From shiyee at shiyee.dk  Sat Jan  5 21:41:04 2008
From: shiyee at shiyee.dk (Mads Chr. Olesen)
Date: Sat, 05 Jan 2008 21:41:04 +0100
Subject: [py-dev] Simple patch to add port argument to SshGateway
Message-ID: <1199565664.6984.3.camel@localhost.localdomain>

Hi there!

Attached is a very simple patch, to allow specifying the port argument
when connecting via a SshGateway.

-- 
Mads Chr. Olesen <shiyee at shiyee.dk>
shiyee.dk
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pyexecnet-sshport.patch
Type: text/x-patch
Size: 1077 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/pytest-dev/attachments/20080105/d20ec14a/attachment.bin>

From lac at openend.se  Sun Jan  6 11:36:02 2008
From: lac at openend.se (Laura Creighton)
Date: Sun, 06 Jan 2008 11:36:02 +0100
Subject: [py-dev] [TIP] Friendly py.test intro
In-Reply-To: Message from Titus Brown <titus@caltech.edu> 
	of "Sun, 06 Jan 2008 00:37:30 PST." <20080106083730.GA1416@caltech.edu> 
References: <20080106083730.GA1416@caltech.edu> 
Message-ID: <200801061036.m06Aa2hg018202@theraft.openend.se>

In a message of Sun, 06 Jan 2008 00:37:30 PST, Titus Brown writes:
>Hi all,
>
>here's a product of GHOP written by novanasa:
>
>	http://teckla.idyll.org/~t/transfer/py.test.html
>
>It's a simple introduction to py.test.
>
>I think with a bit of cleanup it could be quite useful.  Comments are
>welcome (here or directly on the issue, at
>
>   http://code.google.com/p/google-highly-open-participation-psf/issues/d
>etail?id=269
>
>)
>
>Does anyone have any ideas on where to host it?

Codespeak seems logical.  I've forwarded this to the py-dev mailing list,
where people ought to have opinions on where this should go.

Laura

>
>cheers,
>--titus
>
>_______________________________________________
>testing-in-python mailing list
>testing-in-python at lists.idyll.org
>http://lists.idyll.org/listinfo/testing-in-python


From sol at sr-research.com  Thu Jan 10 22:53:43 2008
From: sol at sr-research.com (Sol Simpson)
Date: Thu, 10 Jan 2008 16:53:43 -0500
Subject: [py-dev] issues running py.execnet
Message-ID: <000301c853d3$48dac1c0$04010164@LIAH>

I'm trying to test out py.execnet running one of the examples:

 

import py

gw = py.execnet.PopenGateway()

channel = gw.remote_exec("""

            import os

            channel.send(os.getpid())

            """)

remote_pid = channel.receive()

print remote_pid != py.std.os.getpid()

 

The issue is that the program hangs on the channel.receive() call. In task
manager I can see that a second python.exe has been launched, but that is
it. The only way to exit the app is to kill one of the python.exe in the
task manager.

 

Any insight would be very helpful.

 

I'm running python 2.5.1 on Windows XP SP2.

 

Thank you,

 

Sol

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/pytest-dev/attachments/20080110/65d6e553/attachment.html>

From holger at merlinux.de  Fri Jan 11 17:59:34 2008
From: holger at merlinux.de (holger krekel)
Date: Fri, 11 Jan 2008 17:59:34 +0100
Subject: [py-dev] issues running py.execnet
In-Reply-To: <000301c853d3$48dac1c0$04010164@LIAH>
References: <000301c853d3$48dac1c0$04010164@LIAH>
Message-ID: <20080111165934.GV13357@solar.trillke>

Hi Sol, 

i tried the example with the 0.9.0 py release on my Windows XP box
with Python 2.5.1 - works for me. 

are you also using 0.9.0? 

Could you give me the output of
    
    >>> gw

and

    >>> gw.remote_exec("channel.send(42)").receive()

?

best, 

holger

On Thu, Jan 10, 2008 at 16:53 -0500, Sol Simpson wrote:
> I'm trying to test out py.execnet running one of the examples:
> 
>  
> 
> import py
> 
> gw = py.execnet.PopenGateway()
> 
> channel = gw.remote_exec("""
> 
>             import os
> 
>             channel.send(os.getpid())
> 
>             """)
> 
> remote_pid = channel.receive()
> 
> print remote_pid != py.std.os.getpid()
> 
>  
> 
> The issue is that the program hangs on the channel.receive() call. In task
> manager I can see that a second python.exe has been launched, but that is
> it. The only way to exit the app is to kill one of the python.exe in the
> task manager.
> 
>  
> 
> Any insight would be very helpful.
> 
>  
> 
> I'm running python 2.5.1 on Windows XP SP2.
> 
>  
> 
> Thank you,
> 
>  
> 
> Sol
> 
>  
> 

> _______________________________________________
> py-dev mailing list
> py-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev


-- 
merlinux GmbH       
http://merlinux.de  


From holger at merlinux.de  Fri Jan 11 22:09:54 2008
From: holger at merlinux.de (holger krekel)
Date: Fri, 11 Jan 2008 22:09:54 +0100
Subject: [py-dev] issues running py.execnet
In-Reply-To: <001901c85486$04e7a560$04010164@LIAH>
References: <000301c853d3$48dac1c0$04010164@LIAH>
	<20080111165934.GV13357@solar.trillke>
	<001901c85486$04e7a560$04010164@LIAH>
Message-ID: <20080111210954.GX13357@solar.trillke>

Hi Sol, 

could you maybe edit 

    py/execnet/gateway.py 

in line 29, enabling the writing to the debug file? 
and try the example and send the content 
of the created debug file(s) here? 

Could you also see what ">>> gw" gives before 
channel.receive() is called? 

best, 

holger

On Fri, Jan 11, 2008 at 14:13 -0500, Sol Simpson wrote:
> Hi holger,
> 
> Thanks for the reply.
> 
> Print gw:
> 
> <PopenGateway receiving/sending (0 active channels)>
> 
> Print: gw.remote_exec("channel.send(42)").receive()
> 
> Hangs like before.
> 
> I am running 0.9.0, compiled using setup.py
> 
> I have tried with stackless python 2.5.1 and standard python 2.5.1.
> 
> Please let me know fi there is something else I can try.
> 
> Thanks again,
> 
> Sol Simpson
> SR Research Ltd.
> 
> -----Original Message-----
> From: holger krekel [mailto:holger at merlinux.de] 
> Sent: January 11, 2008 12:00 PM
> To: Sol Simpson
> Cc: py-dev at codespeak.net
> Subject: Re: [py-dev] issues running py.execnet
> 
> Hi Sol, 
> 
> i tried the example with the 0.9.0 py release on my Windows XP box
> with Python 2.5.1 - works for me. 
> 
> are you also using 0.9.0? 
> 
> Could you give me the output of
>     
>     >>> gw
> 
> and
> 
>     >>> gw.remote_exec("channel.send(42)").receive()
> 
> ?
> 
> best, 
> 
> holger
> 
> On Thu, Jan 10, 2008 at 16:53 -0500, Sol Simpson wrote:
> > I'm trying to test out py.execnet running one of the examples:
> > 
> >  
> > 
> > import py
> > 
> > gw = py.execnet.PopenGateway()
> > 
> > channel = gw.remote_exec("""
> > 
> >             import os
> > 
> >             channel.send(os.getpid())
> > 
> >             """)
> > 
> > remote_pid = channel.receive()
> > 
> > print remote_pid != py.std.os.getpid()
> > 
> >  
> > 
> > The issue is that the program hangs on the channel.receive() call. In task
> > manager I can see that a second python.exe has been launched, but that is
> > it. The only way to exit the app is to kill one of the python.exe in the
> > task manager.
> > 
> >  
> > 
> > Any insight would be very helpful.
> > 
> >  
> > 
> > I'm running python 2.5.1 on Windows XP SP2.
> > 
> >  
> > 
> > Thank you,
> > 
> >  
> > 
> > Sol
> > 
> >  
> > 
> 
> > _______________________________________________
> > py-dev mailing list
> > py-dev at codespeak.net
> > http://codespeak.net/mailman/listinfo/py-dev
> 
> 
> -- 
> merlinux GmbH       
> http://merlinux.de  
> 

-- 
merlinux GmbH       
http://merlinux.de  


From holger at merlinux.de  Fri Jan 11 22:27:48 2008
From: holger at merlinux.de (holger krekel)
Date: Fri, 11 Jan 2008 22:27:48 +0100
Subject: [py-dev] issues running py.execnet
In-Reply-To: <002701c85498$1de5c9e0$04010164@LIAH>
References: <000301c853d3$48dac1c0$04010164@LIAH>
	<20080111165934.GV13357@solar.trillke>
	<001901c85486$04e7a560$04010164@LIAH>
	<20080111210954.GX13357@solar.trillke>
	<002701c85498$1de5c9e0$04010164@LIAH>
Message-ID: <20080111212748.GY13357@solar.trillke>

Hi Sol (i am adding py-dev again to CC), 

On Fri, Jan 11, 2008 at 16:22 -0500, Sol Simpson wrote:
> If I put a gw.exit() at the end of the script, then it ends OK.
> 
> Is that all it was, that I did not have and exit at the end of the script?

ah - i initially thought that the "channel.receive()" call 
itself hang and not the try to leave the python interpreter.

Indeed, you should "gw.exit()" - the automatic cleanup semantics 
(at process exit) are not reliable yet throughout python versions, 
i am afraid. 

