From lac at openend.se  Sun Mar  4 10:57:20 2007
From: lac at openend.se (Laura Creighton)
Date: Sun, 04 Mar 2007 10:57:20 +0100
Subject: [py-dev] In case you are not reading python-announce
Message-ID: <200703040957.l249vKcN004066@theraft.openend.se>


------- Forwarded Message

Return-Path: python-announce-list-bounces at python.org
Delivery-Date: Sun Mar  4 08:11:34 2007
From: Titus Brown <titus at caltech.edu>
To: python-announce-list at python.org
Subject: announcing: "testing-in-python" mailing list
Message-ID: <20070304021155.GB8128 at caltech.edu>

Folks,

catalyzed by the great fun we had at PyCon '07, Grig Gheorghiu and I
have created the "testing-in-python" (or "TIP") mailing list.

This list will hopefully serve as a forum for discussing Python testing
tools, testing approaches useful in Python, Web resources for same, and
whatever else people would like to talk about.  ("Proposed: The Death
Star would not have had a single point of failure had it been written in
Python and tested with the Death Star Testing Protocol Tool.  Discuss.")

Grig has a blog post about the list here:

   http://agiletesting.blogspot.com/2007/02/testing-in-python-mailing-list.html

You can sign up for the [mailman] list here:

   http://lists.idyll.org/listinfo/testing-in-python

Subscribers can post to the list at 'tip at lists.idyll.org'.

List archives are available here:

	http://lists.idyll.org/pipermail/testing-in-python/

cheers,
- --titus
- -- 
------- End of Forwarded Message



From holger at merlinux.de  Wed Mar  7 08:25:03 2007
From: holger at merlinux.de (holger krekel)
Date: Wed, 7 Mar 2007 08:25:03 +0100
Subject: [py-dev] [pypy-dev] py lib 0.9.0: py.test, distributed execution,
	microthreads ...
In-Reply-To: <erjs0f$963$1@sea.gmane.org>
References: <20070214165116.GI16146@solar.trillke> <erjs0f$963$1@sea.gmane.org>
Message-ID: <20070307072503.GD1954@solar.trillke>

Hi Martijn! (hi CCed py-dev),

hope you are doing well!  

I am considering doing a py lib 0.9.1 with a particular focus
on packaging and win32 issues.  Could you maybe provide some pointers,
keypoints on what you think would be nice to have in this
area?  I haven't really followed the packaging mechanisms in
the last year and need to catch up, i guess.  Any help is
appreciated. 

best, 

holger

On Thu, Feb 22, 2007 at 11:41 +0100, Martijn Faassen wrote:
> holger krekel wrote:
> > py lib 0.9.0: py.test, distributed execution, greenlets and more
> > ======================================================================
> > 
> > Welcome to the 0.9.0 py lib release - a library aiming to 
> > support agile and test-driven python development on various levels. 
> > 
> > Main API/Tool Features: 
> > 
> > * py.test: cross-project testing tool with many advanced features
> > * py.execnet: ad-hoc code distribution to SSH, Socket and local sub processes
> > * py.magic.greenlet: micro-threads on standard CPython ("stackless-light")
> > * py.path: path abstractions over local and subversion files 
> > * rich documentation of py's exported API 
> > * tested against Linux, OSX and partly against Win32, python 2.3-2.5
> > 
> > All these features and their API have extensive documentation,
> > generated with the new "apigen", which we intend to make accessible
> > for other python projects as well.  
> > 
> > Download/Install:   http://codespeak.net/py/0.9.0/download.html
> 
> Cool, a pylib release!
> 
> Is a cheeseshop entry planned? I checked the download page but there's 
> no mention of it. Perhaps there's a cheeseshop entry that I didn't see? 
> People might want to do 'easy_install pylib'. Having a cheeseshop entry 
> also makes it more easy to use pylib in any buildouts.
> 
> Regards,
> 
> Martijn
> 
> _______________________________________________
> pypy-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/pypy-dev
> 

-- 
merlinux GmbH       Steinbergstr. 42    31139 Hildesheim   
http://merlinux.de  tel +49 5121 20800 75 (fax 77) 


From brian at dorseys.org  Wed Mar  7 23:32:38 2007
From: brian at dorseys.org (Brian Dorsey)
Date: Wed, 7 Mar 2007 14:32:38 -0800
Subject: [py-dev] From python-dev: Pep-364,
	Transitioning to the Py3K standard library
Message-ID: <66e877b70703071432ha95ec34xb480c82e32d9f49@mail.gmail.com>

Hi py-dev folks!

I just noticed that Barry Warsaw and Brett Cannon created a PEP for a
mechanism for allowing modules to be imported under various aliases to
enable the planned standard library re-organization. They recently
announced it with a thread on python-dev.

http://www.python.org/dev/peps/pep-0364/

I know that various py-lib folks have given these topics quite a bit
of thought - at least at the package namespace level. (I'm thinking of
initpkg.py here) So, I thought there might be some interesting
feedback about this PEP from the experience with initpkg.py.


Take care,
  -Brian


From holger at merlinux.de  Thu Mar  8 00:03:34 2007
From: holger at merlinux.de (holger krekel)
Date: Thu, 8 Mar 2007 00:03:34 +0100
Subject: [py-dev] From python-dev: Pep-364,
	Transitioning to the Py3K standard library
In-Reply-To: <66e877b70703071432ha95ec34xb480c82e32d9f49@mail.gmail.com>
References: <66e877b70703071432ha95ec34xb480c82e32d9f49@mail.gmail.com>
Message-ID: <20070307230334.GQ1954@solar.trillke>

Hi Brian, 

On Wed, Mar 07, 2007 at 14:32 -0800, Brian Dorsey wrote:
> I just noticed that Barry Warsaw and Brett Cannon created a PEP for a
> mechanism for allowing modules to be imported under various aliases to
> enable the planned standard library re-organization. They recently
> announced it with a thread on python-dev.
> 
> http://www.python.org/dev/peps/pep-0364/
> 
> I know that various py-lib folks have given these topics quite a bit
> of thought - at least at the package namespace level. (I'm thinking of
> initpkg.py here) So, I thought there might be some interesting
> feedback about this PEP from the experience with initpkg.py.

thanks for the pointer.  Upon first read, i am a bit concerned
that it may complicate the importing mechanism further. 
Also, i wonder if the mechanisms and API introduced with 
PEP-364 maybe are over-general for the purpose of providing 
a new standard namespace.  

BTW, as it happens i was wondering this morning if
the py lib shouldn't provide some helpers for e.g.
StringIO and pickle-methods, maybe in its py.io namespace,
so you could do: 

    out = py.io.StringIO()
    py.io.pickle.dump(out, obj)
    ...
    
without importing anything else than the usual "import py" ahead. 
the "py.io.pickle" and StringIO would preferably be the c-versions
with a fallback to the python one. 

holger


From faassen at startifact.com  Thu Mar  8 17:05:23 2007
From: faassen at startifact.com (Martijn Faassen)
Date: Thu, 8 Mar 2007 17:05:23 +0100
Subject: [py-dev] [pypy-dev] py lib 0.9.0: py.test, distributed execution,
	microthreads ...
In-Reply-To: <20070307072503.GD1954@solar.trillke>
References: <20070214165116.GI16146@solar.trillke> <erjs0f$963$1@sea.gmane.org>
	<20070307072503.GD1954@solar.trillke>
Message-ID: <8928d4e90703080805re73ca68ycafda52ceabf4edd@mail.gmail.com>

Hey Holger,

On 3/7/07, holger krekel <holger at merlinux.de> wrote:
> Hi Martijn! (hi CCed py-dev),

> hope you are doing well!

Yes, still in Singapore actually but coming back this weekend.

> I am considering doing a py lib 0.9.1 with a particular focus
> on packaging and win32 issues.  Could you maybe provide some pointers,
> keypoints on what you think would be nice to have in this
> area?  I haven't really followed the packaging mechanisms in
> the last year and need to catch up, i guess.  Any help is
> appreciated.

I can't really tell you much about win32 issues. Concerning packaging
- it's a fairly huge issue that I only have partial experience with
too, but I'll attempt to give a few pointers based on my experience:

* The minimal thing you'd want is an entry in the cheeseshop which
points to the .tgz of the source somewhere on your site. If your
setup.py is fairly straightforward you may be in luck; tools like
easy_install (and buildout tools like zc.buildout, which can install
packages locally instead of centrally) might very well be able to
download and install the software automatically. I've even had this
happen with lxml, which means a C-compile happens successfully.

* Next, try some command like 'python setup.py sdist upload' to
automatically upload tarballs to the cheeseshop itself.

* the next step would be to look at setuptools, which is an extension
of distutils. This stuff:

http://peak.telecommunity.com/DevCenter/setuptools

It is a lot of stuff and therefore somewhat intimidating. I believe
the minimal way to start using setuptools is to replace your import in
setup.py with setuptools instead of distutils.

* once you use setuptools, you can use it to create an egg, which is
just another way to package Python code for distribution. If it's just
Python code, that's fairly simple as you get a platform independent
egg. If it's C code you'll need to do some more work in the setup.py
and eggs will be platform specific, but I believe setuptools automates
quite a bit. You can use things like 'python setup.py bdist_egg
upload' to upload the egg.

Important in all this is to keep your setup.py relatively
straightforward. Not always possible, but I've seen some setup.py's
that do all kinds of work themselves that breaks  a lot of all this,
and that's a shame. egg downloading and such starts to get nice once
you have automatic dependency management and the ability to install
eggs as a user, using tools like zc.buildout (if you need lots of
cooperating bits and pieces including things like databases and
servers) or workingenv.py (lightweight clever hack). Both
install eggs in some other place than
/usr/lib/python2.4/site-packages, which is the default and rather
ugly.

I hope these pointers help you a bit to get you on your way.

Regards,

Martijn


From holger at merlinux.de  Thu Mar  8 17:14:23 2007
From: holger at merlinux.de (holger krekel)
Date: Thu, 8 Mar 2007 17:14:23 +0100
Subject: [py-dev] [pypy-dev] py lib 0.9.0: py.test, distributed execution,
	microthreads ...
In-Reply-To: <8928d4e90703080805re73ca68ycafda52ceabf4edd@mail.gmail.com>
References: <20070214165116.GI16146@solar.trillke> <erjs0f$963$1@sea.gmane.org>
	<20070307072503.GD1954@solar.trillke>
	<8928d4e90703080805re73ca68ycafda52ceabf4edd@mail.gmail.com>
Message-ID: <20070308161423.GW1954@solar.trillke>

Hey Martijn!

thanks a lot for the pointers, they help - i think
we should go for generating a plain no-tricks setup.py file. 
This way one can be rather specific of what should be 
included and such, i guess.  Regarding win32: 
py.test can run tests remotely, i.e. from a linux
box on a windows box with full feedback.  If
we have a permanent win32 host somewhere that can compile 
C via distutils (i know such things exist, it's just hard 
to get hold of such a configuration, grrr).  And then
finally have all-platform automated ad-hoc testing,
maybe even with PDB interactive prompts etc :)

best & have a nice flight back,

holger

On Thu, Mar 08, 2007 at 17:05 +0100, Martijn Faassen wrote:
> Hey Holger,
> 
> On 3/7/07, holger krekel <holger at merlinux.de> wrote:
> > Hi Martijn! (hi CCed py-dev),
> 
> > hope you are doing well!
> 
> Yes, still in Singapore actually but coming back this weekend.
> 
> > I am considering doing a py lib 0.9.1 with a particular focus
> > on packaging and win32 issues.  Could you maybe provide some pointers,
> > keypoints on what you think would be nice to have in this
> > area?  I haven't really followed the packaging mechanisms in
> > the last year and need to catch up, i guess.  Any help is
> > appreciated.
> 
> I can't really tell you much about win32 issues. Concerning packaging
> - it's a fairly huge issue that I only have partial experience with
> too, but I'll attempt to give a few pointers based on my experience:
> 
> * The minimal thing you'd want is an entry in the cheeseshop which
> points to the .tgz of the source somewhere on your site. If your
> setup.py is fairly straightforward you may be in luck; tools like
> easy_install (and buildout tools like zc.buildout, which can install
> packages locally instead of centrally) might very well be able to
> download and install the software automatically. I've even had this
> happen with lxml, which means a C-compile happens successfully.
> 
> * Next, try some command like 'python setup.py sdist upload' to
> automatically upload tarballs to the cheeseshop itself.
> 
> * the next step would be to look at setuptools, which is an extension
> of distutils. This stuff:
> 
> http://peak.telecommunity.com/DevCenter/setuptools
> 
> It is a lot of stuff and therefore somewhat intimidating. I believe
> the minimal way to start using setuptools is to replace your import in
> setup.py with setuptools instead of distutils.
> 
> * once you use setuptools, you can use it to create an egg, which is
> just another way to package Python code for distribution. If it's just
> Python code, that's fairly simple as you get a platform independent
> egg. If it's C code you'll need to do some more work in the setup.py
> and eggs will be platform specific, but I believe setuptools automates
> quite a bit. You can use things like 'python setup.py bdist_egg
> upload' to upload the egg.
> 
> Important in all this is to keep your setup.py relatively
> straightforward. Not always possible, but I've seen some setup.py's
> that do all kinds of work themselves that breaks  a lot of all this,
> and that's a shame. egg downloading and such starts to get nice once
> you have automatic dependency management and the ability to install
> eggs as a user, using tools like zc.buildout (if you need lots of
> cooperating bits and pieces including things like databases and
> servers) or workingenv.py (lightweight clever hack). Both
> install eggs in some other place than
> /usr/lib/python2.4/site-packages, which is the default and rather
> ugly.
> 
> I hope these pointers help you a bit to get you on your way.
> 
> Regards,
> 
> Martijn
> _______________________________________________
> py-dev mailing list
> py-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev
> 

-- 
merlinux GmbH       Steinbergstr. 42    31139 Hildesheim   
http://merlinux.de  tel +49 5121 20800 75 (fax 77) 


From jcea at argo.es  Sat Mar 10 17:53:17 2007
From: jcea at argo.es (Jesus Cea)
Date: Sat, 10 Mar 2007 17:53:17 +0100
Subject: [py-dev] Why: "RuntimeError: maximum recursion depth exceeded"
Message-ID: <45F2E27D.7070600@argo.es>

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

I'm a new but happy "greenlets" user since a couple of hours ago :-),
but I have found a problem with stack limit. Since greenlet's "switch()"
stores the stack aside, I don't understand the problem.

