Make test fails on time module with Python 1.5.2/Linux2/EGCS! Help!

Alexander Staubo nospam at mop.no
Thu May 13 21:15:45 EDT 1999


In article <3739E881.3ADC365B at lemburg.com>, mal at lemburg.com says...
> Neel Krishnaswami wrote:
> > 
> > In article <MPG.11a2a389f734af92989728 at news.online.no>,
> > Alexander Staubo <nospam at mop.no> wrote:
> > >After doing a "make; make test", I get a failure in testing test_time.py.
> > >Here's the output when I run the test on its own:
> > >
> > >tradecenter:/usr/src/python# python Lib/test/test_time.py
> > >Traceback (innermost last):
> > >  File "Lib/test/test_time.py", line 3, in ?
> > >    time.altzone
> > >AttributeError: altzone
> > >
> > >Maybe Python has a problem with EGCS? That said, the compilator does not
> > >seem to output any errors (I'd be only happy to forward the make output
> > >to anyone interested.) I have not tried making with plain GCC.
> > 
> > I can confirm that it's not limited to egcs, because I'm getting the
> > same error with gcc 2.7.2 and libc5. :) None of the module-level
> > variables in the time module are appearing for me, either.
> 
> Looks like the #ifdefs in Module/timemodule.c are buggy: the
[snip]

Thanks. That seemed to do the trick. I'm going to get all the latest 
files from the CVS, seems like a generally good idea.

In the meantime, however: I run the test and see that the thread test is 
being skipped. What's up with that?

(Note: I _think_ I am using libc5. I didn't set up the system, and I 
don't know how to tell which version I have. Suggestions?)
 
Alexander Staubo
http://www.mop.no/~alex/
mailto:redhand at mop.no




More information about the Python-list mailing list