T-spec meeting 2025-04-03

--- title: T-spec meeting 2025-04-03 tags: ["T-spec", "meeting", "minutes"] date: 2025-04-03 discussion: https://rust-lang.zulipchat.com/#narrow/stream/399173-t-spec/topic/Meeting.202025-04-03 url: https://hackmd.io/sSHEdhmIQBuI6rtS7HJeQA --- Attendees: - Joel Marcey - Pete LeVasseur - Eric Huss - TC - Merlan - Pierre-Emmanuel Patry - Sid Askary - Pietro - Josh Triplett - Jack Huey Regrets: Agenda: - Updates to the agenda? - FLS integration update - `t-spec` at Rust All-hands? - What do we need? - What do we want to accomplish? - Work Session ## Updates to agenda - SIMD target_feature compatibility (Eric) - Personnel Update at Ferrous ## FLS integration update Announcement done FLS transition plan to the project in progress. Hope to share publicly soon after it settles. Discussing in broad strokes. - Repo transfer to Project - Test annotations - Compiler team might have to weigh in Milestone 1: Actually render the document within the Project processes Milestone 2: Integrate the test annotations into the Project - Compiler MCP or RFC - Compiler team has accepted test annotations for the Reference. Should hopefully do the same with the FLS. Pete: Will paragraph ids still be a thing? Joel: Yes, not planning to make big changes to the doc. TC: Project blog post mentions this as well. Josh: URL for spec? TC: It'll be the same prefixes and those for the Reference. Pietro: It'd be helpful for us for this to be a subtree instead of submodule of `rust-lang/rust`. In our fork, we do everything primarily in the monorepo and we only break it out for publishing, so all of our tooling assumes the monorepo. TC: The project has been moving things to josh rather than subtree. Does that work for you? Pietro: Yes, that'd be OK. Pete: We need to discuss how coding guidelines can link to FLS which should link to specific versions of Rust. TC: Let's settle on the name of the repository. I propose we call it `fls`. That's one of the questions in the doc, and we should unblock this. Joel: For now, or forever? TC: For now. We can always make any other changes later. Joel: Yes, let's do that. TC: OK, the name of the repo will be `fls`. Pietro: Change logo and name that is shown spec.ferrocene.dev? Joel: Yes, to Rust Project logo and FLS. Pietro: What should we do about the traceability matrix? We currently have this code in the main repo rather than in the spec repo; it integrates with `compiletest` and bootstrap. TC: How does this compare with what we do on the Reference? Eric Huss: We do it from within the Reference repo and it scrapes the tests. TC: It seems OK to have this in `rust-lang/rust` as long as we namespace it appropriately, e.g. `fls-traceability-matrix`. Pietro: Yes, that makes sense. ## Rust All-Hands [From Mara](https://rust-lang.zulipchat.com/#narrow/channel/399173-t-spec/topic/All-Hands/with/509530756) > Hello @T-spec! I'd like to know your plan at the all-hands. Do you need your own room? For both days, or just a day or a few hours? Any cross-team discussions you'd like to have? We do not need our own room. We may put something forward as part of an unconf process if there is one. We may want to talk to various teams depending on how the integration and publishing of the FLS is going. ## SIMD target_feature compatibility https://github.com/rust-lang/rust/pull/139309 Should this be documented in the Reference? SIMD is not documented now. But probably should be. Should the PR be blocked by documentation. ## Pietro is leaving Ferrous Pietro is leaving Ferrous on May 8. Any FLS integration work that would require Pietro needs to be done before. Pietro will still remain involved with the Project. ## Process for joining the `t-spec` team Josh and Jack would like to join the team. What is the process? `t-spec` team would like to talk about it. ## Work Session ## Chat +merlan #flirora +merlan #flirora says:Hi, I can’t turn on my sound here 8:00 T TC TC says:You can hear us? 8:01 +merlan #flirora +merlan #flirora says:No, I have it turned off because I’m at work 8:01 +merlan #flirora says:So unfortunately, I don’t think I’ll be able to participate in any meaningful manner, so I’ll be leaving now 8:03 T TC TC says:Thanks for stopping in. 8:04 avatar Pierre-Emmanuel Patry Pierre-Emmanuel Patry says:Thanks for joining Merlan! 8:04 +merlan #flirora +merlan #flirora says:You’re welcome 8:04 avatar Josh Triplett Josh Triplett says:(Could I get a link to the agenda doc?) 8:06 PL Pete LeVasseur Pete LeVasseur says: https://hackmd.io/sSHEdhmIQBuI6rtS7HJeQA ❤️ 8:06 avatar Josh Triplett Josh Triplett says:👍 8:08 Pete LeVasseur Pete LeVasseur says:Reminds me of the scientists that bemoan how poorly sometimes science communication happens in regular news! 8:29 avatar Josh Triplett Josh Triplett says:👍 8:29 Sid Askary Sid Askary says:have to drop off... 8:30 Pete LeVasseur Pete LeVasseur says:👍 8:33 Pete LeVasseur says:👍 8:34 success metric of upstreaming: "we didn't break users" sounds just like Rust itself 😃 8:35 Josh Triplett Josh Triplett says:👍 8:37 PL Pete LeVasseur Pete LeVasseur says:Fun Language Specification 8:38