Re: It's not Rust, it's FreeBSD (and LLVM)
- In reply to: Rein Fernhout (Levitating): "Re: It's not Rust, it's FreeBSD (and LLVM)"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Tue, 03 Sep 2024 22:43:03 UTC
On Tue, Sep 3, 2024 at 4:32 PM Rein Fernhout (Levitating) <me@levitati.ng> wrote: > > I actually think Rust could be an opportunity to modernize the FreeBSD > src. > > But if it is decided that no utilities found in FreeBSD will be eligible > for a Rust rewrite then I don't see why it should be part of src. > > I do think Rust is mature enough for adoption into src. These > discussions have been riddled with misconceptions about Rust (and its > use in the Linux kernel). > > Each FreeBSD release could just pin a stable release of the Rust > toolchain, it is common for Rust projects to have a Minimum Supported > Rust Version (MSRV). > Also Rust does have an extremely strong ecosystem. It includes gems like > the Serde serialization framework and the Clap argument parser. > > I would personally like to potentially see parts of FreeBSD rewritten in > rust. > I think it would invite much more contributors in the long term. > However I don't see rust rewrites happen anytime soon. And if we only > allow new programs in Rust to enter src then I don't think it is worth > the burden of supporting a rust toolchain (and its dependencies). +1. We have some utilities of subpar quality that could benefit from a rewrite, even if it must be in C. But it would be easier and less buggy to rewrite them in Rust. As a programmer fluent in both languages, I'm never going to waste my time writing new software (or rewriting old software) in C unless there's a very very good reason. However, the "only new utilities in Rust" limitation isn't completely useless, if it gives us the confidence to allow rewrites later on. -Alan