chores: Bump smol version to 2.0 #221
Conversation
This should fix Restioson#220. The changes include removing smol-potat in favor of smol::block_on. However, this commit does not remove smol-timeout as it is heavily used on tests and the crate seems to be unmaintained. Nontheless tests keep working.
|
Hey, thanks for the PR! Really sorry for getting back to this so late, studies have been all-consuming this year. I wonder if we would want to keep smol 1.0 support under a different feature (e.g. smol-1)? |
As long as there is a release with smol-1 support, I wouldn't bother too much :) |
|
That's fair, since I suppose we haven't really made any other changes |
|
So, almost a year later, I noticed this PR was still open and that I missed your review. Sorry for that. I ran |
|
Apologies - this dropped off my radar too! |
|
Hmm, looks like a build failure - I think smol relies on async_task? |
|
I think it had something to do with Cargo.lock being modified in both branches so it was stuck in an invalid state. I recreated the lock file with the latest dependencies. I would advise removing the lockfile, as it mostly does nothing for libraries, and switch to something like https://crates.io/crates/cargo-minimal-versions to validate the minimum requirements. |
Having the lockfile ensures deterministic CI runs, which is important in the presence of e.g. semver-incompatible bumps etc It is also the official recommendation from the Rust folks: https://blog.rust-lang.org/2023/08/29/committing-lockfiles/ |
Cargo.lock
Outdated
| [[package]] | ||
| name = "async-std" | ||
| version = "1.12.0" | ||
| version = "1.13.1" |
There was a problem hiding this comment.
Why is async-std getting bumped when we are only changing smol? 🤔
|
The diff on the lockfile seems a bit excessive to me. Can you checkout the one from main and redo the bump to make that smaller? Thanks! |
Oh, I guess I was using the old guidelines. Thanks for sharing.
Done. I had to use Rust 1.75 as 1.88 was failing for wasm_bindgen with the following error: |
|
LGTM, happy for you to merge @thomaseizinger when you're happy |
|
Is the CI error reproducable locally? |
3 participants
This should fix #220. The changes include removing smol-potat in favor of smol::block_on. However, this commit does not remove nor update smol-timeout as it is heavily used on tests and the crate seems to be unmaintained. Nonetheless tests keep working.