I've tried Google, without success.

The code:

"""
>>> from py.magic import greenlet
>>> def a() :
...   i=gr.pop()
...   i.switch()
...
>>> gr=[greenlet.getcurrent()]
>>> gr+=[greenlet(a) for i in xrange(100)]
>>> greenlet(a).switch()
()
>>> gr=[greenlet.getcurrent()]
>>> gr+=[greenlet(a) for i in xrange(1000)]
>>> greenlet(a).switch()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
RuntimeError: maximum recursion depth exceeded
"""

I don't understand why I'm getting this error. I'm sure I'm missing
something...

Thanks in advance.

- --
Jesus Cea Avion                         _/_/      _/_/_/        _/_/_/
jcea at argo.es http://www.argo.es/~jcea/ _/_/    _/_/  _/_/    _/_/  _/_/
jabber / xmpp:jcea at jabber.org         _/_/    _/_/          _/_/_/_/_/
                               _/_/  _/_/    _/_/          _/_/  _/_/
"Things are not so easy"      _/_/  _/_/    _/_/  _/_/    _/_/  _/_/
"My name is Dump, Core Dump"   _/_/_/        _/_/_/      _/_/  _/_/
"El amor es poner tu felicidad en la felicidad de otro" - Leibniz
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iQCVAwUBRfLifZlgi5GaxT1NAQLUVgQAiISmf1ZaGrZ+cH6XsQgxlZJywnDhyM8/
JmxtdXVv2i2nTGSVe4pZFFnlrHyEQNCTsbcM4icTjQxyVMbJqkVGC54qXDo8DTWA
3qV5QV2iktaJSNCxNxAKAzv0E4nP2nNdBPNZ1QdzS78WPAsTLtDE82elQ2EuMsBz
rCaJQJL6vaM=
=dKlM
-----END PGP SIGNATURE-----


From jcea at argo.es  Sat Mar 10 20:02:41 2007
From: jcea at argo.es (Jesus Cea)
Date: Sat, 10 Mar 2007 20:02:41 +0100
Subject: [py-dev] Why: "RuntimeError: maximum recursion depth exceeded"
In-Reply-To: <45F2E27D.7070600@argo.es>
References: <45F2E27D.7070600@argo.es>
Message-ID: <45F300D1.3020508@argo.es>

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

I got it!. The issue is all the "parent" links for each greenlet. I need
to change "parent" by hand or use a trampoline to switch greenlets.

What I don't understand now is whey calling directly another greenlet
increase the recursion level, but calling the parent doesn't. More on
it, no cycles are created :-?.

How is the "parent" update logic when switching to an upstack greenlet?.
I can't find this info in the documentation.

Thanks in advance.

Jesus Cea wrote:
> I'm a new but happy "greenlets" user since a couple of hours ago :-),
> but I have found a problem with stack limit. Since greenlet's "switch()"
> stores the stack aside, I don't understand the problem.

- --
Jesus Cea Avion                         _/_/      _/_/_/        _/_/_/
jcea at argo.es http://www.argo.es/~jcea/ _/_/    _/_/  _/_/    _/_/  _/_/
jabber / xmpp:jcea at jabber.org         _/_/    _/_/          _/_/_/_/_/
                               _/_/  _/_/    _/_/          _/_/  _/_/
"Things are not so easy"      _/_/  _/_/    _/_/  _/_/    _/_/  _/_/
"My name is Dump, Core Dump"   _/_/_/        _/_/_/      _/_/  _/_/
"El amor es poner tu felicidad en la felicidad de otro" - Leibniz
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iQCVAwUBRfMA0Jlgi5GaxT1NAQJHwwP/diieEJKPHYCCPaJXjm2RihkyrJNnRv9h
nJRKR42zHqG9TQFBnAzDesSiutZAWuaWKmn3Kh+nyuAMVGChOoOpfCr/W6BNZ5nv
uLaNVbEsm5ThcXtde5b1WllYSzH28uxeu7cOPah8E397+hd4XdLE8tLWdn1tWcQv
Q1adXl2kVHM=
=6LFC
-----END PGP SIGNATURE-----


From collinw at gmail.com  Sun Mar 11 22:04:21 2007
From: collinw at gmail.com (Collin Winter)
Date: Sun, 11 Mar 2007 16:04:21 -0500
Subject: [py-dev] py.test features for a new unittest
Message-ID: <43aa6ff70703111404x293052eer3e2205bc1adee32c@mail.gmail.com>

Hi all,

I'm working on a new design for Python's unittest to (hopefully) be
shipped with Python 3, and I'm interested in incorporating some of
py.test's features. (You can get a good overview of the heart of the
new design from http://oakwinter.com/code/a-new-unittest/.) The new
unittest is focused on ease of extensibility, while still
incorporating enough
user-facing functionality to be useful. To that end, I'm keen to
import py.test's traceback snipping/beautification, concept of
generative tests, print debugging and (possibly) module-level
setup/teardown.

