[Stackless] Still getting crashes with PySide

Christian Tismer tismer at stackless.com
Fri Dec 7 18:53:13 CET 2012


Hi Lars,

what exactly did you load?

The patch from Anselm is from 2012-09-28.

python-2.7.2-stackless.msi is still from 2011-12-12.
As far as I can see, all installers and binary packages are older than 
the patch.

Eek - you would need to build from source I think.

ciao - chris

On 07.12.12 15:48, lars van Gemerden wrote:
> I am more then happy to help solve this problem, but not sure exactly 
> how. For example 
> C:\Python27\Lib\site-packages\PySide\examples\mainwindows\sdi\sdi.py 
> or any other example i tried does not crash on exit for me (at least 
> not now). My own software does not crash on exit either (mostly/only 
> on start-up, but very unpredictably).
>
> Did anyone try the 
> C:\Python27\Lib\site-packages\PySide\examples\mainwindows\sdi\sdi.py 
> example with PyQt?
>
> I could make a version of my own code with PyQt, but i wouldn't know 
> what to test for since sometimes the suspect crashes don't appear for 
> days or weeks.
>
> One question, i read about Anselms patch above. Is it included in the 
> windows distributable on http://www.stackless.com/wiki/Download?
>
> Cheers, Lars
>
> PS: I just downloaded the latest version from the wiki ande 
>  repaired-> no crashes yet :-)
>
>
> On Sun, Dec 2, 2012 at 12:12 PM, lars van Gemerden 
> <lars at rational-it.com <mailto:lars at rational-it.com>> wrote:
>
>     Hi Chris,
>
>     "One thing would be interesting but involves tome work:
>     Can you install PyQt instead and try the same?"
>
>     I am considering whether i could do so efficiently, setting up my
>     libs etc. isn't really my strong suit and i have quite a complex
>     program.
>     Is there some symptom, test, etc. that could help me determine
>     whether a crash is due to the stackless/PySide combination? Maybe
>     the .dll that crashed? In that case I could switch when i am
>     reasonably sure that the PySide code actually triggers the crash.
>     Otherwise i'd have to switch, continue working with PyQt and if no
>     seemingly related crash occurs for say 2 weeks, report back, which
>     seems a bit roundabout.
>
>     Would it be useful if i send the windows error (APPCRASH ... etc.)?
>
>     BTW: Most of the crashes i suspect of being related happen during
>     program startup, so before any user interaction.
>
>     Cheers, Lars
>
>
>     On Sat, Dec 1, 2012 at 3:56 AM, Christian Tismer
>     <tismer at stackless.com <mailto:tismer at stackless.com>> wrote:
>
>         Hi Lars,
>
>         this is good information!
>         Since you don't get crashes from using stackless features, it is
>         pretty clear that the problems must come from things that
>         I changed to support stackless features.
>
>         Good to know.
>         Still, it would be nice if we had something more deterministic to
>         cause a crash, especially a crash that occurs without manual
>         interaction...
>
>         I think to have a closer look into PySide itself, to see if it
>         uses
>         structures which are not officially public.
>
>         One thing would be interesting but involves tome work:
>         Can you install PyQt instead and try the same?
>
>         thanks & cheers - chris
>
>
>
>         On 30.11.12 11:56, lars van Gemerden wrote:
>>         I am also working with a combination of PySide 1.1.2 (same
>>         installer, but the 1.1.1 version had the same problem) and
>>         stackless 2.7 on windows and have to agree that it is pretty
>>         unstable. It is hard to find a pattern in the crashes and i
>>         do not even know whether i have the same problem, only that
>>         the crashes come and go, even when i do not change anything
>>         significant to the code. It also happens when i do not call
>>         any code that uses stackless explicitely (no tasklets are
>>         created in the portion of code that is actually run, apart
>>         from the main one I guess). Sometimes i get "unknown opcode"
>>         exceptions; could that be related?
>>
>>         It would be very helpful and comforting for the future of my
>>         project if this were resolved. If I can help, let me know ...
>>
>>         Cheers, Lars
>>
>>
>>         On Fri, Nov 23, 2012 at 9:49 PM, Richard Tew
>>         <richard.m.tew at gmail.com <mailto:richard.m.tew at gmail.com>> wrote:
>>
>>             On Sat, Nov 24, 2012 at 7:10 AM, Christian Tismer
>>             <tismer at stackless.com <mailto:tismer at stackless.com>> wrote:
>>             >
>>             C:\Python27\Lib\site-packages\PySide\examples\mainwindows\sdi\sdi.py
>>             >
>>             > Richard, can you please tell me the version of Qt and
>>             PySide?
>>             >
>>             > I will try that on OS X, first, because I'm busy with
>>             building PySide
>>             > anyway. What did you install, a win installer package,
>>             a pip thing
>>             > or from source?
>>
>>             I installed PySide-1.1.2.win32-py2.7.exe which I
>>             downloaded from the
>>             PySide website.  It uses QT 4.8 going by the PyPI page.
>>
>>             Yes, sdi.py, like most examples crashes on exit for me.
>>              But every
>>             example I recall trying crashed in much the same way.
>>

-- 
Christian Tismer             :^)   <mailto:tismer at stackless.com>
Software Consulting          :     Have a break! Take a ride on Python's
Karl-Liebknecht-Str. 121     :    *Starship* http://starship.python.net/
14482 Potsdam                :     PGP key -> http://pgp.uni-mainz.de
phone +49 173 24 18 776  fax +49 (30) 700143-0023
PGP 0x57F3BF04       9064 F4E1 D754 C2FF 1619  305B C09C 5A3B 57F3 BF04
       whom do you want to sponsor today?   http://www.stackless.com/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stackless.com/pipermail/stackless/attachments/20121207/373a34eb/attachment.html>


More information about the Stackless mailing list