Proposal
The soft_unstable lint (rust-lang/rust#64266) was added many years ago to warn against the #[bench] attribute and later also used for RustcEncodable & RustcDecodable and for some inner attribute cleanup. For more than a year now, the lint is entirely unused in the compiler. With it comes some infrastructure in the staged_api attributes to mark library items as soft-unstable such that using them on stable triggers the lint instead of the usual hard error.
I think we should remove the lint and the concept of soft-unstable library features. It is tempting to have some common infrastructure for de-stabilizing accidentally stable features, but the problem is that the lint couples all of these features together in terms of whether the lint is warn- or deny-by-default and whether it shows up in dependencies. In practice we want to control this timeline separately for each such de-stabilization. De-stabilizations are also just not common enough to warrant having all this soft-unstable-library-feature tracking in the compiler (in the entire 10 years since Rust 1.0, to my knowledge this was used exactly once, for the rather exceptional case of RustcEncodable & RustcDecodable).
EDIT: #[bench] also involved a library item, so the soft-unstable library item concept was used twice overall.
We should, however, check in with the libs team before actually removing this.
Mentors or Reviewers
If you have a reviewer or mentor in mind for this work, mention them here. You can put your own name here if you are planning to mentor the work.
Process
The main points of the Major Change Process are as follows:
You can read more about Major Change Proposals on forge.
Proposal
The
soft_unstablelint (rust-lang/rust#64266) was added many years ago to warn against the#[bench]attribute and later also used for RustcEncodable & RustcDecodable and for some inner attribute cleanup. For more than a year now, the lint is entirely unused in the compiler. With it comes some infrastructure in the staged_api attributes to mark library items as soft-unstable such that using them on stable triggers the lint instead of the usual hard error.I think we should remove the lint and the concept of soft-unstable library features. It is tempting to have some common infrastructure for de-stabilizing accidentally stable features, but the problem is that the lint couples all of these features together in terms of whether the lint is warn- or deny-by-default and whether it shows up in dependencies. In practice we want to control this timeline separately for each such de-stabilization. De-stabilizations are also just not common enough to warrant having all this soft-unstable-library-feature tracking in the compiler (in the entire 10 years since Rust 1.0, to my knowledge this was used exactly once, for the rather exceptional case of RustcEncodable & RustcDecodable).
EDIT:
#[bench]also involved a library item, so the soft-unstable library item concept was used twice overall.We should, however, check in with the libs team before actually removing this.
Mentors or Reviewers
If you have a reviewer or mentor in mind for this work, mention them here. You can put your own name here if you are planning to mentor the work.
Process
The main points of the Major Change Process are as follows:
@rustbot secondor kickoff a team FCP with@rfcbot fcp $RESOLUTION.You can read more about Major Change Proposals on forge.