Remove the missing_copy_implementations lint.

[nnethercote]

This lint made a certain amount of sense for people writing code during the period leading up to to Rust 1.0. It doesn't make much sense today. There are many valid cases of types that could impl Copy but do not. The lint is extremely low value and moderately complicated. This commit removes it.

r? @lcnr

[rustbot]

rust-analyzer is developed in its own repository. If possible, consider making this change to rust-lang/rust-analyzer instead.

cc @rust-lang/rust-analyzer

[nnethercote]

Like #126018, this will need an FCP.

[lcnr]

r=me on the code changes themselves, however, I am less certain about removing this lint.

I definitely wouldn't be in favor of adding it to rustc right now, but unlike box_pointers, I can at least imagine a somewhat coherent argument for why this lint should exist. There are some people who use it https://github.com/search?q=missing_copy_implementations+language%3ARust&type=code&l=Rust&p=3.

otoh, It does have false positives and if people are unhappy about us removing it, it can get added to clippy (where it belongs)

@rfcbot fcp merge

[rfcbot]

Team member @lcnr has proposed to merge this. The next step is review by the rest of the tagged team members:

• @Aaron1011
• @cjgillot
• @compiler-errors
• @davidtwco
• @eddyb
• @estebank
• @jackh726
• @lcnr
• @matthewjasper
• @michaelwoerister
• @nagisa
• @oli-obk
• @petrochenkov
• @pnkfelix
• @wesleywiser

Concerns:

• lint-actually-useful (Remove the missing_copy_implementations lint. #126293 (comment))

Once a majority of reviewers approve (and at most 2 approvals are outstanding), this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

[nagisa]

Question is if we can add this to clippy from the get-go (are they interested in maintaining such a lint at all? is it technically feasible? is there something that makes this lint uniquely easier to implement in rustc?)

[petrochenkov]

There's a very similar lint called missing_debug_implementations, do you plan to remove it too?

There are many valid cases of types that could impl Copy but do not.

Many of them are private though.
Both missing Copy and missing Debug are supposed to be used by people writing libraries with a stable public interface, and the lints only apply to that public interface.
That seems useful enough.

I'd be fine with putting these lints into clippy if the lints were new.
But since the lints are already in the compiler for long time, then it's better to keep things as is and not break existing code.

[nnethercote]

There's a very similar lint called missing_debug_implementations, do you plan to remove it too?

No. I think the usefulness for public APIs is more obvious there. The motivation for this PR was the fact that the lint description talks about this lint existing to assist with a pre-1.0 behavioural change, which seems to not be of much relevance any more.

[nnethercote]

The API guidelines on interoperability say "crates that define new types should eagerly implement all applicable, common traits" and specifically names Copy, Clone, Eq, PartialEq, Ord, PartialOrd, Hash, Debug, Display, and Default.

The API guidelines on debuggability say "All public types implement Debug" and "If there are exceptions, they are rare."

A couple of possible reactions to that.

• Why do Debug and Copy get a lint when none of the other traits do?
• Having said that, Debug gets mentioned twice, and with arguably stronger language the second time, so if any single trait deserves a lint maybe it's the one?

[bors]

☔ The latest upstream changes (presumably #120924) made this pull request unmergeable. Please resolve the merge conflicts.

[jhpratt]

I have personally found this lint to be useful in the past. I don't agree with the motivation at all.

[apiraino]

@rfcbot concern lint-actually-useful

From comment: I have personally found this lint to be useful in the past. I don't agree with the motivation at all

[oli-obk]

@rfcbot concern lint-actually-useful

From https://github.com/rust-lang/rust/pull/126293#issuecomment-2192690954: I have personally found this lint to be useful in the past. I don't agree with the motivation at all

bot wasn't listening

[nnethercote]

There is enough opposition to this that I will close it.