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
I've seen discussions where people say images should be loaded from somewhere else. I kinda think people should be able to build and view the docs w/o any network access, and that the images should be included or built by the doc build process.
If the images are produced by code, they can be build by the doc builder, and form user guide we can refer to those images, like you do with sphinx-gallery example files.
Images can end up weighing a significant amount and cannot be removed from the git history once merged into main, resulting in an unnecessarily long git clone
I understand there may be issues related to serving objects coming from repositories (cc @mishig25@Pierrci)
It'll be the website repo which gets large, not a repo which is cloned by people, so clone time shouldn't be an issue, and in CI we can shallow clone which fixes the size issue.
Once in a while, if the website repo gets too large, we can collapse its history, since it's not important (ref: scikit-learn/scikit-learn#21171)
Some images are not rendered correclty in the documentation, and a process needs to be set regarding the contribution of assets to the repository.
The text was updated successfully, but these errors were encountered: