The author didn't submit it to HN, so criticizing it being on HN seems unfair to me.
It's a cool piece of alpha-quality software. It may or may not be meant to be used, that's beside the point. As I see it HN isn't a platform for software recommendations, it's for discussing interesting geeky things. Which this definitely is, even if it was completely unusable today.
Sometimes titles are slightly wrong. It's just not that big of a deal, most of the time. It's a messageboard in large part about the pleasure of finding things out rather than the grump of not everything being exactly as you expect.
> It’s a real disservice to the project not to give a raison d’etre in the readme, or any kind of technical motivation / differences.
This. There's a wave of projects whose only value proposition is this vacuous "let's reinvent the wheel in Rust" sales pitch, where nothing of value is proposed beyond throwing around the Rust buzzword.
Is it really necessary to restate the advantages of rewriting in Rust in every such project? Compared to Ruby programs Rust programs are faster, more robust, more maintainable, and easier to install. That's pretty much the same for any Rust rewrite (e.g. uv).
It would be interesting to know if there are other goals though, e.g. UX improvements.
The speed of homebrew has never been limiting factor (for me). I think there are far more important factors in what’s maintainable or not than language, and homebrew is very easy to install.
There has to be more important reasons to replace a mature widely use project like homebrew.
The issue with homebrew is not the programming language. The issue is homebrew itself. It takes forever to install packages, auto-updates without prompting you, etc. Also on my Mac it's been broken for years. It tries to auto-update and gets stuck forever.
I've been using homebrew for years and never did I ever experienced anything I would classify as "broken".
The thread you posted is comical. How many times does anyone run homebrew per day? Or per week? And you still have people complaining about sub-second execution times of a list command? In an app whose happy flow is downloading hundreds of MB off the internet and save it to disk?
> Is it really necessary to restate the advantages of rewriting in Rust in every such project?
I'm with you when the "source" project is C/C++ or something in that realm, but when we're coming from an already memory-safe language I do think some sort of explanation is helpful. I see Homebrew as more of a "glue" application where its own performance isn't exactly critical as it coordinates processes that are much slower so I don't really care if it's a bit faster.
By how much? Is homebrew really so slow, and used so often that an improvement would matter?
> more maintainable
[citation needed], especially for uv, which is a tool useful only for Python developers, so using a different language limits the pool of contributors.
Agreed - without clear performance metrics or feature differentiation in the README, potential users have no compelling reason to switch from a mature ecosystem like Homebrew to an alpha-stage alternative.
