
enough I didn't pay enough attention in the time between
WWDC'05 and the release of OSX/x86. My port of libffi
no longer works due to slight changes in the calling
conventions. I'll be fixing this shortly, and push those
upstreams and to ctypes.

>
> Next, have the python24-fat changes been ported to the
> 2.5 trunk? If not, should it be relatively easy to do
> myself, or should I just wait for the gurus to check
> it in? If it should be doable, at what version and tag
> was python24-fat forked from the tree (so I can pull
> the stock equivalent and make diffs to start from)?

The changes have not been ported to the 2.5 tree yet,
but I intend to do so before the 2.5a1 release.

>
> How exactly are you supposed to properly install a
> framework build? I assume that if you sudo make
> installframeworks, you can't install any modules via
> .pkg files. (At least the docs say not to do this, and
> it seems the most likely reason.) But if you just make
> installframeworks, it fails on /usr/local/{bin,lib}. I
> ended up sudo'ing it and chgrp/chmod -R on the
> frameworks, but there must be a better way. Maybe
> change the makefiles to install -m775/664 instead of
> 755/644, and -gadmin for the framework bits only? But
> that may not be what you want when building a
> package/darwinport/etc.? Without digging through the
> makefile structure, I'm not sure....

Just install using the pkg? My build script in the
python24-fat tree does a chmod after installing, but
that's mostly because its predecessor also did that.

>
> Are trunk builds supposed to mostly work on an Intel
> Mac? By "mostly work" I mean that if something works
> on a linux box off the trunk, and on it's in the 2.4
> official unofficial release candidate universal
> installer and works on my Intel box, it'll also work
> on building off the trunk on my Intel Mac. I'm not
> expecting any better.... But there are a bunch of
> things that seem to work in trunk/linux and
> 2.4.2/universal but not trunk/Intel mac:
> * test_curses is skipped, even though curses appears
> to work
> * all tests that require networking are skipped, even
> though at least
>    some (maybe all) of the relevant modules work
> * test_re fails
> * test_{unicode,unicodedata,codecs} fail

Dunno about these, they should work just fine. I haven't
tried building the trunk on an intel box yet though.

> * test_{aepack,applesingle,macostools} fail

Those are expected to fail, the python24-fat tree contains
bugfixes for this.

>
> I checked one of the tests, and the Unicode stuff does
> seem to be wrong (u'\u2000'.isspace() returns False on
> trunk, True on everything else). I haven't checked the
> rest. No configure settings seem to make a difference
> (unless I disable the relevant modules or totally
> break the build).
>
> Finally: Is there some known place to look for this
> kind of info besides this list, pythonmac.org's wiki,
> the current trunk code, the bug tracker, and random
> googling? Is there any way to follow threads from this
> list's archive without stopping and rooting around at
> each month boundary? Is there a way to search the
> archives other than googling with
> site:http://mail.python.org/pipermail/pythonmac-sig?

Not that I know of.

Ronald

>
> Thanks.
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From jnutting at gmail.com  Wed Mar 22 15:29:13 2006
From: jnutting at gmail.com (Jack Nutting)
Date: Wed, 22 Mar 2006 15:29:13 +0100
Subject: [Pythonmac-SIG] New Python icons.
In-Reply-To: <4CD2004C-D610-467E-A801-7490C6AF5846@yahoo.ca>
References: <mailman.6.1142938802.27684.pythonmac-sig@python.org>
	<4CD2004C-D610-467E-A801-7490C6AF5846@yahoo.ca>
Message-ID: <f84841020603220629v70a43053i28b25a8156e882ad@mail.gmail.com>

On 3/22/06, Brendan Simons <brendansimons at yahoo.ca> wrote:
>
> Chris, And  These look great!  Which one is for what?
>

Without  knowing the original author's intent, I'd guess that the first one
with a terminal-style window in the background is for the python executable
itself, the second one is for .py documents, and the third one is for .pyc
and .pyo files.  So, I'd say that using the second and third ones (or
something similar) for .py and .pyc/.pyo files on the Mac would be the way
to go.

--
// jack
// http://www.nuthole.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060322/163f4150/attachment.htm 

From Chris.Barker at noaa.gov  Wed Mar 22 18:29:33 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Wed, 22 Mar 2006 09:29:33 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <C045EBDF.2D6%rroberts@adobe.com>
References: <C045EBDF.2D6%rroberts@adobe.com>
Message-ID: <4421897D.4060005@noaa.gov>

Read,

Thanks for posting a summary, it's nice to have this in the archives.

A few notes:

Read Roberts wrote:
> To get a separate installation of Python 2.3.x under Mac OSX 10.3,

Why not just use 2.4.1 ? Is there a reason you need a 2.3 version?

> Two wrinkles to be aware of:
> 1) location of command-line 'python'
> In a plain Mac OS X 10.3 installation, the command 'python' works because
> the system environment variable "PATH" contains a reference to "/usr/bin",

This is totally standard *nix stuff.

 > puts the symbolic
> link under "/user/local/bin",

Also standard. "local" more or less means "not installed by the OS vendor"

> which is not in the default PATH environment variable.

which is too bad. /usr/local/bin is such a standard for user-installed 
stuff, I've always wondered why OS vendors don't' just put it in the 
default PATH.

> I added a symbolic "python2.3.5"  in the directory "/usr/bin,
> pointing to the python program under 'Library/Frameworks/Python.framework",
> so I can invoke this python with the command "python2.3.5".

Here's where you've deviated from the standard: As a rule, you shouldn't 
out ANYTHING in /usr/bin. that's Apple's job. As you haven't written 
over anything, it's probably not going to cause any problems, but the 
usual solution is to add /usr/local/bin to your PATH instead, but adding 
a line something like this to your

/Users/YourName/.profile  file:


export PATH="usr/local/bin/:$PATH"



That will tell your shell (if it's the default bash shell) to look in 
/usr/local/bin for executables before the other places it looks.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From loredo at astro.cornell.edu  Wed Mar 22 19:21:06 2006
From: loredo at astro.cornell.edu (Tom Loredo)
Date: Wed, 22 Mar 2006 13:21:06 -0500
Subject: [Pythonmac-SIG] Matplotlib give bus error (10.3.9)
Message-ID: <1143051666.44219592b05db@astrosun2.astro.cornell.edu>


Hi folks,

I've just installed MacPy 2.4.1, the latest numpy & scipy, and
mpl-0.87.2 on a colleague's Panther PowerBook.  numpy & scipy 
installed fine and passed all tests.  For mpl, we installed
zlib from source, and initially installed Freetype2 and libpng
via the i-Installer.  mpl built and installed fine, but gave
a bus error when trying to plot ("import matplotlib" would
work fine, but "import pylab" would give the error).

The .matplotlib/matplotlibrc file we're using does have TkAgg
and numpy specified, and we've verified that Tkinter is
working via a "Hello World" test script.

Suspecting one of the libraries, we then installed Freetype2
and libpng from source.  We deleted all the mpl stuff in
site-packages and the mpl build directory, and built and
installed mpl again from scratch.  We still get a bus error
from pylab.  I am stumped as to what to try next.  I've
copied verbose output below; I hope someone can offer us
some clues.

Yes, I'm also asking about this on the mpl list.

Thanks,
Tom Loredo

$ pythonw simple_plot.py --verbose-debug-annoying
matplotlib data path 
/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-packages/matplotlib/mpl-data
$HOME=/Users/bill
CONFIGDIR=/Users/bill/.matplotlib
loaded rc file /Users/bill/.matplotlib/matplotlibrc
matplotlib version 0.87.2
verbose.level debug-annoying
interactive is False
platform is darwin
loaded modules: ['pylab', '_bisect', '__future__', 'copy_reg', 
'sre_compile', 'distutils', 'itertools', '_sre', '__main__', 'site', 
'__builtin__', 'datetime', 'matplotlib.re', 'matplotlib.tempfile', 
'encodings', 'pytz.datetime', 'shutil', 'distutils.string', 
'dateutil', 'matplotlib.datetime', 'posixpath', '_random', 
'tempfile', 'errno', 'matplotlib.warnings', 'binascii', 
'encodings.codecs', 'sre_constants', 're', 'matplotlib.md5', 
'os.path', 'pytz.sys', '_codecs', 'distutils.sysconfig', 
'encodings.exceptions', 'pytz.sets', 'math', 'fcntl', 'stat', 
'zipimport', 'string', 'warnings', 'encodings.types', 'UserDict', 
'encodings.ascii', 'matplotlib.sys', 'matplotlib', 'distutils.os', 
'sys', 'pytz.tzinfo', 'pytz', 'matplotlib.__future__', 'codecs', 
'distutils.re', 'matplotlib.pytz', 'types', 'md5', 
'matplotlib.dateutil', 'matplotlib.os', 'thread', 'sre', 'bisect', 
'matplotlib.distutils', 'signal', 'distutils.errors', 'random', 
'linecache', 'matplotlib.shutil', 'posix', 'encodings.aliases', 
'sets', 'exceptions', 'sre_parse', 'pytz.bisect', 'distutils.sys', 
'os', 'strop']
numerix numpy 0.9.6
Bus error


-------------------------------------------------
This mail sent through IMP: http://horde.org/imp/

From rroberts at adobe.com  Wed Mar 22 19:53:07 2006
From: rroberts at adobe.com (Read Roberts)
Date: Wed, 22 Mar 2006 10:53:07 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <4421897D.4060005@noaa.gov>
Message-ID: <C046DD13.2FD%rroberts@adobe.com>




On 3/22/06 9:29 AM, "Christopher Barker" <Chris.Barker at noaa.gov> wrote:

> Read,
> 
> Thanks for posting a summary, it's nice to have this in the archives.
> 
> A few notes:
> 
> Read Roberts wrote:
>> To get a separate installation of Python 2.3.x under Mac OSX 10.3,
> 
> Why not just use 2.4.1 ? Is there a reason you need a 2.3 version?
Yes, because my package includes some Python extension modules written in C
that are also used by a font-editing program, FontLab, that has Python 2.3
embedded. I could build my bundle app with Python 2.4, but then I'd need to
maintain a separate build of the extension modules linked with Python 2.3
for this other program. I build under Mac OSX 10.3 for a similar reason,
easier to build programs that will run under Mac OSX 10.3 as well as 10.4.


>> 1) location of command-line 'python'
<snip>
>> I added a symbolic "python2.3.5"  in the directory "/usr/bin,
>> pointing to the python program under 'Library/Frameworks/Python.framework",
>> so I can invoke this python with the command "python2.3.5".
> 
> Here's where you've deviated from the standard: As a rule, you shouldn't
> out ANYTHING in /usr/bin. that's Apple's job. As you haven't written
> over anything, it's probably not going to cause any problems, but the
> usual solution is to add /usr/local/bin to your PATH instead, but adding
> a line something like this to your
> 
> /Users/YourName/.profile  file:
> 
> 
> export PATH="usr/local/bin/:$PATH"
> 
> 
> 
> That will tell your shell (if it's the default bash shell) to look in
> /usr/local/bin for executables before the other places it looks.

Noted. This is certainly a best practice.
 
> -Chris
> 
> 


From janssen at parc.com  Wed Mar 22 19:55:50 2006
From: janssen at parc.com (Bill Janssen)
Date: Wed, 22 Mar 2006 10:55:50 PST
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: Your message of "Wed, 22 Mar 2006 09:29:33 PST."
	<4421897D.4060005@noaa.gov> 
Message-ID: <06Mar22.105557pst."58633"@synergy1.parc.xerox.com>

> > I added a symbolic "python2.3.5"  in the directory "/usr/bin,
> > pointing to the python program under 'Library/Frameworks/Python.framework",
> > so I can invoke this python with the command "python2.3.5".
> 
> Here's where you've deviated from the standard: As a rule, you shouldn't 
> out ANYTHING in /usr/bin. that's Apple's job. As you haven't written 
> over anything, it's probably not going to cause any problems, but the 
> usual solution is to add /usr/local/bin to your PATH instead, but adding 
> a line something like this to your

Yeah, yeah -- this is correct, Chris, and technically works, but
doesn't *really* work on Mac OS X.  Those of us who've come to Macs
from UNIX are comfortable with it, but old-style Mac users (pre-2000)
aren't used to PATH and hacking it, and their .login or .bashrc
scripts tend to be odd barely-working mixtures of whatever other
people have told them to put there.  Once you've got a GUI, you can't
go back.

Read, my installers put links in /usr/bin/, too.  It's really the only
way to go, until Apple provides something like the Windows registry
complete with a programmatic API for it.

Bill

From Chris.Barker at noaa.gov  Wed Mar 22 20:33:57 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Wed, 22 Mar 2006 11:33:57 -0800
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
References: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
Message-ID: <4421A6A5.7070602@noaa.gov>

Bill Janssen wrote:

> Read, my installers put links in /usr/bin/, too.  It's really the only
> way to go, until Apple provides something like the Windows registry
> complete with a programmatic API for it.

AACCK! NOOOOOO!

You sure don't need the nightmare of the registry for Apple to make this 
kind of thing easier.

1) They could just have a standard place for user-installed command line 
apps to go (/usr/local/bin is a fine choice), and PUT IT IN THE DEFAULT 
PATH!

2) you mentioned a programmatic API: In fact, I think there is a plist 
somewhere that sets the default environment variables for all apps, you 
could add a PATH entry (or add to it) programmaticly easily enough.

3) PATH is only relevant to command line apps: if you've never gone back 
from using a GUI -- you'll never need to care about this.

4) I've used a GUI for many years: Macintosh, various flavors of 
Windows, KDE, etc -- I still keep going back to the command line for 
some things.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Wed Mar 22 22:05:50 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 22 Mar 2006 22:05:50 +0100
Subject: [Pythonmac-SIG] New Python icons.
In-Reply-To: <441F1428.6000301@noaa.gov>
References: <441F1428.6000301@noaa.gov>
Message-ID: <A3EA3298-2150-4AC0-AAB9-A3141C847A26@mac.com>


On 20-mrt-2006, at 21:44, Christopher Barker wrote:

> Hi all,
>
> I just noticed on c.l.p that someone has whipped up a set of icons  
> based
> on the new one on the website.
>
> I know Kevin O. is working to come up with something better, but I
> figured it wouldn't hurt to take a look at these:

The document icons look nice. I don't like the first icon, its too much
a windows icon :-).

I still don't like the new official python logo and am anxiously  
awaiting
Kevin's proposal.

Ronald


From pecora at anvil.nrl.navy.mil  Wed Mar 22 22:26:13 2006
From: pecora at anvil.nrl.navy.mil (Louis Pecora)
Date: Wed, 22 Mar 2006 16:26:13 -0500
Subject: [Pythonmac-SIG] I have Python 2.4.1 but no site packages. What did
	I miss?
Message-ID: <4421C0F5.3080001@anvil.nrl.navy.mil>

A few months ago I installed (at the prodding of gurus here) Python
2.4.1.  I tried to use it today (haven't used it for a while since the
install) and it cannot find Numeric.  I checked and I have no 2.4
directory (under /Library) for site packages.  They're all in the 2.3
directory.  Likewise I have no ~/Library/python/2.4 directory only
~/Library/python/2.3 for my .pth file.   Of course, if I run Python 2.3
it finds all the modules and my stuff.

Should I just create the 2.4 directories and copy the stuff from 2.3
into 2.4?

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pecora at anvil.nrl.navy.mil



From bob at redivi.com  Wed Mar 22 22:51:59 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 22 Mar 2006 13:51:59 -0800
Subject: [Pythonmac-SIG] I have Python 2.4.1 but no site packages. What
	did I miss?
In-Reply-To: <4421C0F5.3080001@anvil.nrl.navy.mil>
References: <4421C0F5.3080001@anvil.nrl.navy.mil>
Message-ID: <4655F3C9-C20A-45FF-8F6E-73D1BA0ACF8E@redivi.com>


On Mar 22, 2006, at 1:26 PM, Louis Pecora wrote:

> A few months ago I installed (at the prodding of gurus here) Python
> 2.4.1.  I tried to use it today (haven't used it for a while since the
> install) and it cannot find Numeric.  I checked and I have no 2.4
> directory (under /Library) for site packages.  They're all in the 2.3
> directory.  Likewise I have no ~/Library/python/2.4 directory only
> ~/Library/python/2.3 for my .pth file.   Of course, if I run Python  
> 2.3
> it finds all the modules and my stuff.

site-packages for official builds are in the framework itself.  Only  
Apple's builds have site-packages at /Library/Python.

> Should I just create the 2.4 directories and copy the stuff from 2.3
> into 2.4?

Creating the ~/Library/Python/2.4 directory is optional, you can go  
ahead and do that.  Copying 2.3 stuff over is not going to work though.

-bob


From rsfinn at gmail.com  Wed Mar 22 23:20:12 2006
From: rsfinn at gmail.com (Russell Finn)
Date: Wed, 22 Mar 2006 17:20:12 -0500
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <4421A6A5.7070602@noaa.gov>
References: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
	<4421A6A5.7070602@noaa.gov>
Message-ID: <a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>

On 3/22/06, Christopher Barker <Chris.Barker at noaa.gov> wrote:
> 2) you mentioned a programmatic API: In fact, I think there is a plist
> somewhere that sets the default environment variables for all apps, you
> could add a PATH entry (or add to it) programmaticly easily enough.

Yes, it's ~/.MacOSX/environment.plist; see Apple's QA1067
<http://developer.apple.com/qa/qa2001/qa1067.html>.

Perhaps the new Python installer should set this location instead of
.profile (which doesn't help people using csh/tcsh)?  (Be sure to
include the "standard" path "/usr/bin:/bin:/usr/sbin:/sbin" as well --
the settings in the .plist file are *not* cumulative.)

> 3) PATH is only relevant to command line apps: if you've never gone back
> from using a GUI -- you'll never need to care about this.

But I would guess a number of Python-based GUI applications probably
use command-line utilities under the covers, which is why a global
setting is useful.

-- Russell Finn

From njriley at uiuc.edu  Wed Mar 22 23:36:18 2006
From: njriley at uiuc.edu (Nicholas Riley)
Date: Wed, 22 Mar 2006 16:36:18 -0600
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>
References: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
	<4421A6A5.7070602@noaa.gov>
	<a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>
Message-ID: <20060322223618.GA78522@uiuc.edu>

On Wed, Mar 22, 2006 at 05:20:12PM -0500, Russell Finn wrote:
> On 3/22/06, Christopher Barker <Chris.Barker at noaa.gov> wrote:
> > 2) you mentioned a programmatic API: In fact, I think there is a plist
> > somewhere that sets the default environment variables for all apps, you
> > could add a PATH entry (or add to it) programmaticly easily enough.
> 
> Yes, it's ~/.MacOSX/environment.plist; see Apple's QA1067
> <http://developer.apple.com/qa/qa2001/qa1067.html>.
> 
> Perhaps the new Python installer should set this location instead of
> .profile (which doesn't help people using csh/tcsh)?  (Be sure to
> include the "standard" path "/usr/bin:/bin:/usr/sbin:/sbin" as well --
> the settings in the .plist file are *not* cumulative.)

Please, no.  Editing the shell init files is bad enough (but necessary
to support users who aren't aware of the difference between /usr and
/usr/local); really, there should be a better way of doing this.  Even
worse is changing the path somewhere hardly anyone knows where to
look.

> > 3) PATH is only relevant to command line apps: if you've never gone back
> > from using a GUI -- you'll never need to care about this.
> 
> But I would guess a number of Python-based GUI applications probably
> use command-line utilities under the covers, which is why a global
> setting is useful.

If Mac GUI apps depend on custom environment variables or a paritcular
PATH being set, they're broken.  They should use command-line
utilities inside their bundles or frameworks, which py2app and friends
make very easy to accomplish.

-- 
Nicholas Riley <njriley at uiuc.edu> | <http://www.uiuc.edu/ph/www/njriley>

From stewart.midwinter at gmail.com  Thu Mar 23 00:18:37 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Wed, 22 Mar 2006 16:18:37 -0700
Subject: [Pythonmac-SIG] path locations
Message-ID: <d4c0d4800603221518k723b8fdel763b4928601da20a@mail.gmail.com>

> From: Christopher Barker <Chris.Barker at noaa.gov>
> Here's where you've deviated from the standard: As a rule, you shouldn't
> out ANYTHING in /usr/bin. that's Apple's job. As you haven't written
> over anything, it's probably not going to cause any problems, but the
> usual solution is to add /usr/local/bin to your PATH instead

Thanks for the reminder.  I had a feeling that we shouldn't put stuff
in /sw/bin, and I am used to using /usr/local/bin on my Linux box, but
it's good to know that others are doing this too.

cheers
S

From stewart.midwinter at gmail.com  Thu Mar 23 00:36:50 2006
From: stewart.midwinter at gmail.com (Stewart Midwinter)
Date: Wed, 22 Mar 2006 16:36:50 -0700
Subject: [Pythonmac-SIG] Matplotlib give bus error (10.3.9)
Message-ID: <d4c0d4800603221536k5eb656d3y746d836a5c864f1b@mail.gmail.com>

---------- Forwarded message ----------
From: Tom Loredo <loredo at astro.cornell.edu>

>mpl built and installed fine, but gave
a bus error when trying to plot ("import matplotlib" would
work fine, but "import pylab" would give the error).

Tom, I came across a bus error in an unrelated package, wax (a
pythonic wrapper on wxPython).  I was eventually able to track it down
to a problem with sys.stdout on the Mac. Then I was able to code a
workaround so that the problem statement doesn't get executed on OS X.
 I did that by inserting print statements (print "got to 1" - and so
on) throughout the problem module and noting after which 'milepost'
the exception occurred.  You could try sticking some of those in your
module and see what you get.

cheers
S

From lanceboyle at qwest.net  Thu Mar 23 01:20:45 2006
From: lanceboyle at qwest.net (lanceboyle at qwest.net)
Date: Wed, 22 Mar 2006 17:20:45 -0700
Subject: [Pythonmac-SIG] matplotlib and AquaTerm
Message-ID: <200E6B10-3F9C-41BB-90F4-12485B52C691@qwest.net>

Does matplotlib work with AquaTerm as a backend?

Jerry


From abarnert at yahoo.com  Thu Mar 23 00:31:04 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Wed, 22 Mar 2006 15:31:04 -0800 (PST)
Subject: [Pythonmac-SIG] Questions about various bits of Intel status
In-Reply-To: <4876C93F-50A0-419B-A1A1-EEE5EB265A91@mac.com>
Message-ID: <20060322233104.82254.qmail@web82303.mail.mud.yahoo.com>

Thanks for the quick answers, and it's good to know
that I'm not missing information right in front of my
face....

Me:
> > First, what's up with ctypes and other things that
> > need libffi? 

Ronald:
> PyObjC contains a port of libffi to darwin/x86, but
sadly
> enough I didn't pay enough attention in the time
between
> WWDC'05 and the release of OSX/x86. My port of
libffi
> no longer works due to slight changes in the calling
> conventions. I'll be fixing this shortly, and push
those
> upstreams and to ctypes.

OK, I'll wait....

By the way, I think I remember seeing some other
project with a libffi port around the same time (late
summer). Maybe pnet or mono or something related? Are
you sharing changes, or just duplication each other's
work? For that matter, is there anyone maintaining
libffi itself, or is it just one of those many
projects that Redhat abandoned and nobody's picked it
up even though people still use it?

> > Next, have the python24-fat changes been ported to
the
> > 2.5 trunk?
 
> The changes have not been ported to the 2.5 tree
yet,
> but I intend to do so before the 2.5a1 release.

OK. I can use a pure-Intel build to play around with
2.5 changes, but your 2.4 universal for actual
development.

> > How exactly are you supposed to properly install a
> > framework build?

> Just install using the pkg? My build script in the
> python24-fat tree does a chmod after installing, but
> that's mostly because its predecessor also did that.

Yeah, that works, unless I want to try a patch, play
with recent trunk changes, use different configure
switches, install two different 2.4.2 versions in
parallel, etc.

I guess it's not that big a deal if you have to do
some chmod/chgrp stuff to make frameworkinstall happy,
since anyone who's going to install from source ought
to be able to figure it out. 

What are the actual problems with having a root/wheel
755 framework directory instead of root/admin 775? I
guess it means you can't install modules to
site-packages out of .pkg files? If it's important, it
would be nice if it were easier to do properly. 

I'm guessing it would help the darwinports and fink
guys out, too. Last I checked, fink wouldn't build
Python at all on Intel, and darwinports wouldn't do a
framework install--but even on PPC, they both ended up
with a root/wheel 755 framework directory.

> > * test_curses is skipped, even though curses
appears to work
> > * all tests that require networking are skipped,
even though
> >    at least some (maybe all) of the relevant
modules work
> > * test_re fails
> > * test_{unicode,unicodedata,codecs} fail
> 
> Dunno about these, they should work just fine. I
> haven't tried building the trunk on an intel box yet
though.

I tested a few more things. So far, all of the skipped
tests that I tried manually worked fine, and I'm not
sure why they're skipped. All of the failed tests that
I've tried actually are broken (like the
u'\u2000'.isspace test I mentioned before). And I
couldn't find bugs filed on any of them. Do people
usually file bugs this early in the dev process?

I might be able to figure some of these out; others,
probably not. I know that a Unicode space is supposed
to be a space, but when I disagree with the computer
about how some regexp should be parsed, it's usually
the computer that's right....

So if I can't figure out the problems, should I at
least gather the verbose test reports and file bugs?

> > * test_{aepack,applesingle,macostools} fail
> 
> Those are expected to fail, the python24-fat tree
> contains bugfixes for this.

OK, cool.


From ronaldoussoren at mac.com  Thu Mar 23 10:34:07 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 10:34:07 +0100
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <20060322223618.GA78522@uiuc.edu>
References: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
	<4421A6A5.7070602@noaa.gov>
	<a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>
	<20060322223618.GA78522@uiuc.edu>
Message-ID: <14715718.1143106447434.JavaMail.ronaldoussoren@mac.com>

 
On Wednesday, March 22, 2006, at 11:37PM, Nicholas Riley <njriley at uiuc.edu> wrote:

>On Wed, Mar 22, 2006 at 05:20:12PM -0500, Russell Finn wrote:
>> On 3/22/06, Christopher Barker <Chris.Barker at noaa.gov> wrote:
>> > 2) you mentioned a programmatic API: In fact, I think there is a plist
>> > somewhere that sets the default environment variables for all apps, you
>> > could add a PATH entry (or add to it) programmaticly easily enough.
>> 
>> Yes, it's ~/.MacOSX/environment.plist; see Apple's QA1067
>> <http://developer.apple.com/qa/qa2001/qa1067.html>.
>> 
>> Perhaps the new Python installer should set this location instead of
>> .profile (which doesn't help people using csh/tcsh)?  (Be sure to
>> include the "standard" path "/usr/bin:/bin:/usr/sbin:/sbin" as well --
>> the settings in the .plist file are *not* cumulative.)
>
>Please, no.  Editing the shell init files is bad enough (but necessary
>to support users who aren't aware of the difference between /usr and
>/usr/local); really, there should be a better way of doing this.  Even
>worse is changing the path somewhere hardly anyone knows where to
>look.

Editing shell init files actually works for *updating* the PATH, you can't
update the PATH using environment.plist, only replace it. That's very, very
bad because the user will loose his existing configuration.

BTW. Careful editing of shell files isn't too bad, anyone that really cares
about the content of those files can remove those edits and do the changes
in their own style.

>
>> > 3) PATH is only relevant to command line apps: if you've never gone back
>> > from using a GUI -- you'll never need to care about this.
>> 
>> But I would guess a number of Python-based GUI applications probably
>> use command-line utilities under the covers, which is why a global
>> setting is useful.
>
>If Mac GUI apps depend on custom environment variables or a paritcular
>PATH being set, they're broken.  They should use command-line
>utilities inside their bundles or frameworks, which py2app and friends
>make very easy to accomplish.

Any GUI that depends on the value of PATH to reach command-line tools
is broken. 

Ronald

>
>-- 
>Nicholas Riley <njriley at uiuc.edu> | <http://www.uiuc.edu/ph/www/njriley>
>_______________________________________________
>Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>http://mail.python.org/mailman/listinfo/pythonmac-sig
>
>

From ronaldoussoren at mac.com  Thu Mar 23 10:36:20 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 10:36:20 +0100
Subject: [Pythonmac-SIG] py2app is building semi-standalone package only.
In-Reply-To: <a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>
References: <06Mar22.105557pst.58633@synergy1.parc.xerox.com>
	<4421A6A5.7070602@noaa.gov>
	<a47214750603221420w134fe5e7wa0176c19ad6eb686@mail.gmail.com>
Message-ID: <757185.1143106580445.JavaMail.ronaldoussoren@mac.com>

 
On Wednesday, March 22, 2006, at 11:21PM, Russell Finn <rsfinn at gmail.com> wrote:

>On 3/22/06, Christopher Barker <Chris.Barker at noaa.gov> wrote:
>> 2) you mentioned a programmatic API: In fact, I think there is a plist
>> somewhere that sets the default environment variables for all apps, you
>> could add a PATH entry (or add to it) programmaticly easily enough.
>
>Yes, it's ~/.MacOSX/environment.plist; see Apple's QA1067
><http://developer.apple.com/qa/qa2001/qa1067.html>.
>
>Perhaps the new Python installer should set this location instead of
>.profile (which doesn't help people using csh/tcsh)?  (Be sure to
>include the "standard" path "/usr/bin:/bin:/usr/sbin:/sbin" as well --
>the settings in the .plist file are *not* cumulative.)

That's why my installer will never update this file, its too easy to lose
existing settings. 

And the installer does update the shell init files for whatever is the shell
of the user installing python, it will update the csh environment for anyone
that uses that shell.

Ronald

From pecora at anvil.nrl.navy.mil  Thu Mar 23 12:45:19 2006
From: pecora at anvil.nrl.navy.mil (Louis Pecora)
Date: Thu, 23 Mar 2006 06:45:19 -0500
Subject: [Pythonmac-SIG] I have Python 2.4.1 but no site packages. What
 did I miss?
Message-ID: <44228A4F.7090802@anvil.nrl.navy.mil>

Bob Ippolito wrote:
> On Mar 22, 2006, at 1:26 PM, Louis Pecora wrote:
>
>  
>> A few months ago I installed (at the prodding of gurus here) Python
>> 2.4.1.  I tried to use it today (haven't used it for a while since the
>> install) and it cannot find Numeric.  I checked and I have no 2.4
>> directory (under /Library) for site packages.  They're all in the 2.3
>> directory.  Likewise I have no ~/Library/python/2.4 directory only
>> ~/Library/python/2.3 for my .pth file.   Of course, if I run Python  2.3
>> it finds all the modules and my stuff.
>>     
>
> site-packages for official builds are in the framework itself.  Only  
> Apple's builds have site-packages at /Library/Python.
>
>  
>> Should I just create the 2.4 directories and copy the stuff from 2.3
>> into 2.4?
>>     
>
> Creating the ~/Library/Python/2.4 directory is optional, you can go  
> ahead and do that.  Copying 2.3 stuff over is not going to work though.
>
> -bob
>
>   

Yes, I tried creating the 2.4 directories, but that didn't work.  So 
what am I to do?  Python 2.4.1 cannot find Numeric, wxPython or anything 
in the 2.3 directories that Python 2.3 can find.  import just gives 
Traceback errors.  Python 2.4.1 cannot find my own modules (which my 
.pth file in ~/Library/2.3 points to).
Am I back to messing with the .plist or $PATH or what?  I thought I had 
this all figured out, but it seems whenever I change anything (e.g. 
install 2.4.1) I am back to square one.
Any help appreciated.

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pecora at anvil.nrl.navy.mil


From cwmoad at gmail.com  Thu Mar 23 14:27:09 2006
From: cwmoad at gmail.com (Charlie Moad)
Date: Thu, 23 Mar 2006 08:27:09 -0500
Subject: [Pythonmac-SIG] matplotlib and AquaTerm
In-Reply-To: <200E6B10-3F9C-41BB-90F4-12485B52C691@qwest.net>
References: <200E6B10-3F9C-41BB-90F4-12485B52C691@qwest.net>
Message-ID: <6382066a0603230527y1f432645x5dcef0394a1f0481@mail.gmail.com>

On 3/22/06, lanceboyle at qwest.net <lanceboyle at qwest.net> wrote:
> Does matplotlib work with AquaTerm as a backend?

Recently an AquaTerm backend was donated to matplotlib.  I haven't had
a chance to test or commit it to svn yet though.  I am attaching it
for anyone interested.

- Charlie
-------------- next part --------------
A non-text attachment was scrubbed...
Name: backend_aqt.py
Type: text/x-python
Size: 24331 bytes
Desc: not available
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060323/392a7261/attachment.py 

From pecora at anvil.nrl.navy.mil  Thu Mar 23 18:26:52 2006
From: pecora at anvil.nrl.navy.mil (Louis Pecora)
Date: Thu, 23 Mar 2006 12:26:52 -0500
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages (e.g.
 Numeric, wxPython, etc.)
Message-ID: <4422DA5C.5020801@anvil.nrl.navy.mil>

Info on converting over installed Python packages to non-Apple Python 
for those who are newbies or are not gurus (includes me):

Bob Ippolito was right.  When you upgrade to a non-Apple python you have 
to re-install all the "add-on" packages like Numeric.  You also have to 
set your paths in the Environment.plist in ~/.MacOSX since any .pth 
files you formerly used in the Apple distribution directories will not 
be used in the upgrade.  You can get new versions of packages to install 
from Bob Ippolito's site:  http://pythonmac.org/packages/.  Just be sure 
to choose the Framework builds for the system you have on your machine 
(Panther or Tiger). 

Many thanks to Bob Ippolito for supplying those packages.

If anyone upgrades and needs to use the kinds.py  package (which gives 
platform limits and sizes for floats, ints, etc.) see below for more 
info on how to get this to work in Python 2.4.


Reinstalling the kinds package.  I assume you already have the package 
in your Python files directory.  You need to reinstall it after you 
upgrade to Python 2.4.  Do this by opening the terminal, cd to the kinds 
folder and type

sudo python setup.py install

give the admin password when prompted.

That puts the needed files in the right place down in the /Library 
directory.  But you have to set up a path down in that directory so when 
you run Python it can find the files it needs for kinds (this is in 
addition to adding the path to the original kinds source directory where 
you ran the install script above).  Create the file kinds.pth with the 
single line in it,

kinds

Save and move this to the directory  
Drive/Library/Framework/Python.framework/Version/2.4/lib/python2.4/site-packages.  
That should do it. 


-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pecora at anvil.nrl.navy.mil


From Chris.Barker at noaa.gov  Thu Mar 23 18:53:16 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 23 Mar 2006 09:53:16 -0800
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <4422DA5C.5020801@anvil.nrl.navy.mil>
References: <4422DA5C.5020801@anvil.nrl.navy.mil>
Message-ID: <4422E08C.9060604@noaa.gov>

Lou,

Thanks for the summary, you beat me to it. I'm a  bit surprised that you 
didn't already know this -- I thought is was made pretty darn clear that 
you needed to re-install all your add-on packages when you installed an 
upgraded Python. In fact, that's one of the reasons' not to use Apple's 
Python -- when they upgrade, it you'll need to re-install everything.

I guess we still need to work on the docs.

> Many thanks to Bob Ippolito for supplying those packages.

Bob built many (most?) of them, but some were built by others, and given 
to Bob to put on the site. I say this because:

> Reinstalling the kinds package.  I assume you already have the package 
> in your Python files directory.  You need to reinstall it after you 
> upgrade to Python 2.4.  Do this by opening the terminal, cd to the kinds 
> folder and type
> 
> sudo python setup.py install

Why don't you build a package out of this? If you install Py2app, you 
can run: bdist_mpkg and you'll get a nice package that you can give to 
Bob to put on the site.

When we get the released version of the Universal binary, I hope many 
folks, like you and I, will make packages  and contribute them. I think 
a large collection of ready to install packages for OS-X will do a lot 
to lower the barrier to entry for new Python users.

> Create the file kinds.pth with the 
> single line in it,

This is a bit odd. These days, most packages are built to not need a 
*.pth file. I think all you need to do is have a __init__.py file in the 
package directory. I'd encourage the kinds developers to update their setup.

by the way, what is kinds? the word "kinds" is so common I'm having a 
hard time finding it with Google.

-Chris

-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From rowen at cesmail.net  Thu Mar 23 19:16:26 2006
From: rowen at cesmail.net (Russell E. Owen)
Date: Thu, 23 Mar 2006 10:16:26 -0800
Subject: [Pythonmac-SIG] I have Python 2.4.1 but no site packages. What
	did I miss?
References: <44228A4F.7090802@anvil.nrl.navy.mil>
Message-ID: <rowen-A8C06F.10162623032006@sea.gmane.org>

In article <44228A4F.7090802 at anvil.nrl.navy.mil>,
 Louis Pecora <pecora at anvil.nrl.navy.mil> wrote:

> Yes, I tried creating the 2.4 directories, but that didn't work.  So 
> what am I to do?  Python 2.4.1 cannot find Numeric, wxPython or anything 
> in the 2.3 directories that Python 2.3 can find.  import just gives 
> Traceback errors.  Python 2.4.1 cannot find my own modules (which my 
> .pth file in ~/Library/2.3 points to).
> Am I back to messing with the .plist or $PATH or what?  I thought I had 
> this all figured out, but it seems whenever I change anything (e.g. 
> install 2.4.1) I am back to square one.
> Any help appreciated.

Pure python packages can be simply copied over (as long as they have the 
original .py source files), but all others (e.g. packages with C 
extensions or with source missing) have to be rebuilt when going from 
Python 2.3 to 2.4 or any jump of that magnitude.

Packages you will have to reinstall include numarray, Numeric, numpy and 
PIL (these all have C extensions). You can find a lot of useful prebuilt 
binaries here <http://pythonmac.org/packages/>. Also, many packages 
build with no problems from source (including Numeric, numarray 
and--last time I tried--numpy).

Regarding your own modules: if they are pure python, just put a copy of 
your .pth file in python 2.4's site packages. On my computer the path is:
/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-pack
ages/
but if in doubt, run python 2.4, import sys and print sys.path

-- Russell


From boyle5 at llnl.gov  Thu Mar 23 19:24:32 2006
From: boyle5 at llnl.gov (James Boyle)
Date: Thu, 23 Mar 2006 10:24:32 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
Message-ID: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>

I'm not sure whether to post this to this list or the numpy list but 
I'll start here.

I am running OX 10.3.9. I installed python 2.4.2 using Universal 
MacPython 2.4.2.dmg.

When I try to install numpy 0.9.6 it fails with the message.
.
.

gcc: _configtest.c
gcc: cannot specify -o with -c or -S and multiple compilations
gcc: cannot specify -o with -c or -S and multiple compilations
failure.
.
.
I am running gcc 3.3. I note that the python 2.4.2 was built using gcc 
4.0.1. Could this difference in gcc versions be the problem?
At present I am at a loss as to what to try next.

--Jim

gcc: _configtest.c
gcc: cannot specify -o with -c or -S and multiple compilations
gcc: cannot specify -o with -c or -S and multiple compilations
failure.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 867 bytes
Desc: not available
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060323/99a807c1/attachment.bin 

From hengist.podd at virgin.net  Thu Mar 23 20:13:24 2006
From: hengist.podd at virgin.net (has)
Date: Thu, 23 Mar 2006 19:13:24 +0000
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
Message-ID: <a05200f00c0489f4e8f50@[82.9.20.25]>

Hi all,

Appscript updates are now available at:

http://freespace.virgin.net/hamish.sanderson/appscript.html

This release adds Intel Mac support [1] and changes the license from LGPL to MIT.

Many thanks to Ben Artin for providing Intel compatibility patches.

If anyone has any problems, please let me know.

has

[1] Re. Python 2.4.x on Intel Macs: I'll make up a binary appscript installer once the universal Python 2.4.2 installer is officially released. Meantime you should be able to build from source.
-- 
http://freespace.virgin.net/hamish.sanderson/

From markus at johalla.de  Thu Mar 23 20:05:56 2006
From: markus at johalla.de (Markus)
Date: Thu, 23 Mar 2006 20:05:56 +0100
Subject: [Pythonmac-SIG] egg question
Message-ID: <745EE540-4569-4B43-8472-F624C9A351F9@johalla.de>

Hi!

I've installed 2.4.2 Universal on a machine which has 2.4.1 from fink  
installed. This worked very well. I had to install some packages with  
'python setup.py install' but no problem here.
Then I wanted sqlobject and I tried 'easy_install -U sqlobject' which  
told, that its already installed under /sw/bin/ (the fink path)

Hmm, ok. I reinstalled the setuptools, but it still ends in /sw/bin.

Now I don't know what to do or where to look.

thanks for your help
markus


Homer:/Users/markus/Documents/Python root# python ez_setup.py
Downloading http://cheeseshop.python.org/packages/2.4/s/setuptools/ 
setuptools-0.6a10-py2.4.egg
Processing setuptools-0.6a10-py2.4.egg
creating /Library/Frameworks/Python.framework/Versions/2.4/lib/ 
python2.4/site-packages/setuptools-0.6a10-py2.4.egg
Extracting setuptools-0.6a10-py2.4.egg to /Library/Frameworks/ 
Python.framework/Versions/2.4/lib/python2.4/site-packages
Adding setuptools 0.6a10 to easy-install.pth file
Installing easy_install script to /Library/Frameworks/ 
Python.framework/Versions/2.4/bin

Installed /Library/Frameworks/Python.framework/Versions/2.4/lib/ 
python2.4/site-packages/setuptools-0.6a10-py2.4.egg
Processing dependencies for setuptools==0.6a10
Setuptools version 0.6a10 or greater has been installed.
(Run "ez_setup.py -U setuptools" to reinstall or upgrade.)
Homer:/Users/markus/Documents/Python root# python ez_setup.py -U  
setuptools
Searching for setuptools
Reading http://www.python.org/pypi/setuptools/
Reading http://peak.telecommunity.com/DevCenter/setuptools
Best match: setuptools 0.6a10
Processing setuptools-0.6a10-py2.4.egg
setuptools 0.6a10 is already the active version in easy-install.pth
Installing easy_install script to /Library/Frameworks/ 
Python.framework/Versions/2.4/bin

Using /Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/ 
site-packages/setuptools-0.6a10-py2.4.egg
Processing dependencies for setuptools
Homer:/Users/markus/Documents/Python root# easy_install -U sqlobject
Searching for sqlobject
Reading http://www.python.org/pypi/sqlobject/
Couldn't find index page for 'sqlobject' (maybe misspelled?)
Scanning index of all packages (this may take a while)
Reading http://www.python.org/pypi/
Reading http://www.python.org/pypi/SQLObject/0.7.0
Reading http://sqlobject.org
Best match: SQLObject 0.7.0
Processing SQLObject-0.7.0-py2.4.egg
SQLObject 0.7.0 is already the active version in easy-install.pth
Installing sqlobject-admin script to /sw/bin

Using /sw/lib/python2.4/site-packages/SQLObject-0.7.0-py2.4.egg
Processing dependencies for sqlobject


From pecora at anvil.nrl.navy.mil  Thu Mar 23 20:35:36 2006
From: pecora at anvil.nrl.navy.mil (Louis Pecora)
Date: Thu, 23 Mar 2006 14:35:36 -0500
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <4422E08C.9060604@noaa.gov>
References: <4422DA5C.5020801@anvil.nrl.navy.mil> <4422E08C.9060604@noaa.gov>
Message-ID: <4422F888.5000800@anvil.nrl.navy.mil>

Christopher Barker wrote:
> Why don't you build a package out of this? If you install Py2app, you 
> can run: bdist_mpkg and you'll get a nice package that you can give to 
> Bob to put on the site.
>
>
OK, I tried this.  installed Py2app (2.4.1 Frameworks version) no 
problem.  But now I have trouble using bdist_mpkg.  I tried two ways 
after cd to the folder with the setup.py file, one from the terminal 
prompt and one inside python (the web page is not very clear to me on 
this or I'm just dense).  I get errors both ways:


louispec% python setup.py bdist_mpkg --open
usage: setup.py [global_opts] cmd1 [cmd1_opts] [cmd2 [cmd2_opts] ...]
   or: setup.py --help [cmd1 cmd2 ...]
   or: setup.py --help-commands
   or: setup.py cmd --help

error: invalid command 'bdist_mpkg'
Exit 1

louispec% python
Python 2.4.1 (#2, Mar 31 2005, 00:05:10)
[GCC 3.3 20030304 (Apple Computer, Inc. build 1666)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import bdist_mpkg
 >>> python setup.py bdist_mpkg --open
  File "<stdin>", line 1
    python setup.py bdist_mpkg --open
               ^
SyntaxError: invalid syntax


Any ideas? 


-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pecora at anvil.nrl.navy.mil


From Chris.Barker at noaa.gov  Thu Mar 23 20:46:29 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 23 Mar 2006 11:46:29 -0800
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <4422F888.5000800@anvil.nrl.navy.mil>
References: <4422DA5C.5020801@anvil.nrl.navy.mil> <4422E08C.9060604@noaa.gov>
	<4422F888.5000800@anvil.nrl.navy.mil>
Message-ID: <4422FB15.7010309@noaa.gov>

Louis Pecora wrote:
> Any ideas? 

yup.

Py2App should have installed a script in /usr/local/bin called 
bdist_mpkg. If that's on your PATH, you should be able to run:

bdist_mpkg

in the directory that your setup.py lives in.

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From leknarf at pacbell.net  Thu Mar 23 20:53:47 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Thu, 23 Mar 2006 11:53:47 -0800
Subject: [Pythonmac-SIG] objc methods to pythonese
Message-ID: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>


Apologies in advance for this tangential question.

I'm trying to wrap my head around how Objective-C methods are  
constructed -- and more importantly, how they're converted to Python  
methods.  Doco I've found uses as an example a method with white  
space between the parts of the method name.  Working with Python has  
made me gun shy with respect to white space ;)

Is the white space syntactically significant?  i.e.:

     [rectangle setWidth:width height:height];
                                                  ^

If I understand correctly, this would translate in Python to:

     rectangle = setWidthheight_(width, height)

Correct that the 'h' in the "height" part of the method name is not  
capitalized?



How about the following?

     - (id)tableView:(NSTableView *)tableView
     objectValueForTableColumn(NSTableColumn *)tableColumn
             row:(int)row

Would this translate like so?

     def tableViewobjectValueForTableColumnrow_(tableView,  
tableColumn, row):




Thanks in advance!  Reminds me of a friend's reverse polish notation  
calculator ...
Scott








From zbir at urbanape.com  Thu Mar 23 21:11:50 2006
From: zbir at urbanape.com (Zachery Bir)
Date: Thu, 23 Mar 2006 15:11:50 -0500
Subject: [Pythonmac-SIG] objc methods to pythonese
In-Reply-To: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
References: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
Message-ID: <1A9BA86D-BA1C-4A2F-AF46-B92308F0F051@urbanape.com>

On Mar 23, 2006, at 2:53 PM, Scott Frankel wrote:

> Apologies in advance for this tangential question.
>
> I'm trying to wrap my head around how Objective-C methods are
> constructed -- and more importantly, how they're converted to Python
> methods.  Doco I've found uses as an example a method with white
> space between the parts of the method name.  Working with Python has
> made me gun shy with respect to white space ;)
>
> Is the white space syntactically significant?  i.e.:
>
>      [rectangle setWidth:width height:height];
>                                                   ^
>
> If I understand correctly, this would translate in Python to:
>
>      rectangle = setWidthheight_(width, height)
>
> Correct that the 'h' in the "height" part of the method name is not
> capitalized?

You're not assigning the result of that method to 'rectangle'. You want:

   rectangle.setWidth_height_(width, height)

each ':' gets replaced in the python signature with an underscore,  
and 'rectangle' is the object you're calling setWidth:height: on.

> How about the following?
>
>      - (id)tableView:(NSTableView *)tableView
>      objectValueForTableColumn(NSTableColumn *)tableColumn
>              row:(int)row
>
> Would this translate like so?
>
>      def tableViewobjectValueForTableColumnrow_(tableView,
> tableColumn, row):

That is a method declaration, not a function - don't forget 'self':

   def tableView_objectValueForTableColumn_row_(self, tableView,  
tableColumn, row):

Zac


From ronaldoussoren at mac.com  Thu Mar 23 21:11:15 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 21:11:15 +0100
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
Message-ID: <4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>


On 23-mrt-2006, at 19:24, James Boyle wrote:

> I'm not sure whether to post this to this list or the numpy list  
> but I'll start here.
>
> I am running OX 10.3.9. I installed python 2.4.2 using Universal  
> MacPython 2.4.2.dmg.
>
> When I try to install numpy 0.9.6 it fails with the message.
> .
> .
>
> gcc: _configtest.c
> gcc: cannot specify -o with -c or -S and multiple compilations
> gcc: cannot specify -o with -c or -S and multiple compilations
> failure.
> .
> .
> I am running gcc 3.3. I note that the python 2.4.2 was built using  
> gcc 4.0.1. Could this difference in gcc versions be the problem?
> At present I am at a loss as to what to try next.

Could you send the entire output? Compiling on 10.3.9 is suppossed to  
work.

Ronald

>
> --Jim
>
> gcc: _configtest.c
> gcc: cannot specify -o with -c or -S and multiple compilations
> gcc: cannot specify -o with -c or -S and multiple compilations
> failure.
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Thu Mar 23 21:14:30 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 23 Mar 2006 12:14:30 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
Message-ID: <442301A6.7070305@noaa.gov>

James Boyle wrote:
> I am running gcc 3.3. I note that the python 2.4.2 was built using gcc 
> 4.0.1. Could this difference in gcc versions be the problem?
> At present I am at a loss as to what to try next.

It sure could. The thing to try is using gcc 4.0.1:

sudo gcc_select 4.x

You may need to upgrade your XCode tools to do this.

-Chris






-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Thu Mar 23 21:12:42 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 21:12:42 +0100
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <a05200f00c0489f4e8f50@[82.9.20.25]>
References: <a05200f00c0489f4e8f50@[82.9.20.25]>
Message-ID: <AF63BF7A-F6BB-4C14-BEFD-A84D85537E77@mac.com>


On 23-mrt-2006, at 20:13, has wrote:


>
> [1] Re. Python 2.4.x on Intel Macs: I'll make up a binary appscript  
> installer once the universal Python 2.4.2 installer is officially  
> released. Meantime you should be able to build from source.

It is officialy released, even if Bill hasn't updated the download  
page at python.org yet.

Ronald

> -- 
> http://freespace.virgin.net/hamish.sanderson/
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From leknarf at pacbell.net  Thu Mar 23 21:46:54 2006
From: leknarf at pacbell.net (Scott Frankel)
Date: Thu, 23 Mar 2006 12:46:54 -0800
Subject: [Pythonmac-SIG] objc methods to pythonese
In-Reply-To: <1A9BA86D-BA1C-4A2F-AF46-B92308F0F051@urbanape.com>
References: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
	<1A9BA86D-BA1C-4A2F-AF46-B92308F0F051@urbanape.com>
Message-ID: <075632C2-1D78-46AE-BF96-4627CDBF48C3@pacbell.net>


Thanks for the clarifications!
Scott



On Mar 23, 2006, at 12:06 PM, Jordan Krushen wrote:


> obj-c:
>     [rectangle setWidth:width height:height]
>
> python:
>     rectangle.setWidth_height_(width, height)


On Mar 23, 2006, at 12:11 PM, Zachery Bir wrote:

> You're not assigning the result of that method to 'rectangle'. You  
> want:
>
>   rectangle.setWidth_height_(width, height)
>
> each ':' gets replaced in the python signature with an underscore,  
> and 'rectangle' is the object you're calling setWidth:height: on.
>>
>>
>>      - (id)tableView:(NSTableView *)tableView
>>      objectValueForTableColumn(NSTableColumn *)tableColumn
>>              row:(int)row
>
> That is a method declaration, not a function - don't forget 'self':
>
>   def tableView_objectValueForTableColumn_row_(self, tableView,  
> tableColumn, row):
>
> Zac
>


From ronaldoussoren at mac.com  Thu Mar 23 22:22:33 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 22:22:33 +0100
Subject: [Pythonmac-SIG] Questions about various bits of Intel status
In-Reply-To: <20060322233104.82254.qmail@web82303.mail.mud.yahoo.com>
References: <20060322233104.82254.qmail@web82303.mail.mud.yahoo.com>
Message-ID: <A657CD11-8AE1-4184-B86A-06749F0F4794@mac.com>


On 23-mrt-2006, at 0:31, Andrew Barnert wrote:

> Thanks for the quick answers, and it's good to know
> that I'm not missing information right in front of my
> face....
>
> Me:
>>> First, what's up with ctypes and other things that
>>> need libffi?
>
> Ronald:
>> PyObjC contains a port of libffi to darwin/x86, but
> sadly
>> enough I didn't pay enough attention in the time
> between
>> WWDC'05 and the release of OSX/x86. My port of
> libffi
>> no longer works due to slight changes in the calling
>> conventions. I'll be fixing this shortly, and push
> those
>> upstreams and to ctypes.
>
> OK, I'll wait....
>
> By the way, I think I remember seeing some other
> project with a libffi port around the same time (late
> summer). Maybe pnet or mono or something related? Are
> you sharing changes, or just duplication each other's
> work? For that matter, is there anyone maintaining
> libffi itself, or is it just one of those many
> projects that Redhat abandoned and nobody's picked it
> up even though people still use it?

Libffi is part of the GCC source tree. I haven't tried
merging my port to darwin/x86 yet because of the general
hackishness of it at the time. Now that Apple has an
actual release of OSX/x86 I'll clean up my patch :-)

IIRC libffi is used in the java part of gcc (which isn't
used by Apple)

>
>>> Next, have the python24-fat changes been ported to
> the
>>> 2.5 trunk?
>
>> The changes have not been ported to the 2.5 tree
> yet,
>> but I intend to do so before the 2.5a1 release.
>
> OK. I can use a pure-Intel build to play around with
> 2.5 changes, but your 2.4 universal for actual
> development.
>
>>> How exactly are you supposed to properly install a
>>> framework build?
>
>> Just install using the pkg? My build script in the
>> python24-fat tree does a chmod after installing, but
>> that's mostly because its predecessor also did that.
>
> Yeah, that works, unless I want to try a patch, play
> with recent trunk changes, use different configure
> switches, install two different 2.4.2 versions in
> parallel, etc.
>
> I guess it's not that big a deal if you have to do
> some chmod/chgrp stuff to make frameworkinstall happy,
> since anyone who's going to install from source ought
> to be able to figure it out.
>
> What are the actual problems with having a root/wheel
> 755 framework directory instead of root/admin 775? I
> guess it means you can't install modules to
> site-packages out of .pkg files? If it's important, it
> would be nice if it were easier to do properly.

root/admin 775 makes it easier to install packages when
you're an admin user, you can do away with the call to
sudo.

>
> I'm guessing it would help the darwinports and fink
> guys out, too. Last I checked, fink wouldn't build
> Python at all on Intel, and darwinports wouldn't do a
> framework install--but even on PPC, they both ended up
> with a root/wheel 755 framework directory.

I don't use fink, and use darwinports only for unix-y tools,
IMHO neither should have to use framework installs :-)


>
>>> * test_curses is skipped, even though curses
> appears to work
>>> * all tests that require networking are skipped,
> even though
>>>    at least some (maybe all) of the relevant
> modules work
>>> * test_re fails
>>> * test_{unicode,unicodedata,codecs} fail
>>
>> Dunno about these, they should work just fine. I
>> haven't tried building the trunk on an intel box yet
> though.
>
> I tested a few more things. So far, all of the skipped
> tests that I tried manually worked fine, and I'm not
> sure why they're skipped. All of the failed tests that
> I've tried actually are broken (like the
> u'\u2000'.isspace test I mentioned before). And I
> couldn't find bugs filed on any of them. Do people
> usually file bugs this early in the dev process?

According to buildbot (www.python.org/dev/buildbot/trunk)
the trunk shouldn't have any test failures.

>
> I might be able to figure some of these out; others,
> probably not. I know that a Unicode space is supposed
> to be a space, but when I disagree with the computer
> about how some regexp should be parsed, it's usually
> the computer that's right....
>
> So if I can't figure out the problems, should I at
> least gather the verbose test reports and file bugs?
>
>>> * test_{aepack,applesingle,macostools} fail
>>
>> Those are expected to fail, the python24-fat tree
>> contains bugfixes for this.
>
> OK, cool.
>


Ronald


From ronaldoussoren at mac.com  Thu Mar 23 22:33:07 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 23 Mar 2006 22:33:07 +0100
Subject: [Pythonmac-SIG] objc methods to pythonese
In-Reply-To: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
References: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
Message-ID: <4DE394AE-E574-4CDF-BD0C-F5D7A797C738@mac.com>


On 23-mrt-2006, at 20:53, Scott Frankel wrote:

>
> Apologies in advance for this tangential question.
>
> I'm trying to wrap my head around how Objective-C methods are
> constructed -- and more importantly, how they're converted to Python
> methods.  Doco I've found uses as an example a method with white
> space between the parts of the method name.  Working with Python has
> made me gun shy with respect to white space ;)
>
> Is the white space syntactically significant?  i.e.:
>
>      [rectangle setWidth:width height:height];
>                                                   ^
>
> If I understand correctly, this would translate in Python to:
>
>      rectangle = setWidthheight_(width, height)
>
> Correct that the 'h' in the "height" part of the method name is not
> capitalized?

I have no idea what gave you that impression. The pyobjc documentation
clearly states that to translate a method name from Objective-C to
python you take the canonical ObjC name (all parts including colons
appended and without whitepace), e.g. setWidth:height:, and then
replace colons by underscores: setWidth_height_

>
>
>
> How about the following?
>
>      - (id)tableView:(NSTableView *)tableView
>      objectValueForTableColumn(NSTableColumn *)tableColumn
>              row:(int)row
>
> Would this translate like so?
>
>      def tableViewobjectValueForTableColumnrow_(tableView,
> tableColumn, row):

tableView_objectValueForTableColumn_row_

Ronald

>
>
>
>
> Thanks in advance!  Reminds me of a friend's reverse polish notation
> calculator ...
> Scott
>
>
>
>
>
>
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From jordan at krushen.com  Thu Mar 23 21:06:40 2006
From: jordan at krushen.com (Jordan Krushen)
Date: Thu, 23 Mar 2006 12:06:40 -0800
Subject: [Pythonmac-SIG] objc methods to pythonese
In-Reply-To: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
References: <5C962CAD-46CA-4903-9281-DB0D8EDE817B@pacbell.net>
Message-ID: <b811ba730603231206u228c1928q5d9ff1faa895e376@mail.gmail.com>

On 3/23/06, Scott Frankel <leknarf at pacbell.net> wrote:

> If I understand correctly, this would translate in Python to:
>
>      rectangle = setWidthheight_(width, height)

obj-c:
    [rectangle setWidth:width height:height]

python:
    rectangle.setWidth_height_(width, height)

> Correct that the 'h' in the "height" part of the method name is not
> capitalized?

Don't change capitalization.  Per the docs, just replace :'s with _'s
and concatenate everything.

> How about the following?
>
>      - (id)tableView:(NSTableView *)tableView
>      objectValueForTableColumn(NSTableColumn *)tableColumn
>              row:(int)row

tableView_objectValueForTableColumn_row_(tableView, tableColumn, row)

HTH,

J.

From abarnert at yahoo.com  Fri Mar 24 00:17:02 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Thu, 23 Mar 2006 15:17:02 -0800 (PST)
Subject: [Pythonmac-SIG] Questions about various bits of Intel status
In-Reply-To: <A657CD11-8AE1-4184-B86A-06749F0F4794@mac.com>
Message-ID: <20060323231702.12920.qmail@web82307.mail.mud.yahoo.com>

Andrew Barnert:
> > What are the actual problems with having a
root/wheel
> > 755 framework directory instead of root/admin 775?
I
> > guess it means you can't install modules to
> > site-packages out of .pkg files? If it's
important, it
> > would be nice if it were easier to do properly.

Ronald Oussoren:
> root/admin 775 makes it easier to install packages
when
> you're an admin user, you can do away with the call
to
> sudo.

Yeah, but adding sudo at the command line isn't that
big a deal. 

I was assuming there were more serious problems, like
installing from a .pkg file (you can't launch the
installer as root nearly as easily as you can sudo
python or make). And maybe others that I didn't
anticipate. Or maybe not.

Also, Apple clearly intends things in /Library to be
admin 775, so why violate that unless there's a good
reason?

If it's not that hard to change make frameworkinstall
to do the right thing, why not do it? 

If it'd be easier to understand a patch rather than an
explanation, just let me know.

If you understand what I'm getting at and I'm just
being stupid (which is quite possible), please tell me
what I'm missing.

[re: darwinports and frameworkinstall]
> > I'm guessing it would help the darwinports and
fink
> > guys out, too. Last I checked, fink wouldn't build
> > Python at all on Intel, and darwinports wouldn't
do a
> > framework install--but even on PPC, they both
ended up
> > with a root/wheel 755 framework directory.
> 
> I don't use fink, and use darwinports only for
unix-y tools,
> IMHO neither should have to use framework installs
:-)

OK, my writing was a bit unclear. The "Last I checked"
part was a parenthetical aside; the "root/wheel 755"
part was the main issue.

If the only problems with wheel 755 are with .pkg
files (which you're probably not going to use with
darwinports or fink) and needing sudo (which doesn't
matter, since you always sudo port install), I guess
that's no problem. But it would still be fixed
automatically if make frameworkinstall were changed.

As for the side issue with framework builds:

If darwinports is going to include a python2.4 that
installs Idle.app, and ports like py-appscript, it
needs to either build a framework or find some other
way to get a working pythonw. Otherwise you get an
Idle that won't run, an appscript that can't be used,
etc.

You could argue that darwinports shouldn't have these
ports at all, I guess. But if they have them, the
ports should work.

BTW, darwinports is actually pretty useful for Mac-y
tools as well as unix-y tools. When I first got my
Intel iMac, I was able to install Bwana-Dik, Smultron,
CocoaDialog, etc. in a few minutes, even though none
of them had an official Intel or universal build. No
different from svk, pngcrush, links, and other unix-y
ports.

[re: tests]
> According to buildbot
> (www.python.org/dev/buildbot/trunk)
> the trunk shouldn't have any test failures.

As I said, none of these failures happened on a linux
box, or a PPC Mac, so I assume it's something (or some
things) specific to Intel Macs. Since buildbot doesn't
have an Intel Mac, I wouldn't expect it to find them.

So, should I file bugs?


From janssen at parc.com  Fri Mar 24 00:56:46 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 23 Mar 2006 15:56:46 PST
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: Your message of "Thu, 23 Mar 2006 12:12:42 PST."
	<AF63BF7A-F6BB-4C14-BEFD-A84D85537E77@mac.com> 
Message-ID: <06Mar23.155649pst."58633"@synergy1.parc.xerox.com>

> It is officialy released, even if Bill hasn't updated the download  
> page at python.org yet.

I must be missing something about updating the new website's pages.

I've removed the old comment on the main downloads page several times,
but it keeps showing up.  In fact, looking at the svn history, it
comes back!

------------------------------------------------------------------------
r9570 | bill.janssen | 2006-03-23 15:52:21 -0800 (Thu, 23 Mar 2006) | 1 line

removed old comment about MacPython installer
------------------------------------------------------------------------
...
------------------------------------------------------------------------
r8986 | bill.janssen | 2006-02-27 11:00:46 -0800 (Mon, 27 Feb 2006) | 1 line

remove invalid comment about Mac python installer
------------------------------------------------------------------------

Guess I'll have to ask pydotorg about it.

Bill

From janssen at parc.com  Fri Mar 24 01:09:13 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 23 Mar 2006 16:09:13 PST
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: Your message of "Thu, 23 Mar 2006 12:12:42 PST."
	<AF63BF7A-F6BB-4C14-BEFD-A84D85537E77@mac.com> 
Message-ID: <06Mar23.160914pst."58633"@synergy1.parc.xerox.com>

> It is officialy released, even if Bill hasn't updated the download  
> page at python.org yet.

Ronald, what's the proper URL to advertise?  I've got

http://homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython%202.4.2.dmg

but perhaps you'd like to upload it to pythonmac.org, or somewhere,
before we link to it from the downloads page.

Bill

From bob at redivi.com  Fri Mar 24 01:13:45 2006
From: bob at redivi.com (Bob Ippolito)
Date: Thu, 23 Mar 2006 16:13:45 -0800
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <06Mar23.160914pst."58633"@synergy1.parc.xerox.com>
References: <06Mar23.160914pst."58633"@synergy1.parc.xerox.com>
Message-ID: <34CEA616-4AA4-4231-953B-D7012DD8DB5E@redivi.com>


On Mar 23, 2006, at 4:09 PM, Bill Janssen wrote:

>> It is officialy released, even if Bill hasn't updated the download
>> page at python.org yet.
>
> Ronald, what's the proper URL to advertise?  I've got
>
> http://homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython 
> %202.4.2.dmg
>
> but perhaps you'd like to upload it to pythonmac.org, or somewhere,
> before we link to it from the downloads page.

It really should end up on python.org with all of the other downloads.

-bob


From janssen at parc.com  Fri Mar 24 01:22:51 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 23 Mar 2006 16:22:51 PST
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: Your message of "Thu, 23 Mar 2006 16:13:45 PST."
	<34CEA616-4AA4-4231-953B-D7012DD8DB5E@redivi.com> 
Message-ID: <06Mar23.162254pst."58633"@synergy1.parc.xerox.com>

> On Mar 23, 2006, at 4:09 PM, Bill Janssen wrote:
> 
> >> It is officialy released, even if Bill hasn't updated the download
> >> page at python.org yet.
> >
> > Ronald, what's the proper URL to advertise?  I've got
> >
> > http://homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython 
> > %202.4.2.dmg
> >
> > but perhaps you'd like to upload it to pythonmac.org, or somewhere,
> > before we link to it from the downloads page.
> 
> It really should end up on python.org with all of the other downloads.
> 
> -bob
> 

I can do that.

Bill

From Chris.Barker at noaa.gov  Fri Mar 24 01:38:53 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 23 Mar 2006 16:38:53 -0800
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <4423063F.2070107@anvil.nrl.navy.mil>
References: <4422DA5C.5020801@anvil.nrl.navy.mil> <4422E08C.9060604@noaa.gov>
	<4422F888.5000800@anvil.nrl.navy.mil> <4422FB15.7010309@noaa.gov>
	<4423063F.2070107@anvil.nrl.navy.mil>
Message-ID: <44233F9D.9020000@noaa.gov>

Louis Pecora wrote:
> %  python setup.py bdist_mpkg --open
> 
> at the shell prompt.    Is that right?

I'm sorry I wasn't clear:

 >> bdist_mpkg

at the shell prompt, all by itself (it's a script) should do it. If 
there is a setup.py in the directory you type that in, it should use it.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Chris.Barker at noaa.gov  Fri Mar 24 01:44:55 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 23 Mar 2006 16:44:55 -0800
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <06Mar23.155649pst.58633@synergy1.parc.xerox.com>
References: <06Mar23.155649pst.58633@synergy1.parc.xerox.com>
Message-ID: <44234107.3040203@noaa.gov>

Bill Janssen wrote:
>> It is officialy released, even if Bill hasn't updated the download  
>> page at python.org yet.

Are we really ready to call it the "official" version to use? I just had 
a colleague try it, then he asked me how to get wxPython and numpy 
working with it.

I don't think there is a wxPython build for it yet (will the one for 
2.4.1 work?)

And he couldn't get numpy to build and install right.

I didn't try to debug that for him, instead I suggested to just use 
2.4.1 for a while longer.

If the new universal build is it now, we need to update 
pythonmac.org/packages with some information about what packages will 
work with it, and start building new universal packages!

I know I need:

wxPython
numpy
matplotlib

At the very least, so I'll help with those, though wxPython is beyond my 
skills.

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From janssen at parc.com  Fri Mar 24 02:51:53 2006
From: janssen at parc.com (Bill Janssen)
Date: Thu, 23 Mar 2006 17:51:53 PST
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: Your message of "Thu, 23 Mar 2006 16:44:55 PST."
	<44234107.3040203@noaa.gov> 
Message-ID: <06Mar23.175155pst."58633"@synergy1.parc.xerox.com>

> Are we really ready to call it the "official" version to use? I just had 
> a colleague try it, then he asked me how to get wxPython and numpy 
> working with it.

OK, I'll hold off on it till people start yelling at me to update the
page.  We still haven't seen Kevin's new icon ideas, either.

Bill

From ronaldoussoren at mac.com  Fri Mar 24 08:01:41 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Fri, 24 Mar 2006 08:01:41 +0100
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <06Mar23.162254pst.58633@synergy1.parc.xerox.com>
References: <06Mar23.162254pst.58633@synergy1.parc.xerox.com>
Message-ID: <54E81D36-D2F9-4AA1-958E-DCA10BCA9A6B@mac.com>


On 24-mrt-2006, at 1:22, Bill Janssen wrote:

>> On Mar 23, 2006, at 4:09 PM, Bill Janssen wrote:
>>
>>>> It is officialy released, even if Bill hasn't updated the download
>>>> page at python.org yet.
>>>
>>> Ronald, what's the proper URL to advertise?  I've got
>>>
>>> http://homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython
>>> %202.4.2.dmg
>>>
>>> but perhaps you'd like to upload it to pythonmac.org, or somewhere,
>>> before we link to it from the downloads page.
>>
>> It really should end up on python.org with all of the other  
>> downloads.
>>
>> -bob
>>
>
> I can do that.

But not yet :-). 2.4.3 will be released soon and I am working on a  
universal
binary version of that.

Ronald
>
> Bill


From Chris.Barker at noaa.gov  Fri Mar 24 18:16:54 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Fri, 24 Mar 2006 09:16:54 -0800
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <4423D739.1080208@anvil.nrl.navy.mil>
References: <4422DA5C.5020801@anvil.nrl.navy.mil> <4422E08C.9060604@noaa.gov>
	<4422F888.5000800@anvil.nrl.navy.mil> <4422FB15.7010309@noaa.gov>
	<4423063F.2070107@anvil.nrl.navy.mil> <44233F9D.9020000@noaa.gov>
	<4423D739.1080208@anvil.nrl.navy.mil>
Message-ID: <44242986.7030308@noaa.gov>

Louis Pecora wrote:

>> >> bdist_mpkg
>>
>> at the shell prompt, all by itself (it's a script) should do it. If 
>> there is a setup.py in the directory you type that in, it should use it.
>>
> Did that.  Python can't find bdist_mpkg .

Python can't find it?, or the shell can't find it? There is a script in 
usr/local/bin called bdist_mpkg and a module that should have been 
installed with Py2App also called bdist_mpkg. If you post your output, 
maybe we can figure out what's wrong with your installation.

For me, the bdist_mpkg script looks like this:

#!/usr/local/bin/python
from bdist_mpkg.scripts.script_bdist_mpkg import main
main()

Pretty simple. If I start up Python, I can see where the module is 
installed:

Python 2.4.1 (#2, Mar 31 2005, 00:05:10)
[GCC 3.3 20030304 (Apple Computer, Inc. build 1666)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import bdist_mpkg
 >>> bdist_mpkg.__file__
'/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site-packages/py2app/bdist_mpkg/__init__.pyc'

I suspect that you installed a PY2App that doesn't match the python you 
are using, so it's installed in the wrong place.

I'm really looking forward to when (soon!) we have the Universal 2.4.3 
and we can all just work hard to support that ONE version!

-Chris






-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From pecora at anvil.nrl.navy.mil  Fri Mar 24 19:28:00 2006
From: pecora at anvil.nrl.navy.mil (Louis Pecora)
Date: Fri, 24 Mar 2006 13:28:00 -0500
Subject: [Pythonmac-SIG] Upgrading to Python 2.4 and handling packages
 (e.g. Numeric, wxPython, etc.)
In-Reply-To: <44242986.7030308@noaa.gov>
References: <4422DA5C.5020801@anvil.nrl.navy.mil> <4422E08C.9060604@noaa.gov>
	<4422F888.5000800@anvil.nrl.navy.mil> <4422FB15.7010309@noaa.gov>
	<4423063F.2070107@anvil.nrl.navy.mil> <44233F9D.9020000@noaa.gov>
	<4423D739.1080208@anvil.nrl.navy.mil> <44242986.7030308@noaa.gov>
Message-ID: <44243A30.8050709@anvil.nrl.navy.mil>

Christopher Barker wrote:
> Louis Pecora wrote:
>
>>> >> bdist_mpkg
>>>
>>> at the shell prompt, all by itself (it's a script) should do it. If 
>>> there is a setup.py in the directory you type that in, it should use 
>>> it.
>>>
>> Did that.  Python can't find bdist_mpkg .
>
> Python can't find it?, or the shell can't find it? There is a script 
> in usr/local/bin called bdist_mpkg and a module that should have been 
> installed with Py2App also called bdist_mpkg. If you post your output, 
> maybe we can figure out what's wrong with your installation.
>

Duh.  I tried to run bdist_mpkg in Python.   I just tried it in the 
shell and it worked.  Even got the dist folder with the package 
installer in it!  It ran on my machine and installed kinds in the 
Frameworks/python/..../site-packages folder.  It did not install a new 
kinds.pth, but that was already there and maybe with that __init__ stuff 
you mentioned it's not needed.  I don't know the latter.

So, generally, it seems to work.  Thanks for guiding me through it.

Now, dumb question.  How do I get it to Bob to put it up on the site (I 
assume it will go under Framework build, Tiger 10.4.  I need some more 
guidance on this step.  Also I have no web site to point to for an 
explanation like the other items have.  I might be able to write one, 
but not soon.  Not sure what to do about that.

-- 
Cheers,

Lou Pecora

Code 6362
Naval Research Lab
Washington, DC  20375
USA
Ph:  +202-767-6002
email:  pecora at anvil.nrl.navy.mil


From abarnert at yahoo.com  Fri Mar 24 21:14:58 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Fri, 24 Mar 2006 12:14:58 -0800 (PST)
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <54E81D36-D2F9-4AA1-958E-DCA10BCA9A6B@mac.com>
Message-ID: <20060324201458.29314.qmail@web82302.mail.mud.yahoo.com>

It might be worth putting up something visible anyway.


Right now, if you have an Intel Mac, you don't want
to/know how to compile Python yourself, and you don't
search the archives of this list, your only choices
are ActiveState, Darwinports (non-framework, and other
issues), or Fink (completely broken install). 

The page could say something like this:

> On Mac OS X 10.4 (Tiger), download and run this
installer (which
> gives you a PPC-only version of Python 2.4.1), and
then download
> and run this other installer (which makes your new
Python co-exist
> peacefully with the pre-installed version of
Python). 

> We're working on a Universal build installer for
Python 2.4.2, and
> plan to have it up shortly. In the meantime, if you
have an Intel
> Macintosh, you can download and run this prerelease
installer, but
> be aware that many packages have not yet been
converted to
> universal builds.

--- Ronald Oussoren <ronaldoussoren at mac.com> wrote:

> 
> On 24-mrt-2006, at 1:22, Bill Janssen wrote:
> 
> >> On Mar 23, 2006, at 4:09 PM, Bill Janssen wrote:
> >>
> >>>> It is officialy released, even if Bill hasn't
> updated the download
> >>>> page at python.org yet.
> >>>
> >>> Ronald, what's the proper URL to advertise? 
> I've got
> >>>
> >>>
>
http://homepage.mac.com/ronaldoussoren/.Public/Universal%20MacPython
> >>> %202.4.2.dmg
> >>>
> >>> but perhaps you'd like to upload it to
> pythonmac.org, or somewhere,
> >>> before we link to it from the downloads page.
> >>
> >> It really should end up on python.org with all of
> the other  
> >> downloads.
> >>
> >> -bob
> >>
> >
> > I can do that.
> 
> But not yet :-). 2.4.3 will be released soon and I
> am working on a  
> universal
> binary version of that.
> 
> Ronald
> >
> > Bill
> 
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
>
http://mail.python.org/mailman/listinfo/pythonmac-sig
> 


From rowen at cesmail.net  Fri Mar 24 21:19:23 2006
From: rowen at cesmail.net (Russell E. Owen)
Date: Fri, 24 Mar 2006 12:19:23 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
Message-ID: <rowen-255055.12192324032006@sea.gmane.org>

In article <4D6DFDB5-B044-4637-9CE9-4E886998A43F at mac.com>,
 Ronald Oussoren <ronaldoussoren at mac.com> wrote:

> On 23-mrt-2006, at 19:24, James Boyle wrote:
>... 
> > I am running OX 10.3.9. I installed python 2.4.2 using Universal  
> > MacPython 2.4.2.dmg.
> >
> > When I try to install numpy 0.9.6 it fails ...
> > .
> > I am running gcc 3.3. I note that the python 2.4.2 was built using  
> > gcc 4.0.1. Could this difference in gcc versions be the problem?
> > At present I am at a loss as to what to try next.
> 
> Could you send the entire output? Compiling on 10.3.9 is suppossed to  
> work.

This may not be relevant, but numpy 0.9.6 built fine for me using MacOS 
X 10.3.9, gcc 3.3 and ActivePython 2.4.2.

-- Russell


From markus at johalla.de  Fri Mar 24 22:23:26 2006
From: markus at johalla.de (Markus)
Date: Fri, 24 Mar 2006 22:23:26 +0100
Subject: [Pythonmac-SIG] egg question
In-Reply-To: <745EE540-4569-4B43-8472-F624C9A351F9@johalla.de>
References: <745EE540-4569-4B43-8472-F624C9A351F9@johalla.de>
Message-ID: <A530E998-4FC9-49FE-B6D7-33EC3BC25458@johalla.de>

Hi!

