Stream: wg-traits

Topic: meeting 2019.06.03


nikomatsakis (Jun 03 2019 at 17:59, on Zulip):

Hey @WG-traits! Sync meeting time?

nikomatsakis (Jun 03 2019 at 18:00, on Zulip):

I have a few announcements:

nikomatsakis (Jun 03 2019 at 18:01, on Zulip):

I've not done a good job of giving a lot of feedback or reviews -- in particular I know I owe @Alexander Regueiro a review or 200, @Keith Yeung some feedback on objects, and @Florian Diebold some feedback on fuel or something like that :)

nikomatsakis (Jun 03 2019 at 18:02, on Zulip):

(We do have the zoom call thing in ~30 minutes, if people are around and interested, we could talk some about the floundering / region constraints stuff, or other things.)

nikomatsakis (Jun 03 2019 at 18:02, on Zulip):

How's everybody else going? :)

Alexander Regueiro (Jun 03 2019 at 18:03, on Zulip):

That's alright. :-)

Alexander Regueiro (Jun 03 2019 at 18:03, on Zulip):

I have to jet now, sorry...

Alexander Regueiro (Jun 03 2019 at 18:03, on Zulip):

maybe next week will be around

nikomatsakis (Jun 03 2019 at 18:04, on Zulip):

(I did at least spend an hour or so commenting the relevant code on @Alexander Regueiro's PR :)

Alexander Regueiro (Jun 03 2019 at 18:04, on Zulip):

yeah, that's much appreciated!

Alexander Regueiro (Jun 03 2019 at 18:06, on Zulip):

but to summarise:
1) waiting for some checks on the last few commits on my ATB PR, and a push, both from Niko. then it's "LGTM" I believe.
2) advice on traits upcasting
3) decide what I can do next. maybe fixing impl-trait-in-bindings with notes from Niko, maybe just leaving this to him. alternatively, the invariant lifetimes issue or other minor ones I have on my to-do list.

nikomatsakis (Jun 03 2019 at 18:06, on Zulip):

I'll try to carve out some time for the trait upcasting things too

Alexander Regueiro (Jun 03 2019 at 18:06, on Zulip):

thanks!

nikomatsakis (Jun 03 2019 at 18:06, on Zulip):

@Sunjay Varma we had talked about doing some stuff with specialiation -- do you still feel like you might have time to explore that :)

Alexander Regueiro (Jun 03 2019 at 18:07, on Zulip):

anyway have a good meeting folks, bye!

nikomatsakis (Jun 03 2019 at 18:07, on Zulip):

Also, does it make sense to try and do a zoom call later today?

Sunjay Varma (Jun 03 2019 at 18:08, on Zulip):

@nikomatsakis Yes! I'll be honest I've been a bit stumped on what to do first. I feel like I'm starting to get a good handle on the RFC/existing code, so I guess the next step is to start to develop rules? Do you still want to use the specialization forest that chalk is generating now?

Sunjay Varma (Jun 03 2019 at 18:08, on Zulip):

I forget where exactly we left off on that

Sunjay Varma (Jun 03 2019 at 18:08, on Zulip):

(It's been a few weeks since we last talked about this!)

nikomatsakis (Jun 03 2019 at 18:08, on Zulip):

yeah tbh I don't remember either :)

nikomatsakis (Jun 03 2019 at 18:08, on Zulip):

I've not had time to re-read @Aaron Turon's blog post etc

nikomatsakis (Jun 03 2019 at 18:08, on Zulip):

but I think I was thinking -- ah yes I vaguely remember now --

nikomatsakis (Jun 03 2019 at 18:09, on Zulip):

that I thought the best starting point would be to try and imagine that some oracle was telling us which impl specializes which

nikomatsakis (Jun 03 2019 at 18:09, on Zulip):

and to work instead on augmenting the rules in chalk to do the right thing if we knew that information

nikomatsakis (Jun 03 2019 at 18:09, on Zulip):

and then to worry about the soundness problems, which amount to making certain specializations illegal

nikomatsakis (Jun 03 2019 at 18:09, on Zulip):

not sure how much of that I communicated

nikomatsakis (Jun 03 2019 at 18:10, on Zulip):

Oh, @detrumi -- I see you were working on reorganizing the tests...

Sunjay Varma (Jun 03 2019 at 18:11, on Zulip):

That's a good starting point! I'll see what I can do :)

Sunjay Varma (Jun 03 2019 at 18:12, on Zulip):

I'll reach out with any questions I have as they come up

Sunjay Varma (Jun 03 2019 at 18:12, on Zulip):

Thanks! :D

Sunjay Varma (Jun 03 2019 at 18:12, on Zulip):

/me writes down everything Niko just said so I don't forget

nikomatsakis (Jun 03 2019 at 18:13, on Zulip):

@Sunjay Varma take a look at this comment

nikomatsakis (Jun 03 2019 at 18:14, on Zulip):

I guess some next steps would be:

both of these ought to be no-ops

nikomatsakis (Jun 03 2019 at 18:14, on Zulip):

things get more interesting when we generate Overrides program clauses

nikomatsakis (Jun 03 2019 at 18:15, on Zulip):

(because, until then, Overrides would always be trivially false, so things like not { Overrides(..) } are just immediately true)

Sunjay Varma (Jun 03 2019 at 18:17, on Zulip):

(Sorry, I don't have a ton of time to engage with what you're saying right at this moment, but I am writing it all down so I can review it later and ask questions/talk more if we need to)

nikomatsakis (Jun 03 2019 at 18:21, on Zulip):

I'll drop some more notes in the issue

nikomatsakis (Jun 03 2019 at 18:22, on Zulip):

Also, does it make sense to try and do a zoom call later today?

I'm going to assume this does not make sense today -- which is fine, I'll use the time to write things down and/or do reviews.

Last update: Nov 12 2019 at 16:10UTC