Stream: t-compiler

Topic: weekly meeting 2019-10-31 #54818


pnkfelix (Oct 31 2019 at 09:10, on Zulip):

Hi @T-compiler/meeting ; the triage meeting will be starting in 4 hours 50 minutes

pnkfelix (Oct 31 2019 at 09:12, on Zulip):

(if you are surprised by the start time, it is because the USA and the EU are out-of-sync with respect to handling of daylight savings time.)

pnkfelix (Oct 31 2019 at 09:12, on Zulip):

also, :pumpkin: Happy Halloween! :pumpkin: :ghost: :smiling_devil:

pnkfelix (Oct 31 2019 at 09:13, on Zulip):

I will be doing pre-triage in a parallel topic

pnkfelix (Oct 31 2019 at 09:14, on Zulip):

this weeks WG check-in is from PGO (profile-guided optimization) and Pipelining

pnkfelix (Oct 31 2019 at 09:14, on Zulip):

@mw , are you around to provide an update for PGO ?

pnkfelix (Oct 31 2019 at 09:15, on Zulip):

and for Pipelining, will either of @nnethercote or @Alex Crichton be around to provide an update on that front?

pnkfelix (Oct 31 2019 at 09:37, on Zulip):

:exclamation: one nominated issue for today is "Some features can no longer be controlled by conditional compilation" #65860 ; I recommend all members of @T-compiler review the thread there, because I'm hoping to discuss it today and hopefully document a clearer picture of our policy and values here.

pnkfelix (Oct 31 2019 at 10:06, on Zulip):

(and also as part of that topic, we should talk about whether to revert PR #65742 given the problems users have raised.)

pnkfelix (Oct 31 2019 at 10:07, on Zulip):

((i do not know how many of those problems would be addressed by instead landing PR #65974 though))

mw (Oct 31 2019 at 10:35, on Zulip):

@pnkfelix yes, I'm around. I need to leave ~10 min early though.

Alex Crichton (Oct 31 2019 at 13:11, on Zulip):

@pnkfelix I should be around yeah, but there won't be much of an update to give!

pnkfelix (Oct 31 2019 at 13:52, on Zulip):

(just a friendly reminder to @T-compiler/meeting that the meeting will start in about 8 minutes

pnkfelix (Oct 31 2019 at 14:02, on Zulip):

okay, lets get started @T-compiler/meeting

pnkfelix (Oct 31 2019 at 14:02, on Zulip):

first we'll have five minutes for Announcements

pnkfelix (Oct 31 2019 at 14:02, on Zulip):

I believe the release is a week away

pnkfelix (Oct 31 2019 at 14:02, on Zulip):

which seems noteworthy to me: Not much time left to resolve beta regressions

simulacrum (Oct 31 2019 at 14:03, on Zulip):

Yes, release is 1 week out. I've not checked in on beta regressions this cycle unfortunately.

pnkfelix (Oct 31 2019 at 14:04, on Zulip):

there's 6 open stable-to-beta tickets

pnkfelix (Oct 31 2019 at 14:04, on Zulip):

maybe we'll take time during the meeting to go over them explicitly

pnkfelix (Oct 31 2019 at 14:05, on Zulip):

(heh, just added another one)

pnkfelix (Oct 31 2019 at 14:07, on Zulip):

okay then, lets get started

pnkfelix (Oct 31 2019 at 14:07, on Zulip):

as a general note, T-compiler has 34 open P-high bugs, 17 of which are unassigned

pnkfelix (Oct 31 2019 at 14:08, on Zulip):

so if you're looking for something to do, browse over the latter list

pnkfelix (Oct 31 2019 at 14:08, on Zulip):

having said that, I didn't have a chance today to formally walk over all the P-high issues. I'll do that after the meeting instead

pnkfelix (Oct 31 2019 at 14:09, on Zulip):

So, next up: beta-nominations

pnkfelix (Oct 31 2019 at 14:09, on Zulip):

There is one beta nomination PR: "save-analysis: Account for async desugaring in async fn return types" #65936

pnkfelix (Oct 31 2019 at 14:10, on Zulip):

this fixes #65590

pnkfelix (Oct 31 2019 at 14:10, on Zulip):

which ... from my reading, is a stable-to-beta regression. Does that sounds correct?

nikomatsakis (Oct 31 2019 at 14:11, on Zulip):

I believe so

pnkfelix (Oct 31 2019 at 14:11, on Zulip):

I'm going to assume that it is and tag it as such

nikomatsakis (Oct 31 2019 at 14:11, on Zulip):

although

nikomatsakis (Oct 31 2019 at 14:11, on Zulip):

well not really

nikomatsakis (Oct 31 2019 at 14:11, on Zulip):

because async is not on stable?

pnkfelix (Oct 31 2019 at 14:11, on Zulip):

oh

nikomatsakis (Oct 31 2019 at 14:11, on Zulip):

it's a fix to the feature that will become stable

pnkfelix (Oct 31 2019 at 14:11, on Zulip):

hmm

pnkfelix (Oct 31 2019 at 14:11, on Zulip):

yes yes

pnkfelix (Oct 31 2019 at 14:12, on Zulip):

okay. interesting corner case in our labelling system

pnkfelix (Oct 31 2019 at 14:12, on Zulip):

since thats the sort of case that should absolutely qualify for beta backport, I thiink...

pnkfelix (Oct 31 2019 at 14:12, on Zulip):

(right? well ... maybe ...)

nikomatsakis (Oct 31 2019 at 14:12, on Zulip):

I think it should, but it is not a regression in the technical sense :)

pnkfelix (Oct 31 2019 at 14:12, on Zulip):

right

pnkfelix (Oct 31 2019 at 14:13, on Zulip):

/me muses on a Pending-Feature label (or About-To-Be-Stabilzed)

nikomatsakis (Oct 31 2019 at 14:13, on Zulip):

I guess you can view it as a regression from the general state of "IDE worked for stable rust code w/o ICEs" :)

pnkfelix (Oct 31 2019 at 14:13, on Zulip):

there is that

pnkfelix (Oct 31 2019 at 14:13, on Zulip):

anyway, sounds good. Approved for backport

pnkfelix (Oct 31 2019 at 14:14, on Zulip):

There are no stable nominations, thank goodness

pnkfelix (Oct 31 2019 at 14:15, on Zulip):

so, lets now walk through the stable-to-beta regressions themselves

pnkfelix (Oct 31 2019 at 14:15, on Zulip):

there are 7 of them: stable-to-beta regressions

pnkfelix (Oct 31 2019 at 14:15, on Zulip):

beta regr 1/7: "ptr types not default" #65525

nikomatsakis (Oct 31 2019 at 14:16, on Zulip):

Likely due to #64710 exposing bindgen bug rust-lang/rust-bindgen#1627. Updating bindgen fixes the issue.

nikomatsakis (Oct 31 2019 at 14:16, on Zulip):

(woah, do arbitrary rust-lang/xxx#123 links work?)

pnkfelix (Oct 31 2019 at 14:16, on Zulip):

okay then. weird.

nikomatsakis (Oct 31 2019 at 14:16, on Zulip):

(I guess zulip improved their linkifiers)

nikomatsakis (Oct 31 2019 at 14:16, on Zulip):

seems like we should close

pnkfelix (Oct 31 2019 at 14:17, on Zulip):

okay

mw (Oct 31 2019 at 14:17, on Zulip):

yeah, it's a bug in bindgen, not much the compiler can do, right?

nikomatsakis (Oct 31 2019 at 14:18, on Zulip):

(anyway, it's due to a libs change)

pnkfelix (Oct 31 2019 at 14:18, on Zulip):

closed

mw (Oct 31 2019 at 14:18, on Zulip):

wow, two levels of "not our problem" :P

pnkfelix (Oct 31 2019 at 14:18, on Zulip):

beta regr 2/7: "non-empty glob statement regression 1.39" #65523

pnkfelix (Oct 31 2019 at 14:18, on Zulip):

was this fixed by #65539 ?

nikomatsakis (Oct 31 2019 at 14:19, on Zulip):

it looks like it became a warning

nikomatsakis (Oct 31 2019 at 14:19, on Zulip):

(ah, this is why -- in the lang team meeting when we discussed #62334 -- I thought that there was a regression involved)

pnkfelix (Oct 31 2019 at 14:19, on Zulip):

can I ask someone to identify a mcve for this bug and double-check that its fixed ?

pnkfelix (Oct 31 2019 at 14:20, on Zulip):

(it seems very very likely that it is)

nikomatsakis (Oct 31 2019 at 14:20, on Zulip):

it looks like it became a warning

(my reasoning here is that the message in the crater log is quite specific, and that error became a warning)

pnkfelix (Oct 31 2019 at 14:20, on Zulip):

(it might be as simple as just downloading the relevant crate and making sure it builds now with just a warning instead of an error)

pnkfelix (Oct 31 2019 at 14:20, on Zulip):

anyway it seems very likely to have been fixed

pnkfelix (Oct 31 2019 at 14:20, on Zulip):

so lets move along

pnkfelix (Oct 31 2019 at 14:21, on Zulip):

beta regr 3/7: "array impls <= 32 1.39 regression" #65522

pnkfelix (Oct 31 2019 at 14:21, on Zulip):

believed to be another bindgen bug ... ?

nikomatsakis (Oct 31 2019 at 14:21, on Zulip):

I think so

mw (Oct 31 2019 at 14:21, on Zulip):

do we need a backport of the fix for #65523?

nikomatsakis (Oct 31 2019 at 14:21, on Zulip):

looks like maybe the same bug, no?

nikomatsakis (Oct 31 2019 at 14:22, on Zulip):

do we need a backport of the fix for #65523?

uh, good point, yes

pnkfelix (Oct 31 2019 at 14:22, on Zulip):

lets beta nom that

pnkfelix (Oct 31 2019 at 14:23, on Zulip):

beta-nom 2/1: "resolve: Turn the "non-empty glob must import something" error into a lint" #65539

pnkfelix (Oct 31 2019 at 14:24, on Zulip):

okay accepted for backport

pnkfelix (Oct 31 2019 at 14:25, on Zulip):

coming back to beta regr 3/7: "array impls <= 32 1.39 regression" #65522

pnkfelix (Oct 31 2019 at 14:25, on Zulip):

should we attempt to verify the claim that updating bindgen fixes these cases?

pnkfelix (Oct 31 2019 at 14:25, on Zulip):

or just cross our fingers?

nikomatsakis (Oct 31 2019 at 14:26, on Zulip):

I think we should survey the others

pnkfelix (Oct 31 2019 at 14:26, on Zulip):

I'll assign myself to attempt to verify

nikomatsakis (Oct 31 2019 at 14:26, on Zulip):

but maybe we can indeed make a test

pnkfelix (Oct 31 2019 at 14:27, on Zulip):

beta-regr 4/7: "bindgen changes in 1.39" #65520

nikomatsakis (Oct 31 2019 at 14:27, on Zulip):

heh

pnkfelix (Oct 31 2019 at 14:27, on Zulip):

great, that's already assigned to me

pnkfelix (Oct 31 2019 at 14:27, on Zulip):

(4 hour ago)

pnkfelix (Oct 31 2019 at 14:27, on Zulip):

so I guess I kn ow what I'll be doing instead of traversing the P-high issues after this meeting

pnkfelix (Oct 31 2019 at 14:28, on Zulip):

beta-regr 5/7: "Star Import Precedence Regression in Beta" #65090

nikomatsakis (Oct 31 2019 at 14:28, on Zulip):

I'm going to try and do those verifications live while we type :)

pnkfelix (Oct 31 2019 at 14:28, on Zulip):

I'm going to close #65090 based on points raised in comments over last three weeks

pnkfelix (Oct 31 2019 at 14:30, on Zulip):

beta-regr 6/7: "linking of libtest failed" #64872

pnkfelix (Oct 31 2019 at 14:30, on Zulip):

I've been working on this. I had a PR that fixed it, but @Alex Crichton was rightly concerned about the impact my PR would have elsewhere

pnkfelix (Oct 31 2019 at 14:31, on Zulip):

At this point I am in agreement with @Alex Crichton that the right thing (at least for the short term) is to fix this by reverting the change to the symbol export logic

pnkfelix (Oct 31 2019 at 14:31, on Zulip):

and so I am working on making a PR for that.

pnkfelix (Oct 31 2019 at 14:32, on Zulip):

(I'll probably work on that in parallel with the testing stuff after the meeting is over)

pnkfelix (Oct 31 2019 at 14:32, on Zulip):

But in any case, I'll need to put the PR up soon since it will need to be beta backported. I might steal a minute in tomorrows design meeting to get that approved.

pnkfelix (Oct 31 2019 at 14:33, on Zulip):

beta-regr 7/7: "Undefined symbol _fltused when compiling to x86_64-unknown-uefi" #62785

pnkfelix (Oct 31 2019 at 14:33, on Zulip):

hasn't @oli been trying to land a fix for this for a month?

