rust-lang / compiler-team Goto Github PK
View Code? Open in Web Editor NEWA home for compiler team planning documents, meeting minutes, and other such things.
Home Page: https://rust-lang.github.io/compiler-team/
License: Apache License 2.0
A home for compiler team planning documents, meeting minutes, and other such things.
Home Page: https://rust-lang.github.io/compiler-team/
License: Apache License 2.0
As several of you know, I managed to successfully create a proof-of-concept Rust REPL about a couple of months ago. This necessitated (at least, given a clean and sane approach) a few — essentially minor — changes to rustc.
As discussed with @nikomatsakis and @Mark-Simulacrum, the we thought the best way to proceed with this would be to create a design document that summarises my approach to creating the REPL, along with some technical details. The idea is to get some feedback either before or during the FCP process, from which I can proceed by submitting PRs.
See the first REPL-related PR (currently pending this issue and FCP) for more examples of the sorts of changes I am making (although that will need to be updated somewhat).
Building on the compiler-team contributor RFC, we wish to propose an active mentorship / internship program. This would be a way for contributors to work closely with a member on some non-complex task, helping them to become full members.
Some rough notes:
In recent PRs to create working groups, the check-in schedule hasn't been updated in initially, this is because we've not documented that step.
Propose a plan for how one goes about making "major changes" to a compiler area. Outline would be:
More detailed proposal available here
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
In our recent steering meeting, we made a decision to try and institute a design meeting. There is now some work to get ready for that.
@michaelwoerister and I proposed a 4 week cycle with this general shape:
The scheduling meeting would be used to decide the topics of the remaining 3 meetings.
That bot would assign the label and write something like:
Hey LLVM icebreakers! This bug has been identified as a good
"LLVM ice-breaking candidate". In case it's useful, here are some
[general instructions] for tackling these sorts of bugs. Maybe take a look?
cc @foo @bar @baz ...
RFC 2689 created the concept of compiler-team contributors. We need to "activate" that work:
The basic idea is to talk about the working group system we adopted earlier this year. How's it working out? Thoughts, concerns? I think before the meeting we would do a survey to take the general temperature, and during the meeting we'd shape the agenda based on the results of that meeting.
Some topics we might look at:
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
The expert map is in machine-readable form (thanks @spastorino!) but we need to fill in the directory information.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
See for example:
but if you manually change your url to one of the missing entries like the following week's, the page is there:
https://rust-lang.github.io/compiler-team/minutes/triage-meeting/2019-06-27/
rust-lang/rust
repositoryThis issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
Check list of things to do:
ICEBreaker-Cleanup-Crew
doneWG-traits does not list WG leaders.
See https://rust-lang.github.io/compiler-team/working-groups/traits/#traits-working-group
It probably deviates from the standard template in other ways too.
We have this rust-lang/rust#54818, it may be good to update it and shape it up and move it into https://github.com/rust-lang/compiler-team/tree/master/content/procedures.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
We're going to need to form intelligent thoughts about the roadmap as a team. This meeting would review some of the goals we had from 2019 and kickstart discussion on what our goals might be over the next year.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
Based on current experience with working groups, we should begin to document some best practices with what has been working and what hasn't.
Rust has had a long-standing problem where LLVM optimizes empty loops into UB. This can very easily create real bugs for folks. We've got a pending PR (#59546) that fixes this by inserting llvm.sideeffect
instructions, but until recently it was blocked on getting some lolbench measurements done. These measurements were recently completed (thanks @mati865!). They suggest the impact is "generally" small but non-zero (I put them in a world readable spreadsheet here, feel free to copy and play with it). We need to make a decision whether this is good enough! I'm going to cc the @rust-lang/lang team since I think lang team input would be useful here.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
In our meeting today, we decided to try and cover #60035 in more detail in a follow-up meeting. @nikomatsakis is going to try to loop up with @Zoxc and @michaelwoerister to fill out the hackmd doc above with details on the proposal as well as questions and concerns to highlight.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
Can we make the "section headers" on the webpage clickable? e.g., like "meeting calendar" and the like. It seems that the anchors like #meeting-calendar
exist -- they're just not easy to spot.
Some of them, for example "mw", are indecipherable unless you already know that it's an IRC (?) token that refers to @michaelwoerister
.
Another that I noticed is "acrichto", perhaps there are some more.
It is desirable to have a schedule for working group check-ins so that this can be planned for and delegated, and to ensure that each working group is heard from regularly and succeeds in their goal. However, the current schedule requires regular manual updating and changes when new groups are added, we should devise a new system that requires less manual work to maintain.
Based on discussion on Zulip, work out what else needs changed with the initial version of the experts map and move it into this repository.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
In #136 some content from the experts map was removed. :(
After rust-lang/rust#64566, which was a prerequisite for dataflow-based const validation, there are two different frameworks for dataflow analysis in the compiler. I've written a prototype along with a design document for unifying the two frameworks.
The dataflow framework is an internal API (with the exception of one use in cargo-clippy
), so this is a good time to improve it if desired. Accordingly, I've made some changes in the prototype, like passing Statement
s and Terminator
s directly into the effect
methods, and making the signature of call_return_effect
more similar to the others. If anyone has had difficulties implementing or inspecting the results of a dataflow analysis, it would be awesome to get your feedback here.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
It seems like forge.rust-lang.org is becoming the home for policy documentation and the like. Let's migrate some of the stuff from the compiler-team repository over there.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
It'd be useful to highlight ICE-breakers, bisecters, and other sorts of bugs using a twitter account, ideally automatically. @pnkfelix mentioned the idea of retweeting when things go unassigned.
This is a (small) extension for future consideration, where we allow contributors/members to set a "subtitle" giving more details about their role in the team (e.g., designer, documentation wizard, coder, etc).
After this #103 was merged some minutes ended being missed, probably in the PR created and merged time window.
I guess that if we add back notes added in PRs in between PR #103 (website) and PR #120 (a PR after website was merged) and add back those meeting notes we would be golden.
Still a sanity check in general won't be a bad thing to do.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
Extend existing out-of-tree crates policy to have more concrete steps for creating out-of-tree crates. @michaelwoerister has produced a document describing the steps they took setting up the measureme
crate, incorporate ideas from this.
'gcx, 'tcx
transitionThis issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
The current repository is just a bunch of mark down files in a directory. This is very hip and low-tech, but it'd be nice if we could have automatically rendered content. A partial list follows:
If we ported over to a jekyll setup and publishing via github pages, that would be awesome.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
We need to turn rust-lang/rfcs#2689 into reality.
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
I thought I had setup my own fork already. The relevant commit is 6589685
You can comment on the commit directly and I'll open a PR with the requested changes.
@rust-lang/infra can we have bors + a locked master branch here, too?
I think all the links on the index of the website are giving 404. Giving a quick look at that I think those are missing a docs/ prefix.
Start here https://rust-lang.github.io/compiler-team/ and check Zulip link and the links in the table but I think all the relative links are broken.
Anyway a sanity check in general to the site would be good.
We should have a contributing section on https://rust-lang.github.io/compiler-team/.
It could say:
Do we have content like this on rustc-guide?, maybe we should just have a contributing section that redirects there?.
rustc-guide should be the canonical home for that sort of thing, as it can go into a lot more depth, and we have a lot of that content.
Initial post will be looking for someone to do #73
This issue originally only aimed to unify the labels across compiler team crates, but was generalized to list all of the things listed in the out-of-tree crates policy. The original description is below.
In order to perform triage in the weekly meeting across nominated issues in all compiler-team maintained repositories, we should make sure that there are a consistent set of labels that are used and that we update the triage meeting agenda to point to these searches.
We had the meeting from #82 but we never created any kind of minutes!
This issue corresponds to a meeting proposal for the compiler team
steering meeting. It corresponds to a possible topic of
discussion. You can read more about the steering meeting procedure
here.
These issues are meant to be used as an "announcements channel"
regarding the proposal, and not as a place to discuss the technical
details. Feel free to subscribe to updates. We'll post comments when
reviewing the proposal in meetings or making a scheduling decision.
In the meantime, if you have questions or ideas, ping the proposers
on Zulip (or elsewhere).
Quoting @nikomatsakis from Zulip:
So I was talking to the person who runs readrust.net on discord. I feel like we need better visibility for the summaries and activities of the various compiler working groups. I was envisioning that it would be nice to have a central blog that just contains these sorts of updates and announcements -- basically, the activity of each of the working groups. In fact, I suspect that the blog should be project wide, with the ability to filter by team, by working group, or by other things.
The readrust.net author encouraged me to write-up an issue with more details. Before doing that, I thought I would open up some discussion here.
I've also been wondering about the structure of the weekly meetings -- I was thinking that maybe, instead of pulling from the groups, each group should push updates, content, and questions, and we can review them in the triage meeting.
Originally, I had in mind that we would go round robin because I thought it'd be useful to be able to see if working groups haven't been active with updates in a while. But we can do that separately just by having the updates organized by working group and checking for working groups that haven't posted any comments in a while (or, honestly, just monitoring Zulip). This could be a signal to go check-in and see how things are going and if we can do anything to help.
The link from the issue template to the proposal template seems broken; it links to _proposal.md
, but the template now appears in _index.md
.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.