<@james:fedora.im>
16:00:44
!startmeeting fpc
<@meetbot:fedora.im>
16:00:45
Meeting started at 2026-04-16 16:00:44 UTC
<@meetbot:fedora.im>
16:00:45
The Meeting name is 'fpc'
<@james:fedora.im>
16:00:47
!topic Roll Call
<@salimma:fedora.im>
16:00:50
!hi
<@zodbot:fedora.im>
16:00:53
Michel Lind ☘ UTC+1: Michel Lind (salimma) - he / him / his
<@decathorpe:fedora.im>
16:01:18
!hi
<@zodbot:fedora.im>
16:01:19
Fabio Valentini: Fabio Valentini (decathorpe) - he / him / his
<@decathorpe:fedora.im>
16:01:25
gah meeting No. 3 for today :(
<@james:fedora.im>
16:01:27
!hi
<@zodbot:fedora.im>
16:01:30
James: James Antill (james) - he / him / his
<@james:fedora.im>
16:01:57
Only number 2 for me ... you're winning!
<@tibbs:fedora.im>
16:02:28
Hello.
<@carlwgeorge:fedora.im>
16:03:26
!hi
<@zodbot:fedora.im>
16:03:27
Carl George: Carl George (carlwgeorge) - he / him / his
<@james:fedora.im>
16:04:17
Is getmax on matrix?
<@james:fedora.im>
16:05:06
gotmax23: Hey
<@decathorpe:fedora.im>
16:05:07
gotmax23 ?
<@james:fedora.im>
16:05:21
Yeh, brain melting apparently.
<@gotmax:matrix.org>
16:07:19
!hi
<@zodbot:fedora.im>
16:07:21
gotmax23: Maxwell G (gotmax23) - he / him / his or they / them / theirs
<@james:fedora.im>
16:08:07
No problem ... is there anything you wanted to talk about?
<@james:fedora.im>
16:09:32
The only real ticket we have to talk about is the Forge move one ... and I assume it's going to take up a bunch of time once we start, so now's the time to bring up anything else.
<@gotmax:matrix.org>
16:09:46
I had https://pagure.io/packaging-committee/issue/1531
<@gotmax:matrix.org>
16:10:15
I also did discuss the Forge orgs with Ryan a bit so happy to be here for that discussion also
<@james:fedora.im>
16:10:23
Yeh, I figured something might have prompted that though ... like something you are looking to change/help-with?
<@decathorpe:fedora.im>
16:10:30
I marked https://pagure.io/packaging-committee/issue/1512 as meeting a while ago, it would be good to discuss ..
<@tibbs:fedora.im>
16:10:39
For the forge, we should just actually pick something and move forward.
<@tibbs:fedora.im>
16:11:01
For 1531, we should definitely consider trimming non-active members at this point.
<@tibbs:fedora.im>
16:11:23
I have strong doubts we'll see Xavier again and will probably not see Igor, either.
<@james:fedora.im>
16:11:40
Yeh, igor turned up once last year, I think.
<@james:fedora.im>
16:11:50
xavier less than that.
<@conan_kudo:matrix.org>
16:11:50
!hi
<@zodbot:fedora.im>
16:11:53
Conan Kudo 🤢: Neal Gompa (ngompa) - he / him / his
<@tibbs:fedora.im>
16:12:06
That would get us back to 9 members, which is the historical number.
<@james:fedora.im>
16:12:15
Yeh
<@tibbs:fedora.im>
16:12:19
Assuming I'm actually looking at the correct membership list.
<@tibbs:fedora.im>
16:12:32
And also assuming we add gotmax23
<@gotmax:matrix.org>
16:12:45
I had been thinking of joining for a while and then I found myself participating in the Forge discussions and figured I ought to ask to join officially
<@gotmax:matrix.org>
16:12:56
I had been thinking of joining for a while and then I found myself participating in the Forge migration discussion and figured I ought to ask to join officially
<@tibbs:fedora.im>
16:13:36
I see no reason not to have someone who wants to do things and who obviously understands what we do.
<@james:fedora.im>
16:14:06
Yeh, the one I consult the most: https://fedoraproject.org/wiki/Packaging_Committee_Meeting_Process#On_the_day_of_the_meeting ... doesn't have xavier, as he's never turned up on matrix IIRC.
<@tibbs:fedora.im>
16:15:24
I'm looking at who is currently an admin of https://pagure.io/packaging-committee/
<@gotmax:matrix.org>
16:16:11
There's also https://fedoraproject.org/wiki/Packaging_Committee#Members which seems to match the admins there
<@james:fedora.im>
16:16:24
Do we want to approve this one "quickly"?
<@carlwgeorge:fedora.im>
16:16:39
i'm +1 to replacing igor with max
<@james:fedora.im>
16:16:50
I'm fairly happy to just shrug and let very experienced packagers do stuff like that.
<@conan_kudo:matrix.org>
16:17:21
sure +1
<@james:fedora.im>
16:17:29
Yeh, I'm +1 adding max and removing xavier and igor.
<@tibbs:fedora.im>
16:17:51
+1
<@salimma:fedora.im>
16:17:53
which one are we voting on? let's do membersihp first then go to the ruby thing
<@salimma:fedora.im>
16:17:59
+1 on the membership changes
<@decathorpe:fedora.im>
16:17:59
both +1!
<@tibbs:fedora.im>
16:18:16
+1 (to the membership thing)
<@decathorpe:fedora.im>
16:18:27
though for the rubygem cucumber thing we might need to wordsmith something.
<@james:fedora.im>
16:18:28
two votes at once, for twice the fun ;)
<@salimma:fedora.im>
16:18:29
on the ruby issue I'll note that for python we don't strictly package from pypi, right? often the pypi upload is missing tests and maintainers would switch to the github archive
<@james:fedora.im>
16:19:21
AIUI there's some kind of compatibility attempts/promises ... so we don't intentionally do things which break pypi usage.
<@conan_kudo:matrix.org>
16:19:46
yeah we don't
<@conan_kudo:matrix.org>
16:20:03
there was a couple of attempts at it, but they went badly
<@conan_kudo:matrix.org>
16:20:40
I also, as a general rule, don't _actually_ want to force getting sources from a particular place for an ecosystem
<@decathorpe:fedora.im>
16:20:46
Ruby guidelines are currently more strict, they say sources MUST be from rubygems. we'd need to relax that requirement to something more like what's currently in the Python guidelines, I think, to address this ticket.
<@conan_kudo:matrix.org>
16:20:57
yup
<@salimma:fedora.im>
16:21:17
yeah
<@decathorpe:fedora.im>
16:21:20
as long as the built package is ~equivalent (for relevant definition of equivalent), I think this would be OK
<@salimma:fedora.im>
16:21:26
I'd rather relax the requirement in general than grant exceptions
<@gotmax:matrix.org>
16:22:09
Well, the Python guidelines provide no recommendation at all
<@james:fedora.im>
16:22:20
Yeh, I'm at least +1 for https://pagure.io/packaging-committee/issue/1512#comment-1000092
<@gotmax:matrix.org>
16:22:32
Is there a reason that in all/most/some cases rubygem archives are preferred?
<@james:fedora.im>
16:22:44
And probably fine for relaxing, unless some Ruby SIG person shows up and starts shouting no.
<@conan_kudo:matrix.org>
16:22:45
that is on purpose
<@gotmax:matrix.org>
16:22:53
Bc for Python some times the PyPI sdists are missing tests or have other issues
<@conan_kudo:matrix.org>
16:23:33
I vaguely recall that gem data doesn't always exist anywhere except in gem archives
<@carlwgeorge:fedora.im>
16:23:35
+1 to switching that ruby guideline to SHOULD. it sounds like rust has good reasons to keep theirs as MUST, so i don't think we need to align this approach across the language package managers.
