Make late_bound_lifetime_arguments a hard error.

[cjgillot]

Split from #103448

That lint has been there for 5 years.

Fixes #42868

[rustbot]

r? @oli-obk

(rustbot has picked a reviewer for you, use r? to override)

[compiler-errors]

Should we crater this? It would be nice to see what crates could use fix PRs to reduce the fallout of this.

[cjgillot]

@bors try

[bors]

⌛ Trying commit 270fa94 with merge ab173a644b8654eb56e6d7c92e29be4886d65169...

[bors]

☀️ Try build successful - checks-actions
Build commit: ab173a644b8654eb56e6d7c92e29be4886d65169 (ab173a644b8654eb56e6d7c92e29be4886d65169)

[compiler-errors]

@craterbot check

[craterbot]

👌 Experiment pr-108782 created and queued.
🤖 Automatically detected try build ab173a644b8654eb56e6d7c92e29be4886d65169
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[craterbot]

🚧 Experiment pr-108782 is now running

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[craterbot]

🎉 Experiment pr-108782 is completed!
📊 56 regressed and 1 fixed (257564 total)
📰 Open the full report.

⚠️ If you notice any spurious failure please add them to the blacklist!
ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[cjgillot]

@craterbot check crates=https://crater-reports.s3.amazonaws.com/pr-108782/retry-regressed-list.txt

[craterbot]

👌 Experiment pr-108782-1 created and queued.
🤖 Automatically detected try build ab173a644b8654eb56e6d7c92e29be4886d65169
🔍 You can check out the queue and this experiment's details.

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[craterbot]

🚧 Experiment pr-108782-1 is now running

ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[craterbot]

🎉 Experiment pr-108782-1 is completed!
📊 56 regressed and 0 fixed (56 total)
📰 Open the full report.

⚠️ If you notice any spurious failure please add them to the blacklist!
ℹ️ Crater is a tool to run experiments across parts of the Rust ecosystem. Learn more

[cjgillot]

Crater report:

• use of cssparser 0.17 and 0.25, those versions date back to 2017 and 2019. Latest release 0.27 does not error.
• use of roaring 0.4.2, from 2019. Latest release 0.10 does not error.
• use of scoped pool 1.0, from 2017. There is one PR to fix the warning since 2017, unmerged.
• zerocopy 0.7.0-alpha.3 which errs because it denies the lint, so ok.
• container-broadcast, no github repo, fixed by applying the suggestion.
• easy-csv, 7 year old, fixed by applying the suggestion.
• Sladuca.berkletree, repository archived.

[oli-obk]

Does roaring 0.5 also not error? The two crates that depend on 0.4 have no README and no updates since 5 years.

The only cssparser dependencies for versions at 0.25 or earlier are either explicitly abandoned or have no README and no updates since 5 years.

[oli-obk]

Nominating for lang team FCP. The summary in the tracking issue #42868 is still up to date, so I'm not creating a new summary.

The only regressions are in abandoned crates or outdated (and unused) versions of crates that have been updated to not trigger the future incompat lint.

[nikomatsakis]

@rfcbot fcp merge

Discussed in a (minimally attended) lang-team triage meeting and we are in favor of moving forward with this.

[rfcbot]

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

• @joshtriplett
• @nikomatsakis
• @pnkfelix
• @scottmcm
• @tmandry

Concerns:

• types-team-input (Make late_bound_lifetime_arguments a hard error. #108782 (comment))
• unclear-motivation (Make late_bound_lifetime_arguments a hard error. #108782 (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!

cc @rust-lang/lang-advisors: FCP proposed for lang, please feel free to register concerns.
See this document for info about what commands tagged team members can give me.

[nikomatsakis]

@rustbot labels -I-lang-nominated

[BoxyUwU]

The intention long term is to make all lifetimes late bound, I would have thought this would mean we should be removing this lint not making it a hard error since otherwise you would be unable to ever turbofish lifetime args. not to mention we may end up with late bound type and const parameters, not being able to turbofish any generic arguments seems not great.

[BoxyUwU]

It would probably be a good idea to nominate this for t-types discussion to figure out how this lint interacts with long term plans for late bound generic parameters

[tmandry]

@rfcbot concern types-team-input

Agreed that types team should weigh in here.

[tmandry]

@rust-lang/types, can you comment on this proposal with respect to the following?

It would probably be a good idea to nominate this for t-types discussion to figure out how this lint interacts with long term plans for late bound generic parameters

[jackh726]

We haven't discussed this yet, but I do think we should weigh in here. Likely we'll go through nominated issues as a team at the beginning of next month.

[RalfJung]

I don't understand why we even want to disallow explicitly stating late-bound lifetime parameters. This is the first time I hear of #42868, and that issue description is confusing to me and doesn't sufficiently motivate the change being made.

I posted my confusion/question at #42868 (comment), though now I realize maybe I should have done that here? let me quote it for your convenience:

I have to say I am quite confused by the issue description here, and I don't understand what the underlying issue is. The core of my confusion seems to be captured by this (repeated) statement

the first argument doesn't make much sense because the function is not parameterized by 'a.

I strongly disagree! The function is parameterized by 'a. Sure, this parameter is a type system fiction (the actual runtime address of the function is the same for all choices of 'a), but that doesn't make 'a any less of a parameter. It can still be valuable to explicitly set that parameter to guide inference.

Taking a step back: The fact that some lifetime parameters are type system fictions while others are actually relevant for monomorphization is very confusing and surprising. (I have never fully understood why this is the case; I'll need to dig into this some time.) This will clearly incur some friction when creating function pointers since monomorphization-relevant ("early-bound") lifetime parameters need to be chosen the moment the function pointer is created. I can see that we might want to improve that situation. But reading just this issue I really don't understand the motivation for the current change, since there's a big gap between "unfortunately some lifetime parameters actually matter for monomorphization" and "other lifetime parameters (that don't matter for monomorphization) don't make sense and we should disallow people from explicitly stating them". Which problem is being solved here?

I guess it's time I use my lang team advisor hat for the first time. ;)
@rfcbot concern unclear-motivation

[tmandry]

Since rfcbot hasn't fully implemented the advisors role yet, I'll proxy Ralf's concern.

@rfcbot concern unclear-motivation

[cjgillot]

There is no real motivation on my end besides simplifying the implementation of #103448. Closing for now, until I get back to that PR.