I plan on bring this up on the new testing-in-python list soon, but I
thought I'd ask this more targeted audience first: is there anything
from py.test (or otherwise) that you'd like to see go into a
totally-rewritten unittest design? Is there anything you've tried to
do in terms of extending unittest that the old design made unworkable
or unnecessarily complicated?

Thanks,
Collin Winter


From holger at merlinux.de  Sun Mar 11 22:10:59 2007
From: holger at merlinux.de (holger krekel)
Date: Sun, 11 Mar 2007 22:10:59 +0100
Subject: [py-dev] py.test features for a new unittest
In-Reply-To: <43aa6ff70703111404x293052eer3e2205bc1adee32c@mail.gmail.com>
References: <43aa6ff70703111404x293052eer3e2205bc1adee32c@mail.gmail.com>
Message-ID: <20070311211058.GI4635@solar.trillke>

Hi Collin!

thanks a lot for your mail - it is very nice that you
consider py.test and its ideas for your work.  
Currently, many of us are very busy with PyPy (and py.test :) 
until april, but i'll see to participate and feedback
a bit because i am indeed interested in the topic. 

On Sun, Mar 11, 2007 at 16:04 -0500, Collin Winter wrote:
> I'm working on a new design for Python's unittest to be shipped with
> Python 3, and I'm interested in incorporating some of py.test's
> features. (You can get a good overview of the heart of the new design
> from http://oakwinter.com/code/a-new-unittest/.) The new unittest is
> focused on ease of extensibility, while still incorporating enough
> user-facing functionality to be useful. To that end, I'm keen to
> import py.test's traceback snipping/beautification, concept of
> generative tests, print debugging and (possibly) module-level
> setup/teardown.

sure.  on a sidenote, i consider "generative tests" still somewhat 
of a hack, at least in terms of py.test's own implementation, 
but maybe also conceptually. 
 
> I plan on bring this up on the new testing-in-python list soon, but I
> thought I'd ask you privately first: is there anything (from py.test
> or otherwise) that you'd like to see go into a totally-rewritten
> unittest design? Is there anything you've tried to do in terms of
> extending unittest that the old design made unworkable or
> unnecessarily complicated?

Maybe the main conceptual issue is that unittest.py based
approaches usually mix test code and test configuration. 
For example, configuring different reporters, extra memory
checks or specifying the distribution of tests across multiple
hosts should IMO not require any changes to the test modules
but go to a separate test configuration.  Admittedly, the current
hooks provided by py.test's conftest.py leave a bit to be wished
for, but this conceptual separation is there.  I think it relates
to the question of providing a test tool or a test framework. 

In any case, i am certainly interested to discuss this further,
either here or on an appropriate list. 

best & cheers, 

holger



From brian at dorseys.org  Mon Mar 12 02:07:52 2007
From: brian at dorseys.org (Brian Dorsey)
Date: Sun, 11 Mar 2007 18:07:52 -0700
Subject: [py-dev] Ideas for Google Summer of Code projects?
Message-ID: <66e877b70703111807j3e61e9d6hcee7005df521bc79@mail.gmail.com>

Hi py-lib devs!

In case any students are interested in working on py-lib as a Google
Summer of Code project, I've just offered to be a mentor.

Some students may be interested in contributing, but not know exactly
what to work on, so I thought about posting a few sample ideas on the
wiki page:
http://wiki.python.org/moin/SummerOfCode/Mentors

One is: adding more documentation and examples of conftest.py based
customization of py.test and integrating other python testing tools
with py.test.

I've been out of action for a while, so I also wanted to ask you all:
What additions or changes would you propose for a Summer of Code
project?

Take care,
-Brian


From holger at merlinux.de  Mon Mar 12 07:16:45 2007
From: holger at merlinux.de (holger krekel)
Date: Mon, 12 Mar 2007 07:16:45 +0100
Subject: [py-dev] Ideas for Google Summer of Code projects?
In-Reply-To: <66e877b70703111807j3e61e9d6hcee7005df521bc79@mail.gmail.com>
References: <66e877b70703111807j3e61e9d6hcee7005df521bc79@mail.gmail.com>
Message-ID: <20070312061645.GJ4635@solar.trillke>

Hi Brian!

On Sun, Mar 11, 2007 at 18:07 -0700, Brian Dorsey wrote:
> In case any students are interested in working on py-lib as a Google
> Summer of Code project, I've just offered to be a mentor.

cool, i also listed myself - Maciej and others: before you
subscribe yourself as mentor, consider that you can also 
just participate :) 
 
> Some students may be interested in contributing, but not know exactly
> what to work on, so I thought about posting a few sample ideas on the
> wiki page:
> http://wiki.python.org/moin/SummerOfCode/Mentors
>
> 
> One is: adding more documentation and examples of conftest.py based
> customization of py.test and integrating other python testing tools
> with py.test.
>
> I've been out of action for a while, so I also wanted to ask you all:
> What additions or changes would you propose for a Summer of Code
> project?

I'd definitely recommend to look into both 
py/doc/test.txt and py/doc/future.txt in 
the "future" sections for somewhat more challenging
ideas - i think it's fine to list them with the Wiki
and reference back to particular documentation. 
If i look at the current Wiki list of ideas, they appear
a bit like "all the boring stuff we don't want to do"
which i don't think is too attractive ...

best, 

holger


From holger at merlinux.de  Fri Mar 16 09:28:25 2007
From: holger at merlinux.de (holger krekel)
Date: Fri, 16 Mar 2007 09:28:25 +0100
Subject: [py-dev] py.test high level document / feedback
Message-ID: <20070316082825.GU4635@solar.trillke>

Hi folks! 

i guess i missed out on pointing people here to
a higher level py.test document, that we are to
finalize next week and submit to the EU. 
It'd be very helpful if you could read it
and provide feedback, maybe it helps to 
get a bit of context. Let me know what you
think! 

http://codespeak.net/pypy/extradoc/eu-report/D02.3_Testing_Framework-interim-2007-03-12.pdf

best & thanks,

holger

-- 
merlinux GmbH       Steinbergstr. 42    31139 Hildesheim   
http://merlinux.de  tel +49 5121 20800 75 (fax 77) 


From grig at agilistas.org  Fri Mar 16 17:16:05 2007
From: grig at agilistas.org (Grig Gheorghiu)
Date: Fri, 16 Mar 2007 09:16:05 -0700
Subject: [py-dev] py.test high level document / feedback
In-Reply-To: <20070316082825.GU4635@solar.trillke>
References: <20070316082825.GU4635@solar.trillke>
Message-ID: <dd49cdd20703160916i3a8c7300l4f9027379df08c24@mail.gmail.com>

On 3/16/07, holger krekel <holger at merlinux.de> wrote:
> Hi folks!
>
> i guess i missed out on pointing people here to
> a higher level py.test document, that we are to
> finalize next week and submit to the EU.
> It'd be very helpful if you could read it
> and provide feedback, maybe it helps to
> get a bit of context. Let me know what you
> think!
>
> http://codespeak.net/pypy/extradoc/eu-report/D02.3_Testing_Framework-interim-2007-03-12.pdf
>

Looks good! I would also put more emphasis on the test collection
process, and on the setup/teardown hooks at the different levels.
These were major innovations that py.test brought to the table, and
that other frameworks such as nose are imitating. Maybe a paragraph
for each of them, giving a bit more context.

Grig


From holger at merlinux.de  Sat Mar 17 15:57:31 2007
From: holger at merlinux.de (holger krekel)
Date: Sat, 17 Mar 2007 15:57:31 +0100
Subject: [py-dev] py lib installation issue
In-Reply-To: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
References: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
Message-ID: <20070317145731.GL4635@solar.trillke>

Hi Martijn! (ccing py-dev) 

i think you may try 0.9.0 - the syntax error issue was handled
there. To be honest, i am surprised a bit that it works 
as i didn't really do anything yet (and certainly not for 0.8.0)
except successfully installing VS2003 on a windows laptop (puh 
though!). 

best, 

holger

On Sat, Mar 17, 2007 at 15:34 +0100, Martijn Faassen wrote:
> Hi there,
> 
> When I install the Py lib using easy_install (cool that this works
> now!), I got this
> error:
> 
> Compiling 
> /usr/lib/python2.4/site-packages/py-0.8.0_alpha2-py2.4.egg/py/test/tes
> ting/data/syntax_error.py ...
>  File 
>  "/usr/lib/python2.4/site-packages/py-0.8.0_alpha2-py2.4.egg/py/test/testi
> ng/data/syntax_error.py", line 3
>    this is really not python
>                            ^
> SyntaxError: invalid syntax
> 
> Not really pretty but something one could live with.
> 
> Until I tried to update my ubuntu and had the upgrade process bail out
> with this same error - I wasn't installing Py then of course, but I
> suspect something that installs Python code tries to compile all .py
> packages it can find.
> 
> I'm not sure what this test aims to do but just thought I'd point out
> this installation issue.
> 
> Regards,
> 
> Martijn
> 

-- 
merlinux GmbH       Steinbergstr. 42    31139 Hildesheim   
http://merlinux.de  tel +49 5121 20800 75 (fax 77) 


From holger at merlinux.de  Sat Mar 17 16:20:04 2007
From: holger at merlinux.de (holger krekel)
Date: Sat, 17 Mar 2007 16:20:04 +0100
Subject: [py-dev] SoC proposals?
Message-ID: <20070317152004.GM4635@solar.trillke>

Hi folks, 

if you know someone (or yourself) who is interested 
to think about a SoC proposal in relation to py lib
and py.test - let this list or me now.  I'd be willing
to mentor and discuss, as are others.  You may 
find inspirations from

    http://codespeak.net/py/dist/test.html#future-planned-features-of-py-test
   
or
    http://codespeak.net/py/dist/future.html

and here is the SoC deadline page (applications are due 24th March): 

    http://code.google.com/support/bin/answer.py?answer=60325&topic=10729

best & cheers, 

holger


From faassen at startifact.com  Sat Mar 17 22:36:15 2007
From: faassen at startifact.com (Martijn Faassen)
Date: Sat, 17 Mar 2007 22:36:15 +0100
Subject: [py-dev] py lib installation issue
In-Reply-To: <20070317145731.GL4635@solar.trillke>
References: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
	<20070317145731.GL4635@solar.trillke>
Message-ID: <8928d4e90703171436g42b39f05r23e098c7cbba9a7d@mail.gmail.com>

Hey,

On 3/17/07, holger krekel <holger at merlinux.de> wrote:
> Hi Martijn! (ccing py-dev)
>
> i think you may try 0.9.0 - the syntax error issue was handled
> there. To be honest, i am surprised a bit that it works
> as i didn't really do anything yet (and certainly not for 0.8.0)
> except successfully installing VS2003 on a windows laptop (puh
> though!).

I'm happy to try 0.9, but your cheeseshop entry still lists 0.8:

http://cheeseshop.python.org/pypi/py

Regards,

Martijn


From brian at dorseys.org  Sat Mar 17 22:56:43 2007
From: brian at dorseys.org (Brian Dorsey)
Date: Sat, 17 Mar 2007 14:56:43 -0700
Subject: [py-dev] py lib installation issue
In-Reply-To: <8928d4e90703171436g42b39f05r23e098c7cbba9a7d@mail.gmail.com>
References: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
	<20070317145731.GL4635@solar.trillke>
	<8928d4e90703171436g42b39f05r23e098c7cbba9a7d@mail.gmail.com>
Message-ID: <66e877b70703171456j65961bdcrf943d0b6359b3991@mail.gmail.com>

On 3/17/07, Martijn Faassen <faassen at startifact.com> wrote:
> I'm happy to try 0.9, but your cheeseshop entry still lists 0.8:

Hmm... that's a bit confusing. Looks like that entry is owned by a
user named rocky. Perhaps I don't understand cheeseshop well enough,
but I don't see any easy way to contact particular users.

Take care,
  -Brian


From faassen at startifact.com  Sun Mar 18 01:36:44 2007
From: faassen at startifact.com (Martijn Faassen)
Date: Sun, 18 Mar 2007 01:36:44 +0100
Subject: [py-dev] py lib installation issue
In-Reply-To: <66e877b70703171456j65961bdcrf943d0b6359b3991@mail.gmail.com>
References: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
	<20070317145731.GL4635@solar.trillke>
	<8928d4e90703171436g42b39f05r23e098c7cbba9a7d@mail.gmail.com>
	<66e877b70703171456j65961bdcrf943d0b6359b3991@mail.gmail.com>
Message-ID: <8928d4e90703171736q70e5489ag2ce0e4a42cfe45cc@mail.gmail.com>

Hey,

On 3/17/07, Brian Dorsey <brian at dorseys.org> wrote:
> On 3/17/07, Martijn Faassen <faassen at startifact.com> wrote:
> > I'm happy to try 0.9, but your cheeseshop entry still lists 0.8:
>
> Hmm... that's a bit confusing. Looks like that entry is owned by a
> user named rocky. Perhaps I don't understand cheeseshop well enough,
> but I don't see any easy way to contact particular users.

I might know which Rocky this is. I sent off an email to him asking
whether it's him.

Regards,

Martijn


From holger at merlinux.de  Sun Mar 18 08:56:20 2007
From: holger at merlinux.de (holger krekel)
Date: Sun, 18 Mar 2007 08:56:20 +0100
Subject: [py-dev] [noway@ask.me: [ANN] GCC 4.1.2 installer for Python
	distutils compilation]
Message-ID: <20070318075620.GT4635@solar.trillke>

Hi folks, 

may the below be a good solution to building win extensions? 
I wouldn't like to test this on my "just-got-vs2003 to run"
laptop, so i'd prefer someone else can give it a try. 
If i interpret the announcement right it should produce
extensions that are usable for downloaded CPython windows
binaries.  Would someone give this a try and report back? 

best, 

    holger


----- Forwarded message from Giovanni Bajo <noway at ask.me> -----

To: comp-lang-python-announce at moderators.isc.org
From: Giovanni Bajo <noway at ask.me>
User-Agent: Thunderbird 1.5.0.10 (Windows/20070221)
Subject: [ANN] GCC 4.1.2 installer for Python distutils compilation
Date: Sun, 18 Mar 2007 03:08:03 +0100
X-Spambayes-Classification: ham; 0.00

Hello,

This page:
   http://www.develer.com/oss/GccWinBinaries

contains a friendly Windows installer for GCC 4.1.2 (MinGW binary version), 
with full support for integrating it with Python installations so that it is 
used by distutils to compile Python extensions.