pnkfelix (Oct 31 2019 at 14:34, on Zulip):

/me starts following the chain of PR's

mw (Oct 31 2019 at 14:34, on Zulip):

I'm a bit surprised this is p-high

mw (Oct 31 2019 at 14:34, on Zulip):

sounds like a tier 3 platform

pnkfelix (Oct 31 2019 at 14:34, on Zulip):

PR rust-lang/compiler-builtins#317 is still open

pnkfelix (Oct 31 2019 at 14:35, on Zulip):

at this point it seems very unlikely that this is going to get resolved before the beta->stable promotion

pnkfelix (Oct 31 2019 at 14:36, on Zulip):

but also: should we reassign this? Is @oli generally around right now?

pnkfelix (Oct 31 2019 at 14:36, on Zulip):

(The prioritization is another question too, but maybe we can revisit that next week)

nikomatsakis (Oct 31 2019 at 14:36, on Zulip):

not sure

mw (Oct 31 2019 at 14:37, on Zulip):

did @oli mention being on vacation?

eddyb (Oct 31 2019 at 14:37, on Zulip):

until next week, yes

pnkfelix (Oct 31 2019 at 14:37, on Zulip):

well anyway that is all the beta regressions. It seems like we should not prioritize effort on resolving issue #62785

mw (Oct 31 2019 at 14:37, on Zulip):

https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/vacation

pnkfelix (Oct 31 2019 at 14:38, on Zulip):

next, lets do I-nominated stuff. (There are two things marked S-waiting-on-team, but the only one we might need to look at is also I-nominated)

pnkfelix (Oct 31 2019 at 14:39, on Zulip):

there are 5 nominated issues and PRs

pnkfelix (Oct 31 2019 at 14:40, on Zulip):

I'm going to go out of order though

pnkfelix (Oct 31 2019 at 14:40, on Zulip):

to try to get the quick things out of the way first

pnkfelix (Oct 31 2019 at 14:40, on Zulip):

I-nominated: "Reduce the places where stable annotations are needed" #65515

pnkfelix (Oct 31 2019 at 14:40, on Zulip):

I want the team to come to consensus on whether to do this or not

pnkfelix (Oct 31 2019 at 14:41, on Zulip):

I saw what I would call light pushback from @centril (perhaps one might even have deemed it devil's advocatry)

mw (Oct 31 2019 at 14:41, on Zulip):

/me needs to run. @pnkfelix has my updates on WG-profile-guided-optimization...

pnkfelix (Oct 31 2019 at 14:41, on Zulip):

thanks @mw

nikomatsakis (Oct 31 2019 at 14:42, on Zulip):

I want the team to come to consensus on whether to do this or not

what is the specific proposal -- just enum variants?

pnkfelix (Oct 31 2019 at 14:42, on Zulip):

bascially: Does anyone here object to the proposal here, at least when reestricted to enum variants?

nikomatsakis (Oct 31 2019 at 14:42, on Zulip):

I think my opinion is we should do it

nikomatsakis (Oct 31 2019 at 14:43, on Zulip):

but I'm still curious if there are surprising edge cases

pnkfelix (Oct 31 2019 at 14:43, on Zulip):

I want the team to come to consensus on whether to do this or not

what is the specific proposal -- just enum variants?

indeed, the somewhat open-ended nature of the proposal's phrasing is maybe cause for concern

nikomatsakis (Oct 31 2019 at 14:43, on Zulip):

it seems more like a libs concern to me in some sense

nikomatsakis (Oct 31 2019 at 14:43, on Zulip):

or at least I'd expect feedback from libs team about this

pnkfelix (Oct 31 2019 at 14:43, on Zulip):

good point

nikomatsakis (Oct 31 2019 at 14:43, on Zulip):

like "ah but back when we added FOO BAR, that would have caused a problem"

pnkfelix (Oct 31 2019 at 14:44, on Zulip):

okay. so maybe what we do is: State that at this meeting, no one present from T-compiler objected to doing this for enums

pnkfelix (Oct 31 2019 at 14:44, on Zulip):

but that we are adding T-libs because we think this is in their balliwick

pnkfelix (Oct 31 2019 at 14:44, on Zulip):

sound good?

pnkfelix (Oct 31 2019 at 14:44, on Zulip):

next nomination: "Document query modifiers" #65929

nikomatsakis (Oct 31 2019 at 14:44, on Zulip):

yep. I left a personal comment in that direction

pnkfelix (Oct 31 2019 at 14:45, on Zulip):

okay, @eddyb actually gave feedback here (on #65929)

pnkfelix (Oct 31 2019 at 14:45, on Zulip):

that they want to see the code/API cleaned up before we work too hard on documentation

pnkfelix (Oct 31 2019 at 14:45, on Zulip):

that's all I wanted to hear, at least for now

pnkfelix (Oct 31 2019 at 14:46, on Zulip):

next nomination: "Allow foreign exceptions to unwind through Rust code and Rust panics to unwind through FFI" #65646

pnkfelix (Oct 31 2019 at 14:47, on Zulip):

(I don't actually know what the action is here for the nomination )

pnkfelix (Oct 31 2019 at 14:47, on Zulip):

((i think maybe @centril just wanted to draw attention to the PR, which is fine, but also does not need to take synchronous time now ))

nikomatsakis (Oct 31 2019 at 14:47, on Zulip):

I think it can probably be removed

nikomatsakis (Oct 31 2019 at 14:47, on Zulip):

I was hoping to review that yesterday but didn't get to it

nikomatsakis (Oct 31 2019 at 14:47, on Zulip):

(sorry @Amanieu)

pnkfelix (Oct 31 2019 at 14:48, on Zulip):

/me wonders if we need separate labels for nominations for synchronous discussion versus nominations for general attention ...

pnkfelix (Oct 31 2019 at 14:48, on Zulip):

anyway ....

pnkfelix (Oct 31 2019 at 14:48, on Zulip):

next nomination: "PowerPC C ABI fixes" #64259

pnkfelix (Oct 31 2019 at 14:49, on Zulip):

eddyb nominated this for team discussion, here

pnkfelix (Oct 31 2019 at 14:50, on Zulip):

does this need a design meeting?

pnkfelix (Oct 31 2019 at 14:50, on Zulip):

I'm not prepared to make an off the cuff decision about it. Certainly not in the short time we have left in today's meeting

pnkfelix (Oct 31 2019 at 14:51, on Zulip):

so lets leave it nominated but not attempt to discuss it now, especially since @mw has gone and I havent' seen sign of @eddyb (perhaps due to the time zone disconnect for this week)

pnkfelix (Oct 31 2019 at 14:51, on Zulip):

last nomination: "Some features can no longer be controlled by conditional compilation" #65860

pnkfelix (Oct 31 2019 at 14:52, on Zulip):

do we need to consider backing out PR #65742 ?

nikomatsakis (Oct 31 2019 at 14:52, on Zulip):

I'm not prepared to make an off the cuff decision about it. Certainly not in the short time we have left in today's meeting

(i'm skimming but definitely this is a bit complex -- still, we should unblock it..)

eddyb (Oct 31 2019 at 14:52, on Zulip):

@pnkfelix o/ (without pings it's hard to keep track of what's up on Zulip)

pnkfelix (Oct 31 2019 at 14:52, on Zulip):

PR #65742 only landed 8 days ago.

eddyb (Oct 31 2019 at 14:52, on Zulip):

I've been trying to keep up though

pnkfelix (Oct 31 2019 at 14:52, on Zulip):

@eddyb you don't get pinged by @T-compiler/meeting ??

eddyb (Oct 31 2019 at 14:53, on Zulip):

sorry I forget to press the emoji... Zulip should add "who has read this" indicators, even if opt-in

pnkfelix (Oct 31 2019 at 14:53, on Zulip):

@eddyb (we'll debug the pinging situation later. If I need to ping every member of @T-compiler to get a quorum, I will happily do so.)

eddyb (Oct 31 2019 at 14:54, on Zulip):

I just mean that most of the stuff in the meeting doesn't need me so I can miss individual items unless I'm pinged on them specifically

pnkfelix (Oct 31 2019 at 14:54, on Zulip):

anyway, regarding issue #65860, I do think it would be good for us for decide whether or not we will back out PR #65742 before nightly gets promoted to beta

nikomatsakis (Oct 31 2019 at 14:54, on Zulip):

yeah

nikomatsakis (Oct 31 2019 at 14:54, on Zulip):

cc @centril @simulacrum

pnkfelix (Oct 31 2019 at 14:54, on Zulip):

