-
-
Notifications
You must be signed in to change notification settings - Fork 7.9k
Fix issue with Sphinx 1.3.4 #5872
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Looks like there is another issue with Sphinx 1.3.4 I will take a look tomorrow
|
It looks like that was as simple as removing the :option: key |
I think the docs changes should be backported to 1.5.2 but the changes to Travis might not be wort it to backport |
BTW it would be good to convert the rest of that file to numpydoc/napoleon too |
There seems to be an issue with downloading the Cursive font from Googles github account. I should fix that before merging |
281c2e8
to
92f8c06
Compare
Ensures that we always get the latest version so we spot issues early. The speedup of the wheelhouse is minimal for these packages which are mainly pure python anyway and we are useing the wheel cache for the ones that are not wheels
…with sphinx 1.3.4
92f8c06
to
7f37443
Compare
Looks like the google fonts are back on github so this is ready for review |
DOC: Fix issue with Sphinx 1.3.4
I will cherry pick this back but discard the travis changes |
DOC: Fix issue with Sphinx 1.3.4 Discarded changes to .travis.yml from #5872
backported to 1.5.x as b7dfa4f |
DOC: Fix issue with Sphinx 1.3.4 Discarded changes to .travis.yml from matplotlib#5872
Fixes #5869 by rewriting the offending doc string as numpy doc
I also updated .travis.yaml. This originally started with the aim to ensure that the latest version of Sphinx and other dependencies is always installed by only installing Numpy and Pandas from the wheel house. (The wheel house requires manually rebuilding the wheel when a new version is released to pypi and was thus using an older version of Sphinx) The speed benefits are limited since other packages than numpy and pandas are fast to install and we now have the pip wheel cache on Travis anyway.
I took the opportunity to refactor the Travis config a bit more to collect all the pip installs one place and group relevant sections.