Fix tox.ini

[tkf]

This PR fixes tox.ini. I also suggest to use tox in Travis CI, to prevent from breaking it.

[pelson]

We've lost an -sv here.

[tkf]

Fixed.

[pelson]

Thanks @tkf.

[mdboom]

Looks good. I'd like to hold off on merging this until post 1.3.0rc, since now's probably not a good time to mess with a known-working CI setup.

[dmcdougall]

This needs a rebase before merging. It looks like that tight_bbox test failure is not indicative or problems here, nonetheless a rebase should trigger a re-test anyway.

@tkf Would you be able to do that?

[tkf]

@dmcdougall OK. I'll rebase.

[tkf]

I rebased. Travis CI works except for Python 3.3. I don't know why it failed.

[mdboom]

Yes -- the 3.3 failure is odd. The output seems to just suddenly stop...

[dmcdougall]

I reran py33 and I have problems with distribute. I then remembered this branch was made before the new distribute patch was merged.

@tkf Can you rebase this again so I can rerun the py33 travis job?

[dmcdougall]

I fetched your branch to do some testing myself. I just noticed that each of the test runs is not respecting the correct environment. Notice that python --version is returning 2.7.3 regardless of which py{26,27,32,33} environment is used.

@tkf Do you know how to make Travis respect the correct tox environment?

[dmcdougall]

Oh it's only using python 2.7 to pip install tox. Disregard my previous comment.

[dmcdougall]

Ok, it looks like tox is using and old version of distribute. Can anyone with actual tox experience confirm/deny this?

[tkf]

@dmcdougall You can use this option to use distribute instead of setuptools: http://tox.readthedocs.org/en/latest/config.html#confval-distribute=True|False

[tacaswell]

@tkf Is this still a live PR? If so, it needs a re-base.

[tacaswell]

Closing due to inactivity and the fact that we have striped tox out for just about every where else.