You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We don't really have a great story today for ensuring that our feature dependencies are 100% correct, both in terms of building crates in isolation or non-typical scenarios, and in terms of being truly minimal. There are tools out there that claim to be able to detect these cases, namely https://crates.io/crates/cargo-hack and https://crates.io/crates/cargo-unused-features (and maybe others?). I don't think running these on PR gates makes sense given their likely long runtimes over our large codebase, but maybe running them on a daily job would make sense?
The text was updated successfully, but these errors were encountered:
There's also apparently the unused-crate-dependencies lint, which can get much more precise than our current cargo-machete-ish approach, but requires actually building. Combining it with cargo-hack could be nice, but would probably take some work to get clean.
We don't really have a great story today for ensuring that our feature dependencies are 100% correct, both in terms of building crates in isolation or non-typical scenarios, and in terms of being truly minimal. There are tools out there that claim to be able to detect these cases, namely https://crates.io/crates/cargo-hack and https://crates.io/crates/cargo-unused-features (and maybe others?). I don't think running these on PR gates makes sense given their likely long runtimes over our large codebase, but maybe running them on a daily job would make sense?
The text was updated successfully, but these errors were encountered: