[Stackless] Pre 2.5.2 merge / testing help appreciated

Richard Tew richard.m.tew at gmail.com
Mon Feb 4 16:16:45 CET 2008


On Feb 4, 2008 10:08 AM, Jeff Senn <senn at maya.com> wrote:
> Seems to work on OS-X (10.5.1) -- with a few tiny caveats which I'll
> mention
> for the record, but don't seem terribly relevant to anyone not trying
> to build
> on OS-X.
>
> There were a few problems building it using the build script - but I
> attribute
> that to a non-released version (e.g. one problem was fetching the html
> for doing
> a documentation build from the python site, another was fetching a
> particular version
> of sqlite).
>
> The 2 test weirdnesses were:
>
>    test_subprocess now triggers the OS-X "the Application Unexpectedly
> Quit" dialog
>    box that the system pops up when a process (that is considered an
> "Application") dies for certain reasons.
>    This could be completely normal due to how the test is terminating
> the process.
>
>    test_socket failed once.  But I have no more details since it now
> succeeds every time
>    I run it and it seemed innocuous, so I'm sticking my head back into
> the sand on that one.
>
>       File "/Library/Frameworks/Python.framework/Versions/2.5/lib/
> python2.5/test/test_socket.py", line 122, in _tearDown
>      self.fail(msg)
> AssertionError: (57, 'Socket is not connected')

The key thing to worry about for Stackless, is, do these things happen
when you disable Stackless and compile the same codebase?

Compile the Stackless code base with the define STACKLESS_OFF set to
something, or edit Stackless/stackless.h and comment out the #define
STACKLESS.  Then run the standard Python tests and see if they still
happen.

Cheers,
Richard.




More information about the Stackless mailing list