Skip to content

Conversation

@jtpio
Copy link
Member

@jtpio jtpio commented Nov 19, 2025

https://github.com/jupyterlite/jupyterlite/releases/tag/v0.7.0rc0

  • Update @jupyterlab and @jupyterlite packages
  • Use @jupyterlite/services

@github-actions
Copy link
Contributor

lite-badge 👈 Try it on ReadTheDocs

@jtpio jtpio added this to the 0.7.0 milestone Nov 19, 2025
pyproject.toml Outdated
dev = [
"build",
"hatch",
"jupyterlab >=4.5.0rc0,<4.6.0",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

>=4.5.0,<4.6.0

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

yep, coming

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's kind of a bummer there's no way to reference build-system from optional-dependencies (or the new shiny dependency-groups)

import { KernelMessage } from '@jupyterlab/services';

import { DriveFS } from '@jupyterlite/contents';
import { DriveFS } from '@jupyterlite/services';
Copy link
Contributor

@bollwyvl bollwyvl Nov 19, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

oof, this increases the worker bundle size by 50%. Perhaps a deeper lib/ path would shake better?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What numbers are you seeing?

[tool.jupyter-releaser.hooks]
before-bump-version = [
"python -m pip install 'jupyterlab~=4.5.0rc0'",
"python -m pip install 'jupyterlab~=4.5.0'",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

we actually could start using dependency-groups for this... or at least, that's my understanding (we don't want to install the package-under-dev).

@jtpio jtpio marked this pull request as ready for review November 19, 2025 16:07
@jtpio
Copy link
Member Author

jtpio commented Nov 19, 2025

@bollwyvl do you think these should be addressed in this PR? Or can we look into them separately, maybe before the 0.7.0 final release?

@bollwyvl
Copy link
Contributor

@jupyterlite/pyodide-kernel:   lib/coincident.worker.js      315.4kb
@jupyterlite/pyodide-kernel:   lib/coincident.worker.js.map    1.4mb
@jupyterlite/pyodide-kernel: ⚡ Done in 64ms
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js      316.8kb
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js.map    1.4mb
@jupyterlite/pyodide-kernel: ⚡ Done in 59ms
@jupyterlite/pyodide-kernel:   lib/coincident.worker.js      227.5kb
@jupyterlite/pyodide-kernel:   lib/coincident.worker.js.map    1.0mb
@jupyterlite/pyodide-kernel: ⚡ Done in 47ms
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js      228.9kb
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js.map    1.0mb
@jupyterlite/pyodide-kernel: ⚡ Done in 48ms

@bollwyvl
Copy link
Contributor

... and again, have tried a whole mess of things to get the workers to play nice-nice with module federation so this wasn't even an issue. i had something that kinda worked with a nasty mess:

  • generate a separate bundle entry point per worker
  • push the bundle text blob over the postmessage pipe
  • generate dynamic classes that extend stuff that come out of it the bundle

... but i hated it 👿

@jtpio
Copy link
Member Author

jtpio commented Nov 19, 2025

From https://github.com/jupyterlite/pyodide-kernel/actions/runs/19512304322/job/55854919639?pr=240

 > @jupyterlite/pyodide-kernel:"build:prod"

@jupyterlite/pyodide-kernel:   lib/coincident.worker.js       33.8kb
@jupyterlite/pyodide-kernel:   lib/coincident.worker.js.map  234.1kb
@jupyterlite/pyodide-kernel: ⚡ Done in 57ms
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js      179.6kb
@jupyterlite/pyodide-kernel:   lib/comlink.worker.js.map  929.3kb
@jupyterlite/pyodide-kernel: ⚡ Done in 58ms

It seems it was mostly related to the dynamic import.

@jtpio jtpio requested a review from bollwyvl November 19, 2025 18:37
Copy link
Contributor

@bollwyvl bollwyvl left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

shook down all the reasonable notebooks from upstream on RTD, a couple at a time, and didn't see any flake (though it didn't somehow fix sqlite 😿 )... LGTM!

@jtpio jtpio merged commit 6a748d7 into jupyterlite:main Nov 20, 2025
20 of 21 checks passed
@jtpio jtpio deleted the lite-070rc0 branch November 20, 2025 08:42
@bollwyvl bollwyvl mentioned this pull request Nov 20, 2025
2 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants