T-spec meeting 2024-12-05

--- title: T-spec meeting 2024-12-05 tags: ["T-spec", "meeting", "minutes"] date: 2024-12-05 discussion: https://rust-lang.zulipchat.com/#narrow/stream/399173-t-spec/topic/Meeting.202024-12-05 url: https://hackmd.io/4PHLIz-XRCySsbmve5QYlg --- Attendees: Joel Marcey, Eric Huss, Pierre-Emmanuel Patry, Connor Horman, Sid Askary, Mara, Niko Matsakis, Monadic Cat, TC, pnkfelix Regrets: Agenda - A welcome (2-5 minutes) - Quick FLS Update (5 minutes) - 2025 H1 Project Goals (30 minutes) ## A Welcome nikomatsakis joining spec team NM: thanks! :purple_heart: ## Quick FLS Update Joel: I have contacted Foundation legal to let them know that there has been agreement to a transfer and provided the parameters that should be agreed upon between us and Ferrous regarding things around licensing and copyright and use of the spec by Ferrous, etc. Joel: I have sent an email to Florian, cc'ing my legal, asking Ferrous to have their legal get in contact with the Foundation legal to get on the same page. Waiting for that communication to happen. Should hopefully not be contentious and generally a formality. Joel: Once the agreement occurs, we will start the process of transferring the FLS over to the Project/Foundation (copyright) and begin a marketing campaign about this and Ferrous' willingness and generosity of allowing this to happen. ## 2025 H1 Project Goals These are *proposed* goals, not official yet. ### [Spec Testing](https://github.com/rust-lang/rust-project-goals/blob/main/src/2025h1/spec-testing.md) Niko: Connor, how much have you talked to the compiler team about this goal? Connor: Haven't talked to them yet? Plan to have meeting, but don't know when the meetings yet. pnkfelix: The compiler design meeting would be appropriate. Connor: Also want to speak to lang team as well, maybe in January. Niko: On lang team, there shouldn't be much dissent. ### [FLS Integration](https://github.com/rust-lang/rust-project-goals/blob/main/src/2025h1/spec-fls-integration.md) Connor: Design axioms. Making decisions effectively and efficiently. Connor: I put dates for things to be done. We should be making decisions towards that goal. Joel: Goals will allow that to happen naturally. Niko: Agree. Connor: Do not sacrifice deliberation for speed, but we can be more efficient. Niko: Regarding the FLS integration, should we have a specific lang team ask here. The lang team should be aware that this is happening, maybe via an RFC? Niko: Should we have the item about merging a chapter from the reference? Niko: The big goal is the taking the FLS over from Ferrocene. The process part of this. Niko: Remove item about integrating Reference chapter into the FLS. Niko: What is the lag period between Rust releases and the update of the FLS. TC: Release the FLS the same as the reference and the other books. But put at the top the Rust version that they are up to date. Niko: Line item that says adjust FLS to make it in accordance with Rust 1.88. Eric: This will be difficult without changes to processes. Niko: Why is it going to be hard? Eric: Because there is a change to the compiler and then they set a PR to change the reference, but it takes months to actually hammering out the PR. Niko: The only difference is that Ferrocene prepares a diff and submits it for review. And these are based on what tests exists. Joel: I think all this is part of the spirit of the goal. Figuring out how this process works and if we need to adjust. Sid: Frequency of releases is in sync with the versions of the spec? Niko: Release of compiler on X, compiler release on X+1, spec is released sometime in between. Connor: 6 weeks is the train the whole project runs on, whatever we can get out in that 6 weeks I think we can live with, and we should eventually get closer to matching the release train. On Joel's point, there are two levels of \[process\] integration: Getting the processes for the FLS intact, and Integrating the Project as a whole with the FLS and Specification as a whole. The latter is more "Shiny Future" and isn't in scope for the 2025h1 goal. Niko: Wants to match the cadence of what the FLS does today at first. Sid: Be sure that the FLS needs to be updated at a quicker cadence than it currently is because the safety critical industry is conservative. Niko: Don't want us to get behind. Keep up with the rate of change. Niko's "hard" requirements (from chat): 1. official release of some version N 2. review a round of edits for version N+1 to see how that goes ## Felix Departure pnkfelix is departing the Rust Project starting in 2025, save for maybe the compiler team. pnkfelix: recommend that TC to be the lang team liasion. Connor: Does there need to be an official liasion? pnkfelix: Not necessarily, but there should be overlap. pnkfelix: Was co-lead. Should there be an additional co-lead. Joel (on behalf of everyone): Thank you pnkfelix for your time and service on the `t-spec` team and the Project in general. Best wishes for what comes next. ## Chat Meeting document: https://hackmd.io/4PHLIz-XRCySsbmve5QYlg 8:01 M Mara Mara says:👏 8:04 avatar Connor Horman Connor Horman says:👏 8:05 avatar Niko Matsakis Niko Matsakis says:Thanks ❤️ 8:05 Niko Matsakis says:big endian 8:08 Niko Matsakis says:ALWAYS 8:08 PF (pnk)Felix (pnk)Felix says:Connor: https://rust-lang.github.io/compiler-team/about/steering-meeting/ 8:11 (pnk)Felix says:specifically: https://forge.rust-lang.org/compiler/steering-meeting.html 8:12 (pnk)Felix says:And the next Compiler Team planning meeting is Friday December 13th 8:12 avatar Connor Horman Connor Horman says:I'm not going to be available that day. 8:13 (pnk)Felix (pnk)Felix says:if you can share your schedule with someone else you can be there, then they could act as your proxy w.r.t. planning 8:13 (pnk)Felix says:sorry, "someone else _who_ can be there" I meant to type 8:14 🙁 8:17 avatar Niko Matsakis Niko Matsakis says:👎 8:17 avatar Connor Horman Connor Horman says:😮 8:17 Monadic Cat Monadic Cat says:lol 8:18 Good luck with everything going forward, Felix 8:21 Monadic Cat says:Thank you for being here 8:21 avatar Niko Matsakis Niko Matsakis says:I wouldn't object to ending early, just sayin'.... I'm crazy busy right now. 8:21 Mara Mara says:👍 8:29 As long as we don't end up in the C# spec world where they are 3 major versions behind. 8:47 Niko Matsakis Niko Matsakis says:niko's hard requirements: 1. official release of some version N 2. review a round of edits for version N+1 to see how that goes 8:49 Niko Matsakis says:(ntd) 8:50 Niko Matsakis says:thanks! 8:50 (pnk)Felix (pnk)Felix says:(yeah I'm officially on th hook to take the kids to a doctors appt next Thursday) 8:51 (pnk)Felix says:so I'm officially calling this my last T-spec mtg. 😃 👍 8:51 MC Monadic Cat Monadic Cat says:*everyone stays on* 8:52 Connor Horman Connor Horman says:(My timer is off, but I've got time) 8:55 Sid Askary Sid Askary says:Got to go... 9:00 Monadic Cat Monadic Cat says:I like to think I'm a decent technical writer, I could volunteer some time to help people write if it'd be welcome. 9:07 Connor Horman Connor Horman says:(Reminds me that I should do a stabilization report for unbounded_shifts) 9:08