Just for the record to answer my own question: I deleted the old site  
package directory (and the fink python) and then reinstalled the eggs.
Now it works.

markus

From abarnert at yahoo.com  Sat Mar 25 00:02:25 2006
From: abarnert at yahoo.com (Andrew Barnert)
Date: Fri, 24 Mar 2006 15:02:25 -0800 (PST)
Subject: [Pythonmac-SIG] Two appscript problems, plus a stupid question
Message-ID: <20060324230225.92113.qmail@web82307.mail.mud.yahoo.com>

If this list is the wrong place, I apologize; just let
me know and I'll ask in the right place. Otherwise:

Now that I can use appscript with Python 2.4 on my
Intel box, I've been going back over some of my
iTunes-controlling scripts, and I've run into two
problems and one stupid question.

First, whenever I try to get a date through appscript,
at least from iTunes, I get a "CommandError: long int
too large to convert to int". I'll give the full
traceback below. So that my script that should list
the last N tracks played in iTunes crashes instead.

Second, whenever I ask for a missing value, the result
is something like aem.AEType('gnsm'). I have a bunch
of scripts that just do things like "if artist:" and
now they have to do things like "if isinstance(artist,
unicode):" instead. This used to work (on a Mac with
ancient versions of appscript and MacPython).

Finally, the stupid question: Is there any easy way to
convert a list of references into a single reference
to the list? (To allow a single IPC call instead of
one per reference.)

In my iTunes scripts, I do this by generating a big
filter:
  # slow version
  ids = [281, 1099, 731, 414, 238, ...]
  tracks = [library.file_tracks.ID(id) for id in ids]
  names = [track.name() for track in tracks]

  # fast version
  dids = [168, 142, 57, 39, 277, ...]
  idfilters = [(its.database_ID == did) for did in
dids]
  idfilter = idfilters[0]
  for f in idfilters[1:]:
    idfilter = idfilter.OR(f)
  trackrefs = library.file_tracks.filter(idfilter)
  names = trackrefs.name()

Is there an easier way to get all the names in a
single IPC call? (Also, is there a way I can filter on
the ID instead of having the fetch and keep around the
database_ID for everything?)

Here's what I'm doing this for: My other Mac has the
whole library on shuffle, and I hear something and
think, "Wow, what was that song?" So I do this:

  $ ssh emac
  $ iwhat `ilast 3`
  468: Add N to (X) - Large Number
  693: ADULT. - Hairing Impaired
  241: T. Raumschmiere - R.Ror
  $ iwhat -a 693
  693: ADULT. - Hairing Impaired (D.U.M.E. 5/6, 2005)
02:36 0/100
  $ iplaylist "D.U.M.E." `ibatch 693`
  $ iplay "D.U.M.E."
  $ irate 90
  ...

>From across the room, I figured out which of the
recent songs was the one I wanted, made a playlist of
all the songs in the same import batch, started
playing it, rated the current track (the first one)
4-1/2 stars, etc. (I remembered that I bought
D.U.M.E., ripped it, then had to go out and never
listened to it....)

The ilast and ibatch scripts need dates, the iwhat
script needs to understand missing values, and a bunch
of the scripts either return a list of IDs (ilast,
ibatch, iadd, isearch, ...) or take a list of IDs
(iwhat, iplaylist, irate, ...). Hence the questions.

Finally, here's the traceback I promised:

$ ibatch 142 168
Traceback (most recent call last):
  File "./icontrol-server", line 117, in ?
    dates = trackrefs.played_date()
  File
"/Library/Python/2.3/site-packages/appscript/specifier.py",
line 356, in __call__
    return self.get(*args, **kargs)
  File
"/Library/Python/2.3/site-packages/appscript/specifier.py",
line 204, in __call__
    raise CommandError(self, (args, kargs), e)
appscript.specifier.CommandError: long int too large
to convert to int
        Failed command:
app(u'/Applications/iTunes.app').library_playlists[1].file_tracks.filter(((its.database_ID
== 168).OR(its.database_ID == 142))).played_date.get()


From marc at precipice.org  Sat Mar 25 01:51:16 2006
From: marc at precipice.org (Marc Hedlund)
Date: Fri, 24 Mar 2006 18:51:16 -0600 (CST)
Subject: [Pythonmac-SIG] "mach-o, but wrong architecture"
Message-ID: <Pine.LNX.4.63.0603241831410.22570@crankshaft.bitmechanic.com>


Hey,

I've seen a few messages about this error, but I'm not clear what the 
resolution is (if any).  I'm using MacOS X 10.4.5, Intel iMac, and the 
bundled Python 2.3.5.  For various reasons I don't want to use Python 2.4 
yet.

I had been working along just fine using Python and wxPython, until I hit 
a problem with wxPython today.  I realized that I was using an old version 
of wxPython, 2.5.3.1, and that it didn't support a feature I needed.  So I 
downloaded wxPython 2.6.2.1, along with the TigerPython23Compat.pkg, and 
installed them both.

Now when I try to run python and import wx, I get this error:

>>> import wx
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/__init__.py", line 42, in ?
     from wx._core import *
   File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/_core.py", line 4, in ?
     import _core_
ImportError: dlopen(/Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so, 2): no suitable image found.  Did find:
         /Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so: mach-o, but wrong architecture

I get that wxPython is built for ppc, and that the version of python on my 
machine is a Universal binary, and that those aren't playing well 
together.  My question is, can I do anything about it?  Is there a way to 
call python so that it is forced to run under Rosetta?  My goal is to run 
Python 2.3 and wxPython 2.6 on the Intel iMac.  Is that possible yet?

Thanks much for any help.  A more detailed log is below if it helps.

-M

---begin---

iMac:~ marc$ python
Python 2.3.5 (#1, Dec 25 2005, 07:24:19)
[GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import wxversion
>>> print wxversion.getInstalled()
['2.6-mac-unicode', '2.5.3-mac-unicode']
>>> import wx
Traceback (most recent call last):
   File "<stdin>", line 1, in ?
   File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/__init__.py", line 42, in ?
     from wx._core import *
   File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/_core.py", line 4, in ?
     import _core_
ImportError: dlopen(/Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so, 2): no suitable image found.  Did find:
         /Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so: mach-o, but wrong architecture
>>> ^D
iMac:~ marc$ python
Python 2.3.5 (#1, Dec 25 2005, 07:24:19)
[GCC 4.0.1 (Apple Computer, Inc. build 5250)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import wxversion
>>> wxversion.select('2.5')
>>> import wx
>>> print wx.VERSION_STRING
2.5.3.1
>>> ^D
iMac:~ marc$ file 
/Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so
/Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so: Mach-O bundle ppc
iMac:~ marc$ file 
/System/Library/Frameworks/Python.framework/Versions/2.3/bin/python
/System/Library/Frameworks/Python.framework/Versions/2.3/bin/python: Mach-O universal binary with 2 architectures
/System/Library/Frameworks/Python.framework/Versions/2.3/bin/python (for architecture i386):    Mach-O executable i386
/System/Library/Frameworks/Python.framework/Versions/2.3/bin/python (for architecture ppc):     Mach-O executable ppc

---end---

From hengist.podd at virgin.net  Sat Mar 25 17:49:07 2006
From: hengist.podd at virgin.net (has)
Date: Sat, 25 Mar 2006 16:49:07 +0000
Subject: [Pythonmac-SIG] Two appscript problems, plus a stupid question
Message-ID: <a05200f06c04ae152c628@[82.20.25.20]>

Andrew Barnert wrote:

>If this list is the wrong place, I apologize; just let
>me know and I'll ask in the right place. Otherwise:

Here's fine.

>First, whenever I try to get a date through appscript,
>at least from iTunes, I get a "CommandError: long int
>too large to convert to int".
>Second, whenever I ask for a missing value, the result
>is something like aem.AEType('gnsm').

Oops, more endian problems. Ta for the heads-up. Fixed and uploaded.


>Finally, the stupid question: Is there any easy way to
>convert a list of references into a single reference
>to the list? (To allow a single IPC call instead of
>one per reference.)

Nope. There are one or two apps that will accept lists of references (e.g. Finder) as the direct parameter to a command, but that's about it. With stuff like this you're always limited in what you can do by the individual application.

Incidentally, the interprocess messaging itself is relative cheap on OS X; the main performance hit tends to be the time taken by applications to evaluate references and perform commands. Lots of opportunities for inefficient algorithms to burn cycles; amalgamating many operations into one simply minimises the amount of waste.


>In my iTunes scripts, I do this by generating a big
>filter:
>
>  # fast version
>  dids = [168, 142, 57, 39, 277, ...]
>  idfilters = [(its.database_ID == did) for did in
>dids]
>  idfilter = idfilters[0]
>  for f in idfilters[1:]:
>    idfilter = idfilter.OR(f)
>  trackrefs = library.file_tracks.filter(idfilter)
>  names = trackrefs.name()
>
>Is there an easier way to get all the names in a
>single IPC call?

Not that I can think of. The obvious thing would be to say:

	trackrefs = library.file_tracks.filter(its.database_ID.isin(dids))

but iTunes isn't smart enough to understand the isin() in that filter test, so I think you're stuck with your multiple equality tests.

Another technique is to grab all the IDs with one command and all names/whatever with another, then filter one against the other in Python; what's quicker will depend on the amount of data involved.


>(Also, is there a way I can filter on
>the ID instead of having the fetch and keep around the
>database_ID for everything?)

While iTunes likes to use by-ID specifiers, it makes it damnably hard to get the ID numbers used. Its dictionary doesn't show an 'id' property on any of its classes, although an undocumented one does appear to exist. Feel free to file a bug report on this - it causes problems for AppleScript as well as appscript.

To deal with missing terminology in Python, you either use appscript.dump to output the keyword-code translation tables to an external module and patch that by hand, or else drop into aem to build at least part of the reference using raw codes:

itunes = app('iTunes')
tracksRef = itunes.playlists[1].tracks
trackIDsRef = tracksRef.AS_objspec.property('ID  ')
ids = itunes.get(trackIDsRef)
print ids


>Here's what I'm doing this for: My other Mac has the
>whole library on shuffle, and I hear something and
>think, "Wow, what was that song?" So I do this:
>
>  $ ssh emac
>  $ iwhat `ilast 3`
>[...]

Nice.

HTH

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From hengist.podd at virgin.net  Sat Mar 25 18:03:53 2006
From: hengist.podd at virgin.net (has)
Date: Sat, 25 Mar 2006 17:03:53 +0000
Subject: [Pythonmac-SIG] Two appscript problems, plus a stupid question
Message-ID: <a05200f00c04b269b1b3d@[82.3.80.151]>

Andrew Barnert wrote:

>In my iTunes scripts, I do this by generating a big
>filter:
>[...]

Come to think of it, you can simplify the OR test a bit, since AND() and OR() can take more than one argument:

itunes = app('iTunes')

dids = [66, 67, 68, 69, 70]

idfilters = [its.database_ID == did for did in dids]
idfilter = idfilters[0]
if idfilters[1:]:
    idfilter = idfilter.OR(*idfilters[1:])
print itunes.playlists[1].tracks.filter(idfilter).name()

HTH

has
-- 
http://freespace.virgin.net/hamish.sanderson/

From bob at redivi.com  Sat Mar 25 18:24:22 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sat, 25 Mar 2006 09:24:22 -0800
Subject: [Pythonmac-SIG] "mach-o, but wrong architecture"
In-Reply-To: <Pine.LNX.4.63.0603241831410.22570@crankshaft.bitmechanic.com>
References: <Pine.LNX.4.63.0603241831410.22570@crankshaft.bitmechanic.com>
Message-ID: <80CF1CD6-C6E2-4377-8B45-88922DE0A029@redivi.com>


On Mar 24, 2006, at 4:51 PM, Marc Hedlund wrote:

> I've seen a few messages about this error, but I'm not clear what the
> resolution is (if any).  I'm using MacOS X 10.4.5, Intel iMac, and the
> bundled Python 2.3.5.  For various reasons I don't want to use  
> Python 2.4
> yet.
>
> I had been working along just fine using Python and wxPython, until  
> I hit
> a problem with wxPython today.  I realized that I was using an old  
> version
> of wxPython, 2.5.3.1, and that it didn't support a feature I  
> needed.  So I
> downloaded wxPython 2.6.2.1, along with the  
> TigerPython23Compat.pkg, and
> installed them both.
>
> Now when I try to run python and import wx, I get this error:
>
>>>> import wx
> Traceback (most recent call last):
>    File "<stdin>", line 1, in ?
>    File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/__init__.py",  
> line 42, in ?
>      from wx._core import *
>    File "//Library/Python/2.3/wx-2.6-mac-unicode/wx/_core.py", line  
> 4, in ?
>      import _core_
> ImportError: dlopen(/Library/Python/2.3/wx-2.6-mac-unicode/wx/ 
> _core_.so, 2): no suitable image found.  Did find:
>          /Library/Python/2.3/wx-2.6-mac-unicode/wx/_core_.so: mach- 
> o, but wrong architecture
>
> I get that wxPython is built for ppc, and that the version of  
> python on my
> machine is a Universal binary, and that those aren't playing well
> together.  My question is, can I do anything about it?  Is there a  
> way to
> call python so that it is forced to run under Rosetta?  My goal is  
> to run
> Python 2.3 and wxPython 2.6 on the Intel iMac.  Is that possible yet?

The best way to do this is to install the PPC version of Python 2.4  
and use that to use PPC extensions.  You really don't want to be  
using that build of Python anyway, Apple missed a bunch of things  
when they tried to port it to i386 (and they're very unlikely to  
correct their mistakes because they have a record of never updating  
Python except on major OS releases).

-bob


From marc at precipice.org  Sat Mar 25 18:26:31 2006
From: marc at precipice.org (Marc Hedlund)
Date: Sat, 25 Mar 2006 11:26:31 -0600 (CST)
Subject: [Pythonmac-SIG] "mach-o, but wrong architecture"
In-Reply-To: <80CF1CD6-C6E2-4377-8B45-88922DE0A029@redivi.com>
References: <Pine.LNX.4.63.0603241831410.22570@crankshaft.bitmechanic.com>
	<80CF1CD6-C6E2-4377-8B45-88922DE0A029@redivi.com>
Message-ID: <Pine.LNX.4.63.0603251126010.22739@crankshaft.bitmechanic.com>


>> I get that wxPython is built for ppc, and that the version of python on my
>> machine is a Universal binary, and that those aren't playing well
>> together.  My question is, can I do anything about it?  Is there a way to
>> call python so that it is forced to run under Rosetta?  My goal is to run
>> Python 2.3 and wxPython 2.6 on the Intel iMac.  Is that possible yet?
>
> The best way to do this is to install the PPC version of Python 2.4 and use 
> that to use PPC extensions.  You really don't want to be using that build of 
> Python anyway, Apple missed a bunch of things when they tried to port it to 
> i386 (and they're very unlikely to correct their mistakes because they have a 
> record of never updating Python except on major OS releases).

Got it.  Thanks much for the response.

-M

From ronaldoussoren at mac.com  Sat Mar 25 18:32:36 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sat, 25 Mar 2006 18:32:36 +0100
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <44234107.3040203@noaa.gov>
References: <06Mar23.155649pst.58633@synergy1.parc.xerox.com>
	<44234107.3040203@noaa.gov>
Message-ID: <0172E15D-787C-4DDF-BB99-AFC38C966018@mac.com>


On 24-mrt-2006, at 1:44, Christopher Barker wrote:

> Bill Janssen wrote:
>>> It is officialy released, even if Bill hasn't updated the download
>>> page at python.org yet.
>
> Are we really ready to call it the "official" version to use? I  
> just had
> a colleague try it, then he asked me how to get wxPython and numpy
> working with it.
>
> I don't think there is a wxPython build for it yet (will the one for
> 2.4.1 work?)

wxPython definitely works, I've tested it. That is, if you are on
a PPC box. The existing wxPython binary installer installs a PPC binary,
that won't work with a universal build on intel.

>
> And he couldn't get numpy to build and install right.

The existing binaries for that should work just fine (again on a PPC
box)

>
> I didn't try to debug that for him, instead I suggested to just use
> 2.4.1 for a while longer.
>
> If the new universal build is it now, we need to update
> pythonmac.org/packages with some information about what packages will
> work with it, and start building new universal packages!

Feel free to provide univesal packages, and tell us about software that
doesn't work with universal python.

Even a wishlist for packages would be useful. Keep in mind that most
of us have limited time to work on this.

Ronald

>
> I know I need:
>
> wxPython
> numpy
> matplotlib
>
> At the very least, so I'll help with those, though wxPython is  
> beyond my
> skills.
>
> -Chris
>
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                      		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Sat Mar 25 18:35:09 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sat, 25 Mar 2006 18:35:09 +0100
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <rowen-255055.12192324032006@sea.gmane.org>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
	<rowen-255055.12192324032006@sea.gmane.org>
Message-ID: <8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>


On 24-mrt-2006, at 21:19, Russell E. Owen wrote:

> In article <4D6DFDB5-B044-4637-9CE9-4E886998A43F at mac.com>,
>  Ronald Oussoren <ronaldoussoren at mac.com> wrote:
>
>> On 23-mrt-2006, at 19:24, James Boyle wrote:
>> ...
>>> I am running OX 10.3.9. I installed python 2.4.2 using Universal
>>> MacPython 2.4.2.dmg.
>>>
>>> When I try to install numpy 0.9.6 it fails ...
>>> .
>>> I am running gcc 3.3. I note that the python 2.4.2 was built using
>>> gcc 4.0.1. Could this difference in gcc versions be the problem?
>>> At present I am at a loss as to what to try next.
>>
>> Could you send the entire output? Compiling on 10.3.9 is suppossed to
>> work.
>
> This may not be relevant, but numpy 0.9.6 built fine for me using  
> MacOS
> X 10.3.9, gcc 3.3 and ActivePython 2.4.2.

The universal binary build contains some trickery to enable building  
extensions
on OSX 10.3.9 even though the default compiler flags contain items  
that don't
work with gcc 3.3. Appearently he's running into problems.

I'll be working on a universal build of python 2.4.3c1 after dinner,  
and will look
into this problem as well.

Ronald

>
> -- Russell
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Sat Mar 25 21:06:27 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sat, 25 Mar 2006 21:06:27 +0100
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <p06230900c04b3efc87e3@[192.168.0.2]>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
	<rowen-255055.12192324032006@sea.gmane.org>
	<8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>
	<p06230900c04b3efc87e3@[192.168.0.2]>
Message-ID: <5A3AE4D7-3ACC-4056-B6F1-6B985926EF48@mac.com>


On 25-mrt-2006, at 19:41, Russell E Owen wrote:

> At 6:35 PM +0100 3/25/06, Ronald Oussoren wrote:
>>>> On 23-mrt-2006, at 19:24, James Boyle wrote:
>>>> ...
>>>>> I am running OX 10.3.9. I installed python 2.4.2 using Universal
>>>>> MacPython 2.4.2.dmg.
>>>>> ...
>>
>> The universal binary build contains some trickery to enable  
>> building extensions
>> on OSX 10.3.9 even though the default compiler flags contain items  
>> that don't
>> work with gcc 3.3. Appearently he's running into problems.
>>
>> I'll be working on a universal build of python 2.4.3c1 after  
>> dinner, and will look
>> into this problem as well.
>
> Would it be easier to have separate installers for 10.3.9 (not  
> universal binaries) and 10.4.x (universal binaries, gcc 4 required)?

I'd prefer to have 1 installer for python on OSX, that makes support  
a lot easier.

Ronald

>
> -- Russell


From bob at redivi.com  Sat Mar 25 22:50:29 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sat, 25 Mar 2006 13:50:29 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <5A3AE4D7-3ACC-4056-B6F1-6B985926EF48@mac.com>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
	<rowen-255055.12192324032006@sea.gmane.org>
	<8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>
	<p06230900c04b3efc87e3@[192.168.0.2]>
	<5A3AE4D7-3ACC-4056-B6F1-6B985926EF48@mac.com>
Message-ID: <46FAA65E-D8E6-4713-81FB-2BC5E075C4AD@redivi.com>


On Mar 25, 2006, at 12:06 PM, Ronald Oussoren wrote:

>
> On 25-mrt-2006, at 19:41, Russell E Owen wrote:
>
>> At 6:35 PM +0100 3/25/06, Ronald Oussoren wrote:
>>>>> On 23-mrt-2006, at 19:24, James Boyle wrote:
>>>>> ...
>>>>>> I am running OX 10.3.9. I installed python 2.4.2 using Universal
>>>>>> MacPython 2.4.2.dmg.
>>>>>> ...
>>>
>>> The universal binary build contains some trickery to enable
>>> building extensions
>>> on OSX 10.3.9 even though the default compiler flags contain items
>>> that don't
>>> work with gcc 3.3. Appearently he's running into problems.
>>>
>>> I'll be working on a universal build of python 2.4.3c1 after
>>> dinner, and will look
>>> into this problem as well.
>>
>> Would it be easier to have separate installers for 10.3.9 (not
>> universal binaries) and 10.4.x (universal binaries, gcc 4 required)?
>
> I'd prefer to have 1 installer for python on OSX, that makes support
> a lot easier.

+1.  From my experience on this list, any choice or ambiguity comes  
at great (support) cost.

We need to have one and preferably only one obvious way to use Python  
on Mac OS X.  This means universal binaries, (only) universal  
extensions whenever possible, and support for 10.3.9+ in one installer.

10.2 users should stick with older releases, get in the habit of  
building it themselves, or use fink or darwinports (if they're still  
supporting 10.2).

-bob


From ronaldoussoren at mac.com  Sat Mar 25 23:53:43 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Sat, 25 Mar 2006 23:53:43 +0100
Subject: [Pythonmac-SIG] resizing terminal and curses
In-Reply-To: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>
References: <a5c3d8350603141000k360c608nac2bcf9675eb053c@mail.gmail.com>
Message-ID: <813C53DB-7521-4F89-92F8-355F1AC56F45@mac.com>


On 14-mrt-2006, at 19:00, Nicholas Cole wrote:

> I couldn't find anything in the archives related to this.
>
> Using both Apple's python and a build of python2.4.2, there seems to
> be a bug when using ncurses.  curses.newwin() is supposed to give a
> new window filling the pysical terminal, but if the terminal has been
> resized this does not seem to happen.  Instead, the window seems to be
> the size that the terminal was when curses was initialised.  This is
> not the case on other platforms, where the window created reflects the
> new size of the terminal.  Is this a known bug?
>
> (it seems to affect both xterm and terminal.app)

Linking with my own build of ncurses 5.5 instead of the system curses  
library
seems to solve this issue. I'll see if I can include this in the  
universal
binary build of python 2.4.3.

Ronald

>
> Best wishes,
>
> N.
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From vinogri at mcmaster.ca  Sun Mar 26 04:53:49 2006
From: vinogri at mcmaster.ca (Ivan Vinogradov)
Date: Sat, 25 Mar 2006 21:53:49 -0500
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <0172E15D-787C-4DDF-BB99-AFC38C966018@mac.com>
References: <06Mar23.155649pst.58633@synergy1.parc.xerox.com>
	<44234107.3040203@noaa.gov>
	<0172E15D-787C-4DDF-BB99-AFC38C966018@mac.com>
Message-ID: <5C0E87E9-623B-4E98-8DAC-14A608C6806A@mcmaster.ca>

Some random comments, apologies for snipping.
>>>> It is officialy released, even if Bill hasn't updated the download
>>>> page at python.org yet.

Works just fine on fresh install of Tiger and PPC.
Making this official +1.
Perhaps could use some cosmetic changes but that's a matter of  
personal taste.


> wxPython definitely works, I've tested it. That is, if you are on
> a PPC box. The existing wxPython binary installer installs a PPC  
> binary,
> that won't work with a universal build on intel.

Package from pythonmac works seamlessly here.

>>
>> And he couldn't get numpy to build and install right.
>
> The existing binaries for that should work just fine (again on a PPC
> box)
>

Numpy 0.9.6 from sourceforge now comes as a dmg mpkg; installs and  
works under a minute.


>
>>
>> I know I need:
>>
>> wxPython
>> numpy
>> matplotlib

All those work from existing packages at pythonmac.

--
Regards, Ivan.

From darran at edmstudio.com  Sun Mar 26 20:11:21 2006
From: darran at edmstudio.com (Darran Edmundson)
Date: Sun, 26 Mar 2006 11:11:21 -0700
Subject: [Pythonmac-SIG] python SWIG problem ...
Message-ID: <2D3809DE-E652-4E60-A03F-23D80F197D44@edmstudio.com>


I've been playing around with using SWIG to generate python
bindings on the Mac.  Even for the most trivial SWIG/python examples,
I keep coming up against the linker not being able to resolve
the symbol "PyInstance_Lookup".   This is an internal-use
function defined in /Library/Frameworks/Python.framework/Headers/ 
classobject.h:

$ grep PyInstance classobject.h
} PyInstanceObject;
PyAPI_DATA(PyTypeObject) PyClass_Type, PyInstance_Type, PyMethod_Type;
#define PyInstance_Check(op) ((op)->ob_type == &PyInstance_Type)
PyAPI_FUNC(PyObject *) PyInstance_New(PyObject *, PyObject *,
PyAPI_FUNC(PyObject *) PyInstance_NewRaw(PyObject *, PyObject *);
PyAPI_FUNC(PyObject *) _PyInstance_Lookup(PyObject *pinst, PyObject  
*name);

If I search the shared libraries (no surprise) it isn't defined ...

$ nm /Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/ 
lib-dynload/*.so | grep PyInstance
          U PyInstance_New
          U PyInstance_NewRaw
          U PyInstance_Type
          U PyInstance_Type

Where else could it be and/or any ideas how to resolve
this issue?  I've asked on the SWIG mailing list but to no
avail.  Is there a trick to getting SWIG running on the Mac?

FYI, I'm using a G4 Powerbook, OS X 10.4.5,

$ which python
/sw/bin/python  (Fink?)

$ python
Python 2.3.3 (#1, Jan  6 2005, 15:03:06)
[GCC 3.3 20030304 (Apple Computer, Inc. build 1666)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>>

As always, any help is much appreciated.

Cheers,
Darran.


