Stream: t-compiler

Topic: #54818 weekly meeting 2019-02-28


mw (Feb 28 2019 at 09:36, on Zulip):

Hey, I won't be able to make it to the triage meeting today...

pnkfelix (Feb 28 2019 at 11:37, on Zulip):

okay thanks for letting us know ahead of time @mw. I'll be sure to assign you some extra bugs during triage as retribution. :wink:

pnkfelix (Feb 28 2019 at 11:49, on Zulip):

By the way, @T-compiler/meeting : our weekly meeting will start in about three hours. I'm going to start the pre-triage process a little earlier than normal because I have an errand to run before the meeting starts.

pnkfelix (Feb 28 2019 at 11:50, on Zulip):

and also, inspired by a change I adopted in the NLL meeting last night, I'll be logging the pre-triage process in a different topic, so that this one stays relatively clean prior to the meeting itself.

pnkfelix (Feb 28 2019 at 11:52, on Zulip):

Aformentioned pre-triage will proceed in another topic on this stream.

pnkfelix (Feb 28 2019 at 13:50, on Zulip):

pre-triage is over. All of the actual triage work remaining for the meeting itself is summarized in the beta-nominations and the I-nominations

pnkfelix (Feb 28 2019 at 13:50, on Zulip):

see you all in about 70 minutes

oli (Feb 28 2019 at 14:40, on Zulip):

I will be on and off during the meeting depending on train WiFi quality

pnkfelix (Feb 28 2019 at 15:00, on Zulip):

hi again @T-compiler/meeting

pnkfelix (Feb 28 2019 at 15:00, on Zulip):

/me just managed, seconds ago, to slam his eye socket into the side of a door

pnkfelix (Feb 28 2019 at 15:01, on Zulip):

/me is going to attempt to run the meeting with an ice pack in one hand

pnkfelix (Feb 28 2019 at 15:01, on Zulip):

so as stated above, I think the result of pre-triage is just the aforementioned beta-nominations and the I-nominations

pnkfelix (Feb 28 2019 at 15:01, on Zulip):

which we'll put a hard end to after 30min

pnkfelix (Feb 28 2019 at 15:02, on Zulip):

and then we'll hear from @davidtwco with a check-in from the @WG-compiler-meta

pnkfelix (Feb 28 2019 at 15:02, on Zulip):

and after that, hopefully, we'll have time to also get a checkin from @matklad (I think) from @WG-rls2.0

pnkfelix (Feb 28 2019 at 15:02, on Zulip):

so that being said, beta-nominations

matklad (Feb 28 2019 at 15:03, on Zulip):

hey! Matklad got lost calendar invite says IRC :)

pnkfelix (Feb 28 2019 at 15:03, on Zulip):

first up is "Make migrate mode work at item level granularity" #58788

pnkfelix (Feb 28 2019 at 15:03, on Zulip):

oh sorry about that @matklad

pnkfelix (Feb 28 2019 at 15:03, on Zulip):

whose in charge of the calendar? @nikomatsakis ?

nikomatsakis (Feb 28 2019 at 15:03, on Zulip):

Huh?

pnkfelix (Feb 28 2019 at 15:03, on Zulip):

hey! Matklad got lost calendar invite says IRC :)

this ^

nikomatsakis (Feb 28 2019 at 15:04, on Zulip):

I don't understand what this means :)

pnkfelix (Feb 28 2019 at 15:04, on Zulip):

okay we'll clear it up after the meeting. :smiley:

nikomatsakis (Feb 28 2019 at 15:04, on Zulip):

(That said, I should add that I can add whomever wants to be able to create calendar events)

pnkfelix (Feb 28 2019 at 15:04, on Zulip):

anyway, regarding #58788: it is brand new. I don't think we'll back port it to the stable branch

pnkfelix (Feb 28 2019 at 15:05, on Zulip):

(the bug is causing soundness bugs to be masked as warnings ; but still, they are issuing warnings)

pnkfelix (Feb 28 2019 at 15:05, on Zulip):

anyway, regarding #58788: it is brand new. I don't think we'll back port it to the stable branch

so we don't have to decide right now whether to backport. We could let it bake on master for a week

pnkfelix (Feb 28 2019 at 15:06, on Zulip):

I've added three emoji reactions to the comment above for the PR

pnkfelix (Feb 28 2019 at 15:07, on Zulip):

so if people have an opinion that can be summarized by "backport!", "don't backport!" or "wait to decide"

pnkfelix (Feb 28 2019 at 15:08, on Zulip):

(my inclination is to go ahead and backport :back:)

nikomatsakis (Feb 28 2019 at 15:08, on Zulip):

It seems fairly harmless and seems good to prevent more code w/ errors from being stable

pnkfelix (Feb 28 2019 at 15:08, on Zulip):

looks like most agree. Okay lets plan to backport

pnkfelix (Feb 28 2019 at 15:08, on Zulip):

next up: "Don't promote function calls to nonpromotable things" #58784

pnkfelix (Feb 28 2019 at 15:10, on Zulip):

I'm inclined to backport :back:

pnkfelix (Feb 28 2019 at 15:10, on Zulip):

looks like most agree

pnkfelix (Feb 28 2019 at 15:10, on Zulip):

I'm going to go slightly out of order

pnkfelix (Feb 28 2019 at 15:11, on Zulip):

next: "Include bounds from promoted constants in NLL" #57202

nikomatsakis (Feb 28 2019 at 15:11, on Zulip):

heh I was just thinking about this PR

nikomatsakis (Feb 28 2019 at 15:11, on Zulip):

and thinking I wanted to check up on what happened to it

nikomatsakis (Feb 28 2019 at 15:12, on Zulip):

I guess you reviewed it @pnkfelix :)

pnkfelix (Feb 28 2019 at 15:12, on Zulip):

yeah

nikomatsakis (Feb 28 2019 at 15:12, on Zulip):

hmm I'm not sure if I like this approach :) but we should discuss separately

pnkfelix (Feb 28 2019 at 15:13, on Zulip):

its not my area of expertise with respect to NLL, but it seemed straight-forward enough

pnkfelix (Feb 28 2019 at 15:13, on Zulip):

okay well then maybe we can wait :hourglass: ?

pnkfelix (Feb 28 2019 at 15:13, on Zulip):

(in terms of not deciding this week on whether to backport)

nikomatsakis (Feb 28 2019 at 15:13, on Zulip):

I'm inclined to wait

pnkfelix (Feb 28 2019 at 15:13, on Zulip):

sounds fine with me. we have breathing room here, i think

nikomatsakis (Feb 28 2019 at 15:13, on Zulip):

but we should schedule an actual slot to discuss, together with @Matthew Jasper, since I'm delayed on this PR more than enough

nikomatsakis (Feb 28 2019 at 15:14, on Zulip):

I'll open a topic #t-compiler/wg-nll to settle on a time :)

pnkfelix (Feb 28 2019 at 15:14, on Zulip):

okay

pnkfelix (Feb 28 2019 at 15:14, on Zulip):

final beta backport nomination is "Warning period for detecting nested impl trait" #58608

pnkfelix (Feb 28 2019 at 15:14, on Zulip):

so the reason I saved this one for last

pnkfelix (Feb 28 2019 at 15:14, on Zulip):

is that its more a question of overall policy

pnkfelix (Feb 28 2019 at 15:15, on Zulip):

i.e., if we don't plan to eventually backport this warning period to beta

pnkfelix (Feb 28 2019 at 15:15, on Zulip):

then there's pretty much no reason to have any warning period at all

nikomatsakis (Feb 28 2019 at 15:16, on Zulip):

In the PR, you write:

But: if we do not intend to beta-backport, then this PR probably should not land at all, since I do not think there is much value in having stable-breakage followed by a 6-week warning cycle. Or at least, we could just just as easily just let the trains run, and if there is significant stable breakage reported, then we do both beta- and stable-back ports at that time, and have a point release.

I think I'd rather backport than do a point release

pnkfelix (Feb 28 2019 at 15:16, on Zulip):

well a point release isn't guaranteed. :smile:

nikomatsakis (Feb 28 2019 at 15:16, on Zulip):

Yeah, I know

nikomatsakis (Feb 28 2019 at 15:16, on Zulip):

Also, I think @centril was correct (the other day) in saying that we should add these "future compat warnings" to our "triage process"

pnkfelix (Feb 28 2019 at 15:17, on Zulip):

but in terms of relative risk, I think I agree. Mainly because someone did encounter this issue in the wild

nikomatsakis (Feb 28 2019 at 15:17, on Zulip):

Though I think that need not block this decision, just a note

nikomatsakis (Feb 28 2019 at 15:17, on Zulip):

I'm inclined to backport

pnkfelix (Feb 28 2019 at 15:17, on Zulip):

(its another instance showing that crater runs really aren't a complete model of our audience's code base)

nikomatsakis (Feb 28 2019 at 15:17, on Zulip):

It feels like we should try not to break people without warnings, particularly when we can easily do so

nikomatsakis (Feb 28 2019 at 15:17, on Zulip):

it gives us some credit to spend when we cause breakage that's harder to avoid :)

pnkfelix (Feb 28 2019 at 15:18, on Zulip):

okay. i added the emoji reactions

pnkfelix (Feb 28 2019 at 15:18, on Zulip):

my own opinion was that I was not really inclined to backport

pnkfelix (Feb 28 2019 at 15:19, on Zulip):

in part because I'm not thrilled with the patch itself.

pnkfelix (Feb 28 2019 at 15:19, on Zulip):

(to explain: I was trying to get it into the nightly before it got cut, so I rushed a little)

pnkfelix (Feb 28 2019 at 15:19, on Zulip):

((though clearly I didn't rush enough))

pnkfelix (Feb 28 2019 at 15:20, on Zulip):

anyway

pnkfelix (Feb 28 2019 at 15:20, on Zulip):

it seems like people are generally okay with the idea of backporting this then?

pnkfelix (Feb 28 2019 at 15:20, on Zulip):

(we should wait until it actually successfully lands on master)

nikomatsakis (Feb 28 2019 at 15:20, on Zulip):

in part because I'm not thrilled with the patch itself.

well, I guess changes things a bit? The patch seemed..ok to me

nikomatsakis (Feb 28 2019 at 15:20, on Zulip):

I didn't look super closely

pnkfelix (Feb 28 2019 at 15:21, on Zulip):

I'm not sure there's much of a cleaner way to precisely capture the cases involved

pnkfelix (Feb 28 2019 at 15:21, on Zulip):

but that doesn't mean I'm happy with it.

nikomatsakis (Feb 28 2019 at 15:21, on Zulip):

You're just concerned about the messiness of random boolean flags? Seems fair.

pnkfelix (Feb 28 2019 at 15:22, on Zulip):

yeah, random flags being toggled at seemingly random places that I derived from somewhat careful analysis of the previous control-flow structure of two of the visitors that had been merged into this common visitor

pnkfelix (Feb 28 2019 at 15:22, on Zulip):

how about if we leave it nominated for a week

pnkfelix (Feb 28 2019 at 15:22, on Zulip):

but if no one objects during that week, then we'll approve it for backport

nikomatsakis (Feb 28 2019 at 15:22, on Zulip):

seems ok

pnkfelix (Feb 28 2019 at 15:23, on Zulip):

okay

pnkfelix (Feb 28 2019 at 15:23, on Zulip):

Next, I-nominated

pnkfelix (Feb 28 2019 at 15:23, on Zulip):

we just finished discussing #58608

pnkfelix (Feb 28 2019 at 15:24, on Zulip):

next is " Type inference fails even though all types are known" #58517

pnkfelix (Feb 28 2019 at 15:24, on Zulip):

I nominated this because I triaged it as P-medium but I want people to have a chance to object to that prioritization

pnkfelix (Feb 28 2019 at 15:25, on Zulip):

it also ... surprised me. At leas,t the hypothesis that it was injected by "Implement 1581 (FusedIterator)" #35656 I found interesting.

nikomatsakis (Feb 28 2019 at 15:25, on Zulip):

I have mixed feels

pnkfelix (Feb 28 2019 at 15:25, on Zulip):

but in all honesty there may not be much to discuss until someone goes in and looks at it more carefully

pnkfelix (Feb 28 2019 at 15:26, on Zulip):

does anyone want to volunteer to dissect #58517 ?

nikomatsakis (Feb 28 2019 at 15:26, on Zulip):

It would perhaps be useful to be collecting examples of places where the type checker needs improvement. I am thinking a bit about the proposed work in the context of RLS 2.0 etc on extracting type checker.

pnkfelix (Feb 28 2019 at 15:26, on Zulip):

(dissection could take many forms; even just bisecting to confirm/deny the above hypothesis woudl be useful)

pnkfelix (Feb 28 2019 at 15:27, on Zulip):

well I don't want to block on waiting for a response to that Question

pnkfelix (Feb 28 2019 at 15:28, on Zulip):

I could leave it nominated, or I could promote to P-high, or I could try to assign it to some team WG to look at further

nikomatsakis (Feb 28 2019 at 15:28, on Zulip):

it's been broken a while

nikomatsakis (Feb 28 2019 at 15:28, on Zulip):

feels like it's hard to justify as P-high per se

nikomatsakis (Feb 28 2019 at 15:28, on Zulip):

though I feel like it'd be interesting to have a working group dedicated to tracking down stable-to-stable regressions like this

nikomatsakis (Feb 28 2019 at 15:29, on Zulip):

as well as other ICEs and lower priority bugs

pnkfelix (Feb 28 2019 at 15:29, on Zulip):

yet another option is that I assign to myself to look at

pnkfelix (Feb 28 2019 at 15:29, on Zulip):

but if it stays at P-medium then I don't know if I'll get around to it soon

nikomatsakis (Feb 28 2019 at 15:29, on Zulip):

well I don't know if WG is the right framing but it continues to feel like an interesting role

pnkfelix (Feb 28 2019 at 15:29, on Zulip):

there are other isues I want us to at least hit on

nikomatsakis (Feb 28 2019 at 15:29, on Zulip):

(maybe interesting to consider for "journeypeople..")

pnkfelix (Feb 28 2019 at 15:29, on Zulip):

first is "fn generated by macro exported from crate loses global #![allow(non_snake_case)]" #58502

pnkfelix (Feb 28 2019 at 15:30, on Zulip):

my question here is: what is even the right behavior? Who decides what is the right behavior?

nikomatsakis (Feb 28 2019 at 15:30, on Zulip):

so

nikomatsakis (Feb 28 2019 at 15:31, on Zulip):

We had some discussions related to this, but never reached a conclusion, in the context of Edition Hygiene

pnkfelix (Feb 28 2019 at 15:31, on Zulip):

okay. this may be best delegated to a specification issue

pnkfelix (Feb 28 2019 at 15:31, on Zulip):

not something we try to figure out in this meeting

pnkfelix (Feb 28 2019 at 15:31, on Zulip):

but we don't have a WG-macros yet, right?

nikomatsakis (Feb 28 2019 at 15:31, on Zulip):

Basically talking about how to decide what edition applied when a construct came from multiple sources -- this seems like the same base question to me

nikomatsakis (Feb 28 2019 at 15:31, on Zulip):

We don't

nikomatsakis (Feb 28 2019 at 15:32, on Zulip):

It seems like exploring hygiene was in theory one of the things you were going to try and do, @pnkfelix, as part of that name res work

pnkfelix (Feb 28 2019 at 15:32, on Zulip):

is there an open issue regarding "how to decide what edition applied when a construct came from multiple sources"

pnkfelix (Feb 28 2019 at 15:32, on Zulip):

It seems like exploring hygiene was in theory one of the things you were going to try and do, pnkfelix, as part of that name res work

this is true

nikomatsakis (Feb 28 2019 at 15:32, on Zulip):

is there an open issue regarding "how to decide what edition applied when a construct came from multiple sources"

yes, I have to find it

nikomatsakis (Feb 28 2019 at 15:32, on Zulip):

That said, do we want to consider this from the regression POV?

pnkfelix (Feb 28 2019 at 15:32, on Zulip):

so maybe I'll just assign #58502 to myself and move along

nikomatsakis (Feb 28 2019 at 15:32, on Zulip):

It feels like a grey area

pnkfelix (Feb 28 2019 at 15:33, on Zulip):

well, lets leave it P-high for now

pnkfelix (Feb 28 2019 at 15:34, on Zulip):

okay I wanted a hard limit of 30 minutes

pnkfelix (Feb 28 2019 at 15:34, on Zulip):

there are some P-high I-nominated issues though

pnkfelix (Feb 28 2019 at 15:34, on Zulip):

/me skims to see if any must be addressed today/this week

pnkfelix (Feb 28 2019 at 15:35, on Zulip):

/me skims to see if any must be addressed today/this week

no I think we're okay

pnkfelix (Feb 28 2019 at 15:35, on Zulip):

so let's move along to Working Group check-in!

pnkfelix (Feb 28 2019 at 15:35, on Zulip):

@davidtwco are you here?

nikomatsakis (Feb 28 2019 at 15:36, on Zulip):

(If not, I can fill in)

pnkfelix (Feb 28 2019 at 15:36, on Zulip):

okay lets have @nikomatsakis drive presentation of what WG-meta has been up to

nikomatsakis (Feb 28 2019 at 15:37, on Zulip):

So, we had our first meeting, and we talked about a few issues, but we've been trying to focus mostly on "actionable" things in terms of getting the WG system up and going.

nikomatsakis (Feb 28 2019 at 15:37, on Zulip):

To that end, we created a **template** for each working group that describes its basic structure

nikomatsakis (Feb 28 2019 at 15:38, on Zulip):

Along with a kind of checklist of what to do when creating one

nikomatsakis (Feb 28 2019 at 15:39, on Zulip):

And we started contacting people who had said they might like to setup a working group

nikomatsakis (Feb 28 2019 at 15:39, on Zulip):

(You can see the topics in #t-compiler/wg-meta)

nikomatsakis (Feb 28 2019 at 15:39, on Zulip):

Our next meeting is today

nikomatsakis (Feb 28 2019 at 15:39, on Zulip):

I'm hoping we can use the meeting to draw out our roadmap a bit more clearly -- it seems like the first step is pretty clearly to get templates in place for the first few working groups

nikomatsakis (Feb 28 2019 at 15:39, on Zulip):

But I'd be curious to hear back from others on what the most useful work we could do is :)

nikomatsakis (Feb 28 2019 at 15:40, on Zulip):

We've talked about a few different things, such as writing up guidelines for people running WGs, documenting other procedures, etc

nikomatsakis (Feb 28 2019 at 15:41, on Zulip):

(This reminds me that tomorrow is the steering meeting, and I was hoping to propose some discussion topics related to this for that -- basically questions that arose where I think more broad discussion might be useful)

nikomatsakis (Feb 28 2019 at 15:42, on Zulip):

One thing that interests me is sort of fine-tuning the WG "presentation process". I'd really like to have it be a bit clearer what we're going for here and to have a more finely tuned system for it.

nikomatsakis (Feb 28 2019 at 15:42, on Zulip):

(And doing a better job of notifying the broader community about our progress)

pnkfelix (Feb 28 2019 at 15:42, on Zulip):

So I want to make sure that its as easy as possible for T-compiler liasons to update the NOTES.md

nikomatsakis (Feb 28 2019 at 15:42, on Zulip):

fine-tuning the WG "presentation process"

that is, the process we're doing right now :) such meta meta

pnkfelix (Feb 28 2019 at 15:43, on Zulip):

but I also don't want to end up with just cut-and-pasting the log of a zulip conversation

nikomatsakis (Feb 28 2019 at 15:43, on Zulip):

Yeah, there is definitely some balance between "ease of maintenance" and "usefulness to an outsider"

pnkfelix (Feb 28 2019 at 15:43, on Zulip):

(such chat logs might be useful to archive, perhaps, but do not seem appropriate for NOTES.md)

nikomatsakis (Feb 28 2019 at 15:43, on Zulip):

I think it woudl be very helpful if we are able to -- for each WG -- setup a set of clear next steps and then just be able to talk about what happened to each one

pnkfelix (Feb 28 2019 at 15:44, on Zulip):

during the check-in, you mean?

nikomatsakis (Feb 28 2019 at 15:44, on Zulip):

right

nikomatsakis (Feb 28 2019 at 15:44, on Zulip):

it may well be that those next steps turned out to be irrelevant and other things we're done instead

pnkfelix (Feb 28 2019 at 15:44, on Zulip):

ah that's what you meante by "presentation process", I see

nikomatsakis (Feb 28 2019 at 15:44, on Zulip):

that seems like useful data

nikomatsakis (Feb 28 2019 at 15:45, on Zulip):

ah that's what you meante by "presentation process", I see

it is, though I think it's non-trivial work. It's sometimes hard to plan. But I think if we really want to get more folks involved, it's definitely imporant work. (That is of course not the goal in all WGs, but even there we should be able to plan the coding process a bit..?)

nikomatsakis (Feb 28 2019 at 15:45, on Zulip):

(Question, were we going to talk about RLS 2.0 today too?)

matklad (Feb 28 2019 at 15:45, on Zulip):

we were planing to

nikomatsakis (Feb 28 2019 at 15:45, on Zulip):

I'm just looking at the clock

matklad (Feb 28 2019 at 15:46, on Zulip):

15 minutes seems enough for a "heart-beat"?

matklad (Feb 28 2019 at 15:46, on Zulip):

Or do you want to move it elsewhere? (to the steering mtng?)

nikomatsakis (Feb 28 2019 at 15:46, on Zulip):

Seems ok. We could talk in more depth about the strategy at the steering meeting if desired

nikomatsakis (Feb 28 2019 at 15:46, on Zulip):

I think the heart beat is good

matklad (Feb 28 2019 at 15:47, on Zulip):

:ok:

pnkfelix (Feb 28 2019 at 15:47, on Zulip):

okay great, yeah lets hear from WG-rls2.0 then

matklad (Feb 28 2019 at 15:47, on Zulip):

So, we've had a number of meeting for

matklad (Feb 28 2019 at 15:48, on Zulip):

the strategy one is the most interesting I guess

matklad (Feb 28 2019 at 15:49, on Zulip):

We were thinking about what work to actually do, and two approaches seemed reasonable:

matklad (Feb 28 2019 at 15:49, on Zulip):
matklad (Feb 28 2019 at 15:51, on Zulip):

We figured out that end-to-end is more valuable for an experiemnt (we get more data about hard bits), and can actually help with the second one: instead of integrating the library with rustc (which requires the library to be strintlcy production ready), it can be integrated immediately with rust-analyzer to get fast feedback)

matklad (Feb 28 2019 at 15:52, on Zulip):

The first candidates for a deeper focus are "parsing and syntax trees", and "macro expansion and name resolution"

nikomatsakis (Feb 28 2019 at 15:52, on Zulip):

One thing I want to add: it seems pretty important to me that, whatever we do, we avoid the "grand overarching redesign" danger, where we are not able to show people progress because we're busy doing overarching work. I think the end-to-end LS that focuses on a subset of Rust is actually a great way of doing that, as it lets people play with it early and get use out of it, and we can then focus on expanding that subset of Rust, but it depends a bit on how fast we think we'll be able to get to "usable" (I think the feeling is .. probably quickly?)

nikomatsakis (Feb 28 2019 at 15:52, on Zulip):

(Note: this is a danger that I feel personally prone to :) e.g., it happened to some extent with MIR, chalk, etc, and I'm trying to be a bit sensitive to that.)

matklad (Feb 28 2019 at 15:53, on Zulip):

Yeah, I personally think "second system syndrom" a couple of times a day :)

matklad (Feb 28 2019 at 15:54, on Zulip):

we haven't started macros and nameres work (though @pnkfelix seems to be in charge here :0)

pnkfelix (Feb 28 2019 at 15:54, on Zulip):

in name at least

matklad (Feb 28 2019 at 15:54, on Zulip):

for "parsing and AST", we did a survey of what currently exists in rust-analyzer. There's a youtube video of the surver

matklad (Feb 28 2019 at 15:55, on Zulip):

that I think is all for WG-RLS-2.0?

matklad (Feb 28 2019 at 15:56, on Zulip):

think we'll be able to get to "usable"

shameless plug, but some folks are already using rust-analyzer for working on rustc, because it at least partially works :)

nikomatsakis (Feb 28 2019 at 15:57, on Zulip):

Yeah :)

nikomatsakis (Feb 28 2019 at 15:57, on Zulip):

I don't expect us to define it in this meeting, but I feel like @matklad it would be a good goal to try and be crisp about the subset of Rust we plan to handle and the work needed to get us to a real functional system.

nikomatsakis (Feb 28 2019 at 15:57, on Zulip):

I think the fact that rust-analyzer is in use already is definitely a good evidence-based argument for why the plan can work

pnkfelix (Feb 28 2019 at 15:57, on Zulip):

even if we can't precise define the edges of the subset

nikomatsakis (Feb 28 2019 at 15:57, on Zulip):

I'm also thinking about how we talk about this externally -- I want to get people excited

pnkfelix (Feb 28 2019 at 15:57, on Zulip):

stating which crates are required for minimum viable product (MVP)

pnkfelix (Feb 28 2019 at 15:57, on Zulip):

is a good idea

nikomatsakis (Feb 28 2019 at 15:58, on Zulip):

both because it's good for project but also because it'll lead to more involvement

nikomatsakis (Feb 28 2019 at 15:58, on Zulip):

stating which crates are required for MVP

yes, this

nikomatsakis (Feb 28 2019 at 15:58, on Zulip):

That's probably the best way to "define" the subset in some sense

matklad (Feb 28 2019 at 15:58, on Zulip):

it'll lead to more involvement

I'd love to see more involvement from t-compiler first, in terms of actual code :)

matklad (Feb 28 2019 at 15:59, on Zulip):

