Update scikit-learn website links to use https instead of http #56
Conversation
|
@justinormont in response to your question on Teams, I'd like to figure out ways to improve the docs process so that it is less complicated. In particular, all the .rst files in our source repo are copies of files in the official docs repo (copied and converted from .md to .rst), so the changes Gal made to those files won't have any effect. It may be worthwhile to completely remove those files from the source repo since they are just copies - what do you think? They originally existed before the creation of the docs repo, and now we maintain them because it is the typical python convention to keep docs files in the source repo in the structure that the Sphinx documentation generator expects. So we currently keep two copies of files so that we are compliant with both Microsoft Docs and Sphinx conventions, but we don't currently use Sphinx in any way. There is something nice about the Sphinx convention though - all docs and source are in the same repo (not possible with Microsoft Docs). Up until a month ago, we followed the Sphinx conventions by default, as if there were no Microsoft Docs, and then had a designated person who would do a manual step to produce the copies and build for Microsoft Docs. But last month I proposed that we reverse this process in order to make it possible for people to edit the docs site simply by clicking the edit button on the webpage (we had some requests for this from Chris Lauren and the content developers). Now, we Given all of that, should we let go of the Sphinx compliant copies to make things easier? Are there any other ways we can improve on this system? |
Updating links to the scikit-learn website to use https instead of http.
This fixes #52.
Additionally, use the opportunity to update scikit-learn references in readme to lowercase to follow convention.
There are other http links that we might want to update in the future.