From bob at redivi.com  Sun Mar 26 22:21:18 2006
From: bob at redivi.com (Bob Ippolito)
Date: Sun, 26 Mar 2006 12:21:18 -0800
Subject: [Pythonmac-SIG] python SWIG problem ...
In-Reply-To: <2D3809DE-E652-4E60-A03F-23D80F197D44@edmstudio.com>
References: <2D3809DE-E652-4E60-A03F-23D80F197D44@edmstudio.com>
Message-ID: <D2DD1FB5-F14E-48F2-884B-1089FAD7C063@redivi.com>


On Mar 26, 2006, at 10:11 AM, Darran Edmundson wrote:

>
> I've been playing around with using SWIG to generate python
> bindings on the Mac.  Even for the most trivial SWIG/python examples,
> I keep coming up against the linker not being able to resolve
> the symbol "PyInstance_Lookup".   This is an internal-use
> function defined in /Library/Frameworks/Python.framework/Headers/
> classobject.h:
>
> $ grep PyInstance classobject.h
--
> If I search the shared libraries (no surprise) it isn't defined ...
>
> $ nm /Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/
--
> Where else could it be and/or any ideas how to resolve
> this issue?  I've asked on the SWIG mailing list but to no
> avail.  Is there a trick to getting SWIG running on the Mac?
>
> FYI, I'm using a G4 Powerbook, OS X 10.4.5,
>
> $ which python
> /sw/bin/python  (Fink?)

You're using a Fink Python here, but you're looking at the shared  
library and headers for the Apple distribution of Python.... you need  
to be looking at /sw/include for the header and /sw/lib for the  
shared lib, or you need to stop using Fink's Python.  You're probably  
using a SWIG that was built for one of them, but using it against the  
library/headers of the other.

-bob


From rowen at u.washington.edu  Sat Mar 25 19:41:36 2006
From: rowen at u.washington.edu (Russell E Owen)
Date: Sat, 25 Mar 2006 10:41:36 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
	<rowen-255055.12192324032006@sea.gmane.org>
	<8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>
Message-ID: <p06230900c04b3efc87e3@[192.168.0.2]>

At 6:35 PM +0100 3/25/06, Ronald Oussoren wrote:
>>>On 23-mrt-2006, at 19:24, James Boyle wrote:
>>>...
>>>>I am running OX 10.3.9. I installed python 2.4.2 using Universal
>>>>MacPython 2.4.2.dmg.
>>>>...
>
>The universal binary build contains some trickery to enable building 
>extensions
>on OSX 10.3.9 even though the default compiler flags contain items that don't
>work with gcc 3.3. Appearently he's running into problems.
>
>I'll be working on a universal build of python 2.4.3c1 after dinner, 
>and will look
>into this problem as well.

Would it be easier to have separate installers for 10.3.9 (not 
universal binaries) and 10.4.x (universal binaries, gcc 4 required)?

-- Russell

From boyle5 at llnl.gov  Mon Mar 27 18:00:45 2006
From: boyle5 at llnl.gov (James Boyle)
Date: Mon, 27 Mar 2006 08:00:45 -0800
Subject: [Pythonmac-SIG] Problems installing Numpy using Universal MacPython
Message-ID: <6b85e3b035ec2097ec815bccdb307bd8@llnl.gov>

I posted this to the numpy mailing list - I am re-posting here since it  
is a more complete description of the situation than my previous post  
to this list.

I am running OX 10.3.9. I installed python 2.4.2 using Universal  
MacPython 2.4.2.dmg.
I am on a PowerPC G4. The Universal build is designed to install on the  
Mac PPC and Intel machines.

When I run the numpy 0.9.6 install it apparently fails on the  
configure. The whole trace back is at the end of this posting.
One thing that I noted  was the line:

gcc options: '-arch ppc -arch i386 -isysroot  
/Developer/SDKs/MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double  
-no-cpp-precomp -mno-fused-madd -fno-common -dynamic -DNDEBUG -g'

It appears that the configure routines think that I have two  
architectures - ppc and i386.
At my low level of understanding, all I can do is speculate that the  
Universal install generates some confusion as to the actual cpu.
Somewhere numpy is keying on information that is ambiguous in the  
universal install.

I have installed Numeric, matplotlib and other packages without a  
problem.

--Jim


The trace back from >python setup.py build:

Running from numpy source directory.
Warning: not existing path in numpy/distutils: site.cfg
F2PY Version 2_2236
blas_opt_info:
   FOUND:
     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
     define_macros = [('NO_ATLAS_INFO', 3)]
     extra_compile_args = ['-faltivec',  
'-I/System/Library/Frameworks/vecLib.framework/Headers']

lapack_opt_info:
   FOUND:
     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
     define_macros = [('NO_ATLAS_INFO', 3)]
     extra_compile_args = ['-faltivec']

running build
running config_fc
running build_src
building py_modules sources
creating build
creating build/src
creating build/src/numpy
creating build/src/numpy/distutils
building extension "numpy.core.multiarray" sources
creating build/src/numpy/core
Generating build/src/numpy/core/config.h
customize NAGFCompiler
customize AbsoftFCompiler
customize IbmFCompiler
customize GnuFCompiler
customize GnuFCompiler
customize GnuFCompiler using config
gcc options: '-arch ppc -arch i386 -isysroot  
/Developer/SDKs/MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double  
-no-cpp-precomp -mno-fused-madd -fno-common -dynamic -DNDEBUG -g'
compile options:  
'-I/Library/Frameworks/Python.framework/Versions/2.4/include/python2.4  
-Inumpy/core/src -Inumpy/core/include  
-I/Library/Frameworks/Python.framework/Versions/2.4/include/python2.4  
-c'
gcc: _configtest.c
gcc: cannot specify -o with -c or -S and multiple compilations
gcc: cannot specify -o with -c or -S and multiple compilations
failure.
removing: _configtest.c _configtest.o
Traceback (most recent call last):
   File "setup.py", line 76, in ?
     setup_package()
   File "setup.py", line 69, in setup_package
     setup( **config.todict() )
   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/core.py",  
line 85, in setup
     return old_setup(**new_attr)
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/core.py", line 149, in setup
     dist.run_commands()
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/dist.py", line 946, in run_commands
     self.run_command(cmd)
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/dist.py", line 966, in run_command
     cmd_obj.run()
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/command/build.py", line 112, in run
     self.run_command(cmd_name)
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/cmd.py", line 333, in run_command
     self.distribution.run_command(command)
   File  
"/Library/Frameworks/Python.framework/Versions/2.4//lib/python2.4/ 
distutils/dist.py", line 966, in run_command
     cmd_obj.run()
   File  
"/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
build_src.py", line 84, in run
     self.build_sources()
   File  
"/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
build_src.py", line 99, in build_sources
     self.build_extension_sources(ext)
   File  
"/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
build_src.py", line 209, in build_extension_sources
     sources = self.generate_sources(sources, ext)
   File  
"/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
build_src.py", line 267, in generate_sources
     source = func(extension, build_dir)
   File "numpy/core/setup.py", line 37, in generate_config_h
     raise "ERROR: Failed to test configuration"
ERROR: Failed to test configuration
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: text/enriched
Size: 4674 bytes
Desc: not available
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060327/b2a6cf8e/attachment.bin 

From ronaldoussoren at mac.com  Mon Mar 27 19:07:28 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 27 Mar 2006 19:07:28 +0200
Subject: [Pythonmac-SIG] Problems installing Numpy using Universal
	MacPython
In-Reply-To: <6b85e3b035ec2097ec815bccdb307bd8@llnl.gov>
References: <6b85e3b035ec2097ec815bccdb307bd8@llnl.gov>
Message-ID: <ECDDA57F-F2BD-4C4F-9D63-99DB37B44545@mac.com>


On 27-mrt-2006, at 18:00, James Boyle wrote:

> I posted this to the numpy mailing list - I am re-posting here  
> since it is a more complete description of the situation than my  
> previous post to this list.
>
> I am running OX 10.3.9. I installed python 2.4.2 using Universal  
> MacPython 2.4.2.dmg.
> I am on a PowerPC G4. The Universal build is designed to install on  
> the Mac PPC and Intel machines.
>
> When I run the numpy 0.9.6 install it apparently fails on the  
> configure. The whole trace back is at the end of this posting.
> One thing that I noted  was the line:
>
> gcc options: '-arch ppc -arch i386 -isysroot /Developer/SDKs/ 
> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
> precomp -mno-fused-madd -fno-common -dynamic -DNDEBUG -g'
>
> It appears that the configure routines think that I have two  
> architectures - ppc and i386.
> At my low level of understanding, all I can do is speculate that  
> the Universal install generates some confusion as to the actual cpu.
> Somewhere numpy is keying on information that is ambiguous in the  
> universal install.

The problem is that gcc 3.3 on 10.3.9 doesn't support some of these  
options. Specifcally it doesn't know '-isysroot' and also doesn't  
support the i386 architecture. I'm pretty sure that I had implemented  
support for stripping these arguments on a 10.3 system. Hopefully  
I'll have time to boot into 10.3 tomorrow.

Ronald

>
> I have installed Numeric, matplotlib and other packages without a  
> problem.
>
> --Jim
>
>
> The trace back from >python setup.py build:
>
> Running from numpy source directory.
> Warning: not existing path in numpy/distutils: site.cfg
> F2PY Version 2_2236
> blas_opt_info:
>   FOUND:
>     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
>     define_macros = [('NO_ATLAS_INFO', 3)]
>     extra_compile_args = ['-faltivec', '-I/System/Library/ 
> Frameworks/vecLib.framework/Headers']
>
> lapack_opt_info:
>   FOUND:
>     extra_link_args = ['-Wl,-framework', '-Wl,Accelerate']
>     define_macros = [('NO_ATLAS_INFO', 3)]
>     extra_compile_args = ['-faltivec']
>
> running build
> running config_fc
> running build_src
> building py_modules sources
> creating build
> creating build/src
> creating build/src/numpy
> creating build/src/numpy/distutils
> building extension "numpy.core.multiarray" sources
> creating build/src/numpy/core
> Generating build/src/numpy/core/config.h
> customize NAGFCompiler
> customize AbsoftFCompiler
> customize IbmFCompiler
> customize GnuFCompiler
> customize GnuFCompiler
> customize GnuFCompiler using config
> gcc options: '-arch ppc -arch i386 -isysroot /Developer/SDKs/ 
> MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
> precomp -mno-fused-madd -fno-common -dynamic -DNDEBUG -g'
> compile options: '-I/Library/Frameworks/Python.framework/Versions/ 
> 2.4/include/python2.4 -Inumpy/core/src -Inumpy/core/include -I/ 
> Library/Frameworks/Python.framework/Versions/2.4/include/python2.4 -c'
> gcc: _configtest.c
> gcc: cannot specify -o with -c or -S and multiple compilations
> gcc: cannot specify -o with -c or -S and multiple compilations
> failure.
> removing: _configtest.c _configtest.o
> Traceback (most recent call last):
>   File "setup.py", line 76, in ?
>     setup_package()
>   File "setup.py", line 69, in setup_package
>     setup( **config.todict() )
>   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/ 
> core.py", line 85, in setup
>     return old_setup(**new_attr)
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/core.py", line 149, in setup
>     dist.run_commands()
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/dist.py", line 946, in run_commands
>     self.run_command(cmd)
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/dist.py", line 966, in run_command
>     cmd_obj.run()
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/command/build.py", line 112, in run
>     self.run_command(cmd_name)
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/cmd.py", line 333, in run_command
>     self.distribution.run_command(command)
>   File "/Library/Frameworks/Python.framework/Versions/2.4//lib/ 
> python2.4/distutils/dist.py", line 966, in run_command
>     cmd_obj.run()
>   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
> build_src.py", line 84, in run
>     self.build_sources()
>   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
> build_src.py", line 99, in build_sources
>     self.build_extension_sources(ext)
>   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
> build_src.py", line 209, in build_extension_sources
>     sources = self.generate_sources(sources, ext)
>   File "/Volumes/IBM1/macPython/numpy-0.9.6/numpy/distutils/command/ 
> build_src.py", line 267, in generate_sources
>     source = func(extension, build_dir)
>   File "numpy/core/setup.py", line 37, in generate_config_h
>     raise "ERROR: Failed to test configuration"
> ERROR: Failed to test  
> configuration_______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From robert.kern at gmail.com  Mon Mar 27 19:26:27 2006
From: robert.kern at gmail.com (Robert Kern)
Date: Mon, 27 Mar 2006 11:26:27 -0600
Subject: [Pythonmac-SIG] Problems installing Numpy using
	Universal	MacPython
In-Reply-To: <ECDDA57F-F2BD-4C4F-9D63-99DB37B44545@mac.com>
References: <6b85e3b035ec2097ec815bccdb307bd8@llnl.gov>
	<ECDDA57F-F2BD-4C4F-9D63-99DB37B44545@mac.com>
Message-ID: <44282043.1020603@gmail.com>

Ronald Oussoren wrote:
> On 27-mrt-2006, at 18:00, James Boyle wrote:
> 
> 
>>I posted this to the numpy mailing list - I am re-posting here  
>>since it is a more complete description of the situation than my  
>>previous post to this list.
>>
>>I am running OX 10.3.9. I installed python 2.4.2 using Universal  
>>MacPython 2.4.2.dmg.
>>I am on a PowerPC G4. The Universal build is designed to install on  
>>the Mac PPC and Intel machines.
>>
>>When I run the numpy 0.9.6 install it apparently fails on the  
>>configure. The whole trace back is at the end of this posting.
>>One thing that I noted  was the line:
>>
>>gcc options: '-arch ppc -arch i386 -isysroot /Developer/SDKs/ 
>>MacOSX10.4u.sdk -fno-strict-aliasing -Wno-long-double -no-cpp- 
>>precomp -mno-fused-madd -fno-common -dynamic -DNDEBUG -g'
>>
>>It appears that the configure routines think that I have two  
>>architectures - ppc and i386.
>>At my low level of understanding, all I can do is speculate that  
>>the Universal install generates some confusion as to the actual cpu.
>>Somewhere numpy is keying on information that is ambiguous in the  
>>universal install.
> 
> The problem is that gcc 3.3 on 10.3.9 doesn't support some of these  
> options. Specifcally it doesn't know '-isysroot' and also doesn't  
> support the i386 architecture. I'm pretty sure that I had implemented  
> support for stripping these arguments on a 10.3 system. Hopefully  
> I'll have time to boot into 10.3 tomorrow.

numpy extends distutils, so it may be overriding certain changes you've made to
distutils. We may have to duplicate those changes in numpy.distutils.

-- 
Robert Kern
robert.kern at gmail.com

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco

From ronaldoussoren at mac.com  Mon Mar 27 19:28:16 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Mon, 27 Mar 2006 19:28:16 +0200
Subject: [Pythonmac-SIG] Python 2.4.3c1 universal binaries
Message-ID: <8D77A948-7489-49A1-81BF-BB6E18F92FE4@mac.com>


I have a universal build of the Python 2.4.3c1 release in the public  
folder of my idisk (at http://homepage.mac.com/ronaldoussoren). This  
passes the unittests on OSX 10.4 (both intel and ppc). I haven't  
tested it with 10.3.9 yet.

This also solves the problems Nicolas Cole was having w.r.t. resizing  
terminals and curses applications. This is because this build  
includes it's own version of the curses library that don't suffer  
from the bug in Apple's build. Hopefully this also means that the  
curses module works on 10.3.9.

Ronald

From Chris.Barker at noaa.gov  Mon Mar 27 19:50:33 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 27 Mar 2006 09:50:33 -0800
Subject: [Pythonmac-SIG] Installing numpy 0.9.6 problems
In-Reply-To: <5A3AE4D7-3ACC-4056-B6F1-6B985926EF48@mac.com>
References: <0eb5ac1afed1f1c206f4552600d8d8a4@llnl.gov>
	<4D6DFDB5-B044-4637-9CE9-4E886998A43F@mac.com>
	<rowen-255055.12192324032006@sea.gmane.org>
	<8EE688C7-AF91-4489-AC46-8FFAE8FCED4D@mac.com>
	<p06230900c04b3efc87e3@[192.168.0.2]>
	<5A3AE4D7-3ACC-4056-B6F1-6B985926EF48@mac.com>
Message-ID: <442825E9.5060007@noaa.gov>

Ronald Oussoren wrote:

> I'd prefer to have 1 installer for python on OSX, that makes support  
> a lot easier.

And we would like to by able to use Py2App to build apps that will run 
under both 10.3.9 and 10.4.*

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Chris.Barker at noaa.gov  Mon Mar 27 19:57:35 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 27 Mar 2006 09:57:35 -0800
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: <0172E15D-787C-4DDF-BB99-AFC38C966018@mac.com>
References: <06Mar23.155649pst.58633@synergy1.parc.xerox.com>
	<44234107.3040203@noaa.gov>
	<0172E15D-787C-4DDF-BB99-AFC38C966018@mac.com>
Message-ID: <4428278F.3050005@noaa.gov>

Ronald Oussoren wrote:

> wxPython definitely works, I've tested it. That is, if you are on
> a PPC box. The existing wxPython binary installer installs a PPC binary,
> that won't work with a universal build on intel.

That's good to know,but we really do need to get a Universal build of 
wxPython (and lots of other stuff!) built. It looks like it will beg a 
bit of a project, and it doesn't look like any of the people who really 
know what they are doing (Kevin O, Robin, ??? )have the time and/or 
tools for it now.

I'm going to give a try this week myself. Prepare yourselves for questions!

I'm also going to take this to mean that, in general, all the packages 
built for the 2.4.1 Framework build should work with the Universal build 
On PPC.

Users of the Intel machines are probably still better off using the 
2.4.1 PPC build and related packages, and it will all run under Rosetta.

> Feel free to provide universal packages, and tell us about software that
> doesn't work with universal python.

I'll be working on that this week. I don't have an Intel box to test on 
at the moment, but should be getting one soon.

> Even a wishlist for packages would be useful. 

wxPython
numpy
Numeric
numarray
matplotlib
PIL

I can work on all of those.

Then:
PyGTK : Is there a version that can work entirely fink (or darwinports) 
free?
PyQT: that's going to be fun!

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From janssen at parc.com  Mon Mar 27 21:44:31 2006
From: janssen at parc.com (Bill Janssen)
Date: Mon, 27 Mar 2006 11:44:31 PST
Subject: [Pythonmac-SIG] [ann] Appscript Installer 1.3 released
In-Reply-To: Your message of "Thu, 23 Mar 2006 16:09:13 PST."
	<06Mar23.160914pst."58633"@synergy1.parc.xerox.com> 
Message-ID: <06Mar27.114437pst."58633"@synergy1.parc.xerox.com>

I've added a link in the Wiki page about alternate distribution
pointing to Ronald's 2.4.2 installer.

Bill

From janssen at parc.com  Mon Mar 27 21:36:56 2006
From: janssen at parc.com (Bill Janssen)
Date: Mon, 27 Mar 2006 11:36:56 PST
Subject: [Pythonmac-SIG] Marking the pre-built extensions as Intel-capable...
Message-ID: <06Mar27.113702pst."58633"@synergy1.parc.xerox.com>

I'm looking over the downloads page again, thinking about what we need
to do to support the Universal builds.

First off, the discussion last week about which packages will work
with it seems very important.

Who's got edit access to pythonmac.org?

Could someone please put a note on http://pythonmac.org/packages/ to
say that the packages listed there are all PPC-only?

And do we need a new section for 10.4.2-capable packages, maybe with a
stamp that marks those that are Intel-capable?

Bill

From bob at redivi.com  Mon Mar 27 23:12:25 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 13:12:25 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
In-Reply-To: <06Mar27.113702pst."58633"@synergy1.parc.xerox.com>
References: <06Mar27.113702pst."58633"@synergy1.parc.xerox.com>
Message-ID: <5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>


On Mar 27, 2006, at 11:36 AM, Bill Janssen wrote:

> I'm looking over the downloads page again, thinking about what we need
> to do to support the Universal builds.
>
> First off, the discussion last week about which packages will work
> with it seems very important.
>
> Who's got edit access to pythonmac.org?

Just me, but I'm willing to dole out access to anyone who will commit  
to doing something with those privileges :)

> Could someone please put a note on http://pythonmac.org/packages/ to
> say that the packages listed there are all PPC-only?

Done

> And do we need a new section for 10.4.2-capable packages, maybe with a
> stamp that marks those that are Intel-capable?

That whole section really needs to be restructured to address the  
common questions and issues people have regarding finding the right  
packages for them.  I think I'd like to turn http://pythonmac.org/ 
packages/ into just a launch pad for finding a list of packages for  
the Python that you're interested in.

-bob


From Chris.Barker at noaa.gov  Tue Mar 28 00:55:29 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 27 Mar 2006 14:55:29 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions
	as	Intel-capable...
In-Reply-To: <5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
Message-ID: <44286D61.1070504@noaa.gov>

Bob Ippolito wrote:

> That whole section really needs to be restructured to address the  
> common questions and issues people have regarding finding the right  
> packages for them.

great idea.

> I think I'd like to turn http://pythonmac.org/ 
> packages/ into just a launch pad for finding a list of packages for  
> the Python that you're interested in.

Do you mean that it wouldn't actually host them for download there?

Please no. I think it is a VERY good idea to have a collection there for 
download. Pretty soon, I think we'll be able to call the 2.4.3 Universal 
build the "officially recommended" build, and we can have a collection 
of packages there for it. I now I'll contribute a few, and I'm sure 
others will as well.

One question is: should they be eggs or traditional *.mpkgs?

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From bob at redivi.com  Tue Mar 28 01:18:42 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 15:18:42 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions
	as	Intel-capable...
In-Reply-To: <44286D61.1070504@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
Message-ID: <2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>


On Mar 27, 2006, at 2:55 PM, Christopher Barker wrote:

> Bob Ippolito wrote:
>
>> That whole section really needs to be restructured to address the
>> common questions and issues people have regarding finding the right
>> packages for them.
>
> great idea.
>
>> I think I'd like to turn http://pythonmac.org/
>> packages/ into just a launch pad for finding a list of packages for
>> the Python that you're interested in.
>
> Do you mean that it wouldn't actually host them for download there?

No, I mean that that URL won't directly list any packages.  It'll be  
a listing of package lists and enough information to direct a newish  
user to the right list.

> Please no. I think it is a VERY good idea to have a collection  
> there for
> download. Pretty soon, I think we'll be able to call the 2.4.3  
> Universal
> build the "officially recommended" build, and we can have a collection
> of packages there for it. I now I'll contribute a few, and I'm sure
> others will as well.
>
> One question is: should they be eggs or traditional *.mpkgs?

I'd certainly prefer eggs where possible, but transitionally we're  
definitely going to need *.mpkgs.  Maybe both for now, and/or  
separate pages for .mpkgs and binary eggs?

-bob


From Chris.Barker at noaa.gov  Tue Mar 28 02:00:07 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 27 Mar 2006 16:00:07 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions
	as	Intel-capable...
In-Reply-To: <2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
Message-ID: <44287C87.4020201@noaa.gov>

Bob Ippolito wrote:

> No, I mean that that URL won't directly list any packages.  It'll be a 
> listing of package lists and enough information to direct a newish user 
> to the right list.

Then yes, that's a great idea. I am, as we e-speak, reworking those 
pages with a different structure now. I'm no web designer, but it should 
give us something that someone can make prettieer if they want.

> I'd certainly prefer eggs where possible, but transitionally we're 
> definitely going to need *.mpkgs.  Maybe both for now, and/or separate 
> pages for .mpkgs and binary eggs?

What happens if you double-click on a *.egg?

I like that anyone with OS-X can figure out how to install from a *.mpkg

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From njriley at uiuc.edu  Tue Mar 28 03:01:50 2006
From: njriley at uiuc.edu (Nicholas Riley)
Date: Mon, 27 Mar 2006 19:01:50 -0600
Subject: [Pythonmac-SIG] Marking the pre-built extensions
	as	Intel-capable...
In-Reply-To: <44287C87.4020201@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov>
Message-ID: <20060328010150.GB36580@uiuc.edu>

On Mon, Mar 27, 2006 at 04:00:07PM -0800, Christopher Barker wrote:
> What happens if you double-click on a *.egg?

That'd be an incredibly cool utility to have, and a lot easier to
write than something like PackageManager.

-- 
Nicholas Riley <njriley at uiuc.edu> | <http://www.uiuc.edu/ph/www/njriley>

From daniel at brightfire.com  Tue Mar 28 03:10:42 2006
From: daniel at brightfire.com (daniel at brightfire.com)
Date: Mon, 27 Mar 2006 20:10:42 -0500
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
Message-ID: <20060328011042.7DB5610990F@ws6-4.us4.outblaze.com>

>From: Christopher Barker <Chris.Barker at noaa.gov>
>To: pythonmac-sig at python.org
>Subject: Re: [Pythonmac-SIG] Installing numpy 0.9.6 problems
>Date: Mon, 27 Mar 2006 09:50:33 -0800

>Ronald Oussoren wrote:

>> I'd prefer to have 1 installer for python on OSX, that makes 
>> support  a lot easier.

>And we would like to by able to use Py2App to build apps that will 
>run under both 10.3.9 and 10.4.*

>-Chris

But shouldn't we be able to build -ppc, -i386, and even -ppc64 with gcc 4.0 and then use 'lipo' to paste them to together and let the loader sort them out at run-time? Does the loader on 10.3 know enough to pick '-arch ppc' even with other in there or am I assuming too much? I'd test it myself, but my G5 1.8 Dual has died and my Quad hasn't arrived yet since I am out of town for the week ;-(.

Daniel

From janssen at parc.com  Tue Mar 28 03:08:55 2006
From: janssen at parc.com (Bill Janssen)
Date: Mon, 27 Mar 2006 17:08:55 PST
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
In-Reply-To: Your message of "Mon, 27 Mar 2006 13:12:25 PST."
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com> 
Message-ID: <06Mar27.170859pst."58633"@synergy1.parc.xerox.com>

> > Could someone please put a note on http://pythonmac.org/packages/ to
> > say that the packages listed there are all PPC-only?
> 
> Done

Looks good, thanks.

Bill

From Chris.Barker at noaa.gov  Tue Mar 28 03:10:48 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Mon, 27 Mar 2006 17:10:48 -0800
Subject: [Pythonmac-SIG] The Universal Buld and py2app
In-Reply-To: <2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
Message-ID: <44288D18.6080400@noaa.gov>

Bob,

What's the status of Py2App and the new universal build. I'd like to 
start making packages for it, but I"d really like bdist_mpkg.

Or should I figure out how to make binary eggs instead?

