[Python-Dev] red buildbots on 2.7

["Martin v. Löwis"]

> Bill listed several other failures he saw on the buildbots and I see the
> same set, plus test_posix.

Still, the question would be whether any of these failures can manage to 
block a release. Are they regressions from 2.6? That would make them 
good candidates for release blockers. Except that I still would like to 
see commitment from somebody to fix them or else they can't block the 
release: if "we" don't mean that supporting a platform also means 
volunteering to fix bugs, then I guess "we" should stop declaring the
platform supported. Just wishing that it was supported actually doesn't 
make it so.

If the test failure *isn't* a regression, I think it shouldn't block the 
release.

Regards,
Martin