best,

holger

> -----Original Message-----
> From: holger krekel [mailto:holger at merlinux.de] 
> Sent: January 11, 2008 4:10 PM
> To: Sol Simpson
> Cc: py-dev at codespeak.net
> Subject: Re: [py-dev] issues running py.execnet
> 
> Hi Sol, 
> 
> could you maybe edit 
> 
>     py/execnet/gateway.py 
> 
> in line 29, enabling the writing to the debug file? 
> and try the example and send the content 
> of the created debug file(s) here? 
> 
> Could you also see what ">>> gw" gives before 
> channel.receive() is called? 
> 
> best, 
> 
> holger
> 
> On Fri, Jan 11, 2008 at 14:13 -0500, Sol Simpson wrote:
> > Hi holger,
> > 
> > Thanks for the reply.
> > 
> > Print gw:
> > 
> > <PopenGateway receiving/sending (0 active channels)>
> > 
> > Print: gw.remote_exec("channel.send(42)").receive()
> > 
> > Hangs like before.
> > 
> > I am running 0.9.0, compiled using setup.py
> > 
> > I have tried with stackless python 2.5.1 and standard python 2.5.1.
> > 
> > Please let me know fi there is something else I can try.
> > 
> > Thanks again,
> > 
> > Sol Simpson
> > SR Research Ltd.
> > 
> > -----Original Message-----
> > From: holger krekel [mailto:holger at merlinux.de] 
> > Sent: January 11, 2008 12:00 PM
> > To: Sol Simpson
> > Cc: py-dev at codespeak.net
> > Subject: Re: [py-dev] issues running py.execnet
> > 
> > Hi Sol, 
> > 
> > i tried the example with the 0.9.0 py release on my Windows XP box
> > with Python 2.5.1 - works for me. 
> > 
> > are you also using 0.9.0? 
> > 
> > Could you give me the output of
> >     
> >     >>> gw
> > 
> > and
> > 
> >     >>> gw.remote_exec("channel.send(42)").receive()
> > 
> > ?
> > 
> > best, 
> > 
> > holger
> > 
> > On Thu, Jan 10, 2008 at 16:53 -0500, Sol Simpson wrote:
> > > I'm trying to test out py.execnet running one of the examples:
> > > 
> > >  
> > > 
> > > import py
> > > 
> > > gw = py.execnet.PopenGateway()
> > > 
> > > channel = gw.remote_exec("""
> > > 
> > >             import os
> > > 
> > >             channel.send(os.getpid())
> > > 
> > >             """)
> > > 
> > > remote_pid = channel.receive()
> > > 
> > > print remote_pid != py.std.os.getpid()
> > > 
> > >  
> > > 
> > > The issue is that the program hangs on the channel.receive() call. In
> task
> > > manager I can see that a second python.exe has been launched, but that
> is
> > > it. The only way to exit the app is to kill one of the python.exe in the
> > > task manager.
> > > 
> > >  
> > > 
> > > Any insight would be very helpful.
> > > 
> > >  
> > > 
> > > I'm running python 2.5.1 on Windows XP SP2.
> > > 
> > >  
> > > 
> > > Thank you,
> > > 
> > >  
> > > 
> > > Sol
> > > 
> > >  
> > > 
> > 
> > > _______________________________________________
> > > py-dev mailing list
> > > py-dev at codespeak.net
> > > http://codespeak.net/mailman/listinfo/py-dev
> > 
> > 
> > -- 
> > merlinux GmbH       
> > http://merlinux.de  
> > 
> 
> -- 
> merlinux GmbH       
> http://merlinux.de  
> 

-- 
merlinux GmbH       
http://merlinux.de  


From sol at sr-research.com  Fri Jan 11 22:33:23 2008
From: sol at sr-research.com (Sol Simpson)
Date: Fri, 11 Jan 2008 16:33:23 -0500
Subject: [py-dev] issues running py.execnet
In-Reply-To: <20080111212748.GY13357@solar.trillke>
References: <000301c853d3$48dac1c0$04010164@LIAH>
	<20080111165934.GV13357@solar.trillke>
	<001901c85486$04e7a560$04010164@LIAH>
	<20080111210954.GX13357@solar.trillke>
	<002701c85498$1de5c9e0$04010164@LIAH>
	<20080111212748.GY13357@solar.trillke>
Message-ID: <002b01c85499$9cd4a450$04010164@LIAH>

Thanks very much. 

Sol Simpson
SR Research Ltd.


-----Original Message-----
From: holger krekel [mailto:holger at merlinux.de] 
Sent: January 11, 2008 4:28 PM
To: Sol Simpson
Cc: py-dev at codespeak.net
Subject: Re: [py-dev] issues running py.execnet

Hi Sol (i am adding py-dev again to CC), 

On Fri, Jan 11, 2008 at 16:22 -0500, Sol Simpson wrote:
> If I put a gw.exit() at the end of the script, then it ends OK.
> 
> Is that all it was, that I did not have and exit at the end of the script?

ah - i initially thought that the "channel.receive()" call 
itself hang and not the try to leave the python interpreter.

Indeed, you should "gw.exit()" - the automatic cleanup semantics 
(at process exit) are not reliable yet throughout python versions, 
i am afraid. 

