Travis is broken on master

Calculating the recipes which need to be turned into feedstocks.
Making feedstock for nilearn
Traceback (most recent call last):
  File "/Users/travis/miniconda/bin/conda-smithy", line 9, in <module>
    load_entry_point('conda-smithy==0.10.0', 'console_scripts', 'conda-smithy')()
  File "/Users/travis/miniconda/lib/python3.5/site-packages/conda_smithy/cli.py", line 228, in main
    args.subcommand_func(args)
  File "/Users/travis/miniconda/lib/python3.5/site-packages/conda_smithy/cli.py", line 101, in __call__
    generate_feedstock_content(feedstock_directory, args.recipe_directory, meta)
  File "/Users/travis/miniconda/lib/python3.5/site-packages/conda_smithy/cli.py", line 32, in generate_feedstock_content
    configure_feedstock.main(target_directory)
  File "/Users/travis/miniconda/lib/python3.5/site-packages/conda_smithy/configure_feedstock.py", line 323, in main
    for key, value in file_config.items():
AttributeError: 'list' object has no attribute 'items'
Traceback (most recent call last):
  File ".CI/create_feedstocks.py", line 94, in <module>
    '--feedstock-directory', feedstock_dir])
  File "/Users/travis/miniconda/lib/python3.5/subprocess.py", line 584, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['conda', 'smithy', 'init', '/Users/travis/build/conda-forge/staged-recipes/recipes/nilearn', '--feedstock-directory', '/var/folders/gw/_2jq29095y7b__wtby9dg_5h0000gn/T/tmp1xgk4lu5__feedstocks/nilearn-feedstock']' returned non-zero exit status 1

[jakirkham]

Alright. Hopefully someone else can give it a look then.

I’m sure @pelson will look in the morning (UK time). He has pretty good insight into these sorts of things.

This bug is causing any merged PRs to not have their recipes moved out of staged-recipes/recipes.

Yes, like all services there are downtimes. As this is a community-based one that is in its early stages, patience is very important. Everyone here has day jobs that to varying extents do not involve this. That being said, I’m sure this will be resolved well within a day.

Now those recipes get built any time the CI’s run (glueviz, nilearn and supervisor for now) in addition to the recipes that are actually being tested and added in new PRs.

Actually this is easily fixed. Simply merge your PR’s branch with master and then git rm -r the recipes that need to go. After committing and pushing, I think you will find that everything builds fine in your PR without extra stuff.

[pelson]

It actually turned out to need a v0.10.3. Thank goodness for quick releases 👍 😄