I'm wondering if self-profiling should be its own WG. It has a clear goal, a clear set of people working on it, and it's complex enough to be an ongoing effort for a few months at least. cc @nikomatsakis @Wesley Wiser
@Wesley Wiser I also feel like the self-profiling stuff might make for a nice WG, presuming you're still planning to push on that. Thoughts? I would mainly imagine that we lay out a (brief) roadmap, create a stream, and then check in on in the meetings every once in a while to see how you're doing (not necessarily that it's a big mentoring opportunity, although I wouldn't be surprised if we were able to get a people to hack on things with you that way)
(cc @mw, @eddyb, who else?)
You might want to see this thread.
@mw beat you to the idea by a few hours.
I was thinking it yesterday dang it
it will be great to have such a WG to maintain a list of rustc's Gandalf
:+1: from me!
@nikomatsakis Yep, I'm still pushing on that. @mw actually just r+'d some more changes of mine related to this. I think there's probably enough stuff to work on that if anyone else were interested, we could find stuff for them to do. So a WG makes sense to me.
(a small FYI that I've merged the other topic into this one so there aren't two)
@mw I guess we should probably chat at some point and figure out what the next steps are for the project. I know you have some ideas around improving the performance and representation of data in the profiler. I'd like to see the export format stabilize a bit so we can start building tools to interpret the data.
I'm sure there's other considerations as well
Yes, that makes sense and sounds like a good topic for a first WG meeting :)
@davidtwco could you set up a
@Wesley Wiser do you want to set up the working group in the compiler-team repo?
Similarly, a user group with you and Wesley exists.
@mw Sure! I can do that.