I just mean that most of the stuff in the meeting doesn't need me so I can miss individual items unless I'm pinged on them specifically

... (I think this is missing part of the point of the synchronous meeting...)

pnkfelix (Oct 31 2019 at 14:55, on Zulip):

... (as in, I know meetings are boring and there are seemingly-higher priority items. But what's the point of trying to schedule them if people aren't going to bother participating?)

pnkfelix (Oct 31 2019 at 14:55, on Zulip):

anyway sorry I don't want to get sidetracked

pnkfelix (Oct 31 2019 at 14:56, on Zulip):

We haven't had the WG checkins yet

nikomatsakis (Oct 31 2019 at 14:56, on Zulip):

not sure how much background is worth going into here. This has to do with a change to feature gate syntax like box 22, even if that occurs in a #[cfg] syntax. This affects "stable" code that used such features conditionally -- the correct way to do so after this PR is to put them in a separate file and use a #[cfg(foo)] mod m; statement

simulacrum (Oct 31 2019 at 14:56, on Zulip):

I think it would be wrong to not back it out

nikomatsakis (Oct 31 2019 at 14:56, on Zulip):

but the question at hand is only partly about the specifics

simulacrum (Oct 31 2019 at 14:56, on Zulip):

i.e. we clearly have ongoing debate

pnkfelix (Oct 31 2019 at 14:56, on Zulip):

right. there is a matter of our overall policy and values when it comes to these bugs

pnkfelix (Oct 31 2019 at 14:57, on Zulip):

and in the past I think we have treated crater like any other test: It is a way to expose bugs, but it doesn't prove there won't be a significant impact from a change

simulacrum (Oct 31 2019 at 14:57, on Zulip):

and it is a regression for code, even if not necessarily an unexpected one (I am still unclear on this point)

nikomatsakis (Oct 31 2019 at 14:58, on Zulip):

I think it was expected (by the PR author(s)), and I think some sort of change like this is necessary, but I don't think it was necessarily broadly discussed

pnkfelix (Oct 31 2019 at 14:58, on Zulip):

anyway unfortunately i have not allocated the appropriate amount of time in this meeting to have this discussion. :sad:

pnkfelix (Oct 31 2019 at 14:58, on Zulip):

let me at least copy-and-paste the WG-pgo update from @mw

pnkfelix (Oct 31 2019 at 14:58, on Zulip):

here's the update for the PGO working group:

nikomatsakis (Oct 31 2019 at 14:59, on Zulip):

:tada: for getting it done :)

pnkfelix (Oct 31 2019 at 14:59, on Zulip):

if you have questions about this, you should ping @mw (who, as previously noted, had to run)

pnkfelix (Oct 31 2019 at 15:00, on Zulip):

Separately, @Alex Crichton had mentioned up above that there might not be much to report from WG-pipelining

nikomatsakis (Oct 31 2019 at 15:00, on Zulip):

seems like WG-PGO and WG-pipelining are maybe just "done", no?

pnkfelix (Oct 31 2019 at 15:00, on Zulip):

maybe

pnkfelix (Oct 31 2019 at 15:00, on Zulip):

might be good to close them down

nikomatsakis (Oct 31 2019 at 15:00, on Zulip):

I'm going to open a PR against compiler-team repo to remove them, I think

nikomatsakis (Oct 31 2019 at 15:00, on Zulip):

anyway unfortunately i have not allocated the appropriate amount of time in this meeting to have this discussion. :sad:

how/where should we have that conversation?

pnkfelix (Oct 31 2019 at 15:01, on Zulip):

well, first let me say

pnkfelix (Oct 31 2019 at 15:01, on Zulip):

Thanks to everyone in @T-compiler/meeting for joining

pnkfelix (Oct 31 2019 at 15:01, on Zulip):

anyway unfortunately i have not allocated the appropriate amount of time in this meeting to have this discussion. :sad:

how/where should we have that conversation?

lets maybe open a fresh topic?

nikomatsakis (Oct 31 2019 at 15:02, on Zulip):

OK.

pnkfelix (Oct 31 2019 at 15:02, on Zulip):

like here: https://rust-lang.zulipchat.com/#narrow/stream/131828-t-compiler/topic/.C2.BFrevert.3F.20PR.20Pre-expansion.20gate.20most.20of.20the.20things.20.2365742/near/179539595

Last update: Nov 20 2019 at 01:25UTC