Better File Management and WebAssembly Support Proposal

[Staninna]

I'm making a game for narrrowblast. It has now 50 pictures and a JavaScript file that's 1.4 MB. The platform restricts uploads to a single HTML file, which poses a challenge due to the result being an enormous file when encoding all JavaScript and images into the HTML.

To improve this, I propose that users are allowed to upload a .zip file containing all files for a slide. This would improve user experience by reducing file size and enhancing file management.

Additionally, I'm interested in using Rust for optimizing heavy tasks within my game. But for that I need support for WebAssembly

[luttje]

I'm down for supporting zip uploads with multiple files.

WebAssembly I'm not too sure about. I'd like it if we could keep user content sandboxed and with WA I'm not sure we can at all.

Besides, what is so heavy that you need WA? Plain JS should be sufficient for most use cases imo.

[Staninna]

There are 3 main reasons I want to use WASM/Rust:

1. An AI/ML for players to play against in a single player mode or maybe doubles haven't figured that out (The heavy part)
2. I want to learn more about WASM and what it can/can't do
3. I already made a game similar to this one 1,5 years ago, and I could use that as a base instead of working my way up again

As for whether WASM can be sandboxed like JS within an iframe, I'm uncertain. However, all entry points for the WASM must be invoked from JS. As far as I know

[Staninna]

Plus it grands me access to the large ecosystem within rust or said other language

[luttje]

Thanks for elaborating, very cool project you've got there!

I guess we can eventually consider supporting WASM, but can you perhaps first make a more minimal prototype with just JS? That way other students can also learn from your code, without it being too complex right away. We really need more examples of the invite system, if we want to get more students on board in creating cool stuff with this.

[Staninna]

I could stall this project for a bit and create some good documented examples like a "Boter kaas en eieren" and other smaller games/quizzes because this project will be either way deep and complicated with the whole Pokémon battle logic/system embedded into it

[luttje]

That sounds like a good idea. Especially because that lowers the complexity for fine-tuning and improving the invite system