best,

holger

> -----Original Message-----
> From: holger krekel [mailto:holger at merlinux.de] 
> Sent: January 11, 2008 4:10 PM
> To: Sol Simpson
> Cc: py-dev at codespeak.net
> Subject: Re: [py-dev] issues running py.execnet
> 
> Hi Sol, 
> 
> could you maybe edit 
> 
>     py/execnet/gateway.py 
> 
> in line 29, enabling the writing to the debug file? 
> and try the example and send the content 
> of the created debug file(s) here? 
> 
> Could you also see what ">>> gw" gives before 
> channel.receive() is called? 
> 
> best, 
> 
> holger
> 
> On Fri, Jan 11, 2008 at 14:13 -0500, Sol Simpson wrote:
> > Hi holger,
> > 
> > Thanks for the reply.
> > 
> > Print gw:
> > 
> > <PopenGateway receiving/sending (0 active channels)>
> > 
> > Print: gw.remote_exec("channel.send(42)").receive()
> > 
> > Hangs like before.
> > 
> > I am running 0.9.0, compiled using setup.py
> > 
> > I have tried with stackless python 2.5.1 and standard python 2.5.1.
> > 
> > Please let me know fi there is something else I can try.
> > 
> > Thanks again,
> > 
> > Sol Simpson
> > SR Research Ltd.
> > 
> > -----Original Message-----
> > From: holger krekel [mailto:holger at merlinux.de] 
> > Sent: January 11, 2008 12:00 PM
> > To: Sol Simpson
> > Cc: py-dev at codespeak.net
> > Subject: Re: [py-dev] issues running py.execnet
> > 
> > Hi Sol, 
> > 
> > i tried the example with the 0.9.0 py release on my Windows XP box
> > with Python 2.5.1 - works for me. 
> > 
> > are you also using 0.9.0? 
> > 
> > Could you give me the output of
> >     
> >     >>> gw
> > 
> > and
> > 
> >     >>> gw.remote_exec("channel.send(42)").receive()
> > 
> > ?
> > 
> > best, 
> > 
> > holger
> > 
> > On Thu, Jan 10, 2008 at 16:53 -0500, Sol Simpson wrote:
> > > I'm trying to test out py.execnet running one of the examples:
> > > 
> > >  
> > > 
> > > import py
> > > 
> > > gw = py.execnet.PopenGateway()
> > > 
> > > channel = gw.remote_exec("""
> > > 
> > >             import os
> > > 
> > >             channel.send(os.getpid())
> > > 
> > >             """)
> > > 
> > > remote_pid = channel.receive()
> > > 
> > > print remote_pid != py.std.os.getpid()
> > > 
> > >  
> > > 
> > > The issue is that the program hangs on the channel.receive() call. In
> task
> > > manager I can see that a second python.exe has been launched, but that
> is
> > > it. The only way to exit the app is to kill one of the python.exe in
the
> > > task manager.
> > > 
> > >  
> > > 
> > > Any insight would be very helpful.
> > > 
> > >  
> > > 
> > > I'm running python 2.5.1 on Windows XP SP2.
> > > 
> > >  
> > > 
> > > Thank you,
> > > 
> > >  
> > > 
> > > Sol
> > > 
> > >  
> > > 
> > 
> > > _______________________________________________
> > > py-dev mailing list
> > > py-dev at codespeak.net
> > > http://codespeak.net/mailman/listinfo/py-dev
> > 
> > 
> > -- 
> > merlinux GmbH       
> > http://merlinux.de  
> > 
> 
> -- 
> merlinux GmbH       
> http://merlinux.de  
> 

-- 
merlinux GmbH       
http://merlinux.de  



From holger at merlinux.de  Sat Jan 26 13:39:08 2008
From: holger at merlinux.de (holger krekel)
Date: Sat, 26 Jan 2008 13:39:08 +0100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
Message-ID: <20080126123908.GE1337@solar.trillke>

Hey everybody, 

as some of you already know i'd like to get a py lib/py.test 1.0 out 
in February.  IOW, my sabbatical time which i took after the heavy last EU 
period of PyPy developments mid last year ... is over and i am slowly 
getting back into gear.  

Yesterday, i merged Maciej Fijalkowski's branch into py/trunk which 
unified objects used for distributed and in-process testing.  
I am going to refine the architecture further.  I generally try to 
keep track here: http://codespeak.net/svn/py/trunk/py/doc/TODO.txt 

How could you help with 1.0?  

For starters, it'd be great if you mail me URLs of 
code or projects using py.test and/or point to custom 
conftest.py's (py.test's extension mechanism).  

Feel free to also mail me your favourite annoyance or feature
wish along ... or point to ones from other tools.  This all
increases chances that 1.0 will work to your expectations :) 

I could use contributions or help (or sometimes just hints) in 
the following particular areas: 

- integrating py.test with Twisted 
- packaging for debian, setuptools
- unix/windows cross-platform communication and process
  deployment 

Please feel free to join the #pylib channel on freenode where you 
see commits to the code base and have me and a few other contributors 
hanging out ...  i am happy to discuss and answer questions there! 

I am also particularly interested in collaboration with 
other testing tool authors - let's work on improving everyone's 
testing experiences and on world wide code quality! :) 

best & cheers, 

holger

-- 
py.test: http://pytest.org 
py lib: http://pylib.org 
PyPy: http://codespeak.net/pypy 
merlinux GmbH: http://merlinux.de  


From grig at gheorghiu.net  Sat Jan 26 17:30:51 2008
From: grig at gheorghiu.net (Grig Gheorghiu)
Date: Sat, 26 Jan 2008 08:30:51 -0800 (PST)
Subject: [py-dev] [TIP] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080126123908.GE1337@solar.trillke>
Message-ID: <467999.47168.qm@web54502.mail.re2.yahoo.com>

--- holger krekel <holger at merlinux.de> wrote:

> Hey everybody, 
> 
> 
> Feel free to also mail me your favourite annoyance or feature
> wish along ... or point to ones from other tools.  This all
> increases chances that 1.0 will work to your expectations :) 
> 
> I could use contributions or help (or sometimes just hints) in 
> the following particular areas: 
> 
> - integrating py.test with Twisted 
> - packaging for debian, setuptools
> - unix/windows cross-platform communication and process
>   deployment 
> 

Hi, Holger!

My main request for py.test would be #2 in your list above. I'd like to
be able to say 'easy_install py.test' and be done. From what I see in
the Python community, the more a tool can be easily deployed and
upgraded, and the more it integrates with other standard tools (such as
setuptools) -- the better its chances of being used on a large scale.

Grig




From beacybooks at bigpond.com  Sat Jan 26 23:34:20 2008
From: beacybooks at bigpond.com (Tom Harris)
Date: Sun, 27 Jan 2008 09:34:20 +1100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080126123908.GE1337@solar.trillke>
References: <20080126123908.GE1337@solar.trillke>
Message-ID: <479BB56C.5000902@bigpond.com>

I use py.test for testing embedded systems, and I find that I use 
generative tests a lot. Some means of running setup/teardown methods for 
_each_ test in a generative test would be very useful, I have a few 
ideas for syntax, but none for implementation.

Also I second the request for easy deployment of py.test, as I never use 
the other components of py.* there could even be a distribution with 
these removed.

Tom Harris BeacyBooks <beacybooks at bigpond.com>



holger krekel wrote:

>Hey everybody, 
>
>as some of you already know i'd like to get a py lib/py.test 1.0 out 
>in February.  IOW, my sabbatical time which i took after the heavy last EU 
>period of PyPy developments mid last year ... is over and i am slowly 
>getting back into gear.  
>
>Yesterday, i merged Maciej Fijalkowski's branch into py/trunk which 
>unified objects used for distributed and in-process testing.  
>I am going to refine the architecture further.  I generally try to 
>keep track here: http://codespeak.net/svn/py/trunk/py/doc/TODO.txt 
>
>How could you help with 1.0?  
>
>For starters, it'd be great if you mail me URLs of 
>code or projects using py.test and/or point to custom 
>conftest.py's (py.test's extension mechanism).  
>
>Feel free to also mail me your favourite annoyance or feature
>wish along ... or point to ones from other tools.  This all
>increases chances that 1.0 will work to your expectations :) 
>
>I could use contributions or help (or sometimes just hints) in 
>the following particular areas: 
>
>- integrating py.test with Twisted 
>- packaging for debian, setuptools
>- unix/windows cross-platform communication and process
>  deployment 
>
>Please feel free to join the #pylib channel on freenode where you 
>see commits to the code base and have me and a few other contributors 
>hanging out ...  i am happy to discuss and answer questions there! 
>
>I am also particularly interested in collaboration with 
>other testing tool authors - let's work on improving everyone's 
>testing experiences and on world wide code quality! :) 
>
>best & cheers, 
>
>holger
>
>  
>


From holger at merlinux.de  Sun Jan 27 00:49:44 2008
From: holger at merlinux.de (holger krekel)
Date: Sun, 27 Jan 2008 00:49:44 +0100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
In-Reply-To: <479BB56C.5000902@bigpond.com>
References: <20080126123908.GE1337@solar.trillke>
	<479BB56C.5000902@bigpond.com>
Message-ID: <20080126234944.GH1337@solar.trillke>

Hi Tom, 

On Sun, Jan 27, 2008 at 09:34 +1100, Tom Harris wrote:
> I use py.test for testing embedded systems, and I find that I use 
> generative tests a lot. Some means of running setup/teardown methods for 
> _each_ test in a generative test would be very useful, I have a few 
> ideas for syntax, but none for implementation.

i'd like to think more about generative tests and particularly
their implementation but it's likely out of 1.0 scope.  
But Feel free to post considerations/ideas about generative tests
and how to syntactically express them! 

> Also I second the request for easy deployment of py.test, as I never use 
> the other components of py.* there could even be a distribution with 
> these removed.

py.test itself internally uses many of the py lib's components
(and actually often triggered their development).  

Regarding your usage - in case you have a publically accessible
code base please point me to it. 

best, 

holger

> Tom Harris BeacyBooks <beacybooks at bigpond.com>
> 
> 
> 
> holger krekel wrote:
> 
> >Hey everybody, 
> >
> >as some of you already know i'd like to get a py lib/py.test 1.0 out 
> >in February.  IOW, my sabbatical time which i took after the heavy last EU 
> >period of PyPy developments mid last year ... is over and i am slowly 
> >getting back into gear.  
> >
> >Yesterday, i merged Maciej Fijalkowski's branch into py/trunk which 
> >unified objects used for distributed and in-process testing.  
> >I am going to refine the architecture further.  I generally try to 
> >keep track here: http://codespeak.net/svn/py/trunk/py/doc/TODO.txt 
> >
> >How could you help with 1.0?  
> >
> >For starters, it'd be great if you mail me URLs of 
> >code or projects using py.test and/or point to custom 
> >conftest.py's (py.test's extension mechanism).  
> >
> >Feel free to also mail me your favourite annoyance or feature
> >wish along ... or point to ones from other tools.  This all
> >increases chances that 1.0 will work to your expectations :) 
> >
> >I could use contributions or help (or sometimes just hints) in 
> >the following particular areas: 
> >
> >- integrating py.test with Twisted 
> >- packaging for debian, setuptools
> >- unix/windows cross-platform communication and process
> >  deployment 
> >
> >Please feel free to join the #pylib channel on freenode where you 
> >see commits to the code base and have me and a few other contributors 
> >hanging out ...  i am happy to discuss and answer questions there! 
> >
> >I am also particularly interested in collaboration with 
> >other testing tool authors - let's work on improving everyone's 
> >testing experiences and on world wide code quality! :) 
> >
> >best & cheers, 
> >
> >holger
> >
> >  
> >
> _______________________________________________
> py-dev mailing list
> py-dev at codespeak.net
> http://codespeak.net/mailman/listinfo/py-dev
> 

-- 
py.test: http://pytest.org 
py lib: http://pylib.org 
PyPy: http://codespeak.net/pypy 
merlinux GmbH: http://merlinux.de 


From johnny at johnnydebris.net  Sun Jan 27 10:48:18 2008
From: johnny at johnnydebris.net (Guido Wesdorp)
Date: Sun, 27 Jan 2008 10:48:18 +0100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080126123908.GE1337@solar.trillke>
References: <20080126123908.GE1337@solar.trillke>
Message-ID: <479C5362.7020600@johnnydebris.net>

holger krekel wrote:
> IOW, my sabbatical time which i took after the heavy last EU 
> period of PyPy developments mid last year ... is over and i am slowly 
> getting back into gear.  
>
>   
Welcome back!

> For starters, it'd be great if you mail me URLs of 
> code or projects using py.test and/or point to custom 
> conftest.py's (py.test's extension mechanism).  
>
>   
I wrote a couple of conftests, one which executes unittest.py's TestCase 
instances, and one which runs JavaScript tests using Spidermonkey (the 
latter as part of a set of JS libs to handle common issues in JS and 
browsers):

http://johnnydebris.net/svn/projects/py_unittest/ - unittest integration

http://johnnydebris.net/svn/projects/jsbase/ - the base lib with 
conftest (and some implementations of assert* functions)

Also, I'd like to join development a bit early februari (is a release 
date set already?), I'll make sure to be on #pylib around that time...

Cheers,

Guido


From holger at merlinux.de  Sun Jan 27 22:23:41 2008
From: holger at merlinux.de (holger krekel)
Date: Sun, 27 Jan 2008 22:23:41 +0100
Subject: [py-dev] [TIP] towards 1.0 / code URLS using py.test?
In-Reply-To: <467999.47168.qm@web54502.mail.re2.yahoo.com>
References: <20080126123908.GE1337@solar.trillke>
	<467999.47168.qm@web54502.mail.re2.yahoo.com>
Message-ID: <20080127212341.GM1337@solar.trillke>

Hi Grig! 

On Sat, Jan 26, 2008 at 08:30 -0800, Grig Gheorghiu wrote:
> --- holger krekel <holger at merlinux.de> wrote:
> 
> > Hey everybody, 
> > 
> > 
> > Feel free to also mail me your favourite annoyance or feature
> > wish along ... or point to ones from other tools.  This all
> > increases chances that 1.0 will work to your expectations :) 
> > 
> > I could use contributions or help (or sometimes just hints) in 
> > the following particular areas: 
> > 
> > - integrating py.test with Twisted 
> > - packaging for debian, setuptools
> > - unix/windows cross-platform communication and process
> >   deployment 
> > 
> 
> Hi, Holger!
> 
> My main request for py.test would be #2 in your list above. I'd like to
> be able to say 'easy_install py.test' and be done. From what I see in
> the Python community, the more a tool can be easily deployed and
> upgraded, and the more it integrates with other standard tools (such as
> setuptools) -- the better its chances of being used on a large scale.

surely, you are right.  What happens if you type 

    easy_install py

?  For me this works and installs py.test (which is part of the py lib). 

Is it possible to have something like an alias so that when people 
type "easy_install pytest" or so it also installs the right thing? 

best & cheers, 

holger


From schmir at gmail.com  Mon Jan 28 05:53:23 2008
From: schmir at gmail.com (Ralf Schmitt)
Date: Mon, 28 Jan 2008 05:53:23 +0100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080126123908.GE1337@solar.trillke>
References: <20080126123908.GE1337@solar.trillke>
Message-ID: <932f8baf0801272053p6607671flccf6902f412c1451@mail.gmail.com>

On Jan 26, 2008 1:39 PM, holger krekel <holger at merlinux.de> wrote:

>
> I could use contributions or help (or sometimes just hints) in
> the following particular areas:
>
> - integrating py.test with Twisted


http://systemexit.de/repo/py.test.twisted/ contains a mercurial repository,
which makes testing code running under the twisted reactor possible. (I've
already asked about integrating it into the py library).

Regards,
- Ralf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/pytest-dev/attachments/20080128/70f42706/attachment.html>

From holger at merlinux.de  Mon Jan 28 11:09:44 2008
From: holger at merlinux.de (holger krekel)
Date: Mon, 28 Jan 2008 11:09:44 +0100
Subject: [py-dev] towards 1.0 / code URLS using py.test?
In-Reply-To: <932f8baf0801272053p6607671flccf6902f412c1451@mail.gmail.com>
References: <20080126123908.GE1337@solar.trillke>
	<932f8baf0801272053p6607671flccf6902f412c1451@mail.gmail.com>
Message-ID: <20080128100944.GN1337@solar.trillke>

Hi Ralf, 

On Mon, Jan 28, 2008 at 05:53 +0100, Ralf Schmitt wrote:
> On Jan 26, 2008 1:39 PM, holger krekel <holger at merlinux.de> wrote:
> >
> > I could use contributions or help (or sometimes just hints) in
> > the following particular areas:
> >
> > - integrating py.test with Twisted
> 
> 
> http://systemexit.de/repo/py.test.twisted/ contains a mercurial repository,
> which makes testing code running under the twisted reactor possible. (I've
> already asked about integrating it into the py library).

Yip, in fact i wanted to get back to you - thanks for answering here. 
Your integration looks suprisingly small! I wasn't aware that
using greenlets with twisted works that well, to be honest.  
Definitely makes sense to include this technique - i think we'll 
need to sort out some more issues like interaction with distributed
testing, doctests, ReST tests etc.  (on py-dev). 

best & thanks, 

holger

-- 
py.test: http://pytest.org 
py lib: http://pylib.org 
PyPy: http://codespeak.net/pypy 
merlinux GmbH: http://merlinux.de 


From exarkun at divmod.com  Sat Jan 26 15:27:09 2008
From: exarkun at divmod.com (Jean-Paul Calderone)
Date: Sat, 26 Jan 2008 09:27:09 -0500
Subject: [py-dev] [TIP] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080126123908.GE1337@solar.trillke>
Message-ID: <20080126142709.6859.1607151009.divmod.quotient.1026@ohm>

On Sat, 26 Jan 2008 13:39:08 +0100, holger krekel <holger at merlinux.de> wrote:
>Hey everybody,
>
>as some of you already know i'd like to get a py lib/py.test 1.0 out
>in February.  IOW, my sabbatical time which i took after the heavy last EU
>period of PyPy developments mid last year ... is over and i am slowly
>getting back into gear.

I hope you had an enjoyable and refreshing sabbatical.  Welcome back. :)

>
>Yesterday, i merged Maciej Fijalkowski's branch into py/trunk which
>unified objects used for distributed and in-process testing.
>I am going to refine the architecture further.  I generally try to
>keep track here: http://codespeak.net/svn/py/trunk/py/doc/TODO.txt

Cool.  This means that if I have been using the reporter-merge branch,
I can just use trunk now, right?

>
>How could you help with 1.0?
>
> [snip]
>
>I could use contributions or help (or sometimes just hints) in
>the following particular areas:
>
>- integrating py.test with Twisted

Do you have any particular ideas or goals for this?

Jean-Paul


From exarkun at divmod.com  Mon Jan 28 16:09:38 2008
From: exarkun at divmod.com (Jean-Paul Calderone)
Date: Mon, 28 Jan 2008 10:09:38 -0500
Subject: [py-dev] [TIP]  towards 1.0 / code URLS using py.test?
In-Reply-To: <20080128100944.GN1337@solar.trillke>
Message-ID: <20080128150938.6859.686086216.divmod.quotient.1802@ohm>

On Mon, 28 Jan 2008 11:09:44 +0100, holger krekel <holger at merlinux.de> wrote:
>Hi Ralf,
>
>On Mon, Jan 28, 2008 at 05:53 +0100, Ralf Schmitt wrote:
>> On Jan 26, 2008 1:39 PM, holger krekel <holger at merlinux.de> wrote:
>> >
>> > I could use contributions or help (or sometimes just hints) in
>> > the following particular areas:
>> >
>> > - integrating py.test with Twisted
>>
>>
>> http://systemexit.de/repo/py.test.twisted/ contains a mercurial repository,
>> which makes testing code running under the twisted reactor possible. (I've
>> already asked about integrating it into the py library).

Is it intentional that this integration busy-loops, guaranteeing 100% CPU
usage for the entire test-run?

Jean-Paul


From grig at gheorghiu.net  Mon Jan 28 20:49:25 2008
From: grig at gheorghiu.net (Grig Gheorghiu)
Date: Mon, 28 Jan 2008 11:49:25 -0800 (PST)
Subject: [py-dev] [TIP] towards 1.0 / code URLS using py.test?
In-Reply-To: <20080127212341.GM1337@solar.trillke>
Message-ID: <336645.87454.qm@web54501.mail.re2.yahoo.com>

--- holger krekel <holger at merlinux.de> wrote:

> What happens if you type 
> 
>     easy_install py
> 
> ?  For me this works and installs py.test (which is part of the py
> lib). 
> 

You're right, it works for me too. I wasn't aware that it's already
been setuptools-ified. That's good news!

> Is it possible to have something like an alias so that when people 
> type "easy_install pytest" or so it also installs the right thing? 
> 

Not sure, I'm not really a setuptools expert.

Grig


From schmir at gmail.com  Tue Jan 29 07:02:28 2008
From: schmir at gmail.com (Ralf Schmitt)
Date: Tue, 29 Jan 2008 07:02:28 +0100
Subject: [py-dev] [TIP]  towards 1.0 / code URLS using py.test?
In-Reply-To: <20080128150938.6859.686086216.divmod.quotient.1802@ohm>
References: <20080128100944.GN1337@solar.trillke>
	<20080128150938.6859.686086216.divmod.quotient.1802@ohm>
Message-ID: <932f8baf0801282202s1f9fa7e3l717e100d41f08dda@mail.gmail.com>

On Jan 28, 2008 4:09 PM, Jean-Paul Calderone <exarkun at divmod.com> wrote:

> On Mon, 28 Jan 2008 11:09:44 +0100, holger krekel <holger at merlinux.de>
> wrote:
> >Hi Ralf,
> >
> >On Mon, Jan 28, 2008 at 05:53 +0100, Ralf Schmitt wrote:
> >> On Jan 26, 2008 1:39 PM, holger krekel <holger at merlinux.de> wrote:
> >> >
> >> > I could use contributions or help (or sometimes just hints) in
> >> > the following particular areas:
> >> >
> >> > - integrating py.test with Twisted
> >>
> >>
> >> http://systemexit.de/repo/py.test.twisted/ contains a mercurial
> repository,
> >> which makes testing code running under the twisted reactor possible.
> (I've
> >> already asked about integrating it into the py library).
>
> Is it intentional that this integration busy-loops, guaranteeing 100% CPU
> usage for the entire test-run?
>

Why do you think so? Apparently this isn't true on my machine:

~/py.test.twisted/ cat test_some.py                                 ralf at redok
#! /usr/bin/env py.test.twisted

from twisted.internet import defer, reactor
from twisted.python import failure
from twisted.python import log

def test_deferred():
    d = defer.Deferred()
    def done():
        log.msg("done")
        d.callback(None)

    reactor.callLater(25, done)
    log.msg("Returning %r" % (d,))
    return d

def test_defer_fail():
    def fun():
        log.msg("provoking NameError")
        rsdfg
    return defer.maybeDeferred(fun)
~/py.test.twisted/ time ./py.test.twisted test_some.py              ralf at redok
2008-01-29 07:01:14+0100 [-] Log opened.
============================= test process starts
==============================
executable:   /home/ralf/py25/bin/python  (2.5.2-alpha-0)
using py lib: /home/ralf/py25/lib/python2.5/site-packages/py-0.9.0-py2.5.egg/py
<rev unknown>

test_some.py[2] .F

________________________________________________________________________________
_________________________ entrypoint: test_defer_fail
__________________________

    def fun():
        log.msg("provoking NameError")
E       rsdfg
>       NameError: global name 'rsdfg' is not defined

[/home/ralf/py.test.twisted/test_some.py:20]
- - - - - - - - - - -  test_defer_fail: recorded stdout - - - - - - - - - -
- -
2008-01-29 07:01:39+0100 [-] provoking NameError

________________________________________________________________________________
============= tests finished: 1 passed, 1 failed in 25.00 seconds
==============
2008-01-29 07:01:39+0100 [-] Main loop terminated.
./py.test.twisted test_some.py  0.22s user 0.06s system 1% cpu 25.278 total
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/pytest-dev/attachments/20080129/b2f4e71c/attachment.html>

From exarkun at divmod.com  Tue Jan 29 15:09:07 2008
From: exarkun at divmod.com (Jean-Paul Calderone)
Date: Tue, 29 Jan 2008 09:09:07 -0500
Subject: [py-dev] [TIP]  towards 1.0 / code URLS using py.test?
In-Reply-To: <932f8baf0801282202s1f9fa7e3l717e100d41f08dda@mail.gmail.com>
Message-ID: <20080129140907.6859.19712609.divmod.quotient.2293@ohm>

On Tue, 29 Jan 2008 07:02:28 +0100, Ralf Schmitt <schmir at gmail.com> wrote:
>On Jan 28, 2008 4:09 PM, Jean-Paul Calderone <exarkun at divmod.com> wrote:
>
>> [snip]
>>
>> Is it intentional that this integration busy-loops, guaranteeing 100% CPU
>> usage for the entire test-run?
>>
>
>Why do you think so? Apparently this isn't true on my machine:
>
> [snip]
>
>2008-01-29 07:01:39+0100 [-] Main loop terminated.
>./py.test.twisted test_some.py  0.22s user 0.06s system 1% cpu 25.278 total
>

You're right, of course.  I misread the way Deferreds were handled.  Sorry
for the noise. :)

Jean-Paul


From rpk at mac.com  Thu Jan 31 22:29:35 2008
From: rpk at mac.com (Bob Kuehne)
Date: Thu, 31 Jan 2008 16:29:35 -0500
Subject: [py-dev] py.execnet: hang in channel.receive()
Message-ID: <FDEBD443-B48F-4188-AC1D-259B9C1F89A3@mac.com>

hi gang,

i first saw holger present py.execnet at pycon a few years back, and  
have been waiting to use this cool tool. so now i'm on it, and i've  
got a problem. i'm running this on two macs, both running os x 10.5.1.  
i've installed py via easy_install, so i've got a luverly egg, and  
here's my script that i'm trying to just test with:

--
import py
gw = py.execnet.SshGateway( "localhost" )
channel = gw.remote_exec( 'channel.send( [ [ False, True, False,  
0, ] ] )' )
print channel.receive()
channel.close()
gw.exit()
--

so, on mac numero 1, running python 2.5.1, i see the output i'd expect:

--
Python 2.5.1 (r251:54863, Oct  5 2007, 21:08:09)
[GCC 4.0.1 (Apple Inc. build 5465)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import py
 >>> gw = py.execnet.SshGateway( "localhost" )
 >>> channel = gw.remote_exec( 'channel.send( [ [ False, True, False,  
0, ] ] )' )
 >>> print channel.receive()
[[False, True, False, 0]]
 >>> channel.close()
 >>> gw.exit()
--

but on another mac, also python 2.5.1, i see the same, but the  
'channel.receive()' is hung. if i break there (control-c), i get the  
following:

--
Python 2.5.1 (r251:54863, Oct  9 2007, 20:39:46)
[GCC 4.0.1 (Apple Inc. build 5465)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
 >>> import py
 >>> gw = py.execnet.SshGateway( "localhost" )
 >>> channel = gw.remote_exec( 'channel.send( [ [ False, True, False,  
0, ] ] )' )
 >>> print channel.receive()
^CKilled by signal 2.
Traceback (most recent call last):
   File "<stdin>", line 1, in <module>
   File "/Library/Python/2.5/site-packages/py-0.9.0-py2.5.egg/py/ 
execnet/channel.py", line 172, in receive
     x = queue.get()
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/Queue.py", line 165, in get
     self.not_empty.wait()
   File "/System/Library/Frameworks/Python.framework/Versions/2.5/lib/ 
python2.5/threading.py", line 214, in wait
     waiter.acquire()
KeyboardInterrupt
--

now, the question is what's different. and it's at this point that i  
wave my arms and go, wtf? i really don't know, but i'm wondering aloud  
if anyone else has seen this sort of behavior, and if so, what they  
might have encountered or changed to get it to be better behaved.

thanks,
bob