I’ve wanted to use Hoot for a while now. I’ve done a bit of Wasm via Rust, and the UX is really good. But the DOM stuff is handled via a web-sys JS shim. The Leptos and Dioxus web frameworks are great if you’re building that kind of thing. But the learning curve is rather steep.

The prospect of a Wasm compiler that can itself run in Wasm means Hoot is not just ideal for interactive web pages, but for building arbitrary runtimes. The spritely institute provides other tools (Goblins) for making distributed runtimes.

And if Scheme is not your preferred language, Scheme is an excellent language for building compilers.

Hoot has the potential to make Wasm a lot more accessible.

I recently learned about Guile Hoot via this System Crafters video: https://www.youtube.com/watch?v=8LuQtoy9NLs

Looks awesome and I hope to use it some day in the near future.

This post is from 2023. Please add "(2023)" to the headline.

* This looks awesome. I'm looking forward to proof by demonstrated real-world production-grade project, and experience report of same.

* The awesomeness might be more clear once there are toolkit/framework layers above this low level. Just like a first demo of simple things in Svelte doesn't show you the complexity and bulk of Svelte internals by which that is implemented.

* Maybe someone will have an idea for some killer application of this that makes it appealing to more than just the people who already know they want to use Scheme. In the past, one such idea was to use first-class continuations to implement Web forms UI more easily, and another was to use Scheme to implement rich meta data model in a tractable way that would've been tons more work in the alternative of the day (Java). Or maybe the killer feature doesn't come down to a fundamental strength, since today in theory you can shoehorn JS to do most things, but you come up with an all-around elegant solution that only happens to use Scheme, and it's so great that people like it anyway.

This is cool. I think I'll try to learn it!