Stream: t-compiler

Topic: weekly meeting 2020-01-02 #54818


pnkfelix (Jan 02 2020 at 13:28, on Zulip):

Hi @T-compiler/meeting ; the triage meeting will be starting in 1 hours 32 minutes

pnkfelix (Jan 02 2020 at 13:29, on Zulip):

I will be doing pre-triage in a parallel topic

pnkfelix (Jan 02 2020 at 13:30, on Zulip):

This week we have scheduled checkin's with WG-learning and WG-llvm

pnkfelix (Jan 02 2020 at 13:30, on Zulip):

@Santiago Pastorino are you available to present any checkin stuff from WG-learning?

pnkfelix (Jan 02 2020 at 13:31, on Zulip):

/me just realized that one of the tasks for each T-compiler meeting should probably be to establish an appropriate zulip topics for next week's WG checkins

pnkfelix (Jan 02 2020 at 13:33, on Zulip):

as for WG-llvm, I think its fair to say its status as a WG may have been misplaced. We do need a dedicated zulip stream, but the lack of focused/specific goals has really made it a mostly inactive thing as far as formal WG's go.

Santiago Pastorino (Jan 02 2020 at 14:25, on Zulip):

Santiago Pastorino are you available to present any checkin stuff from WG-learning?

I'm packing to leave for vacations but is WG-learning time again?

Santiago Pastorino (Jan 02 2020 at 14:25, on Zulip):

didn't we do it last time?

pnkfelix (Jan 02 2020 at 14:27, on Zulip):

/me looks

pnkfelix (Jan 02 2020 at 14:28, on Zulip):

I have no idea when it was last visited. I'm just relying on the website to tell me the truth about the checkin schedule.

pnkfelix (Jan 02 2020 at 14:29, on Zulip):

anyway its no problem. @nikomatsakis I assume you can report in from WG-learning instead.

Santiago Pastorino (Jan 02 2020 at 14:29, on Zulip):

@pnkfelix let me check but this seems like the 2000 year problem that just happened in 2020 :joy:

pnkfelix (Jan 02 2020 at 14:30, on Zulip):

omg

Santiago Pastorino (Jan 02 2020 at 14:30, on Zulip):

we did a checkin on 2019-12-05

Santiago Pastorino (Jan 02 2020 at 14:30, on Zulip):

well maybe it's time? I don't think so

Santiago Pastorino (Jan 02 2020 at 14:31, on Zulip):

https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/weekly.20meeting.202019-12-05.20.2354818/near/182673783

pnkfelix (Jan 02 2020 at 14:33, on Zulip):

maybe part of issue is that we skipped last week's meeting

pnkfelix (Jan 02 2020 at 14:33, on Zulip):

so it seems like time has moved faster than it has. Or maybe other way around.

Santiago Pastorino (Jan 02 2020 at 14:35, on Zulip):

in any case, there's not much update from learning WG :)

pnkfelix (Jan 02 2020 at 15:01, on Zulip):

Hi @T-compiler/meeting! Add a :wave: emoji to show you're here :)

pnkfelix (Jan 02 2020 at 15:03, on Zulip):

oh lets have five minutes for ...

pnkfelix (Jan 02 2020 at 15:03, on Zulip):

Announcements

pnkfelix (Jan 02 2020 at 15:03, on Zulip):

Happy New Year

centril (Jan 02 2020 at 15:04, on Zulip):
pnkfelix (Jan 02 2020 at 15:04, on Zulip):

so this isn't exactly an announcement, but this is as good a time as any to ask

pnkfelix (Jan 02 2020 at 15:04, on Zulip):

I want to find out what IDE/editors the rustc-developers use for hacking on rustc

Zoxc (Jan 02 2020 at 15:05, on Zulip):

VS Code + rust_analyzer

pnkfelix (Jan 02 2020 at 15:05, on Zulip):

what's best here, a poll maybe

pnkfelix (Jan 02 2020 at 15:05, on Zulip):

