Stream: wg-async-foundations

Topic: meeting 2020.01.28


tmandry (Jan 28 2020 at 17:53, on Zulip):

hello @WG-async-foundations! reminder, meeting is in 7 minutes

tmandry (Jan 28 2020 at 18:01, on Zulip):

ok, let's start

tmandry (Jan 28 2020 at 18:01, on Zulip):

@WG-async-foundations o/

tmandry (Jan 28 2020 at 18:02, on Zulip):

starting with triage

tmandry (Jan 28 2020 at 18:02, on Zulip):

Uncategorized issues

tmandry (Jan 28 2020 at 18:02, on Zulip):

Unsized locals and async ICE #68543

centril (Jan 28 2020 at 18:03, on Zulip):

doesn't seem like a priority cause it requires nightly

tmandry (Jan 28 2020 at 18:03, on Zulip):

yeah, it depends on an unstable feature

centril (Jan 28 2020 at 18:04, on Zulip):

let's call it P-low as well?

tmandry (Jan 28 2020 at 18:05, on Zulip):

I don't generally assign P-* labels in this meeting

tmandry (Jan 28 2020 at 18:05, on Zulip):

and I'm not entirely sure what the guidance is for those

tmandry (Jan 28 2020 at 18:05, on Zulip):

next up

tmandry (Jan 28 2020 at 18:05, on Zulip):

Bad error message with async main #68523

centril (Jan 28 2020 at 18:05, on Zulip):

How do we mark that something is explicitly low prio for the WG though? Do we need another label Async-Await-Low-Prio

tmandry (Jan 28 2020 at 18:06, on Zulip):

How do we mark that something is explicitly low prio for the WG though?

by not marking it as AsyncAwait-OnDeck or AsyncAwait-Focus

tmandry (Jan 28 2020 at 18:06, on Zulip):

we mark it as triaged, but nothing else

centril (Jan 28 2020 at 18:07, on Zulip):

https://github.com/rust-lang/rust/issues/68523

This can probably be checked wherever we check the fn main

tmandry (Jan 28 2020 at 18:08, on Zulip):

it's a good suggestion, not as important as errors that are just wrong, but

tmandry (Jan 28 2020 at 18:08, on Zulip):

I guess I'll mark as OnDeck since errors are a priority for us

tmandry (Jan 28 2020 at 18:09, on Zulip):

"expected bound lifetime parameter, found concrete lifetime" when passing a 'static reference to a closure that returns a future #68521

centril (Jan 28 2020 at 18:09, on Zulip):

We store the span in https://doc.rust-lang.org/nightly/nightly-rustc/syntax/ast/struct.FnHeader.html#structfield.asyncness but we don't propagate it to HIR

tmandry (Jan 28 2020 at 18:10, on Zulip):

ah, good, would you mind adding a comment to #68523?

centril (Jan 28 2020 at 18:10, on Zulip):

doing so :slight_smile: done

tmandry (Jan 28 2020 at 18:12, on Zulip):

"expected bound lifetime parameter, found concrete lifetime" when passing a 'static reference to a closure that returns a future #68521

it looks like this is basically a confusing error message involving lifetimes and async/await

tmandry (Jan 28 2020 at 18:12, on Zulip):

I've heard lots of reports about "errors around lifetimes" being confusing in general, fwiw

tmandry (Jan 28 2020 at 18:12, on Zulip):

so tagging as OnDeck

tmandry (Jan 28 2020 at 18:13, on Zulip):

ICE cause by tokio::spawn a async function return future opaque type #68477

centril (Jan 28 2020 at 18:14, on Zulip):

has assignee

tmandry (Jan 28 2020 at 18:14, on Zulip):

looks like @Aaron Hill is assigned to this already, so I'll upgrade to Focus so we can check in on it

tmandry (Jan 28 2020 at 18:15, on Zulip):

Rustc doesn't report line numbers for type errors within proc macro async fn within macro_rules macro #68430

tmandry (Jan 28 2020 at 18:16, on Zulip):

okay, so diagnostics, marking as OnDeck

centril (Jan 28 2020 at 18:16, on Zulip):

I suspect this is rather foundational

centril (Jan 28 2020 at 18:16, on Zulip):

so maybe not easy to fix

centril (Jan 28 2020 at 18:16, on Zulip):

ill cc petrochenkov

tmandry (Jan 28 2020 at 18:16, on Zulip):

ok, thanks

tmandry (Jan 28 2020 at 18:17, on Zulip):

that's all of the untriaged issues!

tmandry (Jan 28 2020 at 18:17, on Zulip):

let's check in on what we're working on

tmandry (Jan 28 2020 at 18:17, on Zulip):

Focus issues

tmandry (Jan 28 2020 at 18:18, on Zulip):

Unhelpful error messages from borrow checker with async/await #67765

tmandry (Jan 28 2020 at 18:19, on Zulip):

it looks like discussion on the PR (#67911) has continued, albeit slowly, perhaps

centril (Jan 28 2020 at 18:19, on Zulip):

doesn't seem likely to land right now

tmandry (Jan 28 2020 at 18:20, on Zulip):

ok, it will probably require @nikomatsakis to be involved in any case, and they're out right now

tmandry (Jan 28 2020 at 18:20, on Zulip):

moving on for now

tmandry (Jan 28 2020 at 18:20, on Zulip):

Confusing/incorrect error message with incoherent implementations and async blocks #67651

tmandry (Jan 28 2020 at 18:21, on Zulip):

this was assigned to @nikomatsakis, who I think said he wanted to continue doing some work on it

tmandry (Jan 28 2020 at 18:22, on Zulip):

"async fn presence affects an unrelated error message #66312" is the same status (probable dupe)

tmandry (Jan 28 2020 at 18:22, on Zulip):

async/await: awaiting inside a match block captures borrow too eagerly #57017

tmandry (Jan 28 2020 at 18:23, on Zulip):

has a PR (#65782)

tmandry (Jan 28 2020 at 18:23, on Zulip):

but that PR is blocked on more design discussion

centril (Jan 28 2020 at 18:24, on Zulip):

and probably some lang team involvement

tmandry (Jan 28 2020 at 18:25, on Zulip):

as I mentioned in the past, I think we need to come up with a strategy for unblocking things like this

tmandry (Jan 28 2020 at 18:25, on Zulip):

either dedicated "design discussions" in this WG, or sending them off to another team, or something

centril (Jan 28 2020 at 18:26, on Zulip):

If it is "changes to the spec" (e.g. "allow more code to compile and it's not obvious a bug") I think only the language team could approve such a change

centril (Jan 28 2020 at 18:27, on Zulip):

but preparing material in the WG is still helpful

tmandry (Jan 28 2020 at 18:27, on Zulip):

sure, maybe it's a design discussion to come up with a concise formulation of the problem, which can be reviewed by the lang team

centril (Jan 28 2020 at 18:28, on Zulip):

right

tmandry (Jan 28 2020 at 18:28, on Zulip):

I'm considering cc'ing them on this change, but also not hopeful that it will produce any meaningful result (without more effort)

tmandry (Jan 28 2020 at 18:28, on Zulip):

and would just add to noise, ultimately

tmandry (Jan 28 2020 at 18:29, on Zulip):

Ok, time's almost up, and I want to see if anyone is looking for something to work on?

tmandry (Jan 28 2020 at 18:29, on Zulip):

OnDeck issues

tmandry (Jan 28 2020 at 18:29, on Zulip):

is a good-sized list

centril (Jan 28 2020 at 18:30, on Zulip):

https://github.com/rust-lang/rust/issues/68523 seems to have been snatched just now

tmandry (Jan 28 2020 at 18:30, on Zulip):

excellent

tmandry (Jan 28 2020 at 18:32, on Zulip):
tmandry (Jan 28 2020 at 18:32, on Zulip):

these are a couple I picked out that seem high priority

tmandry (Jan 28 2020 at 18:32, on Zulip):

also @centril, I noticed you assigned yourself to #67204

tmandry (Jan 28 2020 at 18:33, on Zulip):

are you planning to do implementation work on it?

centril (Jan 28 2020 at 18:34, on Zulip):

I tried it for a while but hit some issues with what I think are look-ahead machinery bugs in the parser infra

centril (Jan 28 2020 at 18:34, on Zulip):

(https://github.com/Centril/rust/commits/async-slit-2015)

centril (Jan 28 2020 at 18:34, on Zulip):

so another approach may be necessary

centril (Jan 28 2020 at 18:35, on Zulip):

cc @Esteban K├╝ber

centril (Jan 28 2020 at 18:35, on Zulip):

It might take some time to get to, so if they Esteban wants to steal it from me that's fine

centril (Jan 28 2020 at 18:35, on Zulip):

I'll probably review in that case

tmandry (Jan 28 2020 at 18:36, on Zulip):

okay

tmandry (Jan 28 2020 at 18:38, on Zulip):

well, thanks for coming =)

tmandry (Jan 28 2020 at 18:52, on Zulip):

I've been thinking, I'd like our meetings to include a quick retrospective of what got done in the last week or two, instead of only looking at what's left to do. It's good to celebrate the progress we're making, I think.

To start, here are some issues we closed recently:

Last update: Feb 25 2020 at 04:05UTC