I support this.  I worked a lot with the author to make the spec precise.  (It was pretty confusing and incomplete before.)

TL;DR: the intent is clear and useful; the details are actually surprisingly tricky.  But (like type inference) users won't really care!

Simon

On Tue, 12 Mar 2024 at 16:39, Chris Dornan <chris@chrisdornan.com> wrote:
Folks,

Proposal: Fine-Grained Unused Warnings (#42)
Author: Jakob Brünker
Rendered proposal: https://github.com/JakobBruenker/ghc-proposals/blob/fine-grained-unused/proposals/0000-fine-grained-unused-warnings.rst
Discussion: https://github.com/ghc-proposals/ghc-proposals/pull/434
Recommendation: Acceptance

## Summary

The proposal partitions warning about unused identifiers into 

a) bindings that are truly unused (not mentioned anywhere) and 
b) bindings that are mentioned exclusively in code that is itself (transitively) unused,

and controls the latter with the new flag -Windirectly-unused-binds, which is enabled by default (to preserve existing behaviour).

Everybody seems to be in favour of the proposal in general and it has been extensively revised for clarity and to ensure in interoperates consistently with the existing warning-flags mechanisms.

I propose that we accept this proposal if nobody objects by the start of next week (Monday, 2024-03-18).

Chris
_______________________________________________
ghc-steering-committee mailing list
ghc-steering-committee@haskell.org
https://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-steering-committee