Sorry if this sounds a bit offensive, I just can't find a better words.

pnkfelix (Feb 28 2019 at 15:59, on Zulip):

yeah sadly I have to admit that I haven't tried building rust-analyzer recently

nikomatsakis (Feb 28 2019 at 15:59, on Zulip):

Sounds fair =)

pnkfelix (Feb 28 2019 at 15:59, on Zulip):

got invested in the vfs fixes for Mac OS X

pnkfelix (Feb 28 2019 at 15:59, on Zulip):

then got side-tracked

matklad (Feb 28 2019 at 15:59, on Zulip):

what I am trying to say is that for invovment to be efficient, mentoring must be awesome

matklad (Feb 28 2019 at 15:59, on Zulip):

currently, mentoring is bottlenecked on me and @Florian Diebold :)

nikomatsakis (Feb 28 2019 at 16:00, on Zulip):

Hmm, yes.

nikomatsakis (Feb 28 2019 at 16:00, on Zulip):

I was wondering about it as I wrote it :)

nikomatsakis (Feb 28 2019 at 16:00, on Zulip):

that is -- about the desire to have people involved, but also the desire for us to be ready

nikomatsakis (Feb 28 2019 at 16:03, on Zulip):

So -- we're at 11 -- but I wanted to pose a question. If we were to spend a time on RLS2.0 discussion tomorrow, what would be the best angle? It seems like maybe the most relevant topic is trying to talk about .. hmm .. who will have time to do what? How much involvement is to be expected?

matklad (Feb 28 2019 at 16:03, on Zulip):

I guess the time is up, but I'd love to discuss "how to involve t-compiler with IDE business" question some time :)

pnkfelix (Feb 28 2019 at 16:03, on Zulip):

@nikomatsakis you mean at steering meeting, right?

matklad (Feb 28 2019 at 16:04, on Zulip):

who will have time to do what?

:+1:

Igor Matuszewski (Feb 28 2019 at 16:04, on Zulip):

what time will the meeting be at tomorrow?

pnkfelix (Feb 28 2019 at 16:04, on Zulip):

same time as this meeting

Igor Matuszewski (Feb 28 2019 at 16:04, on Zulip):

great, thanks

matklad (Feb 28 2019 at 16:04, on Zulip):

heh, that means that I might not be able to participate: got a Rust class at this time slot :(

nikomatsakis (Feb 28 2019 at 16:05, on Zulip):

D'oh

nikomatsakis (Feb 28 2019 at 16:05, on Zulip):

the irony of missing steering meeting for rust

nikomatsakis (Feb 28 2019 at 16:05, on Zulip):

that said, there are other things (beyond RLS 2.0) worth talking about

matklad (Feb 28 2019 at 16:05, on Zulip):

I'll be available through the first 20 minutes though

nikomatsakis (Feb 28 2019 at 16:05, on Zulip):

I was going to brainstorm a few ideas and leave a post with some thoughts

nikomatsakis (Feb 28 2019 at 16:05, on Zulip):

If nothing else we can advertise for RLS 2.0 involvement :)

davidtwco (Feb 28 2019 at 16:11, on Zulip):

Apologies for not being able to attend, was dragged into meetings at the last minute.

davidtwco (Feb 28 2019 at 16:12, on Zulip):

I wrote this in advance but wasn't here to post it:

This week, the meta working group held a meeting where we briefly discussed what the goal and scope of the meta working group should be; we then discussed the bootstrap process for working groups - what information we need from working groups, what's the minimum set of guidelines we can establish for the working groups, etc.

Before the meeting, we produced a list of ideas for things that the meta working group might want to think about. nikomatsakis then made a spreadsheet that summarized these more.

The meeting produced a bunch of "components" that a working group would need - a Zulip stream, GitHub group, contacts (leads, compiler team meeting liason, etc.) and things like that. We also discussed what it would mean to be a member of a working group.

We then went away and produced the working group template that other teams can base their directories in the repository on. Niko added some notes about creating new working groups and that process. Minimal definitions for the meta and async working groups were added. Niko then started kicking off some of the working groups in the #t-compiler/wg-meta channel.

pnkfelix (Feb 28 2019 at 16:16, on Zulip):

THanks @davidtwco

pnkfelix (Feb 28 2019 at 16:16, on Zulip):

I have to go now; bye everyone, thanks for attending!

Last update: Nov 16 2019 at 01:10UTC