Stream: t-compiler/wg-prioritization

Topic: t-libs and libs-impl


Santiago Pastorino (May 08 2020 at 18:42, on Zulip):

@pnkfelix we've said that we should consider issues with libs-impl label from now on

Santiago Pastorino (May 08 2020 at 18:42, on Zulip):

I'm going to add those to our procedures

Santiago Pastorino (May 08 2020 at 18:43, on Zulip):

but we've also talked about T-libs

Santiago Pastorino (May 08 2020 at 18:44, on Zulip):

so talking about concrete steps

Santiago Pastorino (May 08 2020 at 18:45, on Zulip):

we've said that we don't want libs nominations, so that means no T-libs and I-nominated and neither libs-impl and I-nominated, right?

Santiago Pastorino (May 08 2020 at 18:45, on Zulip):

unsure if libs-impl and I-nominated together make sense

pnkfelix (May 08 2020 at 18:45, on Zulip):

um

pnkfelix (May 08 2020 at 18:45, on Zulip):

I think libs-impl and I-nominated could be us, and probably is

pnkfelix (May 08 2020 at 18:45, on Zulip):

i mean, it falls back into the classic "what was this nominated for?" problem

Santiago Pastorino (May 08 2020 at 18:45, on Zulip):

yes if happens I guess

pnkfelix (May 08 2020 at 18:46, on Zulip):

but if something is tagged libs-impl, then I assume its (solely?) a library implementation issue

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

I'd add libs-impl and I-nominated

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

yeah

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

ok, I think that's fine

pnkfelix (May 08 2020 at 18:46, on Zulip):

I suppose its possible some issues could be both design and implementation related

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

yeah

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

also, do we want to search for things that are general T-libs?

pnkfelix (May 08 2020 at 18:46, on Zulip):

which could lead to it being tagged with libs-impl while also being nominate for lib design discussion

pnkfelix (May 08 2020 at 18:46, on Zulip):

we'll just have to deal with those cases

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

Santiago Pastorino said:

also, do we want to search for things that are general T-libs?

it's clear to me that we don't want nominations

Santiago Pastorino (May 08 2020 at 18:46, on Zulip):

but do we want something from there?

pnkfelix (May 08 2020 at 18:47, on Zulip):

I want to be careful about wasting our time on general T-libs searches

pnkfelix (May 08 2020 at 18:47, on Zulip):

there are some that may make sense in general

pnkfelix (May 08 2020 at 18:47, on Zulip):

but I would almost prefer

pnkfelix (May 08 2020 at 18:47, on Zulip):

that we try to just let them get tagged with libs-impl separately

Santiago Pastorino (May 08 2020 at 18:48, on Zulip):

yeah

pnkfelix (May 08 2020 at 18:48, on Zulip):

rather than commit to going through the set of T-libs issues each week

Santiago Pastorino (May 08 2020 at 18:48, on Zulip):

I wonder if we should do some re-tagging work

pnkfelix (May 08 2020 at 18:48, on Zulip):

maybe

Santiago Pastorino (May 08 2020 at 18:48, on Zulip):

like things that are T-libs but should really be libs-impl

Santiago Pastorino (May 08 2020 at 18:48, on Zulip):

unsure but may be a lot of work :)

pnkfelix (May 08 2020 at 18:48, on Zulip):

its also possible

Santiago Pastorino (May 08 2020 at 18:48, on Zulip):

we could also let it happen

pnkfelix (May 08 2020 at 18:48, on Zulip):

that we may end up deciding that a libs-design tag is needed

pnkfelix (May 08 2020 at 18:49, on Zulip):

just so we can tag issues with that

pnkfelix (May 08 2020 at 18:49, on Zulip):

to make it clear "we've looked at it, its not implementation work"

Santiago Pastorino (May 08 2020 at 18:49, on Zulip):

yep

pnkfelix (May 08 2020 at 18:49, on Zulip):

but I haven't pulled trigger on that yet

pnkfelix (May 08 2020 at 18:49, on Zulip):

we can wait until we find evidence that its warranted

Santiago Pastorino (May 08 2020 at 18:49, on Zulip):

that would be the best way to go forward really

Santiago Pastorino (May 08 2020 at 18:50, on Zulip):

the other possibility is that everyone is aware that libs-impl exist and we keep starting to tag things with it as we go

Santiago Pastorino (May 08 2020 at 18:50, on Zulip):

so libs issues should have or T-libs or libs-impl tag

Santiago Pastorino (May 08 2020 at 18:50, on Zulip):

the best thing about this approach is that every Rust contributor would be involved in the process :)

Santiago Pastorino (May 08 2020 at 18:51, on Zulip):

but it will take some possible large time to be sure that we don't have things unproperly tagged

pnkfelix (May 08 2020 at 19:02, on Zulip):

Santiago Pastorino said:

so libs issues should have or T-libs or libs-impl tag

well, I do think that everything should, in principle, have at least one T- tag after its been triaged

pnkfelix (May 08 2020 at 19:02, on Zulip):

so maybe libs issues should have either T-libs or libs-impl+T-compiler ?

pnkfelix (May 08 2020 at 19:03, on Zulip):

which I admit is confusing

pnkfelix (May 08 2020 at 19:03, on Zulip):

but the situation itself is confusing

Santiago Pastorino (May 08 2020 at 19:07, on Zulip):

I think I'd go with T-libs and T-libs + libs-impl

Santiago Pastorino (May 08 2020 at 19:08, on Zulip):

as part of the team compiler process we can just search for libs-impl and solve our problems

Santiago Pastorino (May 08 2020 at 19:09, on Zulip):

but tagging libs things with T-compiler doesn't convince me because our situation may end being temporary :shrug:

Last update: Jun 05 2020 at 23:15UTC