Update to jupyterlab 4 #110
Conversation
| "@types/fscreen": "^1.0.1", | ||
| "@types/node": "^10.11.6", | ||
| "@types/three": "^0.141.0", | ||
| "fscreen": "^1.2.0", | ||
| "ipycanvas": "^0.8.2", | ||
| "ipycanvas": "^0.13.3", |
There was a problem hiding this comment.
needed to bump this to get the canvas widget to properly compile
| "jupyterlab~=3.1", | ||
| "hatchling>=1.21.1", | ||
| "jupyterlab>=4.0.0,<5", | ||
| "hatch-nodejs-version>=0.3.2", |
There was a problem hiding this comment.
migration script missedhatch-nodejs-version, but the cookiecutter has it, so added it.
| @@ -50,12 +50,16 @@ Tracker = "https://github.com/yt-project/widgyts/issues" | |||
|
|
|||
| [tool.hatch.build] | |||
| artifacts = [ | |||
| "widgyts/labextension/*.tgz", | |||
There was a problem hiding this comment.
do we want this? not sure. the cookiecutter has it.
| @@ -50,7 +50,7 @@ export class FullscreenButtonView extends DOMWidgetView { | |||
| this.el.classList.add('jupyter-widgets'); | |||
| this.el.classList.add('jupyter-button'); | |||
| this.el.classList.add('widget-button'); | |||
| this.delegate('click', this.el, this._handle_click); | |||
| this.delegate('click', this.el.value, this._handle_click); | |||
There was a problem hiding this comment.
got a compilation error without this complaining about assigning a string output to a HTMLButtonElement, based on some google/SO searching, seemed like this would work. and it does now compile. though just realized I didn't test the button, will go do that now ...
There was a problem hiding this comment.
hmm. ok, ya, i don't even see the fullscreen button any more, so i broke something here.
There was a problem hiding this comment.
I don't think it totally works yet. So let's not worry about this for this PR and fix it at a later time. If I can figure it out I'll update your PR.
| - name: jlpm config | ||
| run: jlpm config set -H enableImmutableInstalls false |
There was a problem hiding this comment.
Can't work out why, but enableImmutableInstalls was set to true in the CI environment, which causes an error when there is no yarn.lock file on install:
YN0028: The lockfile would have been created by this install, which is explicitly forbidden.
but the yarn file should be created here. so setting it to false seems correct. just not sure what caused the change in the CI test enviornment in this PR.
There was a problem hiding this comment.
We may want to commit our yarn.lock file. I'm not sure if it's a universally-accepted practice for jupyterlab extensions ... but maybe I'm just out of the loop or woefully ignorant?!
There was a problem hiding this comment.
ya, i'm not sure either -- maybe we should be committing it? but I can research that separate from this PR.
| "@types/fscreen": "^1.0.1", | ||
| "@types/node": "^10.11.6", | ||
| "@types/three": "^0.141.0", | ||
| "fscreen": "^1.2.0", | ||
| "ipycanvas": "^0.8.2", | ||
| "ipycanvas": "^0.13.3", | ||
| "jupyter-threejs": "^2.3.0" | ||
| }, | ||
| "devDependencies": { |
There was a problem hiding this comment.
these deps were copy/pasted from the current cookiecutter, which is configured for jupyterlab 4. I did remove the jest and babel deps that are in the cookiecutter, because we're not using them anywhere (and don't need to as far as I can tell)
There was a problem hiding this comment.
So there's a possibility that I added them when wasm-bundling wasn't as straightforward as it is now.
This PR bumps us up to jupyterlab 4.
I went through the current cookiecutter (https://github.com/jupyter-widgets/widget-ts-cookiecutter), comparing dependencies and config files and fixed a few compilation errors.
This now compiles locally in a fresh environment for me. It also seems to fix the issue i was having with the canvas viewer in firefox! I can now click and drag without issue (EDIT: maybe it doesn't... freezing again?)
I'm not positive all the changes here are necessary. will go throuhg the PR and annotate with some comments.
Also note this is based off of (and blocked by) #109