Stream: t-compiler/meetings

Topic: planning meeting 2020.07.31


pnkfelix (Jul 31 2020 at 13:50, on Zulip):

Hi @T-compiler/meeting ; our planning meeting will be held in this topic, starting in about 10 minutes or so

pnkfelix (Jul 31 2020 at 13:53, on Zulip):

I'll be trying to follow the steps listed at https://forge.rust-lang.org/compiler/steering-meeting/how-to-run-planning.html

pnkfelix (Jul 31 2020 at 13:54, on Zulip):

(I happen to have a periodically screaming 19 month-old on my lap, so there might be some hiccups along the way.)

pnkfelix (Jul 31 2020 at 13:54, on Zulip):

((or at least typos))

pnkfelix (Jul 31 2020 at 13:55, on Zulip):

in the meantime, this is a good time, or at least effectively your last chance, to get any last-minute proposals posted to https://github.com/rust-lang/compiler-team/issues?q=is%3Aopen+is%3Aissue+label%3Ameeting-proposal

pnkfelix (Jul 31 2020 at 13:58, on Zulip):

(and while I'm sitting here, reflecting on how we don't have very many design meeting proposals, but we do have a number of major change proposals .... I am led to wonder if we might be well-served by allocating any extra Friday meeting slots to discussing MCP's that have been left unseconded ... but perhaps that sort of idea is best left as part of the proposed MCP retrospective.)

pnkfelix (Jul 31 2020 at 14:00, on Zulip):

okay, well, the time has arrived

pnkfelix (Jul 31 2020 at 14:00, on Zulip):

Hi again to @T-compiler/meeting , lets start the planning meeting

pnkfelix (Jul 31 2020 at 14:00, on Zulip):

we'll start off with five minutes for any announcements

pnkfelix (Jul 31 2020 at 14:01, on Zulip):

@simulacrum are we going to have a .2 release today/monday?

pnkfelix (Jul 31 2020 at 14:01, on Zulip):

the context here is issue #74954 "Pattern matching regression 1.45.0 -> 1.45.1 (+nightly)"

pnkfelix (Jul 31 2020 at 14:02, on Zulip):

ah yes there is a PR up, at #74958

pnkfelix (Jul 31 2020 at 14:02, on Zulip):

let me r+ that

pnkfelix (Jul 31 2020 at 14:05, on Zulip):

okay, so, we have three unscheduled meeting proposals

pnkfelix (Jul 31 2020 at 14:06, on Zulip):

shown here

pnkfelix (Jul 31 2020 at 14:06, on Zulip):

the three proposals are

nikomatsakis (Jul 31 2020 at 14:06, on Zulip):

we're not so good at writing the minutes, are we? =)

pnkfelix (Jul 31 2020 at 14:07, on Zulip):
  1. compiler-team#318 "discuss results of 2020 contributor survey"
  2. compiler-team#314 "MCP retrospective"
  3. compiler-team#177 "merge llvm.sideeffect change"
pnkfelix (Jul 31 2020 at 14:08, on Zulip):

it seems likely that the story hasn't changed regarding compiler-team#177 (in the past we've declined to schedule a meeting because we've either needed more data or needed the performance to improve, or both)

pnkfelix (Jul 31 2020 at 14:08, on Zulip):

but let me double-check if anything has changed there

pnkfelix (Jul 31 2020 at 14:08, on Zulip):

nikomatsakis said:

we're not so good at writing the minutes, are we? =)

yeah, I admit partial blame there

nikomatsakis (Jul 31 2020 at 14:08, on Zulip):

at this point maybe we should just link to the zulip topic and call it a day :)

nikomatsakis (Jul 31 2020 at 14:08, on Zulip):

(seems fine...)

pnkfelix (Jul 31 2020 at 14:09, on Zulip):

eh I think there is still value to trying to come up with proper minutes

nikomatsakis (Jul 31 2020 at 14:09, on Zulip):

there is definitely value in trying to distill the conversation..

