[GSoC 2025] Requesting Feedback on Proposal

1

Hi everyone,

I’ve been working on a draft proposal for the CodeWorld GHC Update project, and I’d really appreciate any feedback from the community. You can find the proposal here.

I haven’t been able to get in touch with @cdsmith yet (pinging just in case), so I would appreciate if you can provide feedback since the deadline is frighteningly close. Feel free to leave comments directly on the doc, reply here, or DM me if that’s easier.

Thanks for your time—looking forward to contributing!

Regards,
Arpit Yadav

3 Likes
2

Seems uncontroversial. I don’t see a discussion of risks or possible regressions. Are there any functional or non functional downsides to using WASM?

3

Thanks for your time! I chose to focus on the JavaScript backend since I have some prior experience with it, unlike WASM. That said, @ dmjio:matrix.org suggested using jsaddle for abstraction, which would enable us to target both JS and WASM backends.

There may be some regression with the requirements checker feature in CodeWorld. This relies on Template Haskell, which wasn’t supported by the JS backend in its initial release (though this limitation isn’t mentioned in the GHC 9.12.1 docs). There could also be performance regressions, but I lack the expertise to comment on this preliminary stage. Given these factors, I’m leaning toward proceeding with a dual deployment approach.

4

Perhaps you will find the new support for “GHCi browser mode” useful when working on this project.

3 Likes