On Wed, 2010-01-13 at 13:20 +0000, Tim Foster wrote:
> 'pargs -e' on run.py and the coverage processes seemed to show the
> correct python path though and I haven't been able to track down what's
> happening yet having dug around the tests/ directory a bit.
I should also say, pfiles on the same processes show them loading stuff
from /usr/lib/python2.6/... so that does certainly explain why we're
seeing the error, just not the cause.
I dug a bit deeper, and found that coverage itself is actually screwing
about with sys.path. Commenting out those lines from
proto/root_i386/usr/lib/python2.6/vendor-packages/coverage/execfile.py
caused tests to pass (!)
It shouldn't be clobbering sys.path[0], it should be inserting that into
the path ahead of what's there I think.
cheers,
tim
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
