Stream: t-compiler/wg-meta

Topic: rewording the requirements for design meetings


nikomatsakis (Sep 13 2019 at 12:38, on Zulip):

I was thinking today that we should "lighten up" the "design meeting rules". It seems clear that the current rules are "too onerous", and I think we should bias towards "just come with an interesting idea, as many details as you want". If that becomes an issue, we'll worry about it then.

nikomatsakis (Sep 13 2019 at 12:38, on Zulip):

And no this has nothing to do with me still trying to figure out just how to prep for a discussion about rust-analyzer/chalk a few hours before meeting ;)

pnkfelix (Sep 13 2019 at 12:42, on Zulip):

well ... I will admit that we seem to set the bar too high for people to post proposals

pnkfelix (Sep 13 2019 at 12:43, on Zulip):

but at the same time: This is a meeting where we are striving to get as much of the team synchronously present (and actively engaged). So it makes sense for there to be some up front effort put in, to allow the other participants to usefully prep before hand.

pnkfelix (Sep 13 2019 at 12:44, on Zulip):

if you make the proposals free form, then why not just devote the time to going through T-compiler I-nominated issues?

pnkfelix (Sep 13 2019 at 12:45, on Zulip):

(because github issues are not appropriate medium for communicating design ideas, perhaps ....)

nikomatsakis (Sep 13 2019 at 12:47, on Zulip):

Yeah, I know. I mean maybe the answer is we just have to try a bit harder.

nikomatsakis (Sep 13 2019 at 12:48, on Zulip):

Still, I think we can kind of "work our way" there a bit?

pnkfelix (Sep 13 2019 at 12:48, on Zulip):

I guess my point is, when you say:

If that becomes an issue, we'll worry about it then.

but it may not be easy to tell if it becomes an issue. If people disengage from or skip the meeting because they do not understand what is being proposed, we may or may not be able to detect that

nikomatsakis (Sep 13 2019 at 12:48, on Zulip):

Fair.

nikomatsakis (Sep 13 2019 at 12:48, on Zulip):

if you make the proposals free form, then why not just devote the time to going through T-compiler I-nominated issues?

I guess I think this is .. sort of reasonable?

nikomatsakis (Sep 13 2019 at 12:49, on Zulip):

more to the point, nomination occurs for many reasons

nikomatsakis (Sep 13 2019 at 12:49, on Zulip):

but the reason of "needs closer attention" feels to me pretty clearly ill-served today

nikomatsakis (Sep 13 2019 at 12:49, on Zulip):

So @Zoxc pointed me at this internals post that they wrote some time back -- which I completely missed at the time.

nikomatsakis (Sep 13 2019 at 12:50, on Zulip):

I'd like to have a design meeting to discuss that

pnkfelix (Sep 13 2019 at 12:50, on Zulip):

yeah maybe we could, during the thursday meetings, enqueue topics for the coming friday meetings

nikomatsakis (Sep 13 2019 at 12:50, on Zulip):

seems plausible yeah to spin out a "meeting proposal" as a way of handling an issue that can't be resolved with a few minutes of discussion

nikomatsakis (Sep 13 2019 at 12:50, on Zulip):

I guess I just think that some amount of prep and thought will occur "semi-naturally"

nikomatsakis (Sep 13 2019 at 12:50, on Zulip):

and/or maybe we can give some structure to "encourage" it without being forbidding to start

nikomatsakis (Sep 13 2019 at 12:50, on Zulip):

I guess I just think that some amount of prep and thought will occur "semi-naturally"

this is obviously false :P

nikomatsakis (Sep 13 2019 at 12:51, on Zulip):

anyway I don't know the right answer but I don't think we're hitting it yet :)

pnkfelix (Sep 13 2019 at 12:51, on Zulip):

seems plausible yeah to spin out a "meeting proposal" as a way of handling an issue that can't be resolved with a few minutes of discussion

I'd be willing to give this a shot, okay.

Last update: Nov 11 2019 at 22:00UTC