Direct download link:
   http://www.develer.com/~rasky/gcc-4.1.2-mingw-setup.exe


Who needs this package?

  * People who wants to use FLOSS tools to develop Python extensions.
  * People who wants to use the recent GCC 4.1.2 to develop Python extensions, 
given that it easily outperforms the 4-years-old Visual Studio .NET 2003.


What's special about this?

  * mingw.org still has GCC 3.4.2, so go figure. Also, you need to compose 
other packages together. This is a single installer with everything inside.
  * By default, MinGW GCC links with MSVCRT.DLL, and not MSVCR71.DLL (used by 
Python 2.4 and 2.5). Fixing this is pretty complicate, and there's much 
confusion (Google turns up red herrings). This package handles everything for 
you, and it just works.

-- 
Giovanni Bajo
Develer S.r.l.
http://www.develer.com
-- 
http://mail.python.org/mailman/listinfo/python-announce-list

        Support the Python Software Foundation:
        http://www.python.org/psf/donations.html


----- End forwarded message -----

-- 
merlinux GmbH       Steinbergstr. 42    31139 Hildesheim   
http://merlinux.de  tel +49 5121 20800 75 (fax 77) 


From faassen at startifact.com  Sun Mar 18 16:19:40 2007
From: faassen at startifact.com (Martijn Faassen)
Date: Sun, 18 Mar 2007 16:19:40 +0100
Subject: [py-dev] py lib installation issue
In-Reply-To: <8928d4e90703171736q70e5489ag2ce0e4a42cfe45cc@mail.gmail.com>
References: <8928d4e90703170734x7b631246na003bf6a022dfa10@mail.gmail.com>
	<20070317145731.GL4635@solar.trillke>
	<8928d4e90703171436g42b39f05r23e098c7cbba9a7d@mail.gmail.com>
	<66e877b70703171456j65961bdcrf943d0b6359b3991@mail.gmail.com>
	<8928d4e90703171736q70e5489ag2ce0e4a42cfe45cc@mail.gmail.com>
Message-ID: <8928d4e90703180819m1c795654o4639b53aec1a7063@mail.gmail.com>

Hey,

On 3/18/07, Martijn Faassen <faassen at startifact.com> wrote:

> On 3/17/07, Brian Dorsey <brian at dorseys.org> wrote:
> > On 3/17/07, Martijn Faassen <faassen at startifact.com> wrote:
> > > I'm happy to try 0.9, but your cheeseshop entry still lists 0.8:
> >
> > Hmm... that's a bit confusing. Looks like that entry is owned by a
> > user named rocky. Perhaps I don't understand cheeseshop well enough,
> > but I don't see any easy way to contact particular users.
>
> I might know which Rocky this is. I sent off an email to him asking
> whether it's him.

The Rocky I thought it might be wasn't it, so I'm at a loss who
manages that entry.

Regards,

Martijn


From fijal at genesilico.pl  Sun Mar 18 22:04:52 2007
From: fijal at genesilico.pl (Maciek Fijalkowski)
Date: Sun, 18 Mar 2007 22:04:52 +0100
Subject: [py-dev] Generative tests
Message-ID: <45FDA974.6090603@genesilico.pl>

I would like to re-think a bit generative tests and here are my ideas.

1. There should be *no* setup_function called for generator, rather 
setup_generator and setup_generated_function called in a) test 
collection (before generator itself is called) b) run of a single function

2. Distributed testing should treat generators as single function, ie 
not split them between hosts. This will solve a lot of issues, also 
regarding speed.

What do you think?

PS. Names are rather ad-hoc, I'm happy with different choice.


From faassen at startifact.com  Fri Mar 23 23:21:36 2007
From: faassen at startifact.com (Martijn Faassen)
Date: Fri, 23 Mar 2007 23:21:36 +0100
Subject: [py-dev] unwedging the cheeseshop for py
Message-ID: <eu1jtg$m44$1@sea.gmane.org>

Hi there,

It was nice to discover a cheeseshop entry for the 'py' library that 
appears to work on my system (at least py.path):

http://cheeseshop.python.org/pypi/py

Unfortunately, it's stuck at 0.8alpha 2, and the cheese shop owner is 
the user 'rocky'. We don't know who 'rocky' is. Some py developer needs 
to contact the Cheeseshop authorities (who are they? I guess contacting 
richardjones at optusnet.com.au is a good bet?) and ask for 
maintenership/ownership for that entry. Then they can hopefully easily 
upload the latest version.

I think Holger might be a good person to contact them, as luckily he's 
listed as the author, so there's a good chance he'll be believed.

I'd be nice if we could figure out who this 'rocky' person is...

Regards,

Martijn