By the way, is it possible to have the new universal build co-exist with 
the 2.4.1 framework build? At the moment, everything seems to be 
working, at least on my PPC machine, but it makes me a bit nervous to 
just write over what I've got working.

A clarification for others:

AFAICT, on a PPC Mac, the new universal build can be used with old, PC 
only, packages for 2.4.1. This means that for those of us on PPC 
machine, we can have all the old packages while we test and build new ones.

Very cool.

-Chris






-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From janssen at parc.com  Tue Mar 28 03:13:42 2006
From: janssen at parc.com (Bill Janssen)
Date: Mon, 27 Mar 2006 17:13:42 PST
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
In-Reply-To: Your message of "Mon, 27 Mar 2006 16:00:07 PST."
	<44287C87.4020201@noaa.gov> 
Message-ID: <06Mar27.171344pst."58633"@synergy1.parc.xerox.com>

> > I'd certainly prefer eggs where possible, but transitionally we're 
> > definitely going to need *.mpkgs.  Maybe both for now, and/or separate 
> > pages for .mpkgs and binary eggs?
> 
> What happens if you double-click on a *.egg?
> 
> I like that anyone with OS-X can figure out how to install from a *.mpkg

Without any experience with "eggs" (I believe that they are another
Phillip Eby brainstorm?), can I suggest that a standard command-line
way of simply wrapping an egg as an mpkg would be a "good thing to
have"?  That way, contributors could just build eggs, and the website
could serve up either a raw egg or a wrapped egg as an mpkg.

Is there a specification for the "egg" format?

Bill

From bob at redivi.com  Tue Mar 28 03:29:02 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 17:29:02 -0800
Subject: [Pythonmac-SIG] The Universal Buld and py2app
In-Reply-To: <44288D18.6080400@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44288D18.6080400@noaa.gov>
Message-ID: <D41D8AA7-E465-494C-A7A0-11689DBC8C67@redivi.com>


On Mar 27, 2006, at 5:10 PM, Christopher Barker wrote:

> Bob,
>
> What's the status of Py2App and the new universal build. I'd like  
> to start making packages for it, but I"d really like bdist_mpkg.

Not Yet.  I have some patches I need to integrate and test.  I'll try  
and get to it this weekend, maybe sooner.

> Or should I figure out how to make binary eggs instead?

That's what works now on universal.

> By the way, is it possible to have the new universal build co-exist  
> with the 2.4.1 framework build? At the moment, everything seems to  
> be working, at least on my PPC machine, but it makes me a bit  
> nervous to just write over what I've got working.

No, they don't coexist.  They install to the same spot.

> A clarification for others:
>
> AFAICT, on a PPC Mac, the new universal build can be used with old,  
> PC only, packages for 2.4.1. This means that for those of us on PPC  
> machine, we can have all the old packages while we test and build  
> new ones.
>
> Very cool.

That's correct.

-bob


From bob at redivi.com  Tue Mar 28 03:33:25 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 17:33:25 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
	Intel-capable...
In-Reply-To: <442889F6.6070805@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov> <442889F6.6070805@noaa.gov>
Message-ID: <D1617966-D221-424B-B61F-A3E9D9F6B11F@redivi.com>


On Mar 27, 2006, at 4:57 PM, Christopher Barker wrote:

> Christopher Barker wrote:
>
>> Then yes, that's a great idea. I am, as we e-speak, reworking  
>> those pages with a different structure now. I'm no web designer,  
>> but it should give us something that someone can make prettier if  
>> they want.
>
> First I added a link to the packages page at the top of the main  
> pythonmac.org page.
>
> I've enclosed a simple rearrangement of the pythonmac.org/packages  
> page. What I've done is create a separate page for each supported  
> build. As the packages built for 10.3 will also work on 10.4, I  
> double listed them, so that a 10.4 user could just go to that page,  
> and see everything that they can use.
>
> I also put a link in  to download the python builds themselves, why  
> not be able to get it all from one place?
>
> Ultimately, we really should have some little semi-automated tool  
> that would make it easy to add a new package to the list, updated  
> the page. For now, doing by hand is not too big a deal, and I  
> don't' want to mess with tools before we settle on a page layout.
>
> The pages are enclosed as a zip file. Ii hope anyone who is  
> interested will give feedback.

Cool, I'll take a look at this probably tomorrow.  The package list  
is currently automated using a dumb little script.. check this:
# the input data
http://pythonmac.org/packages/packages.txt
# the script
http://undefined.org/python/packages.py.txt

-bob



From bob at redivi.com  Tue Mar 28 03:38:08 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 17:38:08 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions
	as	Intel-capable...
In-Reply-To: <44287C87.4020201@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov>
Message-ID: <31B2CF7C-8CEC-4CDB-8C29-9F5783F010B8@redivi.com>


On Mar 27, 2006, at 4:00 PM, Christopher Barker wrote:

> Bob Ippolito wrote:
>
>> No, I mean that that URL won't directly list any packages.  It'll  
>> be a
>> listing of package lists and enough information to direct a newish  
>> user
>> to the right list.
>
> Then yes, that's a great idea. I am, as we e-speak, reworking those
> pages with a different structure now. I'm no web designer, but it  
> should
> give us something that someone can make prettieer if they want.
>
>> I'd certainly prefer eggs where possible, but transitionally we're
>> definitely going to need *.mpkgs.  Maybe both for now, and/or  
>> separate
>> pages for .mpkgs and binary eggs?
>
> What happens if you double-click on a *.egg?

Nothing yet, but it wouldn't be that hard to make a little app that  
associates itself with .egg

> I like that anyone with OS-X can figure out how to install from a  
> *.mpkg

Sure.  Ideally bdist_mpkg will produce *.mpkg installers that just  
contain eggs (for bootstrapping maybe), but that's not currently the  
case.

-bob


From bob at redivi.com  Tue Mar 28 03:44:13 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 17:44:13 -0800
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
In-Reply-To: <20060328011042.7DB5610990F@ws6-4.us4.outblaze.com>
References: <20060328011042.7DB5610990F@ws6-4.us4.outblaze.com>
Message-ID: <9328CD34-4AC6-4AFA-BBC8-FEFA96B073B7@redivi.com>


On Mar 27, 2006, at 5:10 PM, daniel at brightfire.com wrote:

>> From: Christopher Barker <Chris.Barker at noaa.gov>
>> To: pythonmac-sig at python.org
>> Subject: Re: [Pythonmac-SIG] Installing numpy 0.9.6 problems
>> Date: Mon, 27 Mar 2006 09:50:33 -0800
>
>> Ronald Oussoren wrote:
>
>>> I'd prefer to have 1 installer for python on OSX, that makes
>>> support  a lot easier.
>
>> And we would like to by able to use Py2App to build apps that will
>> run under both 10.3.9 and 10.4.*

That will work as long as the apps don't include anything that  
strongly links to 10.4 specific features (e.g. xattr).  We've taken  
steps to make sure that the right thing happens with Python itself,  
but the onus is on third party extensions to also do the right thing  
when relevant.  For most libraries, there shouldn't be an issue.

> But shouldn't we be able to build -ppc, -i386, and even -ppc64 with  
> gcc 4.0 and then use 'lipo' to paste them to together and let the  
> loader sort them out at run-time? Does the loader on 10.3 know  
> enough to pick '-arch ppc' even with other in there or am I  
> assuming too much? I'd test it myself, but my G5 1.8 Dual has died  
> and my Quad hasn't arrived yet since I am out of town for the  
> week ;-(.

ppc64 is definitely not going to work.  You'd need a vanilla unix  
build with no Mac specific stuff to do ppc64, because it only really  
can talk to libSystem.

I don't know what you're talking about with regard to "loader".  If  
you're asking whether 10.3 can run universal binaries, then sure --  
Mac OS X has always had support for fat mach-o files, it's just the  
toolchain that's changed such that you can reasonably make them now.

-bob


From bob at redivi.com  Tue Mar 28 03:45:38 2006
From: bob at redivi.com (Bob Ippolito)
Date: Mon, 27 Mar 2006 17:45:38 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
In-Reply-To: <06Mar27.171344pst."58633"@synergy1.parc.xerox.com>
References: <06Mar27.171344pst."58633"@synergy1.parc.xerox.com>
Message-ID: <AE3A7C9E-908A-410F-AEF3-E885D901C3A2@redivi.com>


On Mar 27, 2006, at 5:13 PM, Bill Janssen wrote:

>>> I'd certainly prefer eggs where possible, but transitionally we're
>>> definitely going to need *.mpkgs.  Maybe both for now, and/or  
>>> separate
>>> pages for .mpkgs and binary eggs?
>>
>> What happens if you double-click on a *.egg?
>>
>> I like that anyone with OS-X can figure out how to install from a  
>> *.mpkg
>
> Without any experience with "eggs" (I believe that they are another
> Phillip Eby brainstorm?), can I suggest that a standard command-line
> way of simply wrapping an egg as an mpkg would be a "good thing to
> have"?  That way, contributors could just build eggs, and the website
> could serve up either a raw egg or a wrapped egg as an mpkg.
>
> Is there a specification for the "egg" format?

http://peak.telecommunity.com/DevCenter/PythonEggs
http://peak.telecommunity.com/DevCenter/EasyInstall
http://peak.telecommunity.com/DevCenter/PkgResources
http://peak.telecommunity.com/DevCenter/setuptools

-bob


From aleaxit at gmail.com  Tue Mar 28 03:54:45 2006
From: aleaxit at gmail.com (Alex Martelli)
Date: Mon, 27 Mar 2006 17:54:45 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
Message-ID: <e8a0972d0603271754v5aa9e54dm6a6ef281123cf5fc@mail.gmail.com>

> Without any experience with "eggs" (I believe that they are another
> Phillip Eby brainstorm?), can I suggest that a standard command-line

Yep, they are.

> Is there a specification for the "egg" format?

http://peak.telecommunity.com/DevCenter/PythonEggs

They're zipfiles containing additional metadata, though it's hard to
find all details and formalspecs on the site...


Alex

From ronaldoussoren at mac.com  Tue Mar 28 09:38:11 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 09:38:11 +0200
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
In-Reply-To: <20060328011042.7DB5610990F@ws6-4.us4.outblaze.com>
References: <20060328011042.7DB5610990F@ws6-4.us4.outblaze.com>
Message-ID: <2231A925-636E-402F-B6CC-47CE0D42F4F8@mac.com>


On 28-mrt-2006, at 3:10, daniel at brightfire.com wrote:

>> From: Christopher Barker <Chris.Barker at noaa.gov>
>> To: pythonmac-sig at python.org
>> Subject: Re: [Pythonmac-SIG] Installing numpy 0.9.6 problems
>> Date: Mon, 27 Mar 2006 09:50:33 -0800
>
>> Ronald Oussoren wrote:
>
>>> I'd prefer to have 1 installer for python on OSX, that makes
>>> support  a lot easier.
>
>> And we would like to by able to use Py2App to build apps that will
>> run under both 10.3.9 and 10.4.*
>
>> -Chris
>
> But shouldn't we be able to build -ppc, -i386, and even -ppc64 with  
> gcc 4.0 and then use 'lipo' to paste them to together and let the  
> loader sort them out at run-time? Does the loader on 10.3 know  
> enough to pick '-arch ppc' even with other in there or am I  
> assuming too much? I'd test it myself, but my G5 1.8 Dual has died  
> and my Quad hasn't arrived yet since I am out of town for the  
> week ;-(.

Sigh, the universal python build already runs on 10.3.9 and supports  
ppc and i386. I don't have a PPC64 box so cannot even try to support  
this. Supporting this in the main build would also require more  
special casing because ppc64 processes can only use a small part of  
the OSX ABI.

Ronald

>
> Daniel
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 28 09:50:39 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 09:50:39 +0200
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
	Intel-capable...
In-Reply-To: <20060328010150.GB36580@uiuc.edu>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov> <20060328010150.GB36580@uiuc.edu>
Message-ID: <A8EE5398-6B9D-466F-AFD0-4E78181427B9@mac.com>


On 28-mrt-2006, at 3:01, Nicholas Riley wrote:

> On Mon, Mar 27, 2006 at 04:00:07PM -0800, Christopher Barker wrote:
>> What happens if you double-click on a *.egg?
>
> That'd be an incredibly cool utility to have, and a lot easier to
> write than something like PackageManager.

Before anyone starts coding, what should happen when you double-click
on an egg? I'd say this should bring up a dialog that allows you to
install the egg and possibly set some options. Installation will be done
using easy_install.

Ronald

>
> --  
> Nicholas Riley <njriley at uiuc.edu> | <http://www.uiuc.edu/ph/www/ 
> njriley>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From ronaldoussoren at mac.com  Tue Mar 28 11:50:48 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 11:50:48 +0200
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
	Intel-capable...
In-Reply-To: <2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
Message-ID: <41F369A5-D50C-49E3-97F9-52C8BBC771FC@mac.com>


On 28-mrt-2006, at 1:18, Bob Ippolito wrote:

>
> On Mar 27, 2006, at 2:55 PM, Christopher Barker wrote:
>
>> Bob Ippolito wrote:
>>
>>> That whole section really needs to be restructured to address the
>>> common questions and issues people have regarding finding the right
>>> packages for them.
>>
>> great idea.
>>
>>> I think I'd like to turn http://pythonmac.org/
>>> packages/ into just a launch pad for finding a list of packages for
>>> the Python that you're interested in.
>>
>> Do you mean that it wouldn't actually host them for download there?
>
> No, I mean that that URL won't directly list any packages.  It'll be
> a listing of package lists and enough information to direct a newish
> user to the right list.
>
>> Please no. I think it is a VERY good idea to have a collection
>> there for
>> download. Pretty soon, I think we'll be able to call the 2.4.3
>> Universal
>> build the "officially recommended" build, and we can have a  
>> collection
>> of packages there for it. I now I'll contribute a few, and I'm sure
>> others will as well.
>>
>> One question is: should they be eggs or traditional *.mpkgs?
>
> I'd certainly prefer eggs where possible, but transitionally we're
> definitely going to need *.mpkgs.  Maybe both for now, and/or
> separate pages for .mpkgs and binary eggs?

For casual users it is easier to install mpkgs, hence we'll need
mpkgs until someone writes the tool that allows one to double-click
on eggs to install them. I'm tempted to do so myself just to avoid
packaging stuff twice :-)

Mpkgs have another advantage: they allow you to include documentation
and examples into the package. None of the existing packages on  
pythonmac.org
(except for pyobjc and possibly py2app) actually use this possibility,
therefore I'd say this is not a very important advantage.

Ronald


From daniel at brightfire.com  Tue Mar 28 17:46:34 2006
From: daniel at brightfire.com (daniel at brightfire.com)
Date: Tue, 28 Mar 2006 10:46:34 -0500
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
Message-ID: <20060328154634.9B48A9F036@ws6-2.us4.outblaze.com>


> ----- Original Message -----
> From: "Ronald Oussoren" <ronaldoussoren at mac.com>
> To: daniel at brightfire.com
> Subject: Re: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
> Date: Tue, 28 Mar 2006 09:38:11 +0200
> 
> Sigh, the universal python build already runs on 10.3.9 and 
> supports  ppc and i386. I don't have a PPC64 box so cannot even try 
> to support  this. Supporting this in the main build would also 
> require more  special casing because ppc64 processes can only use a 
> small part of  the OSX ABI.
> 
> Ronald

Not saying it should, I was just making the point that FAT binaries can be constructed in more than even two ways. So it seems the numpy build is treating the 'arch' flags as command line options that can be logically combined and not as seperate binary build selectors as it should. Sorry for trying to help. I am approriately chastised now.


From daniel at brightfire.com  Tue Mar 28 17:49:35 2006
From: daniel at brightfire.com (daniel at brightfire.com)
Date: Tue, 28 Mar 2006 10:49:35 -0500
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
Message-ID: <20060328154936.57E17EE425@ws6-1.us4.outblaze.com>


> ----- Original Message -----
> From: "Bob Ippolito" <bob at redivi.com>
> To: daniel at brightfire.com
> Subject: Re: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
> Date: Mon, 27 Mar 2006 17:44:13 -0800
> 
> 
> 
> On Mar 27, 2006, at 5:10 PM, daniel at brightfire.com wrote:
> 
> >> From: Christopher Barker <Chris.Barker at noaa.gov>
> >> To: pythonmac-sig at python.org
> >> Subject: Re: [Pythonmac-SIG] Installing numpy 0.9.6 problems
> >> Date: Mon, 27 Mar 2006 09:50:33 -0800
> >
> >> Ronald Oussoren wrote:
> >
> >>> I'd prefer to have 1 installer for python on OSX, that makes
> >>> support  a lot easier.
> >
> >> And we would like to by able to use Py2App to build apps that will
> >> run under both 10.3.9 and 10.4.*
> 
> That will work as long as the apps don't include anything that  
> strongly links to 10.4 specific features (e.g. xattr).  We've taken 
>   steps to make sure that the right thing happens with Python 
> itself,  but the onus is on third party extensions to also do the 
> right thing  when relevant.  For most libraries, there shouldn't be 
> an issue.
> 
> > But shouldn't we be able to build -ppc, -i386, and even -ppc64 
> > with  gcc 4.0 and then use 'lipo' to paste them to together and 
> > let the  loader sort them out at run-time? Does the loader on 
> > 10.3 know  enough to pick '-arch ppc' even with other in there or 
> > am I  assuming too much? I'd test it myself, but my G5 1.8 Dual 
> > has died  and my Quad hasn't arrived yet since I am out of town 
> > for the  week ;-(.
> 
> ppc64 is definitely not going to work.  You'd need a vanilla unix  
> build with no Mac specific stuff to do ppc64, because it only 
> really  can talk to libSystem.

 I understand--just making a point that the kernel checks the header of the binary to determine how to have it loaded and by which dynamic linker. May be I am stating the obvious, but that,  and not combining then with 'lipo' is my assesment of the incorrect configuration of the numpy build 'recipe'

> I don't know what you're talking about with regard to "loader".  If 
>   you're asking whether 10.3 can run universal binaries, then sure 
> --  Mac OS X has always had support for fat mach-o files, it's just 
> the  toolchain that's changed such that you can reasonably make 
> them now.

Well I sacrificed recision for brevity. So...I meant the appropriate dynamic linker for the binary as I mentioned above--I was just trying to be terse. Correct me if I am wrong, but still glossing over some details and simplifying, the kernel checks the header of the binary  to determine which dynamic linker to use to load the binary's segments. Then they are loaded/overlayed into the process space according to their type by the appropriate linker/segment loader/thing. 'dylibs'  are loaded with the timing of the loading depending upon the binding choice, pre-binding being no longer possible in 10.4 and just-in-time being the recommendation now as I recall.

So it's just some drudgery to compile for both formats, use 'lipo' to combine them and we have a universal app. Right? Not trying trivialize the work involved--just stating that I think the tools are there and its possible.

Daniel

> -bob

>


From ronaldoussoren at mac.com  Tue Mar 28 17:55:28 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 17:55:28 +0200
Subject: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
In-Reply-To: <20060328154634.9B48A9F036@ws6-2.us4.outblaze.com>
References: <20060328154634.9B48A9F036@ws6-2.us4.outblaze.com>
Message-ID: <83D6A3DA-B513-45FA-90BE-311D1FAE1C6E@mac.com>


On 28-mrt-2006, at 17:46, daniel at brightfire.com wrote:

>
>> ----- Original Message -----
>> From: "Ronald Oussoren" <ronaldoussoren at mac.com>
>> To: daniel at brightfire.com
>> Subject: Re: [Pythonmac-SIG] Pythonmac-SIG Digest, Vol 35, Issue 51
>> Date: Tue, 28 Mar 2006 09:38:11 +0200
>>
>> Sigh, the universal python build already runs on 10.3.9 and
>> supports  ppc and i386. I don't have a PPC64 box so cannot even try
>> to support  this. Supporting this in the main build would also
>> require more  special casing because ppc64 processes can only use a
>> small part of  the OSX ABI.
>>
>> Ronald
>
> Not saying it should, I was just making the point that FAT binaries  
> can be constructed in more than even two ways. So it seems the  
> numpy build is treating the 'arch' flags as command line options  
> that can be logically combined and not as seperate binary build  
> selectors as it should. Sorry for trying to help. I am approriately  
> chastised now.

My previous message was a bit too sharp, I hadn't intended to punish  
anyone :-)

The numpy build can be failing for several reasons:

* Using a too old build of universal macpython, early builds didn't  
support building extensions on 10.3

* My patch for distutils is buggy

* numpy's custom distutils code is interfering with my changes

If you're using Xcode 2.1 or later on OSX 10.4 you can several -arch  
flags and binaries will be build for all named architectures. This is  
very convenient, otherwise we'd have to teach distutils about the  
compile+lipo dance you mentioned and I don't want to perform the  
necessary surgery on distutils to get it to do that if I don't have  
to :-)

Ronald

From Chris.Barker at noaa.gov  Tue Mar 28 18:20:06 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 28 Mar 2006 08:20:06 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
 Intel-capable...
In-Reply-To: <A8EE5398-6B9D-466F-AFD0-4E78181427B9@mac.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov> <20060328010150.GB36580@uiuc.edu>
	<A8EE5398-6B9D-466F-AFD0-4E78181427B9@mac.com>
Message-ID: <44296236.2030001@noaa.gov>

Ronald Oussoren wrote:

> Before anyone starts coding, what should happen when you double-click
> on an egg? I'd say this should bring up a dialog that allows you to
> install the egg and possibly set some options. Installation will be done
> using easy_install.

That sounds good to me.

Does easy-install come with the new Universal build package? Or is a 
good first step to make a package out of that?

I'd like to start populating the world with Universal packages. I see a 
couple possible routes:

1) wait for Bob to finish Py2App (bdist_mpkg): I think he's indicated 
he's hoping to get that done this weekend.

2) build *.eggs, and hope someone comes up with a app to associate them 
with later. (and it the meantime they can be installed with the command 
line easy-install

3) just build them for myself, and wait until this settles out some 
more. Maybe start a Wiki page with a list of packages and an indication 
of what was needed to do to build them.

-Chris

---
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Chris.Barker at noaa.gov  Tue Mar 28 18:26:34 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 28 Mar 2006 08:26:34 -0800
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
 Intel-capable...
In-Reply-To: <41F369A5-D50C-49E3-97F9-52C8BBC771FC@mac.com>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<41F369A5-D50C-49E3-97F9-52C8BBC771FC@mac.com>
Message-ID: <442963BA.4050706@noaa.gov>

Ronald Oussoren wrote:

> For casual users it is easier to install mpkgs, hence we'll need
> mpkgs until someone writes the tool that allows one to double-click
> on eggs to install them. I'm tempted to do so myself just to avoid
> packaging stuff twice :-)
> 
> Mpkgs have another advantage: they allow you to include documentation
> and examples into the package. None of the existing packages on 
> pythonmac.org
> (except for pyobjc and possibly py2app) actually use this possibility,
> therefore I'd say this is not a very important advantage.

Doesn't the wxPython package?

Anyway, another advantage of eggs is that they can be versioned, so that 
users can have different versions of a package installed at once.

Couldn't we put a *.egg into a *.mpkg? and get the best of both worlds?

-Chris



-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Tue Mar 28 19:18:56 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 19:18:56 +0200
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
	Intel-capable...
In-Reply-To: <442963BA.4050706@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<41F369A5-D50C-49E3-97F9-52C8BBC771FC@mac.com>
	<442963BA.4050706@noaa.gov>
Message-ID: <CDD68E5F-E35F-493F-BADC-D7F8A835D848@mac.com>


On 28-mrt-2006, at 18:26, Christopher Barker wrote:

> Ronald Oussoren wrote:
>
>> For casual users it is easier to install mpkgs, hence we'll need
>> mpkgs until someone writes the tool that allows one to double-click
>> on eggs to install them. I'm tempted to do so myself just to avoid
>> packaging stuff twice :-)
>>
>> Mpkgs have another advantage: they allow you to include documentation
>> and examples into the package. None of the existing packages on
>> pythonmac.org
>> (except for pyobjc and possibly py2app) actually use this  
>> possibility,
>> therefore I'd say this is not a very important advantage.
>
> Doesn't the wxPython package?
>
> Anyway, another advantage of eggs is that they can be versioned, so  
> that
> users can have different versions of a package installed at once.
>
> Couldn't we put a *.egg into a *.mpkg? and get the best of both  
> worlds?

There's currently two ways to easily build redistributable packages for
OSX: the oldest is bdist_mpkg, the newer is bdist_egg.

bdist_mpkg is/was part of py2app and includes a script that allows you
to build a .mpkg for every python package that includes a setup.py  
script.

bdist_egg is part of setuptools and AFAIK doesn't include a tool for  
building
eggs for python packages that don't support it, but there is an easy way
to build them.

There is not yet an easy way to place an egg into a .mpkg, although  
Bob has
indicated that he wants to restructure bdist_mpkg to to this.

Ronald

>
> -Chris
>
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                      		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From janssen at parc.com  Tue Mar 28 19:26:01 2006
From: janssen at parc.com (Bill Janssen)
Date: Tue, 28 Mar 2006 09:26:01 PST
Subject: [Pythonmac-SIG] Marking the pre-built extensions as
	Intel-capable...
In-Reply-To: Your message of "Tue, 28 Mar 2006 08:26:34 PST."
	<442963BA.4050706@noaa.gov> 
Message-ID: <06Mar28.092610pst."58633"@synergy1.parc.xerox.com>

> Couldn't we put a *.egg into a *.mpkg? and get the best of both worlds?

That would be my vote.

