Lang/RfL meeting 2025-07-16

Attendance

People: TC, tmandry, Miguel, Alice Ryhl, BennoLossin, Boqun Feng, Gary Guo, Josh Triplett, Wesley Wiser, Xiang

2025H2 project goals

Language goals:

https://hackmd.io/5giQebPEROKWzxoB78lVIQ

Tyler: Ding, do you see any possibility of getting arbitrary_self_types and derive_coerce_pointee without Deref: Receiver?

Xiang: Seems unlikely; we are blocked on the new trait solver for a workaround. https://github.com/rust-lang/rust/pull/138952

Compiler goals:

https://hackmd.io/3HtawTviT3-u6dbdmfVhrA

Tyler: Is the standard for stability of compiler flags higher than it needs to be for users that drive rustc directly?

Wesley: There has been some hesitation to stabilize flags without a wider story for how they can be deployed to the broader Rust community. We are working on ways to solve this from two different angles. With build-std work it looks like we are close to figuring this out. Similar for target-features.

Alice: Out of curiosity, how is build-std progressing?

Wesley: There is a pre-RFC being shopped around. It's very large. It has an extensive plan, but the current feedback is that it's too large for an MVP. I can share if anyone is interested.

Ongoing work updates

Ding: We have a Pre-RFC for the supertrait-item-in-subtrait-impl work. I would need to add two modifications to the RFC to incorporate the requests in the comment t-lang thread.

Tracking

Tracking issue

For next time

Miguel: Might want something like consteval: Functions that can only run at compile time. Not very urgent.

Tyler: That sounds related to the second half of this project goal: https://github.com/rust-lang/rust-project-goals/pull/311

Miguel: Also, cfgs in core came up again so we can disable certain items there (?)