pnkfelix (Jul 31 2020 at 14:09, on Zulip):

but you do have a point that it may be making this step less efficient

pnkfelix (Jul 31 2020 at 14:09, on Zulip):

maybe we could have a queue of "undistilled meetings" tracked somewhere other than this list of issues

pnkfelix (Jul 31 2020 at 14:10, on Zulip):

(and then close each of these issues when they get inserted onto the queue)

pnkfelix (Jul 31 2020 at 14:10, on Zulip):

or we could, you know, try harder to write up those summaries. :smile:

nikomatsakis (Jul 31 2020 at 14:11, on Zulip):

heh we can always searching for -label:meeting-scheduled

nikomatsakis (Jul 31 2020 at 14:11, on Zulip):

but I think it's good to be reminded...

pnkfelix (Jul 31 2020 at 14:11, on Zulip):

regarding compiler-team#177

nikomatsakis (Jul 31 2020 at 14:11, on Zulip):

btw, I will be on vacation next Friday

pnkfelix (Jul 31 2020 at 14:11, on Zulip):

there was a blog post posted

pnkfelix (Jul 31 2020 at 14:12, on Zulip):

though I cannot find it on the blog itself...?

pnkfelix (Jul 31 2020 at 14:12, on Zulip):

ah, here it is: https://blog.rust-lang.org/inside-rust/2020/03/19/terminating-rust.html

pnkfelix (Jul 31 2020 at 14:13, on Zulip):

so was there anything interesting on the associated internals thread ... ?

nikomatsakis (Jul 31 2020 at 14:13, on Zulip):

I don't recall anything coming from it

nikomatsakis (Jul 31 2020 at 14:15, on Zulip):

I tend to feel like we should close the meeting proposal.

nikomatsakis (Jul 31 2020 at 14:15, on Zulip):

I don't really know that there's a lot to discuss

pnkfelix (Jul 31 2020 at 14:15, on Zulip):

well

nikomatsakis (Jul 31 2020 at 14:15, on Zulip):

I think the consensus of how to address the issue was to add some targeted code to detect the apparently common case of loop { }

pnkfelix (Jul 31 2020 at 14:15, on Zulip):

its good to be aware that its still an outstanding problem

nikomatsakis (Jul 31 2020 at 14:15, on Zulip):

sure, but we have a lot of those

pnkfelix (Jul 31 2020 at 14:15, on Zulip):

true

nikomatsakis (Jul 31 2020 at 14:15, on Zulip):

and I'm not sure what makes this one so special

pnkfelix (Jul 31 2020 at 14:16, on Zulip):

well, I think I myself want to read the posts on internals before I close this...

pnkfelix (Jul 31 2020 at 14:17, on Zulip):

lets leave it open and see what happens in four weeks when it comes up again

pnkfelix (Jul 31 2020 at 14:17, on Zulip):

(I certainly don't think we're at a point where we would schedule a design meeting for it.)

nikomatsakis (Jul 31 2020 at 14:17, on Zulip):

OK

pnkfelix (Jul 31 2020 at 14:17, on Zulip):

That leaves:

  1. compiler-team#318 "discuss results of 2020 contributor survey"
  2. compiler-team#314 "MCP retrospective"
pnkfelix (Jul 31 2020 at 14:18, on Zulip):

I think we can have meetings for both of those topics

pnkfelix (Jul 31 2020 at 14:19, on Zulip):

@nikomatsakis I assume you do not think its too soon to have an MCP retrospective, given that you proposed it?

nikomatsakis (Jul 31 2020 at 14:20, on Zulip):

I'm not sure :) I didn't think so

nikomatsakis (Jul 31 2020 at 14:20, on Zulip):

I guess "retrospective" might not be the right term

nikomatsakis (Jul 31 2020 at 14:20, on Zulip):

but sort of "check-in on how people feel it's been going"

pnkfelix (Jul 31 2020 at 14:20, on Zulip):

assuming we're generally agreed that both meetings are worthwhile, I would propose that we do compiler-team#318 before compiler-team#314, because the analysis for compiler-team#318 is already done, while we still would want to do a fresh survey for compiler-team#314

nikomatsakis (Jul 31 2020 at 14:21, on Zulip):

sounds right

pnkfelix (Jul 31 2020 at 14:21, on Zulip):

but, oh, @nikomatsakis did just say a few messages up that they would be on vacation

pnkfelix (Jul 31 2020 at 14:21, on Zulip):

(or was that not in this topic? I cannot find it now ...)

pnkfelix (Jul 31 2020 at 14:22, on Zulip):

okay, next Friday @nikomatsakis is on vacation

pnkfelix (Jul 31 2020 at 14:22, on Zulip):

@nikomatsakis but you are planning to be around for the other Fridays?

nikomatsakis (Jul 31 2020 at 14:22, on Zulip):

I was just checking

nikomatsakis (Jul 31 2020 at 14:22, on Zulip):

I'm not available on Aug 7 or Aug 21

nikomatsakis (Jul 31 2020 at 14:22, on Zulip):

which leaves Aug 14 and Aug 28

nikomatsakis (Jul 31 2020 at 14:23, on Zulip):

(or schedule for when I'm not around, but these sound like topics I'm interestedin :)

pnkfelix (Jul 31 2020 at 14:23, on Zulip):

isn't 28th a planning meeting slot?

nikomatsakis (Jul 31 2020 at 14:23, on Zulip):

oh, yes

nikomatsakis (Jul 31 2020 at 14:23, on Zulip):

I wasn't thinking :)

pnkfelix (Jul 31 2020 at 14:23, on Zulip):

(plan, design design design, plan, ...)

pnkfelix (Jul 31 2020 at 14:23, on Zulip):

so okay

nikomatsakis (Jul 31 2020 at 14:24, on Zulip):

it's vacation time :beach:

pnkfelix (Jul 31 2020 at 14:24, on Zulip):

well, maybe we do just the contributor survey results, on August 14th

pnkfelix (Jul 31 2020 at 14:24, on Zulip):

and we leave the MCP "retrospective" for the next cycle?

pnkfelix (Jul 31 2020 at 14:24, on Zulip):

since I agree that it would be good to have @nikomatsakis at both such meetings?

nikomatsakis (Jul 31 2020 at 14:25, on Zulip):

seems fine

pnkfelix (Jul 31 2020 at 14:25, on Zulip):

this does mean we leave two friday slots unused this cycle. But I think that's okay. Let's take it easy

nikomatsakis (Jul 31 2020 at 14:25, on Zulip):

I don't think there is a rush to do the MCP meeting

pnkfelix (Jul 31 2020 at 14:26, on Zulip):

(I was again musing internally about whether to allocate some of those friday slots to discuss any MCP's that have gone unseconded)

nikomatsakis (Jul 31 2020 at 14:26, on Zulip):

interesting thought!

pnkfelix (Jul 31 2020 at 14:26, on Zulip):

(but I think that can wait. Or rather, if someone is worried that an MCP isn't getting the attention it needs, then maybe that someone can promote it to a design meeting proposal.)

nikomatsakis (Jul 31 2020 at 14:27, on Zulip):

I have been thinking that we should go through the MCPs that have been open for a while and decide how to work with them.

nikomatsakis (Jul 31 2020 at 14:27, on Zulip):

(Either close or try to at least find someone to ping)

pnkfelix (Jul 31 2020 at 14:27, on Zulip):

Basically, I think the idea of promoting unseconded MCPs to design meeting proposals is not terrible; but I am not sure whether it should be an automatic process, versus something that someone has to push through

pnkfelix (Jul 31 2020 at 14:28, on Zulip):

it seems like if there doesn't exist someone willing to do such promotion (with associated leg work of writing the meeting proposal, for example), then that is a sign that we don't have the workforce to support the proposal

pnkfelix (Jul 31 2020 at 14:28, on Zulip):