Bill

From ronaldoussoren at mac.com  Tue Mar 28 19:27:34 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Tue, 28 Mar 2006 19:27:34 +0200
Subject: [Pythonmac-SIG] Marking the pre-built extensions	as
	Intel-capable...
In-Reply-To: <44296236.2030001@noaa.gov>
References: <06Mar27.113702pst.58633@synergy1.parc.xerox.com>
	<5FB132E3-C499-4C2A-A927-CAEAE1839855@redivi.com>
	<44286D61.1070504@noaa.gov>
	<2CDC34C5-909B-4642-93EA-9BC73B05156D@redivi.com>
	<44287C87.4020201@noaa.gov> <20060328010150.GB36580@uiuc.edu>
	<A8EE5398-6B9D-466F-AFD0-4E78181427B9@mac.com>
	<44296236.2030001@noaa.gov>
Message-ID: <B3969A28-9020-488D-A47F-B9C0CD090F86@mac.com>


On 28-mrt-2006, at 18:20, Christopher Barker wrote:

> Ronald Oussoren wrote:
>
>> Before anyone starts coding, what should happen when you double-click
>> on an egg? I'd say this should bring up a dialog that allows you to
>> install the egg and possibly set some options. Installation will  
>> be done
>> using easy_install.
>
> That sounds good to me.
>
> Does easy-install come with the new Universal build package? Or is a
> good first step to make a package out of that?
>
> I'd like to start populating the world with Universal packages. I  
> see a
> couple possible routes:
>
> 1) wait for Bob to finish Py2App (bdist_mpkg): I think he's indicated
> he's hoping to get that done this weekend.
>
> 2) build *.eggs, and hope someone comes up with a app to associate  
> them
> with later. (and it the meantime they can be installed with the  
> command
> line easy-install
>
> 3) just build them for myself, and wait until this settles out some
> more. Maybe start a Wiki page with a list of packages and an  
> indication
> of what was needed to do to build them.

The universal build does not include setuptools. I'd include some  
bootstrap
code in the tool that is associated with .egg files.

I'm going to build eggs for all software that I use, with mpkgs for  
software
that doesn't support eggs (like twisted).

Ronald

>
> -Chris
>
> ---
> Christopher Barker, Ph.D.
> Oceanographer
>                                      		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From dgrassi at lexar.com  Tue Mar 28 20:20:47 2006
From: dgrassi at lexar.com (Dan Grassi)
Date: Tue, 28 Mar 2006 13:20:47 -0500
Subject: [Pythonmac-SIG] SQLite & python
Message-ID: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>

Hi,

I am totally confused by all the versions and name conflicts of py- 
sqlite and sqlite not to mention the python version confusion.  What  
I really want to do is use both sqlite version 2 and sqlite version 3  
since I have existing sqlite version2 DBs and new CoreData apps that  
use sqlite version 3.  From a version/nomenclature standpoint there  
is a huge mess eg: py-sqlite works with either sqlite 2 or sqlite 3  
and py-sqlite2 works with sqlite version 3.

I have been using py-sqlite with sqlite version 2 successfully.

First problem I see is that there are two version of Python  on my  
Tiger system:
/usr/bin/python version 2.3.5
/usr/local/bin/python version 2.4.1

Second problem:
The (Apple) installed version of sqlite3 is 3.0.8.6
py-sqlite 2.1 requires sqlite 3.2.2
py-sqlite 2.0 requires sqlite 3.1
py-sqlite 1.1 requires sqlite 3.x
So it seems that py-sqlite 1.1 is the one I need.

If I install it my installed py-sqlite that used the python sqlite  
version 2 gets overwritten, not nice.  It also seems to get  
instaslled in the python 2.3 library, also not nice.

If I go to darwin ports it wants to upgrade my python to 2.4.2 first.


Does anyone have a solution?


All in all the whole Python database (lack of) mess is ridiculous!  I  
know that Guido does not feel that databases have any place in python  
but I don't get it, why does the "batteries included" not include  
database support, databases like MySQL, PostgreSQL & SQLite?


At this point I am ready to mostly give up on Python after over 10  
years of use and evangelizing.


Dan

From delza at livingcode.org  Tue Mar 28 21:52:12 2006
From: delza at livingcode.org (Dethe Elza)
Date: Tue, 28 Mar 2006 11:52:12 -0800
Subject: [Pythonmac-SIG] SQLite & python
In-Reply-To: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>
References: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>
Message-ID: <24d517dd0603281152o3029f0ffmae7688ee13edb644@mail.gmail.com>

On 3/28/06, Dan Grassi <dgrassi at lexar.com> wrote:
> Hi,
>
> I am totally confused by all the versions and name conflicts of py-
> sqlite and sqlite not to mention the python version confusion.  What
> I really want to do is use both sqlite version 2 and sqlite version 3
> since I have existing sqlite version2 DBs and new CoreData apps that
> use sqlite version 3.  From a version/nomenclature standpoint there
> is a huge mess eg: py-sqlite works with either sqlite 2 or sqlite 3
> and py-sqlite2 works with sqlite version 3.
>
> I have been using py-sqlite with sqlite version 2 successfully.
>
> First problem I see is that there are two version of Python  on my
> Tiger system:
> /usr/bin/python version 2.3.5

The above is Apple-supplied Python.  Apple uses it for internal
scripting stuff, you should pretend it isn't there.  Ignore it, don't
use it, but whatever you do, don't try to remove or replace it.

> /usr/local/bin/python version 2.4.1

This is the one you want.  If your shell doesn't already point to this
by default when you type "python" on the command line, it should.  Try
"which python" to see what your default is.

> Second problem:
> The (Apple) installed version of sqlite3 is 3.0.8.6
> py-sqlite 2.1 requires sqlite 3.2.2
> py-sqlite 2.0 requires sqlite 3.1
> py-sqlite 1.1 requires sqlite 3.x
> So it seems that py-sqlite 1.1 is the one I need.

To complicate matters, there is also APSW (Another Python SQLite
Wrapper) which can be used with any version of SQLite.

http://www.rogerbinns.com/apsw.html

> If I install it my installed py-sqlite that used the python sqlite
> version 2 gets overwritten, not nice.  It also seems to get
> instaslled in the python 2.3 library, also not nice.

Probably your PATH environment variable points to the Apple-installed
python before the current python.  You can change this, probably in
the ".bash-profile" file in your home directory.  Make sure
/usr/local/bin comes before /usr/bin in the PATH.

> If I go to darwin ports it wants to upgrade my python to 2.4.2 first.

Why do you need DarwinPorts?  You have both Python and SQLite
installed already.  Just download the wrapper (pysqlite or apsw) and
install it using "python setup.py build; sudo python setup.py install"
 after making sure the correct python is in your PATH.

> Does anyone have a solution?

See above.

> All in all the whole Python database (lack of) mess is ridiculous!  I
> know that Guido does not feel that databases have any place in python
> but I don't get it, why does the "batteries included" not include
> database support, databases like MySQL, PostgreSQL & SQLite?

Python's batteries do include BerkelyDB "out of the box", as well as
gdbm and/or dbm, including a built-in portable implementation of dbm
for platforms that don't have it already.  There are wrappers for
PostgreSQL, MySQL, Oracle, SQLite, etc. There are object-relational
mappings like SQLObject. The python world has a very rich and lively
database environment.

> At this point I am ready to mostly give up on Python after over 10
> years of use and evangelizing.

That would be too bad.  It's just getting really good.

--Dethe

From egb at usno.navy.mil  Tue Mar 28 22:38:00 2006
From: egb at usno.navy.mil (Eric G. Barron)
Date: Tue, 28 Mar 2006 15:38:00 -0500
Subject: [Pythonmac-SIG] SQLite & python
In-Reply-To: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>
References: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>
Message-ID: <13FA3F95-E1C6-4939-A0CD-2C047E95C648@usno.navy.mil>


On 28 Mar 2006, at 1320, Dan Grassi wrote:

> All in all the whole Python database (lack of) mess is ridiculous!  I
> know that Guido does not feel that databases have any place in python
> but I don't get it, why does the "batteries included" not include
> database support, databases like MySQL, PostgreSQL & SQLite?

Funny you mention that...there's a related discussion on python-dev  
that just began this morning.

http://mail.python.org/pipermail/python-dev/2006-March/062905.html

Guido hasn't weighed in yet, though.

From Chris.Barker at noaa.gov  Tue Mar 28 23:18:20 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Tue, 28 Mar 2006 13:18:20 -0800
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
Message-ID: <4429A81C.9010805@noaa.gov>

Hi all,

I'm embarking on the project of building the various packages I need 
with the new Universal Build. I've set up a Wiki page to keep track of 
what I (and others) have done. Please take a look if you are interested, 
and add your own work.


http://pythonmac.org/wiki/UniversalPackages

-Chris




-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Wed Mar 29 20:33:45 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 29 Mar 2006 20:33:45 +0200
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
In-Reply-To: <4429A81C.9010805@noaa.gov>
References: <4429A81C.9010805@noaa.gov>
Message-ID: <EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>


On 28-mrt-2006, at 23:18, Christopher Barker wrote:

> Hi all,
>
> I'm embarking on the project of building the various packages I need
> with the new Universal Build. I've set up a Wiki page to keep track of
> what I (and others) have done. Please take a look if you are  
> interested,
> and add your own work.


To answer to the title of the message, the problem with numpy on  
10.3.9 is indead caused by numpy patching distutils. This interferes  
with my changes. I have slightly modified my changes to distutils and  
numpy now builds correctly. This change will be in the universal  
build for python 2.4.3.

BTW. from my limited testing it seems that numpy binaries build on  
10.4 also work on 10.3.9. I don't actually use numpy myself, so don't  
know how to properly test it.

Ronald



From rowen at cesmail.net  Wed Mar 29 20:54:51 2006
From: rowen at cesmail.net (Russell E. Owen)
Date: Wed, 29 Mar 2006 10:54:51 -0800
Subject: [Pythonmac-SIG] How to get MacOS X version?
Message-ID: <rowen-AB194C.10545129032006@sea.gmane.org>

How can I get the version of MacOS X within a python program?
(e.g. 10.4.5 -- something that is recognizable, rather than the kernel 
version).

In theory I can get it from gestalt, but I haven't figured out how to do 
it via the Carbon package, or even if it's possible.

Any hints?

-- Russell

P.S. I'm going to be trying to figure out the same thing for unix and 
even Windows. I think I know how to do those, but if you happen to know 
or have sample code, that'd be great.


From robert.kern at gmail.com  Wed Mar 29 21:14:50 2006
From: robert.kern at gmail.com (Robert Kern)
Date: Wed, 29 Mar 2006 13:14:50 -0600
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
In-Reply-To: <EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>
References: <4429A81C.9010805@noaa.gov>
	<EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>
Message-ID: <442ADCAA.1070306@gmail.com>

Ronald Oussoren wrote:
> On 28-mrt-2006, at 23:18, Christopher Barker wrote:
> 
>>Hi all,
>>
>>I'm embarking on the project of building the various packages I need
>>with the new Universal Build. I've set up a Wiki page to keep track of
>>what I (and others) have done. Please take a look if you are  
>>interested,
>>and add your own work.
> 
> To answer to the title of the message, the problem with numpy on  
> 10.3.9 is indead caused by numpy patching distutils. This interferes  
> with my changes. I have slightly modified my changes to distutils and  
> numpy now builds correctly. This change will be in the universal  
> build for python 2.4.3.

Excellent! Thank you! Is there a source repository where these changes are being
checked in? I think numpy can take the responsibility of keeping up with your
changes.

> BTW. from my limited testing it seems that numpy binaries build on  
> 10.4 also work on 10.3.9. I don't actually use numpy myself, so don't  
> know how to properly test it.

numpy.test() will run the test suite.

-- 
Robert Kern
robert.kern at gmail.com

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco

From ronaldoussoren at mac.com  Wed Mar 29 21:19:40 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 29 Mar 2006 21:19:40 +0200
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
In-Reply-To: <qnk64lx6pw4.fsf@arbutus.physics.mcmaster.ca>
References: <4429A81C.9010805@noaa.gov>
	<EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>
	<qnk64lx6pw4.fsf@arbutus.physics.mcmaster.ca>
Message-ID: <27675B9C-AFDF-44E8-A19F-527DC1AF23FE@mac.com>


On 29-mrt-2006, at 20:56, David M. Cooke wrote:
>
> Numpy's unit tests can be run by
>
>>>> import numpy
>>>> numpy.test()

Duh, how obvious can it get. Thanks for the tip!

Ronald

>
> --  
> |>|\/|<
> /--------------------------------------------------------------------- 
> -----\
> |David M. Cooke                      http:// 
> arbutus.physics.mcmaster.ca/dmc/
> |cookedm at physics.mcmaster.ca


From cookedm at physics.mcmaster.ca  Wed Mar 29 20:56:43 2006
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: Wed, 29 Mar 2006 13:56:43 -0500
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
In-Reply-To: <EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com> (Ronald Oussoren's
	message of "Wed, 29 Mar 2006 20:33:45 +0200")
References: <4429A81C.9010805@noaa.gov>
	<EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>
Message-ID: <qnk64lx6pw4.fsf@arbutus.physics.mcmaster.ca>

Ronald Oussoren <ronaldoussoren at mac.com> writes:

> On 28-mrt-2006, at 23:18, Christopher Barker wrote:
>
>> Hi all,
>>
>> I'm embarking on the project of building the various packages I need
>> with the new Universal Build. I've set up a Wiki page to keep track of
>> what I (and others) have done. Please take a look if you are  
>> interested,
>> and add your own work.
>
>
> To answer to the title of the message, the problem with numpy on  
> 10.3.9 is indead caused by numpy patching distutils. This interferes  
> with my changes. I have slightly modified my changes to distutils and  
> numpy now builds correctly. This change will be in the universal  
> build for python 2.4.3.
>
> BTW. from my limited testing it seems that numpy binaries build on  
> 10.4 also work on 10.3.9. I don't actually use numpy myself, so don't  
> know how to properly test it.

Numpy's unit tests can be run by

>>> import numpy
>>> numpy.test()

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke                      http://arbutus.physics.mcmaster.ca/dmc/
|cookedm at physics.mcmaster.ca

From bob at redivi.com  Wed Mar 29 21:27:12 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 29 Mar 2006 11:27:12 -0800
Subject: [Pythonmac-SIG] How to get MacOS X version?
In-Reply-To: <rowen-AB194C.10545129032006@sea.gmane.org>
References: <rowen-AB194C.10545129032006@sea.gmane.org>
Message-ID: <ADB6B9D0-5D12-4495-B2B4-65E0291204E5@redivi.com>


On Mar 29, 2006, at 10:54 AM, Russell E. Owen wrote:

> How can I get the version of MacOS X within a python program?
> (e.g. 10.4.5 -- something that is recognizable, rather than the kernel
> version).
>
> In theory I can get it from gestalt, but I haven't figured out how  
> to do
> it via the Carbon package, or even if it's possible.
>
> Any hints?

platform.mac_ver() -- this is almost definitely broken on Python  
2.3.5 on Intel though, because Carbon.* and gestalt are mostly broken  
there.  The reliable way that has always and should always work is to  
parse either /usr/bin/sw_vers or /System/Library/CoreServices/ 
SystemVersion.plist.

> P.S. I'm going to be trying to figure out the same thing for unix and
> even Windows. I think I know how to do those, but if you happen to  
> know
> or have sample code, that'd be great.

Probably also in the platform module...

-bob


From ronaldoussoren at mac.com  Wed Mar 29 22:12:40 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Wed, 29 Mar 2006 22:12:40 +0200
Subject: [Pythonmac-SIG] Bulding numpy with the Universal Build
In-Reply-To: <442ADCAA.1070306@gmail.com>
References: <4429A81C.9010805@noaa.gov>
	<EC52B3B7-385B-47BD-9796-CDBFEEE3C827@mac.com>
	<442ADCAA.1070306@gmail.com>
Message-ID: <5B647764-C3CF-4EA7-819C-8FC9A40F0E04@mac.com>


On 29-mrt-2006, at 21:14, Robert Kern wrote:

> Ronald Oussoren wrote:
>> On 28-mrt-2006, at 23:18, Christopher Barker wrote:
>>
>>> Hi all,
>>>
>>> I'm embarking on the project of building the various packages I need
>>> with the new Universal Build. I've set up a Wiki page to keep  
>>> track of
>>> what I (and others) have done. Please take a look if you are
>>> interested,
>>> and add your own work.
>>
>> To answer to the title of the message, the problem with numpy on
>> 10.3.9 is indead caused by numpy patching distutils. This interferes
>> with my changes. I have slightly modified my changes to distutils and
>> numpy now builds correctly. This change will be in the universal
>> build for python 2.4.3.
>
> Excellent! Thank you! Is there a source repository where these  
> changes are being
> checked in? I think numpy can take the responsibility of keeping up  
> with your
> changes.

There's a repository at http://svn.pythonmac.org/python24/python24- 
fat. There is
no need to keep up with my changes to distutils though, numpy  
currently compiles
both on 10.3.9 and 10.4 without any changes.

Ronald


From kquirk at solidworks.com  Wed Mar 29 23:35:30 2006
From: kquirk at solidworks.com (Kent Quirk)
Date: Wed, 29 Mar 2006 16:35:30 -0500
Subject: [Pythonmac-SIG] How to get MacOS X version?
Message-ID: <AD85EF0CD3800548B3B22C832F08788E029DDDCF@CORP-MAIL6.solidworks.swk>

-----Original Message-----
From: pythonmac-sig-bounces at python.org
[mailto:pythonmac-sig-bounces at python.org] On Behalf Of Russell E. Owen
Sent: Wednesday, March 29, 2006 1:55 PM

How can I get the version of MacOS X within a python program?
(e.g. 10.4.5 -- something that is recognizable, rather than the kernel 
version).

<snip>

P.S. I'm going to be trying to figure out the same thing for unix and 
even Windows. I think I know how to do those, but if you happen to know 
or have sample code, that'd be great.

-----------

On Windows, you can use sys.getwindowsversion, which will return
something that looks like:

(5, 1, 2600, 2, 'Service Pack 2')

The first 3 values define the windows version; the fourth is generated
by python code as a general NT vs CE vs ME identifier.

Here's some sample code (table data cheerfully borrowed from the Inno
Setup installer tool on Windows).

##########################
import sys

win_versions = {
        '4.0.950':'Windows 95',
        '4.0.1111':'Windows 95 OSR 2 & OSR 2.1',
        '4.0.1212':'Windows 95 OSR 2.5',
        '4.1.1998':'Windows 98',
        '4.1.2222':'Windows 98 Second Edition',
        '4.9.3000':'Windows Me',
        '4.0.1381':'Windows NT 4.0',
        '5.0.2195':'Windows 2000',
        '5.1.2600':'Windows XP',
        '5.2.3790':'Windows Server 2003',
    }

version = win_versions["%d.%d.%d" % sys.getwindowsversion()[:3]]
print version
##########################

Note that this tool does NOT identify 64-bit versions of Windows
(they'll show up as either XP or Server 2K3). I don't know how you'd do
that.


From saggau at gmail.com  Thu Mar 30 01:28:46 2006
From: saggau at gmail.com (Saggau)
Date: Wed, 29 Mar 2006 18:28:46 -0500
Subject: [Pythonmac-SIG] NSUncaughtRuntimeErrorException and
	NSArrayController
Message-ID: <e7a9b3d30603291528s229d6e50h15c03c701abdb68f@mail.gmail.com>

Hi folks.  I've done a load of googling after getting SIG(variable) errors
when adding and removing objects to/from an arrayController in pyobjc
and can't seem
to figure this one out.  I've tried the pyobjc list and have no
response so far.
Apologies for cross-posting and please let me know if I'm asking the
question poorly, vaguely,
or on the wrong mailing list.

I set Signals.dumpStackOnFatalSignal() as well as
Debugging.installVerboseExceptionHandler() and then I start adding python
(subclassing NSObject) objects to my NSArrayController.  After a variable
number of additions and/or subtractions, I get the stack trace below.  (there
are no other errors, including those related to KVC compliance, etc)

Intuition is telling me that I may be in trouble because I'm instantiating
python singleton objects (subclassing the class directly below) as
attributes of the object that I'm adding to the NSArrayController, but I
could be way off here.

*Update: I removed all singletons from my code and have the same behavior.
*Update: this happens whether or not I have the objects delete on
removal from the
controller.

class scAbstractSingleton(object):
   def __new__(cls, *args, **kwds):
      it = cls.__dict__.get("__it__")
      if it is not None: return it
      cls.__it__ = it = object.__new__(cls)
      it.init(*args, **kwds)
      return it

   def init(self, *args, **kwds):
      pass

Is there a way post-mortem to get the object that *was* at 0x114d740 in the
trace below?  Is there a way to recursively enumerate all attributes of a
given object to do a print(theAttribute) on instantiation?  That could be
crazy...  Any cool pdb tricks I should know?

I assume that figuring out which object (0x114d740 in the trace below) was
freed is a good place to start, but I don't know if that's possible in
retrospect, as that object is freed.  I can see the memory referenced in
gdb, but I don't see anything that would help...at least not to my
inexperienced eyes.  Who knows the proper voodoo?

Thanks in advance to the cocoa/pythonistas!

Cheers,
Jonathan Saggau

2006-03-27 21:10:28.897 Trigger[436] *** ObjC exception
'NSUncaughtRuntimeErrorException' (reason: 'objc: FREED(id): message retain
sent to freed object=0x114d740') discarded
Stack trace (most recent call last):
  __dyld_start (in dyld)
  0x00002414 (in Trigger)
  0x00003e50 (in Trigger)
  0x00003db0 (in Trigger)
  _PyRun_SimpleFileExFlags (in Python)
  _PyRun_FileExFlags (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_EvalCodeEx (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyEval_EvalCodeEx (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_GetFuncDesc (in Python)
  __PyUnicodeUCS2_IsAlpha (in Python)
  _PyRun_FileExFlags (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_EvalCodeEx (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyEval_EvalCodeEx (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_GetFuncDesc (in Python)
  _objc_NSApplicationMain (in _AppKit.so) (_AppKit.m:129)
  _NSApplicationMain (in AppKit)
  -[NSApplication run] (in AppKit)
  -[NSApplication sendEvent:] (in AppKit)
  -[NSWindow sendEvent:] (in AppKit)
  -[NSControl mouseDown:] (in AppKit)
  -[NSButtonCell trackMouse:inRect:ofView:untilMouseUp:] (in AppKit)
  -[NSCell trackMouse:inRect:ofView:untilMouseUp:] (in AppKit)
  -[NSCell _sendActionFrom:] (in AppKit)
  -[NSControl sendAction:to:] (in AppKit)
  -[NSApplication sendAction:to:from:] (in AppKit)
  _ffi_closure_ASM (in _objc.so) (darwin_closure.S:95)
  _ffi_closure_helper_DARWIN (in _objc.so) (ffi_darwin.c:703)
  _method_stub (in _objc.so) (libffi_support.m:479)
  _PyObject_Call (in Python)
  _pysel_call (in _objc.so) (selector.m:941)
  _PyObject_Call (in Python)
  _PyFunction_SetClosure (in Python)
  _PyEval_EvalCodeEx (in Python)
  _PyEval_EvalCode (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyEval_GetFuncDesc (in Python)
  _PyObject_Call (in Python)
  _objcsel_call (in _objc.so) (selector.m:517)
  _PyObjCFFI_Caller (in _objc.so) (libffi_support.m:1300)
  _ffi_call (in _objc.so) (ffi_darwin.c:404)
  _ffi_call_DARWIN (in _objc.so) (darwin.S:121)
  -[NSArrayController _insertObject:atArrangedObjectIndex:objectHandler:]
(in AppKit)
  -[NSArrayController didChangeValuesForArrangedKeys:objectKeys:indexKeys:]
(in AppKit)
  -[NSController _notifyObserversForKeyPath:change:] (in AppKit)
  -[NSObject(NSKeyValueObservingPrivate) _notifyObserversForKeyPath:change:]
(in Foundation)
  -[NSSelectionBinder observeValueForKeyPath:ofObject:change:context:] (in
AppKit)
  -[NSValueBinder _observeValueForKeyPath:ofObject:context:] (in AppKit)
  -[NSSelectionBinder
_adjustObject:mode:observedController:observedKeyPath:context:editableState:adjustState:]
(in AppKit)
  -[_NSModelObservingTracker
setIndexReferenceModelObjectArray:clearAllModelObjectObserving:] (in AppKit)
  ___objc_error (in libobjc.A.dylib)
2006-03-27 21:10:28.899 Trigger[436] objc: FREED(id): message retain sent to
freed object=0x114d740



--
The extent to which an individual can resist being blindly led by tradition
is a good measure of his vitality.  - Harry Partch

There is a computer disease that anybody who works with computers knows
about. It's a very serious disease and it interferes completely with the
work. The trouble with computers is that you 'play' with them! - Richard P.
Feynman

Few people are capable of expressing with equanimity opinions which differ
from the prejudices of their social environment.  Most people are not even
capable of forming such opinions.  - Einstein
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060329/4ac5bda3/attachment.htm 

From bob at redivi.com  Thu Mar 30 01:46:50 2006
From: bob at redivi.com (Bob Ippolito)
Date: Wed, 29 Mar 2006 15:46:50 -0800
Subject: [Pythonmac-SIG] NSUncaughtRuntimeErrorException and
	NSArrayController
In-Reply-To: <e7a9b3d30603291528s229d6e50h15c03c701abdb68f@mail.gmail.com>
References: <e7a9b3d30603291528s229d6e50h15c03c701abdb68f@mail.gmail.com>
Message-ID: <074D71FD-BF0D-438C-A9F8-7E0AB96951A5@redivi.com>


On Mar 29, 2006, at 3:28 PM, Saggau wrote:

> Hi folks.  I've done a load of googling after getting SIG(variable)  
> errors
> when adding and removing objects to/from an arrayController in  
> pyobjc and can't seem
> to figure this one out.  I've tried the pyobjc list and have no  
> response so far.
>
> Apologies for cross-posting and please let me know if I'm asking  
> the question poorly, vaguely,
> or on the wrong mailing list.
>
> I set Signals.dumpStackOnFatalSignal() as well as
> Debugging.installVerboseExceptionHandler
> () and then I start adding python
> (subclassing NSObject) objects to my NSArrayController.  After a  
> variable
> number of additions and/or subtractions, I get the stack trace  
> below.  (there
> are no other errors, including those related to KVC compliance, etc)
>
>
> Intuition is telling me that I may be in trouble because I'm  
> instantiating
> python singleton objects (subclassing the class directly below) as
> attributes of the object that I'm adding to the NSArrayController,  
> but I
>
> could be way off here.
>
> *Update: I removed all singletons from my code and have the same  
> behavior.
> *Update: this happens whether or not I have the objects delete on  
> removal from the
> controller.
>
> class scAbstractSingleton(object):
>
>    def __new__(cls, *args, **kwds):
>       it = cls.__dict__.get("__it__")
>       if it is not None: return it
>       cls.__it__ = it = object.__new__(cls)
>       it.init(*args, **kwds)
>       return it
>
>
>    def init(self, *args, **kwds):
>       pass
>
> Is there a way post-mortem to get the object that *was* at  
> 0x114d740 in the
> trace below?  Is there a way to recursively enumerate all  
> attributes of a
> given object to do a print(theAttribute) on instantiation?  That  
> could be
>
> crazy...  Any cool pdb tricks I should know?
>
> I assume that figuring out which object (0x114d740 in the trace  
> below) was
> freed is a good place to start, but I don't know if that's possible in
> retrospect, as that object is freed.  I can see the memory  
> referenced in
>
> gdb, but I don't see anything that would help...at least not to my
> inexperienced eyes.  Who knows the proper voodoo?

