Stream: t-compiler/wg-prioritization

Topic: should we consider I-unsound issues


Santiago Pastorino (May 21 2020 at 17:56, on Zulip):

I was wondering if we should consider I-unsound as a way to call for prioritization

Santiago Pastorino (May 21 2020 at 17:56, on Zulip):

more or less this https://github.com/rust-lang/rust/issues?q=is%3Aopen+label%3A%22I-unsound+%F0%9F%92%A5%22+label%3A%22T-compiler%22+-label%3A%22P-critical%22+-label%3A%22P-high%22+-label%3A%22P-medium%22+-label%3A%22P-low%22

Santiago Pastorino (May 21 2020 at 17:56, on Zulip):

there are 15 I-unsound issues for T-compiler

Santiago Pastorino (May 21 2020 at 17:57, on Zulip):

wonder if we should prioritize them all and starting now, if we should automatically tag with I-prioritize

Santiago Pastorino (May 21 2020 at 17:57, on Zulip):

@WG-prioritization thoughts?

o0Ignition0o - Jeremy Lempereur (May 21 2020 at 17:58, on Zulip):

I think we should. I unsound seems like something to watch closely

LeSeulArtichaut (May 21 2020 at 17:59, on Zulip):

Do we want to include feature-gated unsoundness, i.e. with a requires-nightly label?

LeSeulArtichaut (May 21 2020 at 18:02, on Zulip):

Other than those, yes, I also think we should prioritize them all

simulacrum (May 21 2020 at 18:06, on Zulip):

hm I would actually say that they do not generally need to be something we're checking in on

simulacrum (May 21 2020 at 18:06, on Zulip):

a one time pass though makes sense perhaps

simulacrum (May 21 2020 at 18:06, on Zulip):

maybe surfacing them in monthly reports or something, but not beyond that

simulacrum (May 21 2020 at 18:06, on Zulip):

if we had 0 then a new one would be more important

simulacrum (May 21 2020 at 18:06, on Zulip):

I think surfacing new ones is important in general

simulacrum (May 21 2020 at 18:06, on Zulip):

but historical ones, well, if we've not solved them by now telling t-compiler in triage meetings seems unlikely to help

Santiago Pastorino (May 21 2020 at 18:29, on Zulip):

yeah, I was looking mainly to add a search to triagebot

Santiago Pastorino (May 21 2020 at 18:30, on Zulip):

maybe as @LeSeulArtichaut pointed out, we should skip requires-nightly

Santiago Pastorino (May 21 2020 at 18:30, on Zulip):

so this search https://github.com/rust-lang/rust/issues?q=is%3Aopen+label%3A%22I-unsound+%F0%9F%92%A5%22+label%3A%22T-compiler%22+-label%3A%22P-critical%22+-label%3A%22P-high%22+-label%3A%22P-medium%22+-label%3A%22P-low%22+-label%3A%22requires-nightly%22

Last update: Jun 05 2020 at 22:20UTC