You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Actually I'm not sure why we provide the intersphinx mapping in here for so many optional dependencies but missed some others (e.g. why h5py but not pandas)?
I suggest to only keep python, numpy and astropy here, and the rest can go into docs/conf.py for the package repositories. If we keep it however, consider adding more mappings.
The text was updated successfully, but these errors were encountered:
I'd be more inclined to include pandas than remove existing ones. If we remove ones we'll probably have to create v2 of the default configuration, and I'd rather not create new configurations lightly.
I'm actually fine by adding more, pandas, ipython and pytest that core also uses. From those probably pandas is the only one useful for multiple packages. I would maybe consider adding scikit-image, too.
Also, at some point when we have the docs set up, I think it would make sense to have it there for all the astropy coordinated packages, too (though I'm not sure how sphinx behaves when it has the mapping of a released version for a package it's actually building).
Actually I'm not sure why we provide the intersphinx mapping in here for so many optional dependencies but missed some others (e.g. why
h5py
but notpandas
)?I suggest to only keep python, numpy and astropy here, and the rest can go into
docs/conf.py
for the package repositories. If we keep it however, consider adding more mappings.The text was updated successfully, but these errors were encountered: