Stream: t-compiler

Topic: #54818 weekly meeting 2018-11-15


nagisa (Nov 15 2018 at 13:53, on Zulip):

The weekly meeting commences in slightly more than an hour.

nagisa (Nov 15 2018 at 14:55, on Zulip):

@T-compiler no minna-san, meeting in 5 minutes!

nagisa (Nov 15 2018 at 15:00, on Zulip):

Since @pnkfelix is on a vacation I’ll be driving this and possibly the next meeting. Lets begin, shall we.

nagisa (Nov 15 2018 at 15:00, on Zulip):

Going through the list in #54818, as usual.

nikomatsakis (Nov 15 2018 at 15:00, on Zulip):

one sec :)

nikomatsakis (Nov 15 2018 at 15:00, on Zulip):

I wanted to add something

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

we should go through P-high, but it's also a good idae for us to check the Rust 2018 Release milestone list

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

perhaps after

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

that's all :)

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

though of course those things ought to be P-high'd if applicable, not sure if they have been :)

nagisa (Nov 15 2018 at 15:01, on Zulip):

@nikomatsakis spoiler alert: there are no issues for T-compiler on that list :slight_smile:

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

heh that's... wrong

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

but that means I have to fix the list ;)

nagisa (Nov 15 2018 at 15:01, on Zulip):

(that aren’t covered by other categories)

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

yep ok

nikomatsakis (Nov 15 2018 at 15:01, on Zulip):

that's fine

nagisa (Nov 15 2018 at 15:02, on Zulip):

# P-high T-compiler

8 issues, 2 issues without assignees, 3 issues meanigfully updated since the last meeting.

nagisa (Nov 15 2018 at 15:02, on Zulip):

Lifetimes bug on beta 1.31: the associated type may not live long enough #55756: @pnkfelix bisected and crafted a reduced test case. @nikomatsakis then investigated this further…

nagisa (Nov 15 2018 at 15:02, on Zulip):

please tell about your investigation, @nikomatsakis :slight_smile:

nikomatsakis (Nov 15 2018 at 15:03, on Zulip):

tl;dr I know what the problem is, hope to have a fix today

nagisa (Nov 15 2018 at 15:03, on Zulip):

/me assigns @nikomatsakis

nikomatsakis (Nov 15 2018 at 15:04, on Zulip):

I'm .. not entirely happy with the fix I have in mind :P but it's ok. It basically "ports" the NLL rule to lexical; the rule being that whenever a type T appears appears at some position P, all regions in that type must include that position P

nikomatsakis (Nov 15 2018 at 15:04, on Zulip):

(we currently require that T: P, which I believed to be equivalent, but it's not quite, in the case of associated types, which is the part I am not entirely happy with)

nikomatsakis (Nov 15 2018 at 15:05, on Zulip):

in practice, I think it will be ok, and we have some room to tweak how we handle this once lazy norm lands (and once we start making some of the improvements to region inference I want to do)

nagisa (Nov 15 2018 at 15:06, on Zulip):

I’m having a hard time understanding where the differences come from at all… ah, normalization strikes again

nagisa (Nov 15 2018 at 15:06, on Zulip):

okay.

nikomatsakis (Nov 15 2018 at 15:06, on Zulip):

the next one on the list is

NLL error on closure, but not on equivalent function #55526

nikomatsakis (Nov 15 2018 at 15:07, on Zulip):

which I had not seen yet actually

nagisa (Nov 15 2018 at 15:07, on Zulip):

Yes, NLL error on closure, but not on equivalent function #55526: @pnkfelix had no opportunity to look at this. No meanigful changes since the last week?

nikomatsakis (Nov 15 2018 at 15:07, on Zulip):

no, I will assign to myself to investigate

nikomatsakis (Nov 15 2018 at 15:07, on Zulip):

as pnkfelix is on PTO

nikomatsakis (Nov 15 2018 at 15:07, on Zulip):

but it looks like it's not a release blocker since it's only a warning

Esteban Küber (Nov 15 2018 at 15:07, on Zulip):

Those two are not equivalent, the fn _has_ a return type

Esteban Küber (Nov 15 2018 at 15:07, on Zulip):

the closure doesn't

nikomatsakis (Nov 15 2018 at 15:08, on Zulip):

yeah, it may well be legitimate

nagisa (Nov 15 2018 at 15:08, on Zulip):

Okay, moving on.

nagisa (Nov 15 2018 at 15:09, on Zulip):

regression: stack overflow on macosx with xcode 6.4 #55471: Stacker efforts are progressing, we did a crater run recently, no regressions there, can merge. Still not X-platform (only x86-64 windows/linux/osx), though, and my efforts on more portability that are progressing fairly slowly at the current time. @Oli, does your PR account for the supported platforms?

oli (Nov 15 2018 at 15:09, on Zulip):

not yet, I need to get MSVC working, too

nikomatsakis (Nov 15 2018 at 15:09, on Zulip):

do we know about perf impact now?

oli (Nov 15 2018 at 15:09, on Zulip):

1% on a few benchmarks

oli (Nov 15 2018 at 15:09, on Zulip):

3-4% on a few generally fluctuating ones

nagisa (Nov 15 2018 at 15:09, on Zulip):

On the PR I said that perf impact (as long as it is not too major) is secondary to making things work at all

nikomatsakis (Nov 15 2018 at 15:10, on Zulip):

ok I just remembered @Oli saying there was a bit hit initially

oli (Nov 15 2018 at 15:10, on Zulip):

yes, that has been reduced

oli (Nov 15 2018 at 15:11, on Zulip):

there was stack thrashing and bad scaling for large stack uses

oli (Nov 15 2018 at 15:11, on Zulip):

both gone now

nagisa (Nov 15 2018 at 15:11, on Zulip):

In general I’d be comfortable even with rolling this out incrementally, as my "portable" stack manipulation library won’t materialize over a single weekend for sure.

Wesley Wiser (Nov 15 2018 at 15:11, on Zulip):

(latest perf results)

oli (Nov 15 2018 at 15:11, on Zulip):

after fixing MSVC we can roll this out for a few platforms

nagisa (Nov 15 2018 at 15:12, on Zulip):

@Oli did you implement some sort of checking for support and correspondingly handling platforms where stacker does not support the platform?

nagisa (Nov 15 2018 at 15:12, on Zulip):

I don’t recall seeing such code last time I looked at the PR (admittedly quite a while ago)

oli (Nov 15 2018 at 15:12, on Zulip):

yes, but msvc should be handled, but it still keeps overflowing the stack in tests

nagisa (Nov 15 2018 at 15:13, on Zulip):

Okay, so progress here is clear, @Oli keeps working on MSVC support, I’m working on more portability architecture wise and T-compiler is generally comfortable with this approach, right?

nagisa (Nov 15 2018 at 15:14, on Zulip):

Moving on, then, [1.31 beta] Trait bound is not satisfied #54467: @pnkfelix said “Maybe we need to reconsider backporting PR #54701 ... will investigate...”

nagisa (Nov 15 2018 at 15:14, on Zulip):

It doesn’t seem that the investigation results, if they exist, are public.

nikomatsakis (Nov 15 2018 at 15:14, on Zulip):

hmm

nikomatsakis (Nov 15 2018 at 15:14, on Zulip):

I'd be comfortable with a backport there

nikomatsakis (Nov 15 2018 at 15:15, on Zulip):

I could do it today

nagisa (Nov 15 2018 at 15:16, on Zulip):

I added you as an assignee

nikomatsakis (Nov 15 2018 at 15:16, on Zulip):

reading the issue more closely

nikomatsakis (Nov 15 2018 at 15:16, on Zulip):

well, ok, I will investigate

nikomatsakis (Nov 15 2018 at 15:16, on Zulip):

probably @pnkfelix didn't get to it

nagisa (Nov 15 2018 at 15:16, on Zulip):

Though a question on whether we are comfortable with the backport is also here.

nikomatsakis (Nov 15 2018 at 15:17, on Zulip):

yeah

nagisa (Nov 15 2018 at 15:17, on Zulip):

Moving on, Rustc does not warn about use with paths incompatible with uniform_paths for edition 2018 #53797: no meaningful changes since the last week?

nikomatsakis (Nov 15 2018 at 15:17, on Zulip):

not that I know of; might be fixed by @Vadim Petrochenkov's PR

nikomatsakis (Nov 15 2018 at 15:17, on Zulip):

https://github.com/rust-lang/rust/pull/55884

nikomatsakis (Nov 15 2018 at 15:17, on Zulip):

which is worth discussing later :) in particular the unorthodox "target beta first" approach

nagisa (Nov 15 2018 at 15:18, on Zulip):

Mhm, it is beta-nominated.

nagisa (Nov 15 2018 at 15:18, on Zulip):

ICE when running cargo doc on typenum at librustc/traits/structural_impls.rs:178 #52873: appears to have no meaningful changes since the last week, although there are some comments by @pnkfelix on process issues.

nagisa (Nov 15 2018 at 15:18, on Zulip):

Most notably:

nagisa (Nov 15 2018 at 15:19, on Zulip):

On the subject of closing: Can we close this issue in light of #55738? Or do we need to wait until the 1.30.1 point release has officially been deployed?

nikomatsakis (Nov 15 2018 at 15:19, on Zulip):

I feel like we can close once the backport PR lands?

nagisa (Nov 15 2018 at 15:19, on Zulip):

the point release has been deployed either way, so it seems to me like we should close regardless

nikomatsakis (Nov 15 2018 at 15:19, on Zulip):

yeah, moot point

nagisa (Nov 15 2018 at 15:19, on Zulip):

Closed.

nagisa (Nov 15 2018 at 15:20, on Zulip):

Next is [Rust 2018] rustdoc doesn't link "pub use whatever_crate;" #52509: it seems that T-rustdoc claims this does not block Rust2018.

nagisa (Nov 15 2018 at 15:20, on Zulip):

This may or may not make this not P-high.

nagisa (Nov 15 2018 at 15:21, on Zulip):

And otherwise be more T-rustdoc than T-compiler.

nagisa (Nov 15 2018 at 15:22, on Zulip):

How about removing T-compiler so we don’t get back to this the following weeks, now that it is not as important to us?

nikomatsakis (Nov 15 2018 at 15:23, on Zulip):

sounds good

nagisa (Nov 15 2018 at 15:23, on Zulip):

Okay, removed T-compiler.

nagisa (Nov 15 2018 at 15:23, on Zulip):

Next up is rewrite ... to ..= as an idiom lint for Rust 2018 edition #51043: @varkor posted a PR implementing such a lint (beta-nominated!). Has been removed from the Rust2018 milestone.

nagisa (Nov 15 2018 at 15:25, on Zulip):

Since we’ll be discussing beta-nominated issues next, feel free to also decide whether we should backport the PR or not :slight_smile:

nikomatsakis (Nov 15 2018 at 15:26, on Zulip):

I'm in favor of backporting

nikomatsakis (Nov 15 2018 at 15:26, on Zulip):

it's small, low risk

nikomatsakis (Nov 15 2018 at 15:26, on Zulip):

idiom lints are not enabled by default

nikomatsakis (Nov 15 2018 at 15:26, on Zulip):

but we do encourage people to try them out, seems ok

nagisa (Nov 15 2018 at 15:26, on Zulip):

I’m also in favour.

nagisa (Nov 15 2018 at 15:27, on Zulip):

Is there anybody against?

nagisa (Nov 15 2018 at 15:28, on Zulip):

Okay, I marked it as beta-accepted.

nagisa (Nov 15 2018 at 15:28, on Zulip):

Going through the list… next is the 2018 milestone in which the only issues for T-compiler are also P-high and thus covered.

nagisa (Nov 15 2018 at 15:29, on Zulip):

# beta-nominated

nagisa (Nov 15 2018 at 15:29, on Zulip):

xLTO: Don't pass --plugin-opt=thin to LLD. That's not supported anymore. #55947: tiny patch, no tests.

mw (Nov 15 2018 at 15:29, on Zulip):

it's actually not very important to backport this since everything is unstable anyway

nagisa (Nov 15 2018 at 15:30, on Zulip):

I believe there are users of lld, though

nagisa (Nov 15 2018 at 15:30, on Zulip):

embedded, perhaps.

mw (Nov 15 2018 at 15:30, on Zulip):

yeah, but this only affects xLTO

nagisa (Nov 15 2018 at 15:30, on Zulip):

it (the features involved) being unstable also lowers the risk of the backport, so I’m somewhat in favour.

mw (Nov 15 2018 at 15:31, on Zulip):

there's basically no risk, since no one is using xLTO :)

nagisa (Nov 15 2018 at 15:31, on Zulip):

That too.

Wesley Wiser (Nov 15 2018 at 15:31, on Zulip):

The patch itself doesn't look risky at all

nagisa (Nov 15 2018 at 15:31, on Zulip):

Lets do it this way: +1 this comment if in favour, -1 if against :slight_smile:

nagisa (Nov 15 2018 at 15:32, on Zulip):

Seems unlikely to get 5 -1s now, so marking as beta accepted, I guess

nikomatsakis (Nov 15 2018 at 15:32, on Zulip):

/me tries to add more than one :-1:

nikomatsakis (Nov 15 2018 at 15:32, on Zulip):

/me fails

nikomatsakis (Nov 15 2018 at 15:33, on Zulip):

ok, I didn't really try

nagisa (Nov 15 2018 at 15:34, on Zulip):

[beta] resolve: Implement uniform paths 2.0 #55884: part of Rust2018 milestone. HUGE change. Has not landed yet.

nagisa (Nov 15 2018 at 15:34, on Zulip):

So this is the PR that @nikomatsakis has alluded to earlier.

nagisa (Nov 15 2018 at 15:35, on Zulip):

which is worth discussing later :slight_smile: in particular the unorthodox "target beta first" approach

nikomatsakis (Nov 15 2018 at 15:35, on Zulip):

indeed

nikomatsakis (Nov 15 2018 at 15:35, on Zulip):

so from what I understand, landing this change would mean that -- briefly -- there is a change that exists only on beta but not nightly

nikomatsakis (Nov 15 2018 at 15:35, on Zulip):

which isn't our usual way of doing things

nagisa (Nov 15 2018 at 15:35, on Zulip):

So about this beta-first approach, what does it get us over landing the PR to master first, or even concurrently to both master and beta?

nikomatsakis (Nov 15 2018 at 15:35, on Zulip):

but @Vadim Petrochenkov explains their logic on the PR

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

Why beta:

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

I think the tl;dr is that it's not ready for nightly yet :)

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

but we do (in their opinion, and perhaps mine) really want it for Rust 2018

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

I'm a bit nervous because it's a large change

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

coming late

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

I'm not esp. familiar with the code

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

otoh I of course know that @Vadim Petrochenkov tends to be very careful :)

nikomatsakis (Nov 15 2018 at 15:36, on Zulip):

I'd love to hear what @eddyb thinks

nagisa (Nov 15 2018 at 15:37, on Zulip):

Well, isn’t paths 2.0 absolutely necessary for 2018 if we want to make it complete?

nagisa (Nov 15 2018 at 15:37, on Zulip):

or am I confusing it somewhat with the module system rework?

nikomatsakis (Nov 15 2018 at 15:37, on Zulip):

I guess there is a "lang team" aspect of this too

nikomatsakis (Nov 15 2018 at 15:37, on Zulip):

it's sort of complex :(

nikomatsakis (Nov 15 2018 at 15:37, on Zulip):

so the scenario is this

nikomatsakis (Nov 15 2018 at 15:37, on Zulip):

we started out with one change ("anchored paths") -- that was implemented for a long time

nikomatsakis (Nov 15 2018 at 15:38, on Zulip):

the main thing is that for use foo::bar, foo must be a crate

nikomatsakis (Nov 15 2018 at 15:38, on Zulip):

then we experimented with this idea of uniform paths, where foo can be a crate, but it might also be something from the local module, and we error in the case of ambiguity

nikomatsakis (Nov 15 2018 at 15:38, on Zulip):

we wound up concluding that this was kind of better

nikomatsakis (Nov 15 2018 at 15:38, on Zulip):

however, the method of implementing was perhaps problematic

nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

and so @Vadim Petrochenkov has this new implementation, which is also more general

nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

in particular I think it handles cases like this:

fn foo() {
    enum Bar { X, Y }
    use Bar::*:
}
nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

where the use Bar::* is not equivalent to use self::Bar::*

nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

as that would draw from the module

nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

in fact, we have previously had no way to do that at all

nikomatsakis (Nov 15 2018 at 15:39, on Zulip):

(to import from a fn-local item)

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

looks like crater check status is Running (75%)

Wesley Wiser (Nov 15 2018 at 15:40, on Zulip):

I think the tl;dr is that it's not ready for nightly yet :)

I'm confused, if the PR isn't ready for nightly, how can it be ok to land on beta?

Or do you mean "ready" in the sense that it hasn't been rebased onto master yet?

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

yes, that's what I mean

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

the code was written against beta first

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

and needs to be "ported" to nightly

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

I don't know how hard that will be

nikomatsakis (Nov 15 2018 at 15:40, on Zulip):

presumably not very

Wesley Wiser (Nov 15 2018 at 15:40, on Zulip):

Got it. Thanks!

nikomatsakis (Nov 15 2018 at 15:41, on Zulip):

so I guess the assumption is:

nagisa (Nov 15 2018 at 15:41, on Zulip):

Would be amusing to see a scenario where this lands into beta but then we perpetually are unable to forward-port the PR in time

nikomatsakis (Nov 15 2018 at 15:41, on Zulip):

if the crater run comes out clean

nikomatsakis (Nov 15 2018 at 15:42, on Zulip):

do we land this

nikomatsakis (Nov 15 2018 at 15:42, on Zulip):

I guess @eddyb did review and wrote

I won't pretend to fully understand the inner workings of early_resolve_ident_in_lexical_scope, but if it worked before for macros, this PR LGTM.

nikomatsakis (Nov 15 2018 at 15:43, on Zulip):

they are certainly the other person who understands resolve best

oli (Nov 15 2018 at 15:43, on Zulip):

I'm in favour but with stomach aches

nikomatsakis (Nov 15 2018 at 15:43, on Zulip):

I think that's roughly where I am

nagisa (Nov 15 2018 at 15:43, on Zulip):

do we want perhaps to use the FCP approach?

nagisa (Nov 15 2018 at 15:43, on Zulip):

I would love to give it a thorough look

nikomatsakis (Nov 15 2018 at 15:43, on Zulip):

I'm fine with that but don't want to delay too much

nikomatsakis (Nov 15 2018 at 15:43, on Zulip):

we did want to get all things on beta landed by next wednesday

nikomatsakis (Nov 15 2018 at 15:43, on Zulip):

seeing as U.S. thanksgiving is Thu-Fri of next week

nagisa (Nov 15 2018 at 15:44, on Zulip):

Hmm. Okay, I’m fine with landing without FCP then

nikomatsakis (Nov 15 2018 at 15:44, on Zulip):

but by all means do take a look

nagisa (Nov 15 2018 at 15:44, on Zulip):

I can always do that post-land and complain loudly later

nikomatsakis (Nov 15 2018 at 15:44, on Zulip):

I plan to do some reviewing today too

nagisa (Nov 15 2018 at 15:45, on Zulip):

Okay, lets move on 15 minutes left

nagisa (Nov 15 2018 at 15:45, on Zulip):

Fix ICE in return_type_impl_trait #55800: fairly small change, tested.

nagisa (Nov 15 2018 at 15:45, on Zulip):

I don’t see any reason to not backport

nagisa (Nov 15 2018 at 15:46, on Zulip):

Except perhaps for:

     // HACK: `type_of_def_id()` will fail on these (#55796), so return None
nikomatsakis (Nov 15 2018 at 15:46, on Zulip):

seems ok

nikomatsakis (Nov 15 2018 at 15:46, on Zulip):

not wild about the fix

nikomatsakis (Nov 15 2018 at 15:46, on Zulip):

not @Esteban Küber's fault per se

nikomatsakis (Nov 15 2018 at 15:46, on Zulip):

I'd like to get away from using HIR map -- I think at minimum these tests could be done on the def-path

eddyb (Nov 15 2018 at 15:46, on Zulip):

(I am for the uniform paths refactor landing btw, but I cannot vouch for the implementation fully, btw)

nikomatsakis (Nov 15 2018 at 15:47, on Zulip):

but anyway

nikomatsakis (Nov 15 2018 at 15:47, on Zulip):

(and maybe the setup can be otherwise cleaned)

nagisa (Nov 15 2018 at 15:47, on Zulip):

Since this fix has already landed to master, we’ll have to live with it one way or another

nagisa (Nov 15 2018 at 15:47, on Zulip):

unless it somehow gets improved before the next beta releases

Esteban Küber (Nov 15 2018 at 15:48, on Zulip):

@nagisa The proper fix would be for type_of_def_id to return Option<DefId>, but that was a much larger PR :)

nagisa (Nov 15 2018 at 15:48, on Zulip):

so I’m in favour.

nagisa (Nov 15 2018 at 15:48, on Zulip):

Anybody against?

nagisa (Nov 15 2018 at 15:48, on Zulip):

/me waits a little while

Esteban Küber (Nov 15 2018 at 15:49, on Zulip):

/me abstains

nagisa (Nov 15 2018 at 15:50, on Zulip):

/me wrote the summary comment and is marking as beta-accepted

nagisa (Nov 15 2018 at 15:50, on Zulip):

Next up is Check for negative impls when finding auto traits #55356: small change, tested.

nagisa (Nov 15 2018 at 15:51, on Zulip):

This fixes an ICE

nagisa (Nov 15 2018 at 15:51, on Zulip):

Although that ICE is a stable-to-stable regression

nagisa (Nov 15 2018 at 15:51, on Zulip):

so it does not seem to be critical to backport

nikomatsakis (Nov 15 2018 at 15:52, on Zulip):

I think it also only affects unstable code

nikomatsakis (Nov 15 2018 at 15:52, on Zulip):

that is using impl !Send explicitly

nikomatsakis (Nov 15 2018 at 15:52, on Zulip):

I'm (weakly) inclined not to backport

nagisa (Nov 15 2018 at 15:53, on Zulip):

Is there anybody strongly in favour?

nikomatsakis (Nov 15 2018 at 15:53, on Zulip):

(well, maybe it would affect code using code from libstd based on UnsafeCell...? not sure)

mw (Nov 15 2018 at 15:53, on Zulip):

Is there a workaround?

nikomatsakis (Nov 15 2018 at 15:53, on Zulip):

but it sounds obscure from the issue (#55321)

mw (Nov 15 2018 at 15:54, on Zulip):

hm, if it only ICEs on unstable code then beta/stable can't compile it anyway...

mw (Nov 15 2018 at 15:54, on Zulip):

(^^^note to myself)

nagisa (Nov 15 2018 at 15:55, on Zulip):

I guess the consensus is here.

nagisa (Nov 15 2018 at 15:55, on Zulip):

Next is resolve: Filter away macro prelude in modules with #[no_implicit_prelude] on 2018 edition #55630: Is destined for T-lang, undergoing FCP, @nrc needs to mark their checkbox.

nagisa (Nov 15 2018 at 15:56, on Zulip):

oh nrc is no longer on T-compiler, are they?

nagisa (Nov 15 2018 at 15:56, on Zulip):

Then that is all for beta-nominated issues.

nagisa (Nov 15 2018 at 15:56, on Zulip):

/me double checks

nikomatsakis (Nov 15 2018 at 15:56, on Zulip):

we don't need 100% for FCP anymore :)

nikomatsakis (Nov 15 2018 at 15:56, on Zulip):

nrc is still on lang team, but I think they don't object

nikomatsakis (Nov 15 2018 at 15:57, on Zulip):

(note that it already merged to nightly)

nikomatsakis (Nov 15 2018 at 15:57, on Zulip):

I think we should backport

nikomatsakis (Nov 15 2018 at 15:57, on Zulip):

it's a breaking change to Rust 2018 otherwise I thikn?

nikomatsakis (Nov 15 2018 at 15:57, on Zulip):

(plus very small etc etc)

nagisa (Nov 15 2018 at 15:57, on Zulip):

I don’t mind accepting for backport

nagisa (Nov 15 2018 at 15:59, on Zulip):

beta-accepted.

nagisa (Nov 15 2018 at 15:59, on Zulip):

That’s all for the beta-nominated list.

nagisa (Nov 15 2018 at 15:59, on Zulip):

We have no stable-nominated issues/PRs!

nagisa (Nov 15 2018 at 15:59, on Zulip):

There is 1 stable-to-beta regression Name _ defined multiple times #55811: @petrochenkov self-assigned this.

nikomatsakis (Nov 15 2018 at 16:00, on Zulip):

P-high I guess

nagisa (Nov 15 2018 at 16:01, on Zulip):

The last thing on my list is a I-nominated issue that has been nominated for a loong time and had no activity for a month

nagisa (Nov 15 2018 at 16:01, on Zulip):

(edition lint: migrating extern crate with #[macro_use] #52043). But since we are over time, I’m inclined to call this meeting over instead :slight_smile:

nagisa (Nov 15 2018 at 16:02, on Zulip):

Thanks all for participating!

nikomatsakis (Nov 15 2018 at 16:02, on Zulip):

thanks @nagisa for running

nikomatsakis (Nov 15 2018 at 16:02, on Zulip):

oh btw

nikomatsakis (Nov 15 2018 at 16:03, on Zulip):

next week Thu is Thanksgiving here in the US

nikomatsakis (Nov 15 2018 at 16:03, on Zulip):

I at least won't be around, jfyi

nikomatsakis (Nov 15 2018 at 16:03, on Zulip):

also, steering meeting tomorrow

nagisa (Nov 15 2018 at 16:03, on Zulip):

Same time as the T-compiler meeting?

nikomatsakis (Nov 15 2018 at 16:04, on Zulip):

yes

nikomatsakis (Nov 15 2018 at 16:04, on Zulip):

see also this internals post and the replies, where @mw and I were brainstorming a bit what to discuss :)

Last update: Nov 16 2019 at 02:30UTC