0.1.0 Release Plan

This is an issue for the 0.1.0 release planning.

We still have the MVP board for the things we wanted to tackle: https://github.com/jupyterlite/jupyterlite/projects/1 (although this was mostly before we have the jupyterlite CLI)

Having this in an issue makes it more visible and it can also be pinned.

Dates

There will be a talk at PyData Global 2021 about JupyterLite: https://pydata.org/global2021/schedule/presentation/113/jupyterlite-jupyter-webassembly-python/

Maybe we can aim for a 0.1.0b0 just before the event, somewhere around Wednesday 27 October. We could then stay in beta for a little while, or just proceed with rc and then final one and two weeks after.

Issues

• https://github.com/jupyterlite/jupyterlite/pull/371
• https://github.com/jupyterlite/jupyterlite/pull/310
• https://github.com/jupyterlite/jupyterlite/issues/382
• A look at the issues tagged with bug

Nice to have

• Not strictly related to the main jupyterlite repo, but it would be great to have Xeus kernels packaged and release:
  • https://github.com/jupyterlite/xeus-kernel for the common package other xeus kernels can depend on
  • https://github.com/DerThorsten/jupyterlite_xeus_lua (to be moved to the jupyterlite org later) for a xeus-based Lua kernel
  • https://github.com/DerThorsten/xeus-wren following the same approach for a different language

[bollwyvl]

I feel like a proper release would also divest this repo of the renamed pyodide kernel stuff (either before or after #310, which I do intend to bring home soon)… not because I dislike it or anything, but so that that side of the house is more free to make changes moving forward, and to make space for other python kernel implementations, if need be.

In addition to formalizing the ipylite packaging/API, really only #45 (which will be mostly covered by #310, minus some MathJax convenience stuff) is what’s holding back the robolite implementation, which is already one of the more impressive jupyterlite demos. We intend to pitch (part of) that for the upcoming RoboCon 2022.

[jtpio]

welp, seems like lab 3.4.0rc0 and pyodide 0.22.1a0 dropped in the last 24… seems worthy to grab those features.

yes these would be great to get in! And should hopefully be just a matter of bumping dependencies.

whether we want to lose the hey, you installed this with --pre, you get exactly what you asked for retort

• if folk start making more demands (without contributions) about how this project isn’t meeting their exact production needs, i don’t know how much fun it’s going to be to continue contributing

Right. Although a 0.1.0 can still be interpreted as very early. Also no problem to leave some kind of disclaimer in the docs or the README to mention things are still moving fast and likely to change in future releases. As always it’s an iterative process and folks making demands just have to deal with it, or help by making contributions.