Announcing Rust 1.85.0 and Rust 2024 | Hacker News Reader

Announcing Rust 1.85.0 and Rust 2024

blog.rust-lang.org

84 pointshaakon15 comments

I did “cargo fix —-edition” like the guide said and it added the word “ref” before variable names in certain places, which made the compiler complain. So I had to go through and remove them all.

So it may be easier to fix things manually than rely on cargo fix.

Hi! Does anyone know why I still, after updating `rustup` and the `edition` in `Cargo.toml`, need to run `cargo +nightly fmt `? I have a `rustfmt.toml` that contains this:

  imports_granularity = "Crate"
  group_imports = "StdExternalCrate"

I assumed those would be in, as they've been nightly features for years, and we just got a new edition.

> I assumed those would be in,

A new edition doesn't mean features are automatically stabilized.

I agree with these two specifically, it would be great to have them in stable. They're not yet though: https://github.com/rust-lang/rustfmt/issues/5083

Thank you for the explanation!

No problem. I have to use nightly for this reason too, haha.

Huge! Async function traits resolve one of the biggest pain points for async rust.

Updated a 16k LOC project and had no issues at all.

Unfortunate that they didn't ship the parallel front-end yet, solves my biggest issue with Rust right now.

This is the hugest release of Rust this far.

We are in 2025 already. Some other tool release their version N at the end of the year N-1.

On the other hand, people will be able to use Rust 2024 in production in a few month, if not now. While most C++ codebase are still stuck with C++17 or older.

Just like many C code bases are stuck on C89, many Python in Python 2, Java in Java 8, .NET in .NET Framework,...

Not that every compiler is up to date, but at very least clang, gcc and vc++ have already quite good coverage of C++20, and clang/vc++ modules mostly work, good enough that I only use modules in my hobby projects.

It starts by projects, companies, actually wanting to adopt new toolchains.

From what I understand, when the schedule started to slip, the team made a decision to not do a death march just so it happened two months ago instead of today. This seems healthy to me.

The code was frozen already in 2024, it just took some time to actually release it.

Same with C23 and C++23, things take time to iron out.

Reply on news.ycombinator.com