Skip to content

Conversation

@merryman
Copy link
Member

@merryman merryman commented Feb 7, 2025

Can be merged once #1704 is merged.

@linusha linusha force-pushed the feature/source-map-support branch from 2c0c30a to df6e577 Compare February 24, 2025 11:26
@merryman merryman marked this pull request as ready for review February 24, 2025 12:55
@merryman merryman requested a review from linusha February 24, 2025 14:27
@merryman merryman force-pushed the feature/source-map-support branch from df6e577 to a0c0f2a Compare June 30, 2025 12:48
Copy link
Contributor

@linusha linusha left a comment

Choose a reason for hiding this comment

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

Alright, I took the liberty of removing some (very few) unused lines of code.
I also raised some points that I found to be concerning below.

I checked the tests, one fails:
Screenshot from 2025-07-09 18-31-32

Additionally, I developed a bit inside of lively and everything worked. I also checked building a project (the website), which worked. However, one thing on the website is off (bottom most picture):
image

If you are sure this is unrelated to these changes, that's also fine with me. We should however note that somewhere and fix it.

Pending the issues raised below and the failing test, this seems good to me, @merryman. Any other pointers as to where I should apply some pressure to check whether this is good to go?

Comment on lines +15 to +17
"rollup": "4.27.3",
"@rollup/rollup-linux-x64-gnu": "4.27.3",
"@rollup/rollup-darwin-arm64": "4.27.3",
Copy link
Contributor

Choose a reason for hiding this comment

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

Do we risk problems here on windows machines?

Copy link
Member Author

Choose a reason for hiding this comment

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

We may better test that manually on our trusty windows setup, but doesn't windows use this obscure linux subsystem which should be fine with linux-x64?

Copy link
Contributor

Choose a reason for hiding this comment

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

That might be how it works, I don't remember anymore from the top of my head honestly, what happens there under the hood. I'll note that down and test it eventually though!

@merryman
Copy link
Member Author

merryman commented Jul 22, 2025

Alright, I took the liberty of removing some (very few) unused lines of code. I also raised some points that I found to be concerning below.

I checked the tests, one fails: Screenshot from 2025-07-09 18-31-32

Interesting, it appears this test only crashes in fast loaded environments. I will look into that.
This has now been fixed.

Additionally, I developed a bit inside of lively and everything worked. I also checked building a project (the website), which worked. However, one thing on the website is off (bottom most picture): image

This issue appears on main as well, so i would say it is unrelated to these changes.

@merryman merryman force-pushed the feature/source-map-support branch from 15eeb91 to 0ed2d5f Compare July 22, 2025 14:51
@merryman
Copy link
Member Author

Pending the issues raised below and the failing test, this seems good to me, @merryman. Any other pointers as to where I should apply some pressure to check whether this is good to go?

Verifying the debugging with source-maps works OK, would be a good test ^^.

merryman added 19 commits July 23, 2025 15:58
Merge the var and ref replacement and the import export traversal into single passes.
The key is here to make the node functions parametrized, so that depending on the underlying AST system different nodes can be created by the same transform.
Was needed due to to the changes in the class transform that preserves the class name as a binding during class initialization.
Since babel.js has a bunch of circular imports it has weird importing specific code that checks if modules have been already initialized. We introduce a method called bulletProofNamespace() which helps us to adjust these error checking codes such that they work in split bundles.
The transpiler initialization needed to be adjusted.
@merryman merryman force-pushed the feature/source-map-support branch from 0ed2d5f to 4a40a38 Compare July 23, 2025 13:58
@linusha linusha force-pushed the feature/source-map-support branch from 4a40a38 to 63f0d90 Compare July 24, 2025 13:05
Copy link
Contributor

@linusha linusha left a comment

Choose a reason for hiding this comment

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

Verifying the debugging with source-maps works OK, would be a good test ^^.

I did that! And it works :)

I will keep testing this on a windows machine on my list and hopefully do that next week. Regardless, my proposal would be to continue merging this for now, and afterwards also merging #1711.

@merryman merryman merged commit a3a56c3 into main Jul 24, 2025
3 checks passed
@merryman merryman deleted the feature/source-map-support branch July 24, 2025 15:03
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants