Stream: t-compiler/wg-nll

Topic: status-of-borrowck-blog-post


nikomatsakis (Jun 12 2018 at 17:38, on Zulip):

@pnkfelix et al.: so I started writing up a blog post on the status of borrowck. You can see the draft in this gist. I'd love to get some feedback on the content, but I'm really writing because I realized at the end that I think we could/should maybe do a bit more to make it easy to figure out how to "hop in" and contribute. I'm not sure what to suggest, really? GitHub issues are an option, though I worry about notifications getting lost there — it's increasingly hard to keep up, at least for me personally. Coming to the meeting is an option, but it won't work for all -- should we have an internals thread that we actually keep up to date? Rely on people to read the backlog? Not sure, would appreciate thoughts.

nikomatsakis (Jun 12 2018 at 17:39, on Zulip):

cc @Aaron Turon :point_up:

Keith Yeung (Jun 12 2018 at 18:26, on Zulip):

:thumbs_down:, we should replace all mentions of Polonius to :frog:

Keith Yeung (Jun 12 2018 at 18:27, on Zulip):

that gets people excited

lqd (Jun 12 2018 at 18:35, on Zulip):

maybe keeping this internals thread up to date with the weekly meeting update from the Paper document, like a "this week in NLL" that would also contain GH issues or small descriptions about topics where help is needed

nikomatsakis (Jun 12 2018 at 18:37, on Zulip):

Yeah, I was thinking of something similar

nikomatsakis (Jun 12 2018 at 18:37, on Zulip):

:thumbs_down:, we should replace all mentions of Polonius to :frog:

i can't tell how serious you are here — I could invest a bit of time writing up more about 'phase 2', but I think I'd rather push that to a follow-up post

nikomatsakis (Jun 12 2018 at 18:38, on Zulip):

I can certainly mention datafrog :)

nikomatsakis (Jun 12 2018 at 18:38, on Zulip):

but I don't really want to just get people excited about Polonius — That's not gonna ship for a while, I want to focus on what will ship, which is already exciting enough

Keith Yeung (Jun 12 2018 at 19:51, on Zulip):

on a more serious note -- would people get a bit disappointed that we're not shipping the changes for :frog:? if i were not in WG-nll, i would assume something went wrong in the implementation of NLL, and we had to take a step back from implementing it fully

Keith Yeung (Jun 12 2018 at 19:54, on Zulip):

well, i guess it's a good thing that we kept the NLL term relatively vague, so any kind of new analysis can count as NLL

lqd (Jun 12 2018 at 20:32, on Zulip):

for the performance part of the post, I guess we could mention the NLL dashboard as well — and in general the rustc-perf benchmarks — as a good start to the "NLL performance work trail of breadcrumbs" ? :) (also: in the post, the sidenote about inflateseems unused)

nikomatsakis (Jun 12 2018 at 22:27, on Zulip):

on a more serious note -- would people get a bit disappointed that we're not shipping the changes for :frog:?

part of the point of the post is to set expectations appropriately, I would say.

nikomatsakis (Jun 13 2018 at 14:24, on Zulip):

ok I totally rewrote the post: https://gist.github.com/nikomatsakis/74e541ec03e6b777040a8fb8ad6a2580

Santiago Pastorino (Jun 13 2018 at 14:26, on Zulip):

need to read it :)

Santiago Pastorino (Jun 13 2018 at 14:26, on Zulip):

will do today

davidtwco (Jun 13 2018 at 14:28, on Zulip):

Looks good!

nikomatsakis (Jun 13 2018 at 14:33, on Zulip):

added a few more details and things

lqd (Jun 13 2018 at 14:39, on Zulip):

very nice ! tiny typo on "map is borrowed to produce" and "map is no longer referenced" -> should be vec instead of map

Last update: Nov 22 2019 at 00:10UTC