/poll What IDE/editor do you use

centril (Jan 02 2020 at 15:05, on Zulip):

/me uses rust-analyzer + VSCode, hooked up via sshfs to my ubuntu server

pnkfelix (Jan 02 2020 at 15:05, on Zulip):

Oh, lets leave off whichj rust support system

pnkfelix (Jan 02 2020 at 15:06, on Zulip):

sorry I did not make this lear

pnkfelix (Jan 02 2020 at 15:06, on Zulip):

clear

pnkfelix (Jan 02 2020 at 15:06, on Zulip):

the reason why I'm asking

pnkfelix (Jan 02 2020 at 15:06, on Zulip):

is because I wwant to know which platforms we should put effort into improving the rust-analyzer experience on

pnkfelix (Jan 02 2020 at 15:06, on Zulip):

in order to encourage dogfooding of rust-analyzer

centril (Jan 02 2020 at 15:06, on Zulip):

we have stats for users I think

simulacrum (Jan 02 2020 at 15:06, on Zulip):

hm, isn't rust-analyzer universal via LSP anyway? (or pretty much so?)

simulacrum (Jan 02 2020 at 15:07, on Zulip):

unless I misunderstood platform

pnkfelix (Jan 02 2020 at 15:07, on Zulip):

I just want to make sure the barrier for entry is truly minimized

pnkfelix (Jan 02 2020 at 15:07, on Zulip):

maybe @simulacrum you are asserting that LSP integration is trivial for all platforms of note? Honestly I do not know either way here

centril (Jan 02 2020 at 15:08, on Zulip):

@matklad would ;)

Zoxc (Jan 02 2020 at 15:08, on Zulip):

There's Rust specific extensions on top of LSP which are important too

nikomatsakis (Jan 02 2020 at 15:08, on Zulip):

I'm interested in a survey of rustc devs regardless; I think it would be good to include material in the rustc-guide for how to setup as well

simulacrum (Jan 02 2020 at 15:09, on Zulip):

I guess -- maybe I should ask -- what are we interested in?

centril (Jan 02 2020 at 15:09, on Zulip):

it's very easy for VSCode + RA; just install RA and you're basically good to go

simulacrum (Jan 02 2020 at 15:09, on Zulip):

Editor? OS?

pnkfelix (Jan 02 2020 at 15:10, on Zulip):

For me the questions are: 1. Which IDE/editor do you use to hack on rustc; 2. Do you use rust-analyzer? 3. If yes, how does your workflow with it compare to prior to RA? 4. If no, why not?

pnkfelix (Jan 02 2020 at 15:11, on Zulip):

lets plan to do the survey via email or something

pnkfelix (Jan 02 2020 at 15:11, on Zulip):

because I think relying on responses at meeting is going to omit contributors.

pnkfelix (Jan 02 2020 at 15:11, on Zulip):

(niko had already advised me to do this via email. I just, didn't.)

pnkfelix (Jan 02 2020 at 15:11, on Zulip):

okay next up

simulacrum (Jan 02 2020 at 15:12, on Zulip):

(one note -- we should maybe send it to some wider group, I guess -- not sure that compiler@ is good enough)

centril (Jan 02 2020 at 15:12, on Zulip):

compiler-contributors @ ?

pnkfelix (Jan 02 2020 at 15:12, on Zulip):

here is today's hackmd agenda.

nikomatsakis (Jan 02 2020 at 15:12, on Zulip):

compiler-team@rust-lang.org includes contributors

nikomatsakis (Jan 02 2020 at 15:12, on Zulip):

jfyi

nikomatsakis (Jan 02 2020 at 15:12, on Zulip):

(the -private variation does not)

simulacrum (Jan 02 2020 at 15:13, on Zulip):

right, I meant that even contributors is pretty limited

simulacrum (Jan 02 2020 at 15:13, on Zulip):

maybe an inside rust blog post would be a better fit?

