Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Last three recipes have failed on merge into master

See original GitHub issue

Seems like we’ve hit an api limit in:

https://travis-ci.org/conda-forge/staged-recipes/builds/196331121

GitHub API Rate Limit Info:

---------------------------

Currently remaining 5000 out of 5000.

Will reset in 0:59:59.349455.

Calculating the recipes which need to be turned into feedstocks.

No local changes to save

Current branch master is up to date.

No stash found.

On branch master

Your branch is up-to-date with 'origin/master'.

nothing to commit, working directory clean

GitHub API Rate Limit Info:

---------------------------

Currently remaining 5000 out of 5000.

Will reset in 0:59:58.597998.

Issue Analytics

State:
Created 7 years ago
Comments:7 (7 by maintainers)

Top GitHub Comments

1reaction

minrkcommented, Jan 30, 2017

It looks like the ipympl recipe is in a folder called jupyter-matplotlib and build is failing because conda-forge/ipympl-feedstock was created, but a later part of the code is checking for and failing to find conda-forge/jupyter-matplotlib-feedstock.

I opened #2312 to fix this by renaming the jupyter-matplotlib folder to ipympl, so the two match, which should let things start building again. #2313 should fix the bug in earnest, but that should get more careful review, since I’m not as confident that there aren’t other points where a mismatch could cause breakdown.

0reactions

jakirkhamcommented, Feb 9, 2017

It is a good point @minrk, but it definitely slips below the radar. Really this should be handled programmatically. Either the linter should be catching the directory/package name mismatches as this is trivial to do programmatically and it is easy for human reviewers to miss or we should just always use the package name as the directory name shouldn’t really matter. Oddly enough I think adding this to the linter is easier.

xref: https://github.com/conda-forge/staged-recipes/issues/1073