(but on the flip side, I don't want bureaucracy to become a barrier)

pnkfelix (Jul 31 2020 at 14:29, on Zulip):

Anyway, it sounds like we are agree that we'll schedule a meeting for August 14th for compiler-team#318 "discuss results of 2020 contributor survey"

nikomatsakis (Jul 31 2020 at 14:29, on Zulip):

+1

pnkfelix (Jul 31 2020 at 14:29, on Zulip):

in the 30 minutes we have remaining

pnkfelix (Jul 31 2020 at 14:29, on Zulip):

we can either:

pnkfelix (Jul 31 2020 at 14:29, on Zulip):
  1. Go over the nominated issues that we didn't have time for in yesterdays meeting
pnkfelix (Jul 31 2020 at 14:30, on Zulip):
  1. review performance issues that we barely skimmed in yesterdays meeting
pnkfelix (Jul 31 2020 at 14:30, on Zulip):

or 3. review MCP's that have gone unseconded

pnkfelix (Jul 31 2020 at 14:30, on Zulip):

I'm inclined to go with 1., go over the nominated issues

pnkfelix (Jul 31 2020 at 14:31, on Zulip):

but its possible that the people who attend this meeting are distinct enough from the people that attend the thursday meeting that we would be better off just letting the nominations wait until thursday?

nikomatsakis (Jul 31 2020 at 14:32, on Zulip):

maybe, we hvae 6 folks here, but they're all compiler contributors etc

nikomatsakis (Jul 31 2020 at 14:32, on Zulip):

not exactly a massive quorum but could do worse :)

pnkfelix (Jul 31 2020 at 14:32, on Zulip):

skimming over the nominated issues

pnkfelix (Jul 31 2020 at 14:32, on Zulip):

they all seem important enough to warrant looking at now versus waiting a week

pnkfelix (Jul 31 2020 at 14:32, on Zulip):

I'm looking at this list here: https://hackmd.io/pdiqlL2ZTfawYz_QHz0dRQ#Nominated-Issues

pnkfelix (Jul 31 2020 at 14:32, on Zulip):

so lets see

pnkfelix (Jul 31 2020 at 14:33, on Zulip):

I-nom: "Compiler doesn't terminate with --release" rust#74384

- Compiler doesn't "terminate" when nesting 3 arrays.
- "It seems like this is an LLVM bug."
- Nominated to assign it
- On our last meeting we briefly mention it but wasn't picked up yet

pnkfelix (Jul 31 2020 at 14:34, on Zulip):

a theory was posted that this is an LLVM bug, but I'm not clear on what observations led to that hypothesis

pnkfelix (Jul 31 2020 at 14:34, on Zulip):

also, LLVM ICE-breakers (is there a newer term for this) could be pinged in that case?

nikomatsakis (Jul 31 2020 at 14:34, on Zulip):

seems like a good set of folks to ping,y es

nikomatsakis (Jul 31 2020 at 14:34, on Zulip):

"LLVM notification group"

nikomatsakis (Jul 31 2020 at 14:35, on Zulip):

I imagine the basis is just that release triggers it, but I guess we could grab a backtrace with relative ease

pnkfelix (Jul 31 2020 at 14:35, on Zulip):

I'm feeling frustrated that the forge has a section called "Notification Groups" that does not list the notification groups nor the syntax to ping them

pnkfelix (Jul 31 2020 at 14:36, on Zulip):

where am I suppose to find this? I thought I actually fixed this at some point, on some site

pnkfelix (Jul 31 2020 at 14:36, on Zulip):

ah rustc-dev-guide

pnkfelix (Jul 31 2020 at 14:36, on Zulip):

https://rustc-dev-guide.rust-lang.org/notification-groups/about.html

pnkfelix (Jul 31 2020 at 14:36, on Zulip):

ETOOMANYSITES

nikomatsakis (Jul 31 2020 at 14:37, on Zulip):

it's @rustbot ping llvm I think :)

pnkfelix (Jul 31 2020 at 14:37, on Zulip):

(this actually does seem like areal problem. Since there's so many sites, there is no "single search box" that one could use to search all of them)

pnkfelix (Jul 31 2020 at 14:38, on Zulip):

/me wonders if there's a way we could make a single search box that indexes all of the sites

nikomatsakis (Jul 31 2020 at 14:38, on Zulip):

(heh we've been debating whether forge or rustc-dev-guide is better for this sort of content; I still think forge is the right place for the more "procedural" things, in part because often they are kind of cross-team, but I do see the point that having a "single source of truth" is great...)

pnkfelix (Jul 31 2020 at 14:39, on Zulip):

anyway, regarding rust#74384: I'm already planning to do some LLVM work (trying to isolate a .bc file to post as bug on LLVM's own issue tracker)

pnkfelix (Jul 31 2020 at 14:39, on Zulip):

for a different bug

pnkfelix (Jul 31 2020 at 14:39, on Zulip):

and I imagine something similar is what we'd want to do here

pnkfelix (Jul 31 2020 at 14:39, on Zulip):

let me assign rust#74384 to myself

pnkfelix (Jul 31 2020 at 14:39, on Zulip):

as part of that effort

pnkfelix (Jul 31 2020 at 14:40, on Zulip):

next

Félix Fischer (Jul 31 2020 at 14:40, on Zulip):

Hi, I'm a second late but I like the idea of a single source of truth. Gives some relief that I haven't missed somewhere to check

pnkfelix (Jul 31 2020 at 14:41, on Zulip):

yeah; do we have any web programming experts who could hack something like that up?

pnkfelix (Jul 31 2020 at 14:41, on Zulip):

(says the man who works for Mozilla)

Félix Fischer (Jul 31 2020 at 14:41, on Zulip):

(XD)

pnkfelix (Jul 31 2020 at 14:41, on Zulip):

anyway we can leave that for later

pnkfelix (Jul 31 2020 at 14:41, on Zulip):

moving on

pnkfelix (Jul 31 2020 at 14:41, on Zulip):

I-nom: "error: could not compile gkrust since Rust 1.43 on SPARC Solaris" rust#74551

- Unassigned I-unsound stable to stable regression
- Nominated by @ecstatic-morse, they think this is concerning and may affect tier 1 platforms

nikomatsakis (Jul 31 2020 at 14:42, on Zulip):

did we have a bisection?

pnkfelix (Jul 31 2020 at 14:42, on Zulip):

How are we going to investigate this?

pnkfelix (Jul 31 2020 at 14:43, on Zulip):

The bug description does have a bisection

nikomatsakis (Jul 31 2020 at 14:43, on Zulip):

The commit you bisected to (part of #69676) re-enabled the optimization around drop elaboration in #68528, which could have plausibly caused this kind of issue

nikomatsakis (Jul 31 2020 at 14:43, on Zulip):

(yes)

nikomatsakis (Jul 31 2020 at 14:43, on Zulip):

I'm at a bit of a loss

pnkfelix (Jul 31 2020 at 14:43, on Zulip):

should we consider re-disabling the optimization around drop elaboratino

pnkfelix (Jul 31 2020 at 14:43, on Zulip):

as a precaution?

nikomatsakis (Jul 31 2020 at 14:43, on Zulip):

we have no evidence to suggest that it definitely is a problem anywhere but SPARC

pnkfelix (Jul 31 2020 at 14:44, on Zulip):

true

nikomatsakis (Jul 31 2020 at 14:44, on Zulip):

I guess I would be inclined to wait and see if we see any further reports

nikomatsakis (Jul 31 2020 at 14:44, on Zulip):

or we can make a more minimal example

nikomatsakis (Jul 31 2020 at 14:44, on Zulip):

and/or to try and trace the LVM compilation and see where things 'go wrong'

pnkfelix (Jul 31 2020 at 14:45, on Zulip):

does anyone here have access to a SPARC Solaris machine...?

pnkfelix (Jul 31 2020 at 14:45, on Zulip):

that has been my concern with tracking this down this whole time

pnkfelix (Jul 31 2020 at 14:46, on Zulip):

There's not much we can do directly, from what I understand, if no one has a SPARC

Wesley Wiser (Jul 31 2020 at 14:47, on Zulip):

The GCC CompileFarm has a sparc machine. I assume it would be running Solaris: https://gcc.gnu.org/wiki/CompileFarm

Wesley Wiser (Jul 31 2020 at 14:47, on Zulip):

If anyone has access to that.

Félix Fischer (Jul 31 2020 at 14:47, on Zulip):

I had no idea Solaris was still a thing (yeah I don't have a sparc machine either)

pnkfelix (Jul 31 2020 at 14:47, on Zulip):

Wesley Wiser said:

The GCC CompileFarm has a sparc machine. I assume it would be running Solaris: https://gcc.gnu.org/wiki/CompileFarm

hmm, yes, I think @nagisa mentioned this in the past. I might have even requested access, lets see

pnkfelix (Jul 31 2020 at 14:48, on Zulip):

yes, I do have access

pnkfelix (Jul 31 2020 at 14:49, on Zulip):

or at least I have an email that told me my access was granted.

pnkfelix (Jul 31 2020 at 14:49, on Zulip):

okay, well

pnkfelix (Jul 31 2020 at 14:49, on Zulip):

I guess I'll self-assign this to see if I can replicate atop the build farm

pnkfelix (Jul 31 2020 at 14:49, on Zulip):

though I'm not sure if the bug as described there will be that easy to duplicate

pnkfelix (Jul 31 2020 at 14:50, on Zulip):

since its described as happening atop a full firefox build. :sad:

Wesley Wiser (Jul 31 2020 at 14:50, on Zulip):

oof

pnkfelix (Jul 31 2020 at 14:50, on Zulip):

I'll self-assign

Félix Fischer (Jul 31 2020 at 14:50, on Zulip):

A full... firefox build. Ouch

pnkfelix (Jul 31 2020 at 14:51, on Zulip):

next

Félix Fischer (Jul 31 2020 at 14:51, on Zulip):

(Thank you for self-assigning that, sounds like a lot of compiling time)

pnkfelix (Jul 31 2020 at 14:51, on Zulip):

I-nom: "ICE with the (foo @ ..,) pattern" rust#74702

- I-ICE, I-nominated for discussion, see Zulip topic
- No assignee yet

pnkfelix (Jul 31 2020 at 14:54, on Zulip):

I'm a little nervous that we just had a whole thing where a change to the pattern code caused us to issue a broken stable release

pnkfelix (Jul 31 2020 at 14:54, on Zulip):

so this seems like an area where we need to ... at the very least, tread carefully?

nikomatsakis (Jul 31 2020 at 14:54, on Zulip):

I dn't quite understand

pnkfelix (Jul 31 2020 at 14:55, on Zulip):

(at least, that's my understanding of what went down with PR #74557 and issue #74954)

nikomatsakis (Jul 31 2020 at 14:55, on Zulip):

that said, I think @Vadim Petrochenkov's comment makes sense, and it seems likely that this can be fixed relatively easily

pnkfelix (Jul 31 2020 at 14:55, on Zulip):

I'm trying to say that we had a recent incident where someone fixed an ICE in the pattern code and we thought it was good for a backport, but we were incorrect.

nikomatsakis (Jul 31 2020 at 14:55, on Zulip):

well I don't know I didn't dig too deeply into why the ICE is occurring exactly

Félix Fischer (Jul 31 2020 at 14:55, on Zulip):

Maybe there are more tests to make so that this regression can be detected before merging a PR

nikomatsakis (Jul 31 2020 at 14:56, on Zulip):

I see what you mean by inject a bug -- we did a stable backport, you're saying

pnkfelix (Jul 31 2020 at 14:56, on Zulip):

right

pnkfelix (Jul 31 2020 at 14:56, on Zulip):

I had tried to ensure we would cover our bases with that backport

pnkfelix (Jul 31 2020 at 14:56, on Zulip):

by adding a delay_span_bug

pnkfelix (Jul 31 2020 at 14:56, on Zulip):

but apparently it actually injected a dynamic semantics bug

pnkfelix (Jul 31 2020 at 14:57, on Zulip):

(again, that's my current understanding; I haven't dived deeply into what actually happened with the code of the PR itself)

pnkfelix (Jul 31 2020 at 14:57, on Zulip):

anyway

pnkfelix (Jul 31 2020 at 14:57, on Zulip):

I'm just expressing a bit of nervousness

pnkfelix (Jul 31 2020 at 14:57, on Zulip):

but that nervousness is really only about backporting changes here

pnkfelix (Jul 31 2020 at 14:57, on Zulip):

We should go full steam ahead on trying to make appropriate fixes on nightly

pnkfelix (Jul 31 2020 at 14:58, on Zulip):

so, yeah

pnkfelix (Jul 31 2020 at 14:58, on Zulip):

anyway, I think that the reason that this was nominated

pnkfelix (Jul 31 2020 at 14:58, on Zulip):

was to discuss the point that @Vadim Petrochenkov made, which was that we should be applying the fixes here to AST lowering, not resolve

nikomatsakis (Jul 31 2020 at 14:59, on Zulip):

ok

pnkfelix (Jul 31 2020 at 15:00, on Zulip):

which, as @nikomatsakis said, probably makes sense?

nikomatsakis (Jul 31 2020 at 15:00, on Zulip):

seems ok but I'm not sure exactly what it means

nikomatsakis (Jul 31 2020 at 15:00, on Zulip):

I guess they mean that we should resolve to an actual variable

nikomatsakis (Jul 31 2020 at 15:00, on Zulip):

but we have to report the error (or delay_span_bug) later on and give that variable a type like ty::errr?

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

hmm, yeah maybe

pnkfelix (Jul 31 2020 at 15:03, on Zulip):

its an interesting point: Like, there's a design philosophy question here, right?

pnkfelix (Jul 31 2020 at 15:03, on Zulip):

anyway

pnkfelix (Jul 31 2020 at 15:04, on Zulip):

lets leave this nominated for now

pnkfelix (Jul 31 2020 at 15:04, on Zulip):

if only because there's other work related to #74954 that remains to be done (namely changes to beta and nightly; only stable has been fixed, If i understand correctlyh)

pnkfelix (Jul 31 2020 at 15:05, on Zulip):

and the fallout from finding the right fixes to these ICEs might inform further conversation of rust#74702

pnkfelix (Jul 31 2020 at 15:05, on Zulip):

okay then that's our hour of meeting time

pnkfelix (Jul 31 2020 at 15:05, on Zulip):

thank you to everyone in @T-compiler/meeting for attending!

mark-i-m (Aug 04 2020 at 20:09, on Zulip):

Hmm... unfortunately, I might be traveling on Aug 14 :confused:

nikomatsakis (Aug 04 2020 at 21:39, on Zulip):

@pnkfelix :point_up:

pnkfelix (Aug 04 2020 at 23:20, on Zulip):

hmmm...

pnkfelix (Aug 04 2020 at 23:21, on Zulip):

sounds like we should perhaps leave compiler-team#318 to the next design meeting cycle then?

nikomatsakis (Aug 05 2020 at 09:55, on Zulip):

guess so

pnkfelix (Aug 05 2020 at 13:19, on Zulip):

okay. I'll go cancel the meeting

pnkfelix (Aug 05 2020 at 13:22, on Zulip):

I don't think a new blog post is warranted, though. :smile:

mark-i-m (Aug 05 2020 at 20:29, on Zulip):

Sorry about that!

Last update: Nov 25 2020 at 03:00UTC