Closed joergbrech closed 1 year ago
Thanks for submitting your first pull request! You are awesome! :hugs:
If you haven't done so already, check out EBP's Code of Conduct and our Contributing Guide, as this will greatly help the review process.
Welcome to the EBP community! :tada:
Good catch - do you think we can deprecate kernelName or should we keep it?
I honestly don't know :shrug: I wasn't brave enough to deprecate it.
@stevejpurves would you be willing to provide guidance on whether we can remove/deprecate kernelName
?
I would keep it -- I believe that a kernel launch can fail if the (default) named kernel is not available on the server/service, and if for some reason a kernel named python3
is not there then that would block usage, even though the public binder may have that kernel available.
@stevejpurves so name
and kernelName
serve two different purposes then? Could you link to a doc (or provide briefly here) so that we can cross-link in the sphinx-thebe docs to help others understand when/how to use?
ah looking at the comment in the linked issue, it seems that name
is the new correct option after thebe 0.6, and kernelName
is an old / deprecated option:
Does that sound correct to you @stevejpurves ?
I am not entirely sure, but I believe this could fix https://github.com/executablebooks/jupyter-book/issues/1173 also (with a sphinx-thebe release and a bump in the jb dependencies).
Yep i think you are right. Let's cross link that issue in the PR as well so that it closes it. We can make a release quickly!
@choldgraf I can't actually find any reference to the two existing side by side outside of thebe
.
AskernelName
has no effect in the current thebe release, and as thebe is moving to not using the Juptyerlab sessions at all - it looks like deprecating kernelName
would be ok here after all, and as a breaking change make it clearer for sphinx-thebe
users to switch to /usename
.
In the meantime, thebe 0.8.x
needs an update to provide the backwards compatbility
Ah ok, thanks for this clarification - I'd then suggest:
make kernelName rename itself to name if given, and raise a warning.
@choldgraf I am afraid I don't understand enough about the interplay of thebe/sphinx/jupyter etc. What do you mean by this? Doesn't sphinx-thebe just parse thebe-kernel
from the Markdown metadata and copy it to the thebe html config? Where would kernelName
be given, so that it can be renamed?
@joergbrech ah I'm being silly - kernelName
is chosen by sphinx-thebe
, not the user. I think we can just deprecate the use of kernelName
entirely by deleting that line. I added comments to make that clearer.
I think we should still open an issue about the soon-to-be-deprecated-again API. @stevejpurves would you be willing to explain in a new issue what we'll need to change?
@choldgraf yes I will - how about you go ahead and release this change first, then I can sit down and test the jb|sphinx-thebe|thebe together and make sure I have a clear picture.
@choldgraf just a friendly ping. Are we ready to merge?
Congrats on your first merged pull request in this project! :tada:
Thank you for contributing, we are very proud of you! :heart:
Yep - thanks for the ping. If any of you are interested in having merge rights on this repository, I am happy to provide them. I clearly do not have enough time to maintain this in addition to all the other stuff I have to do, and I don't want to block progress on myself!
Yep - thanks for the ping. If any of you are interested in having merge rights on this repository, I am happy to provide them. I clearly do not have enough time to maintain this in addition to all the other stuff I have to do, and I don't want to block progress on myself!
Thanks, but as this is just a small side project, I don't see myself contributing often. Also I know too little about the whole jupyter-sphinx-thebe ecosystem to be of much value. Once this is in jupyter book I will be happy for a while 🙂
Fixes #59.
See also https://github.com/executablebooks/thebe/issues/349.
Follow-up question: Can we have some kind of doctests verifying that all code blocks in the documentation don't error?
Edit from @choldgraf : i think this also closes #43 and closes #53
Edit: Closes https://github.com/executablebooks/jupyter-book/issues/1173