jupyter/enhancement-proposals

Discussion about Improving the Notebook Experience for Text-based Workflows

aronchick opened this issue · 4 comments

Hi!

Based on community feedback, I've taken a Google Doc and converted it to markdown so that we can have a discussion here. I think it's too soon to call it a JEP, is more a discussion before the Pre-Submission to a JEP.

I'd love to:

  1. Find a place to submit it here so that people can talk about it using Git commenting/flow
  2. Submit the PR so that I'm not the blocking factor on it and people can submit issues or submit new PRs from a branch instead of just having me respond to comments.

After a JEP shepherd has been identified, I'd prefer to have her/him take over the process.

Can anyone recommend what I should do here?

For those not following along what happens in the forum https://discourse.jupyter.org/t/jupyter-and-github-alternative-file-format/4972 is the thread that resulted in the MD document. It is a (very) long thread, but it contains a lot of interesting points. It is probably worth reading it before commenting on the MD as your idea/thought might have already been mentioned there.

Thanks @betatim!

I'm happy to do a PR and submit the markdown, but I wasn't sure where you all would like it. Just let me know!

I don't know what is in the gdoc right now but it did contain a table towards the bottom with "features" as columns and existing solutions as rows.

I think a good first thing to discuss and agree on is a list of features (seeded by the one in the gdoc) which sets constraints for the second step which is discussing existing formats and the third step of discussing how to modify existing solutions/create completely new solutions.

To pick a random one: is "a notebook is one file" a useful constraint to add under the premise of "improve notebook experience for text-based workflows"? I think agreeing on some of these will help constraint discussions to a finite solution space, which would be a good thing (tm). We can always iterate them but without any guard rails I'd expect a discussion to veer off course.

This issue has been mentioned on Jupyter Community Forum. There might be relevant details there:

https://discourse.jupyter.org/t/proposed-jep-investigate-alternate-optional-file-formats/5073/14