centril (Jan 02 2020 at 15:13, on Zulip):

maybe github issue?

pnkfelix (Jan 02 2020 at 15:13, on Zulip):

oh there was one more announcement: I think my report future incompat RFC is near its "final form"

pnkfelix (Jan 02 2020 at 15:13, on Zulip):

maybe an inside rust blog post would be a better fit?

interesting idea.

pnkfelix (Jan 02 2020 at 15:14, on Zulip):

okay so, today's agenda

pnkfelix (Jan 02 2020 at 15:14, on Zulip):

we have a lot of beta noms to get through, I assume in part because we skipped a week's meeting.

pnkfelix (Jan 02 2020 at 15:14, on Zulip):

beta-nom: "Do not ICE on lifetime error involving closures" #67687

centril (Jan 02 2020 at 15:15, on Zulip):

a bit heavy

nikomatsakis (Jan 02 2020 at 15:15, on Zulip):

seems pretty harmless to me

pnkfelix (Jan 02 2020 at 15:16, on Zulip):

does it have tests for generator case?

pnkfelix (Jan 02 2020 at 15:16, on Zulip):

I agree it seems harmless. Just surprised that of the two main cases it added, I think it only tests one?

pnkfelix (Jan 02 2020 at 15:16, on Zulip):

but that's more a comment on the PR itself, rather than the choice on whether to backport

pnkfelix (Jan 02 2020 at 15:17, on Zulip):

anyway I figure beta-approved.

matklad (Jan 02 2020 at 15:17, on Zulip):

hm, isn't rust-analyzer universal via LSP anyway? (or pretty much so?)

@simulacrum 80% so, as we do indeed use custom extensions, which we really should be upstreaming into the protocol...

pnkfelix (Jan 02 2020 at 15:17, on Zulip):

beta-nom: "Treat extern statics just like statics in the “const pointer to static” representation" #67630

pnkfelix (Jan 02 2020 at 15:18, on Zulip):

not landed on master

pnkfelix (Jan 02 2020 at 15:18, on Zulip):

we can wait until next week on this?

centril (Jan 02 2020 at 15:18, on Zulip):

discussion on PR doesn't feel very confidence-inspiring

pnkfelix (Jan 02 2020 at 15:19, on Zulip):

lets just hold off making a decision for now

centril (Jan 02 2020 at 15:19, on Zulip):

waiting sgtm... @nikomatsakis did t-lang have mtg today btw?

pnkfelix (Jan 02 2020 at 15:19, on Zulip):

beta-nom: "Use the correct type for static qualifs" #67621

nikomatsakis (Jan 02 2020 at 15:19, on Zulip):

waiting sgtm... nikomatsakis did t-lang have mtg today btw?

yes

nikomatsakis (Jan 02 2020 at 15:19, on Zulip):

I at least didn't cancel it

nikomatsakis (Jan 02 2020 at 15:20, on Zulip):

who will show I don't know :)

centril (Jan 02 2020 at 15:21, on Zulip):

@Matthew Jasper is this somehow related to default binding modes?

centril (Jan 02 2020 at 15:21, on Zulip):

(ugh... but statics use identifiers, not patterns... so what am I saying)

nikomatsakis (Jan 02 2020 at 15:22, on Zulip):

doesn't look like it :)

pnkfelix (Jan 02 2020 at 15:22, on Zulip):

the code itself here is maybe subtle, but this change seems like purely a good thing?

nikomatsakis (Jan 02 2020 at 15:23, on Zulip):

yes, the change seemed a bit subtle, but smallish and I figured I trusted @Matthew Jasper on this point

pnkfelix (Jan 02 2020 at 15:24, on Zulip):

do you understand why the test needed to use && instead of just & ?

pnkfelix (Jan 02 2020 at 15:24, on Zulip):

oh I guess the thing being promoted is the &Option<String>

pnkfelix (Jan 02 2020 at 15:24, on Zulip):

okay I guess i can follow along with it.

