rfcs VS buck2

I think with Cranelift's investment into an e-graph based optimizer (https://github.com/bytecodealliance/rfcs/blob/main/accepted/cranelift-egraph.md) they are well positioned to have quite competitive performance as a backend.

They're also actively working in this area, for example the recently added equality saturation framework and the pattern matching DSL it builds on.

There's an RFC here: https://github.com/bytecodealliance/rfcs/pull/28 and Saúl Cabrera, the person who is leading this effort and implementing the compiler tier, has a work-in-progress draft PR here: https://github.com/bytecodealliance/wasmtime/pull/4907

We discussed that a bunch in the RFC: https://github.com/bytecodealliance/rfcs/pull/14 . The conclusion was, in short, that the current Wasmtime production users didn't yet require an LTS release process, and the maintenance of an LTS is pretty onerous, so we would come up with one in the future as those requirements become more clear: https://github.com/bytecodealliance/rfcs/pull/14#discussion\_r708638804

If you want something similar thats written in Rust instead of Java, look at Buck2 https://github.com/facebook/buck2

There is clearly desire for more and better tooling in the Path space (camino is the ninth-most-downloaded FS crate, relative-path helps here, Buck2 created their own path library as well) but the cost of implementation is tremendous and, in my opinion, preventing healthy experimentation in this space.

Rust's niche is speed and reliability, which means being pretty good at many domains. That might sound like it can be replaced, but not actually. Let's take an example of a developer tool - ruff. It's a linter for Python that aims to replace a bunch of tools written in Python - Pylint, Flake8 and others. The sales pitch is - it does the same thing as the others, except it's 10-100x faster than the others. You could rewrite it in another language, but people aren't replacing ruff unless it's comparable in speed. Do you get it? Speed is a critical feature in some contexts and it's hard to beat Rust here. Not impossible, but hard. Other examples include turbopack, swc, buck2.

Maybe, with the caveat that coordination/synchronization in a distributed build is a non-trivial overhead for Rust builds under buck2.

Also see (Why Buck2)[https://github.com/facebookincubator/buck2/blob/main/docs/why.md], I especially find "Buck2 is remote execution first" interesting.