Use NSZombieEnabled.

http://developer.apple.com/technotes/tn2004/tn2124.html

-bob


From rowen at cesmail.net  Thu Mar 30 02:18:36 2006
From: rowen at cesmail.net (Russell E. Owen)
Date: Wed, 29 Mar 2006 16:18:36 -0800
Subject: [Pythonmac-SIG] How to get MacOS X version?
References: <rowen-AB194C.10545129032006@sea.gmane.org>
	<ADB6B9D0-5D12-4495-B2B4-65E0291204E5@redivi.com>
Message-ID: <rowen-750F57.16183629032006@sea.gmane.org>

In article <ADB6B9D0-5D12-4495-B2B4-65E0291204E5 at redivi.com>,
 Bob Ippolito <bob at redivi.com> wrote:

> On Mar 29, 2006, at 10:54 AM, Russell E. Owen wrote:
> 
> > How can I get the version of MacOS X within a python program?
> > (e.g. 10.4.5 -- something that is recognizable, rather than the kernel
> > version)....
> 
> platform.mac_ver() -- this is almost definitely broken on Python  
> 2.3.5 on Intel though, because Carbon.* and gestalt are mostly broken  
> there.  The reliable way that has always and should always work is to  
> parse either /usr/bin/sw_vers or /System/Library/CoreServices/ 
> SystemVersion.plist.

Thank you very much! I had just found platform.py via google and was 
starting to study it, but didn't realize it was part of the standard 
library until I got your email.

I'm also glad to know gestalt exists. (I had looked for that, but it's 
not listed in the global module index.)

Also thanks to Kent Quirk for the handy Windows code.

Regards,

-- Russell


From Chris.Barker at noaa.gov  Thu Mar 30 02:33:59 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Wed, 29 Mar 2006 16:33:59 -0800
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <4429A81C.9010805@noaa.gov>
References: <4429A81C.9010805@noaa.gov>
Message-ID: <442B2777.9080904@noaa.gov>

Hi all,

I've been working on a new Wiki page for info about packages on the new 
Universal build:

http://pythonmac.org/wiki/UniversalPackages

Something is up with the formatting:

I have both:

=== A heading ===

= A less important heading =

But they both look the same.

Am I doing something wrong?

-CHB





-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Louis.Wicker at noaa.gov  Tue Mar 28 20:35:18 2006
From: Louis.Wicker at noaa.gov (Louis Wicker)
Date: Tue, 28 Mar 2006 12:35:18 -0600
Subject: [Pythonmac-SIG] F2PY, F90, and Mac os x Tiger: linking problem
Message-ID: <6E4372C0-7B9D-43B5-8388-8E774BF1A0A3@noaa.gov>

All:

last year (twice) I posted a question on several groups looking for  
an answer about an error with trying linking F2PY with XLF 8.1 on Mac  
OS X Tiger.  I have been busy with other developments, but
I thought I would post that it is now working (after I finally gave a  
couple things a try).

1.  Thanks to Bob Ippolito for the suggestion about the Xcode not  
being in sync with the Tiger install.  That solved the __mh_header  
problem.

2.  I got the hello.f example work with the following command line  
with XLF:

f2py --fcompiler=ibm -c -m hello hello.f -L/opt/ibmcmp/xlf/8.1/lib - 
lxlf90 -lxlopt -lxl -lxlfmath

I was getting an error message with out directly linking in the xlf  
libs when I tried to import the shared object file hello.so,

ImportError: Failure linking new module: hello.so: Symbol not found:  
__xlfBeginIO

pretty sure not all those libs are needed - but anyway.

regards,

Lou

------------------------------------------------------------------------ 
----
|  Dr. Louis J. Wicker
|  Research Scientist, National Severe Storms Lab
| 1313 Halley Cir, Norman, OK 73069
| E-mail:   Louis.Wicker at noaa.gov
| HTTP:  www.nssl.noaa.gov/~lwicker
| Phone:    (405) 366-0416
| Fax:       (405) 366-0472
|
|
| Two wrongs don't make a right, but three left's do...
------------------------------------------------------------------------ 
----
|
| "The contents  of this message are mine personally and
| do not reflect any position of  the Government or NOAA."
|
------------------------------------------------------------------------ 
----



From delza at livingcode.org  Thu Mar 30 18:54:58 2006
From: delza at livingcode.org (Dethe Elza)
Date: Thu, 30 Mar 2006 08:54:58 -0800
Subject: [Pythonmac-SIG] NSUncaughtRuntimeErrorException and
	NSArrayController
In-Reply-To: <074D71FD-BF0D-438C-A9F8-7E0AB96951A5@redivi.com>
References: <e7a9b3d30603291528s229d6e50h15c03c701abdb68f@mail.gmail.com>
	<074D71FD-BF0D-438C-A9F8-7E0AB96951A5@redivi.com>
Message-ID: <24d517dd0603300854u5dff7c55w4cba5c136a09760d@mail.gmail.com>

> Use NSZombieEnabled.

And I, for one, welcome our new Zombie Overlords.  I'm proud to use a
platform, a language, and a library which are Zombie-ready. My only
worry is that in today's world, the zombies may not be able to find
enough brains to go around, or worse, will only find brains among the
python community.

%-)

--Dethe

From mantha at chem.unr.edu  Thu Mar 30 19:59:34 2006
From: mantha at chem.unr.edu (Jordan Mantha)
Date: Thu, 30 Mar 2006 09:59:34 -0800
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <442B2777.9080904@noaa.gov>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
Message-ID: <442C1C86.1040401@chem.unr.edu>

Christopher Barker wrote:
> Hi all,
> 
> I've been working on a new Wiki page for info about packages on the new 
> Universal build:
> 
> http://pythonmac.org/wiki/UniversalPackages
> 
> Something is up with the formatting:
> 
> I have both:
> 
> === A heading ===
> 
> = A less important heading =
> 
> But they both look the same.
> 
> Am I doing something wrong?
> 

I think you want is:

= A Heading =

== A less important heading ==

BTW, I'm pretty excited about that wiki page. I recently got my first
Mac (17" Intel iMac) and have had a lot of difficulties with the apps on
that wiki page (I a few data analysis scripts using SciPy). It seemed
like anything using Fortran was doomed to failure. I finally got Numpy
and Scipy (from SVN) installed using gfortran (gfortran-intel-bin.tar.gz
from http://hpc.sourceforge.net/) and the  Universal MacPython 2.4.3c1
build from Ronald Oussoren. I followed the directions at
http://www.scipy.org/Installing_SciPy/Mac_OS_X . I can update the wiki
page when I get a chance.

-Jordan Mantha

From Chris.Barker at noaa.gov  Thu Mar 30 22:17:41 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 30 Mar 2006 12:17:41 -0800
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <442C1C86.1040401@chem.unr.edu>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu>
Message-ID: <442C3CE5.80202@noaa.gov>



Jordan Mantha wrote:
> I think you want is:
> 
> = A Heading =
> 
> == A less important heading ==

Thanks. Did someone go fix that?

> have had a lot of difficulties with the apps on
> that wiki page (I a few data analysis scripts using SciPy). 

SciPy is a challenge!

>  I finally got Numpy
> and Scipy (from SVN) installed using gfortran (gfortran-intel-bin.tar.gz
> from http://hpc.sourceforge.net/) and the  Universal MacPython 2.4.3c1
> build from Ronald Oussoren. I followed the directions at
> http://www.scipy.org/Installing_SciPy/Mac_OS_X . I can update the wiki
> page when I get a chance.

Great, please do so. It does look like it's really going to be a trick 
to get a Universal build of SciPy but maybe separate Intel and PPC 
builds wouldn't be too bad.

As soon as Bob gets bdist_mpkg working with the Universal build, we can 
start making packages for all to use.

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From ronaldoussoren at mac.com  Thu Mar 30 22:23:54 2006
From: ronaldoussoren at mac.com (Ronald Oussoren)
Date: Thu, 30 Mar 2006 22:23:54 +0200
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <442C3CE5.80202@noaa.gov>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu> <442C3CE5.80202@noaa.gov>
Message-ID: <0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>


On 30-mrt-2006, at 22:17, Christopher Barker wrote:

>
>
> Jordan Mantha wrote:
>> I think you want is:
>>
>> = A Heading =
>>
>> == A less important heading ==
>
> Thanks. Did someone go fix that?
>
>> have had a lot of difficulties with the apps on
>> that wiki page (I a few data analysis scripts using SciPy).
>
> SciPy is a challenge!
>
>>  I finally got Numpy
>> and Scipy (from SVN) installed using gfortran (gfortran-intel- 
>> bin.tar.gz
>> from http://hpc.sourceforge.net/) and the  Universal MacPython  
>> 2.4.3c1
>> build from Ronald Oussoren. I followed the directions at
>> http://www.scipy.org/Installing_SciPy/Mac_OS_X . I can update the  
>> wiki
>> page when I get a chance.
>
> Great, please do so. It does look like it's really going to be a trick
> to get a Universal build of SciPy but maybe separate Intel and PPC
> builds wouldn't be too bad.

I'd prefer not to have seperate Intel and PPC binaries. Would merging  
seperate
Intel and PPC builds be of any use?

Ronald
>
> As soon as Bob gets bdist_mpkg working with the Universal build, we  
> can
> start making packages for all to use.
>
> -Chris
>
>
> -- 
> Christopher Barker, Ph.D.
> Oceanographer
>                                      		
> NOAA/OR&R/HAZMAT         (206) 526-6959   voice
> 7600 Sand Point Way NE   (206) 526-6329   fax
> Seattle, WA  98115       (206) 526-6317   main reception
>
> Chris.Barker at noaa.gov
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig


From Chris.Barker at noaa.gov  Fri Mar 31 00:38:24 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 30 Mar 2006 14:38:24 -0800
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu> <442C3CE5.80202@noaa.gov>
	<0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>
Message-ID: <442C5DE0.2070204@noaa.gov>

Ronald Oussoren wrote:
> I'd prefer not to have seperate Intel and PPC binaries. Would merging 
> seperate
> Intel and PPC builds be of any use?

Yes, though I'm not sure what you mean by merging.

Do you mean lipo'ing them together? or having separate builds in one *.mpkg?

Ideally, we could get it all truly Universal, then one could build 
Universal Py2App  bundles with it and all.

I think the big trick with SciPy is the Fortran components. SciPy, 
Fortran and gcc 4 have been having trouble for a while. If that's fixed, 
I have no idea about how to build Universal binaries with gfortran or 
g95. Does anyone know?


While I'm on the topic, I'm looking into building Universal binaries of 
some other packages that depend on other libs. For instance, matplotlib 
depends on libpng and libfreetype. In the past, I've built static libs 
of those, and then linked MPL against them, getting a nice 
self-contained package. However, I have no idea how I would go about 
building Universal static libs, or two sets of libs, or????

Any suggestions?

-Chris





-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From charles.hartman at conncoll.edu  Fri Mar 31 00:54:20 2006
From: charles.hartman at conncoll.edu (Charles Hartman)
Date: Thu, 30 Mar 2006 17:54:20 -0500
Subject: [Pythonmac-SIG] Wiki Formatting
In-Reply-To: <442C5DE0.2070204@noaa.gov>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu> <442C3CE5.80202@noaa.gov>
	<0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>
	<442C5DE0.2070204@noaa.gov>
Message-ID: <D7C5E791-F458-4467-81EA-F5788EF1FB8A@conncoll.edu>

Chris,

Are you still planning to work on wxPython? That would be great -- I  
can't really use the Universal Python till that works too. I'd be  
glad to help, and I'm not lazy -- but I sure am ignorant, so I don't  
think I have much to contribute to this.

Charles Hartman



On Mar 30, 2006, at 5:38 PM, Christopher Barker wrote:

> While I'm on the topic, I'm looking into building Universal  
> binaries of
> some other packages that depend on other libs. For instance,  
> matplotlib
> depends on libpng and libfreetype. In the past, I've built static libs
> of those, and then linked MPL against them, getting a nice
> self-contained package. However, I have no idea how I would go about
> building Universal static libs, or two sets of libs, or????
>
> Any suggestions?
>
> -Chris

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060330/39d46d6e/attachment.htm 

From dgrassi at lexar.com  Fri Mar 31 01:01:29 2006
From: dgrassi at lexar.com (Dan Grassi)
Date: Thu, 30 Mar 2006 18:01:29 -0500
Subject: [Pythonmac-SIG] PyObjC
In-Reply-To: <13FA3F95-E1C6-4939-A0CD-2C047E95C648@usno.navy.mil>
References: <2067FFEC-10B3-4AF5-9E35-060A5ABC4624@lexar.com>
	<13FA3F95-E1C6-4939-A0CD-2C047E95C648@usno.navy.mil>
Message-ID: <85077387-5B03-4F83-B647-A323BD5D2A98@lexar.com>

Hi,

I had embedded python under 10.3.x but under 10.4 it fails on import
from Foundation import *
It seems that the problem is that it is launching the wrong python.

sys.path: ['/System/Library/Frameworks/Python.framework/Versions/2.3/ 
lib/python23.zip'...

I have installed
MacPython-OSX-2.4.1-1.dmg
and
TigerPython24Fix-r2.mpkg

I am including:
/Library/Frameworks/Python.framework/
which has:
/Library/Frameworks/Python.framework/Versions/2.4/lib/python2.4/site- 
packages/PyObjC/

What am I missing that I need to setup?

TIA,

Dan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2600 bytes
Desc: not available
Url : http://mail.python.org/pipermail/pythonmac-sig/attachments/20060330/38e9b26b/attachment.bin 

From kevino at theolliviers.com  Fri Mar 31 01:24:22 2006
From: kevino at theolliviers.com (Kevin Ollivier)
Date: Thu, 30 Mar 2006 15:24:22 -0800
Subject: [Pythonmac-SIG] wxPython universal (was Re:  Wiki Formatting)
In-Reply-To: <D7C5E791-F458-4467-81EA-F5788EF1FB8A@conncoll.edu>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu> <442C3CE5.80202@noaa.gov>
	<0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>
	<442C5DE0.2070204@noaa.gov>
	<D7C5E791-F458-4467-81EA-F5788EF1FB8A@conncoll.edu>
Message-ID: <036B848B-2024-48BA-80D3-AA4670B85486@theolliviers.com>

Hi Charles,

On Mar 30, 2006, at 2:54 PM, Charles Hartman wrote:

> Chris,
>
> Are you still planning to work on wxPython? That would be great --  
> I can't really use the Universal Python till that works too. I'd be  
> glad to help, and I'm not lazy -- but I sure am ignorant, so I  
> don't think I have much to contribute to this.

I've been able to build wxPython as Universal using the Universal  
Python build from about a week ago, and it works on Tiger, but on  
Panther certain things that are supposed to be initialized at load  
time aren't being initialized. (e.g. wxArtProvider) Weirder yet, when  
I create a C++ app against the exact same DLLs, the initialization  
does happen and everything works fine, so this seems to have  
something to do with the wrappers, either how things are being linked  
in or perhaps some difference in initialization from a straight PPC  
build....

I'll try and take some time out in the not too distant future to  
track this down, and in the meantime some folks are working to get  
the lipo'd binary solution to work for wxWidgets, which could  
possibly resolve the issue. (Though considering the C++ builds work  
fine on Panther, I'm beginning to really suspect there's something  
going on with the wrappers....)

In the meantime, if someone has some debugging tips as to where best  
to start poking around to track this down, it would be greatly  
appreciated. :-)

Thanks,

Kevin

> Charles Hartman
>
>
>
> On Mar 30, 2006, at 5:38 PM, Christopher Barker wrote:
>
>> While I'm on the topic, I'm looking into building Universal  
>> binaries of
>> some other packages that depend on other libs. For instance,  
>> matplotlib
>> depends on libpng and libfreetype. In the past, I've built static  
>> libs
>> of those, and then linked MPL against them, getting a nice
>> self-contained package. However, I have no idea how I would go about
>> building Universal static libs, or two sets of libs, or????
>>
>> Any suggestions?
>>
>> -Chris
>
> _______________________________________________
> Pythonmac-SIG maillist  -  Pythonmac-SIG at python.org
> http://mail.python.org/mailman/listinfo/pythonmac-sig

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060330/1ebdfb6f/attachment-0001.htm 

From Chris.Barker at noaa.gov  Fri Mar 31 03:14:56 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 30 Mar 2006 17:14:56 -0800
Subject: [Pythonmac-SIG] Has anyone built PIL with the Universal Python?
Message-ID: <442C8290.4000205@noaa.gov>

Hi all,

I just sat down to build PIL for the Universal Python, and I had a thought:

1) it needs a couple extra libs, in particular:

libjpeg
libfreetype

I have, in the past, compiled that kind of thing as a static lib, and 
linked it to the package. However, as we may end up needing the same 
libs in various packages, maybe we should make a "libs package", that 
might even install into the Python tree, and they can all use that. 
libfreetype is needed for both matplotlib and PIL, for instance.

Thoughts?

-Chris





-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From Chris.Barker at noaa.gov  Fri Mar 31 03:16:24 2006
From: Chris.Barker at noaa.gov (Christopher Barker)
Date: Thu, 30 Mar 2006 17:16:24 -0800
Subject: [Pythonmac-SIG] wxPython universal (was Re:  Wiki Formatting)
In-Reply-To: <036B848B-2024-48BA-80D3-AA4670B85486@theolliviers.com>
References: <4429A81C.9010805@noaa.gov> <442B2777.9080904@noaa.gov>
	<442C1C86.1040401@chem.unr.edu> <442C3CE5.80202@noaa.gov>
	<0C75392C-17CA-422F-9378-FCCF42AF6A15@mac.com>
	<442C5DE0.2070204@noaa.gov>
	<D7C5E791-F458-4467-81EA-F5788EF1FB8A@conncoll.edu>
	<036B848B-2024-48BA-80D3-AA4670B85486@theolliviers.com>
Message-ID: <442C82E8.8000004@noaa.gov>

Kevin Ollivier wrote:

> I've been able to build wxPython as Universal using the Universal Python 
> build from about a week ago, and it works on Tiger, but on Panther 
> certain things that are supposed to be initialized at load time aren't 
> being initialized.

I, for one, think it would be a good start to get a Tiger only one out 
there. Did you need to do anything special to get that working?

-Chris


-- 
Christopher Barker, Ph.D.
Oceanographer
                                     		
NOAA/OR&R/HAZMAT         (206) 526-6959   voice
7600 Sand Point Way NE   (206) 526-6329   fax
Seattle, WA  98115       (206) 526-6317   main reception

Chris.Barker at noaa.gov

From wescpy at gmail.com  Fri Mar 31 03:30:11 2006
From: wescpy at gmail.com (w chun)
Date: Thu, 30 Mar 2006 17:30:11 -0800
Subject: [Pythonmac-SIG] Problem w/Tkinter on 2.5 (Panther)
Message-ID: <78b3a9580603301730p37691ea1x90b7d2ce7c34e90a@mail.gmail.com>

Russell suggested that I xPost to you folks on the MacSIG list...

http://mail.python.org/pipermail/python-list/2006-March/333614.html

thanks in advance!
-- wesley
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
"Core Python Programming", Prentice Hall, (c)2007,2001
    http://corepython.com

wesley.j.chun :: wescpy-at-gmail.com
cyberweb.consulting : silicon valley, ca
http://cyberwebconsulting.com

From jaonary at free.fr  Fri Mar 31 18:35:12 2006
From: jaonary at free.fr (Jaonary Rabarisoa)
Date: Fri, 31 Mar 2006 18:35:12 +0200
Subject: [Pythonmac-SIG] Numpy Scipy and Darwinports
Message-ID: <604B6946-5478-40FB-895E-251A83D437B2@free.fr>

Hi all,
I'd like to build numpy and scipy with the gcc 4.1 provided by  
darwinport (I fact, they need a fortran compiler). How can I do to  
change the compiler to be used with distutils ? I've tried something  
like this :

python setup.py build --compiler=gcc-dp-4.1 ( all the file installed  
with darwinports have the suffix -dp-4.1)

but I didn't work.

Thank you for your advice.

Best Regards,

Jaonary

From cookedm at physics.mcmaster.ca  Fri Mar 31 22:55:03 2006
From: cookedm at physics.mcmaster.ca (David M. Cooke)
Date: Fri, 31 Mar 2006 15:55:03 -0500
Subject: [Pythonmac-SIG] Numpy Scipy and Darwinports
In-Reply-To: <604B6946-5478-40FB-895E-251A83D437B2@free.fr> (Jaonary
	Rabarisoa's message of "Fri, 31 Mar 2006 18:35:12 +0200")
References: <604B6946-5478-40FB-895E-251A83D437B2@free.fr>
Message-ID: <qnklkuq5o7s.fsf@arbutus.physics.mcmaster.ca>

Jaonary Rabarisoa <jaonary at free.fr> writes:

> Hi all,
> I'd like to build numpy and scipy with the gcc 4.1 provided by  
> darwinport (I fact, they need a fortran compiler). How can I do to  
> change the compiler to be used with distutils ? I've tried something  
> like this :
>
> python setup.py build --compiler=gcc-dp-4.1 ( all the file installed  
> with darwinports have the suffix -dp-4.1)
>
> but I didn't work.

Use the CC environment variable:

CC=gcc-dp-4.1 python setup.py build

However, note that gfortran will *not* work with scipy: we don't know
why, but we believe to be something to do with gfortran not being
stable yet. I tested it recently; still no go.

You'll need g77; you're probably better off using the gcc 3.4 from
darwinports. I don't have my laptop on me right now, but I believe
that's what I used. Or, use Apple's 3.3 and g77 from
http://hpc.sourceforge.net/; this works fine. Apple's 4.0 will not work.

-- 
|>|\/|<
/--------------------------------------------------------------------------\
|David M. Cooke                      http://arbutus.physics.mcmaster.ca/dmc/
|cookedm at physics.mcmaster.ca

From artificiallystupid at yahoo.com  Fri Mar 31 22:49:52 2006
From: artificiallystupid at yahoo.com (Johnston Jiaa)
Date: Fri, 31 Mar 2006 12:49:52 -0800 (PST)
Subject: [Pythonmac-SIG] Apple Remote "Mouse"
Message-ID: <20060331204952.54614.qmail@web50407.mail.yahoo.com>

I recently bought a Macbook Pro from Apple.  As it comes with a remote, I thought it would be great to use it as a mouse when not in Front Row.  The fast forward button would move the cursor to the left, the volume increase would move it up the screen, etc and the play button would serve as a "click."

Is there any way to manipulate the cursor position on the screen using Python?  It is greatly appreciated if someone can point in the direction of where to look or what to search as no immediately useful hits came up from google.

I am absolutely clueless about how to implement any of this.  It would be awesome if someone could briefly explain, or maybe point me to a tutorial of any sort, how to write a program that could detect the button pressed on the remote and accordingly move the cursor up, down, side-to-side.  Knowing how to disable the volume control of the "up and down" function of the anticipated remote would also be great, if anyone has time to elaborate how this can be done.

Thank you for reading my message, and I hope you do not regard me as being a waste of your time.



Johnston Jiaa (artificiallystupid at yahoo.com)

		
---------------------------------
Blab-away for as little as 1?/min. Make  PC-to-Phone Calls using Yahoo! Messenger with Voice.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.python.org/pipermail/pythonmac-sig/attachments/20060331/b7942d1c/attachment.htm 

From robert.kern at gmail.com  Fri Mar 31 23:06:33 2006
From: robert.kern at gmail.com (Robert Kern)
Date: Fri, 31 Mar 2006 15:06:33 -0600
Subject: [Pythonmac-SIG] Numpy Scipy and Darwinports
In-Reply-To: <qnklkuq5o7s.fsf@arbutus.physics.mcmaster.ca>
References: <604B6946-5478-40FB-895E-251A83D437B2@free.fr>
	<qnklkuq5o7s.fsf@arbutus.physics.mcmaster.ca>
Message-ID: <442D99D9.4060301@gmail.com>

David M. Cooke wrote:
> Jaonary Rabarisoa <jaonary at free.fr> writes:
> 
> 
>>Hi all,
>>I'd like to build numpy and scipy with the gcc 4.1 provided by  
>>darwinport (I fact, they need a fortran compiler). How can I do to  
>>change the compiler to be used with distutils ? I've tried something  
>>like this :
>>
>>python setup.py build --compiler=gcc-dp-4.1 ( all the file installed  
>>with darwinports have the suffix -dp-4.1)
>>
>>but I didn't work.
> 
> Use the CC environment variable:
> 
> CC=gcc-dp-4.1 python setup.py build
> 
> However, note that gfortran will *not* work with scipy: we don't know
> why, but we believe to be something to do with gfortran not being
> stable yet. I tested it recently; still no go.

IIRC, someone found that gfortran 4.1 will work if d1mach.f is compiled with
only -O, not -O2. Possibly -ffloat-store, too, I don't remember.

In any case, numpy.distutils doesn't let you set the name of the fortran
compiler through an environment variable, I don't think. I recommend making a
symlink named "gfortran" to the actual executable. Then:

CC=gcc-dp-4.1 python setup.py config_fc --fcompiler=gnu95 build_src build_clib
build_ext build

-- 
Robert Kern
robert.kern at gmail.com

"I have come to believe that the whole world is an enigma, a harmless enigma
 that is made terrible by our own mad attempt to interpret it as though it had
 an underlying truth."
  -- Umberto Eco