pnkfelix (Jan 02 2020 at 15:24, on Zulip):

beta-accepted

pnkfelix (Jan 02 2020 at 15:25, on Zulip):

beta-nom: "Don’t suppress move errors for union fields" #67314

pnkfelix (Jan 02 2020 at 15:26, on Zulip):

am I wrong to be nervous about code using get_or_insert having unexpectred effects...

pnkfelix (Jan 02 2020 at 15:27, on Zulip):

oh I guess its fine

pnkfelix (Jan 02 2020 at 15:27, on Zulip):

(the state being updated is newly added here, not preexisting)

centril (Jan 02 2020 at 15:27, on Zulip):

am I wrong to be nervous about code using get_or_insert having unexpectred effects...

Stateful code in general gives me that feeling :P

pnkfelix (Jan 02 2020 at 15:27, on Zulip):

okay I loaded enough of this back into my mental cache to restore confidence personally

nikomatsakis (Jan 02 2020 at 15:28, on Zulip):

I found it a bit tricky to review, which is why I asked for a comment (when reviewing, the comment is in the PR)

nikomatsakis (Jan 02 2020 at 15:28, on Zulip):

but I think it's harmless enough :)

pnkfelix (Jan 02 2020 at 15:28, on Zulip):

beta approved I figure.

pnkfelix (Jan 02 2020 at 15:28, on Zulip):

beta-nom: "Do not ICE on unnamed future" #67289

pnkfelix (Jan 02 2020 at 15:30, on Zulip):

/me really needs to figure out Zulip API... gotta script this stuff...

pnkfelix (Jan 02 2020 at 15:30, on Zulip):

yeah this is truly harmless

nikomatsakis (Jan 02 2020 at 15:30, on Zulip):

adding the emojis is annoying, yeah

pnkfelix (Jan 02 2020 at 15:30, on Zulip):

beta approved

pnkfelix (Jan 02 2020 at 15:31, on Zulip):

beta-nom: "Fix up Command Debug output when arg0 is specified." #67219

centril (Jan 02 2020 at 15:31, on Zulip):

wrong team?

pnkfelix (Jan 02 2020 at 15:31, on Zulip):

or both teams

pnkfelix (Jan 02 2020 at 15:31, on Zulip):

why are we on this

pnkfelix (Jan 02 2020 at 15:31, on Zulip):

@simulacrum ^ ?

simulacrum (Jan 02 2020 at 15:31, on Zulip):

hm, I forget what my logic was

centril (Jan 02 2020 at 15:31, on Zulip):

that is the question

simulacrum (Jan 02 2020 at 15:32, on Zulip):

I guess I was thinking "someone should see this" :)

pnkfelix (Jan 02 2020 at 15:32, on Zulip):

heh

simulacrum (Jan 02 2020 at 15:32, on Zulip):

I'm not sure libs has a process inplace for approving things

pnkfelix (Jan 02 2020 at 15:32, on Zulip):

gotcha

pnkfelix (Jan 02 2020 at 15:32, on Zulip):

the only issue I can see

pnkfelix (Jan 02 2020 at 15:32, on Zulip):

is the formatting of the self.program

pnkfelix (Jan 02 2020 at 15:32, on Zulip):

putting it in square brackets

pnkfelix (Jan 02 2020 at 15:33, on Zulip):

a beta backport here is essentially instantly committing to that choice

pnkfelix (Jan 02 2020 at 15:33, on Zulip):

versus allowing some play over course of time in nightly, e.g. letting someone claim it should not have any annotation, or a different one...

centril (Jan 02 2020 at 15:33, on Zulip):

isn't it unstable?

simulacrum (Jan 02 2020 at 15:34, on Zulip):

we are currently printing rustc rustc arg arg arg vs. rustc arg arg arg

simulacrum (Jan 02 2020 at 15:34, on Zulip):

(on beta)

centril (Jan 02 2020 at 15:34, on Zulip):

a beta backport here is essentially instantly committing to that choice

We don't guarantee the output of Debug impls generally

pnkfelix (Jan 02 2020 at 15:34, on Zulip):

yeah okay we've gotta do something about the regression

pnkfelix (Jan 02 2020 at 15:34, on Zulip):

but for that should we revert the original PR ?

pnkfelix (Jan 02 2020 at 15:34, on Zulip):

hmm

simulacrum (Jan 02 2020 at 15:35, on Zulip):

Ah I see I actually wrote down my reasoning: "I tagged with T-compiler and T-libs; but I feel fairly confident that we just want to backport this. If T-compiler decides not to approve we'll want to ping T-libs as they don't generally notice otherwise I think."

nikomatsakis (Jan 02 2020 at 15:35, on Zulip):

heh

nikomatsakis (Jan 02 2020 at 15:35, on Zulip):

this is specifically the Debug impl output only?

pnkfelix (Jan 02 2020 at 15:36, on Zulip):

well I guess we can leave the final decision here in the hands of T-libs. I personally would be more inclined to revert the PR that added the feature to Command.

nikomatsakis (Jan 02 2020 at 15:36, on Zulip):

if so, it seems ... odd but "ok"

nikomatsakis (Jan 02 2020 at 15:36, on Zulip):

I mean not odd just kind of random, not an established convention

pnkfelix (Jan 02 2020 at 15:36, on Zulip):

and independnently we should figure out where we are exposing the use of the Debug format from within rustc (or cargo ?)

simulacrum (Jan 02 2020 at 15:36, on Zulip):

okay, I will ping libs and try and get it raised in time with folks

pnkfelix (Jan 02 2020 at 15:37, on Zulip):

I can write a comment with my own preference here on the issue, just to have it recorded.

nikomatsakis (Jan 02 2020 at 15:37, on Zulip):

(but yes I agree it doesn't really seem like 'our call')

pnkfelix (Jan 02 2020 at 15:37, on Zulip):

I'll try to remember to do so post meeting.

pnkfelix (Jan 02 2020 at 15:37, on Zulip):

okay that's all the beta-noms

pnkfelix (Jan 02 2020 at 15:37, on Zulip):

there's just one stable nom

pnkfelix (Jan 02 2020 at 15:38, on Zulip):

stable-nom "Do not ICE on lifetime error involving closures" #67687

pnkfelix (Jan 02 2020 at 15:39, on Zulip):

already noted that this seemed a little bit heavy weight

centril (Jan 02 2020 at 15:39, on Zulip):

is it just me or is hackmd having a fit?

pnkfelix (Jan 02 2020 at 15:39, on Zulip):

(e.g. can the injected .typeck_tables_of(self.mir_def_id) cause new errors to arise? Its the kind of thing I don't mind landing on beta, but the bar for stable is higher...)

pnkfelix (Jan 02 2020 at 15:40, on Zulip):

is it just me or is hackmd having a fit?

i'm getting that too

pnkfelix (Jan 02 2020 at 15:41, on Zulip):

anyway I'm inclined to say we don't stable-approve #67687, but I'm open to counter-arguments.

centril (Jan 02 2020 at 15:41, on Zulip):

I feel similarly inclined but not strongly

pnkfelix (Jan 02 2020 at 15:41, on Zulip):

@nikomatsakis do you now if the lookuip I mentioned above is guaranteed to not panic?

Esteban Küber (Jan 02 2020 at 15:41, on Zulip):

I'm ok with not approving it for stable

Esteban Küber (Jan 02 2020 at 15:42, on Zulip):

Noting that the pack of generator test wouldn't affect stable

pnkfelix (Jan 02 2020 at 15:42, on Zulip):

okay, I figure if the PR author acquesces then we don't need to debate. :)

nikomatsakis (Jan 02 2020 at 15:42, on Zulip):

nikomatsakis do you now if the lookuip I mentioned above is guaranteed to not panic?

you mean the bug! call?

pnkfelix (Jan 02 2020 at 15:42, on Zulip):

no, the .typeck_tables_of(self.mir_def_id) call

pnkfelix (Jan 02 2020 at 15:42, on Zulip):

I guess that itself wont panic

nikomatsakis (Jan 02 2020 at 15:42, on Zulip):

that is guaranteed to be ok

nikomatsakis (Jan 02 2020 at 15:42, on Zulip):

iiuc

nikomatsakis (Jan 02 2020 at 15:42, on Zulip):

since we are in borrowck

pnkfelix (Jan 02 2020 at 15:42, on Zulip):

and likewise .node_type(fn_hir_id) ?

nikomatsakis (Jan 02 2020 at 15:43, on Zulip):

though whether the bug! could trigger .. I think no but I wouldn't be shocked to be wrong

Esteban Küber (Jan 02 2020 at 15:43, on Zulip):

I think the only case that would panic is if we have very malformed code that has gone through multiple layers of recovery.

centril (Jan 02 2020 at 15:43, on Zulip):

...which is not unheard of... ;)

pnkfelix (Jan 02 2020 at 15:43, on Zulip):

how bad is the diagnostic when we hit this case

Esteban Küber (Jan 02 2020 at 15:44, on Zulip):

Ice

pnkfelix (Jan 02 2020 at 15:44, on Zulip):

pretty bad, no span

Esteban Küber (Jan 02 2020 at 15:44, on Zulip):

Without any info

pnkfelix (Jan 02 2020 at 15:44, on Zulip):

right

pnkfelix (Jan 02 2020 at 15:44, on Zulip):

can we add a span to that bug! ?

pnkfelix (Jan 02 2020 at 15:44, on Zulip):

(and if we did, would I be more inclined to support backport... hmm....)

Esteban Küber (Jan 02 2020 at 15:44, on Zulip):

Which makes me think we could dump all non emitted disgnotivbuilders when we hit an Ice

centril (Jan 02 2020 at 15:45, on Zulip):

I'm mostly thinking that we need to discuss other things

centril (Jan 02 2020 at 15:45, on Zulip):

(on the mtg)

pnkfelix (Jan 02 2020 at 15:45, on Zulip):

okay

pnkfelix (Jan 02 2020 at 15:46, on Zulip):

How about if we wait a week to decide here too?

nikomatsakis (Jan 02 2020 at 15:46, on Zulip):

seems ok

pnkfelix (Jan 02 2020 at 15:46, on Zulip):

the lack of diagnostic info here makes me pause from just turning down backport outright

pnkfelix (Jan 02 2020 at 15:46, on Zulip):

okay

centril (Jan 02 2020 at 15:46, on Zulip):

(does anyone recall the agenda now that hackmd seems borked? :P)

pnkfelix (Jan 02 2020 at 15:46, on Zulip):

of course, we have lost the hackmd now

pnkfelix (Jan 02 2020 at 15:46, on Zulip):

but there's still the standing agenda I generated that from (linked in zulip topic)

centril (Jan 02 2020 at 15:47, on Zulip):

let's discuss the async regression perhaps?

pnkfelix (Jan 02 2020 at 15:47, on Zulip):

nominated: "Major async/await compiler performance regression" #67706

pnkfelix (Jan 02 2020 at 15:47, on Zulip):

PR #65244 was identified as cause, and there was a trial revert PR posted

pnkfelix (Jan 02 2020 at 15:48, on Zulip):

trial revert is PR #67768

centril (Jan 02 2020 at 15:48, on Zulip):

Same PR also caused #67757

nikomatsakis (Jan 02 2020 at 15:48, on Zulip):

seems clear we should revert

centril (Jan 02 2020 at 15:48, on Zulip):

I think so too; unfortunate :frown:

pnkfelix (Jan 02 2020 at 15:48, on Zulip):

does anyone present object to reverting PR #65244 ?

nikomatsakis (Jan 02 2020 at 15:48, on Zulip):

yeah, ideally somebody would work to figure out and get it landed

nikomatsakis (Jan 02 2020 at 15:49, on Zulip):

I guess we can kick that to #wg-async-foundations (cc @tmandry, we should discuss...)

centril (Jan 02 2020 at 15:49, on Zulip):

did we forget to perftest the PR...? seems like we did an oopsie here

pnkfelix (Jan 02 2020 at 15:49, on Zulip):

the revert was perftested?

pnkfelix (Jan 02 2020 at 15:50, on Zulip):

or you mean the original PR #65244 ?

centril (Jan 02 2020 at 15:50, on Zulip):

no I mean the original PR

pnkfelix (Jan 02 2020 at 15:50, on Zulip):

to be fair, its not like we perftest everything

pnkfelix (Jan 02 2020 at 15:50, on Zulip):

I guess we may want guidelines about that

centril (Jan 02 2020 at 15:51, on Zulip):

yeah maybe; though no biggie, we figured it out in time after

pnkfelix (Jan 02 2020 at 15:51, on Zulip):

heuristics for what kinds of changes in past have caused perf issues. (At least, I think we've had similar problems when adding core traits in the past?)

nikomatsakis (Jan 02 2020 at 15:51, on Zulip):

(side note, I think part of the problem was the combination of impl + "do we want to do this", which led to a long delay, etc etc)

nikomatsakis (Jan 02 2020 at 15:51, on Zulip):

at least I think for me that tends to lead to "suboptimal reviewing"

nikomatsakis (Jan 02 2020 at 15:51, on Zulip):

anyway I think it's fine, this is what nightly is for :)

centril (Jan 02 2020 at 15:51, on Zulip):

great point @nikomatsakis

simulacrum (Jan 02 2020 at 15:51, on Zulip):

yes I would not expect perftest necessarily on a PR like this

simulacrum (Jan 02 2020 at 15:51, on Zulip):

(certainly e.g. I didn't suggest it, though was aware of the PR)

simulacrum (Jan 02 2020 at 15:52, on Zulip):

otoh, we should be somewhat eager to revert I think given (unexpected) perf failures

simulacrum (Jan 02 2020 at 15:52, on Zulip):

which roughly seems consensus anyway

pnkfelix (Jan 02 2020 at 15:52, on Zulip):

okay I left a comment

pnkfelix (Jan 02 2020 at 15:52, on Zulip):

I also wanted us to discuss a beta regression

pnkfelix (Jan 02 2020 at 15:52, on Zulip):

nominated: "not Send due to await retainment" #67611

pnkfelix (Jan 02 2020 at 15:53, on Zulip):

mostly I wanted to make sure someone gets assigned to look into this.

pnkfelix (Jan 02 2020 at 15:53, on Zulip):

@nikomatsakis do you think we can "just" delegate to the Async-Await WG ?

nikomatsakis (Jan 02 2020 at 15:54, on Zulip):

/me looks

pnkfelix (Jan 02 2020 at 15:54, on Zulip):

if we do not know if anyone has time

pnkfelix (Jan 02 2020 at 15:54, on Zulip):

then I will assign this to myself

pnkfelix (Jan 02 2020 at 15:54, on Zulip):

actually, let me just assign it to myself, if only to follow up and find someone else to actually investigate

nikomatsakis (Jan 02 2020 at 15:54, on Zulip):

I think it'd be ok to delegate but we should leave a comment indicating the urgency

nikomatsakis (Jan 02 2020 at 15:55, on Zulip):

that said, it won't be looked at likely until next tue

nikomatsakis (Jan 02 2020 at 15:55, on Zulip):

(if we just do that)

pnkfelix (Jan 02 2020 at 15:55, on Zulip):

okay. I self assigned. If I find time in next few days then I'll look at it

nikomatsakis (Jan 02 2020 at 15:55, on Zulip):

either way we should add the A-async-await label, and discuss probably in #wg-async-foundations

nikomatsakis (Jan 02 2020 at 15:55, on Zulip):

(I could try to look a bit tomorrow if desired too)

nikomatsakis (Jan 02 2020 at 15:56, on Zulip):

I wonder if this too is related to #65244?

pnkfelix (Jan 02 2020 at 15:56, on Zulip):

what tag do I add so that WG-Async-Await will triage this themselves?

nikomatsakis (Jan 02 2020 at 15:56, on Zulip):

it will already happen

nikomatsakis (Jan 02 2020 at 15:56, on Zulip):

it has the A-async-await label

nikomatsakis (Jan 02 2020 at 15:56, on Zulip):

and it does NOT have the AsyncAwait-triaged label

pnkfelix (Jan 02 2020 at 15:56, on Zulip):

okay

nikomatsakis (Jan 02 2020 at 15:56, on Zulip):

that's the key combination :)

pnkfelix (Jan 02 2020 at 15:57, on Zulip):

okay

pnkfelix (Jan 02 2020 at 15:57, on Zulip):

So

pnkfelix (Jan 02 2020 at 15:57, on Zulip):

the only other thing is WG checkins

pnkfelix (Jan 02 2020 at 15:57, on Zulip):

@Santiago Pastorino said that somethnig may be funky with the WG checkin scheduling

pnkfelix (Jan 02 2020 at 15:58, on Zulip):

but according to our own web site, today is WG-llvm and WG-learning

pnkfelix (Jan 02 2020 at 15:58, on Zulip):

I'll just assert that there's nothing to report from WG-llvm (other than my own insight into tracking down a particular LLVM bug)

nikomatsakis (Jan 02 2020 at 15:58, on Zulip):

I don't think much new has happened with learning, we're about the land the "intro to Ty" chapter

pnkfelix (Jan 02 2020 at 15:59, on Zulip):

((tl;dr: sometimes you can use clang rather than trying to construct a .ll input)

nikomatsakis (Jan 02 2020 at 15:59, on Zulip):

well I shouldn't say "we"

nikomatsakis (Jan 02 2020 at 15:59, on Zulip):

as I just reviewed it :P

pnkfelix (Jan 02 2020 at 15:59, on Zulip):

that is part of landing process

nikomatsakis (Jan 02 2020 at 15:59, on Zulip):

I'm not really the best to give a report though

pnkfelix (Jan 02 2020 at 15:59, on Zulip):

anyway sounds great

nikomatsakis (Jan 02 2020 at 15:59, on Zulip):

but I think that's exciting news regardless

pnkfelix (Jan 02 2020 at 15:59, on Zulip):

@Santiago Pastorino said not much had happened

nikomatsakis (Jan 02 2020 at 16:00, on Zulip):

yeah, I do think there was some bug in the scheduling, as I definitely recall doing llvm/learning recently

nikomatsakis (Jan 02 2020 at 16:00, on Zulip):

or maybe as you said it's just b/c holidays

pnkfelix (Jan 02 2020 at 16:00, on Zulip):

okay then. great meeting all, we actually got through all the beta nominations! That's an accomplsihment.

pnkfelix (Jan 02 2020 at 16:00, on Zulip):

of sorts

centril (Jan 02 2020 at 16:00, on Zulip):

/me is really excited about the crate-splitting that has been going on, e.g. by @cjgillot :slight_smile:

nikomatsakis (Jan 02 2020 at 16:00, on Zulip):

btw-- an announcement-- we have the next planning meeting scheduled for tomorrow

pnkfelix (Jan 02 2020 at 16:00, on Zulip):

oh right

pnkfelix (Jan 02 2020 at 16:00, on Zulip):

gotta get those proposal PR's posted, all!

pnkfelix (Jan 02 2020 at 16:01, on Zulip):

anyway, thanks for attending everyone in @T-compiler/meeting !

nikomatsakis (Jan 02 2020 at 16:02, on Zulip):

/me goes to do some reviewing

Last update: Jan 21 2020 at 09:15UTC