<@conan_kudo:matrix.org>
16:23:41
but that is probably not true these days
<@decathorpe:fedora.im>
16:23:55
that sounds good to me
<@conan_kudo:matrix.org>
16:24:01
I'm not sure if it should be a MUST for rust either... but today is not the day I want to argue that
<@carlwgeorge:fedora.im>
16:25:14
no opinion on that, i just mean i'm fine letting it be decided per-stack versus making a blanket thing for everybody
<@decathorpe:fedora.im>
16:25:19
we *can* talk about that at some point - though for now, all uses I've seen in packages that circumvent the MUST rule are ... bad
<@conan_kudo:matrix.org>
16:26:24
yeah, the reason the rule is there is to avoid really bad things
<@conan_kudo:matrix.org>
16:26:49
but the main thing I'm thinking of is having one source package for a collection of crates
<@conan_kudo:matrix.org>
16:27:01
interdependent crates are messy
<@salimma:fedora.im>
16:27:13
yeah let's modify it for Ruby, not in general. but not a per-package opt out either
<@conan_kudo:matrix.org>
16:27:27
and when they are from one git repo on github... I'd rather do that
<@salimma:fedora.im>
16:27:31
well we do allow packaging a workspace all at once
<@salimma:fedora.im>
16:27:33
(for Rust)
<@conan_kudo:matrix.org>
16:27:34
but again, some other time
<@decathorpe:fedora.im>
16:27:34
(for that I'm waiting for the cargo RFC which would add nested crates as first-class feature, rather than botch it in packaging ...)
<@salimma:fedora.im>
16:27:39
it's just you can't do rust-* crates that way
<@conan_kudo:matrix.org>
16:27:51
yeah, fish is like this
<@salimma:fedora.im>
16:27:53
oh nice
<@gotmax:matrix.org>
16:27:55
Yeah the current experience of packaging workspaces as a bunch of separate packages is a bit tedious
<@decathorpe:fedora.im>
16:28:17
tedious, yes, but for now still the best choice lacking better supported alternatives IMO
<@decathorpe:fedora.im>
16:28:27
anyway, off-topic for Ruby
<@tibbs:fedora.im>
16:31:11
I'm a big fan of "whatever makes the most sense" and in this case, relaxing the ruby guideline seems to make sense. I don't know if it makes sense to have a specific-case exception or to relax it for all ruby things.
<@tibbs:fedora.im>
16:31:30
I don't know a whole lot about the ruby ecosystem.
<@james:fedora.im>
16:31:58
I'm fine just doing s/MUST/SHOULD/ in ruby, if people who know more think that's fine. Still +1 for the exceptions in the ticket, if not.
<@gotmax:matrix.org>
16:32:19
Can we +1 this specific case to unblock them and ask them to them submit changed to the guidelines?
<@gotmax:matrix.org>
16:32:44
Whether that should be SHOULD instead of MUST or MUST except in certain cases
<@gotmax:matrix.org>
16:32:55
Can we +1 this specific case to unblock them and ask them to them propose changes to the guidelines?
<@decathorpe:fedora.im>
16:33:17
that sounds good to me. unblock work and move actual guidelines changes to tomorrow's problems :D
<@james:fedora.im>
16:33:40
That's fair, my only concern is they'll just go away once their problem is solved.
<@decathorpe:fedora.im>
16:34:14
I doubt that this will be the case with music being involved :)
<@james:fedora.im>
16:34:20
Everyone good with this?
<@gotmax:matrix.org>
16:34:49
I'm also fine with saying we're generally okay with this but please file a PR first
<@gotmax:matrix.org>
16:35:22
But this is a four month old ticket so it would be nice to have some decision by now
<@salimma:fedora.im>
16:35:30
yeah I think asking for a PR first is probalby ok, but we can take over wordsmithing if necessary
<@salimma:fedora.im>
16:35:40
the delay is not ideal, definitely
<@gotmax:matrix.org>
16:36:13
<@gotmax:matrix.org>
16:36:13
> One think I'd like to note that if this approach does not raise any concerns, I'd like to follow with similar approach with other packages from Cucumber family.
<@gotmax:matrix.org>
16:36:13
I see
<@conan_kudo:matrix.org>
16:36:27
so I think we need to just do the general change
<@gotmax:matrix.org>
16:36:31
So it sounds like this isn't only about the one rubygem-cucumber-messages package that's the subject of the ticket
<@james:fedora.im>
16:38:58
Yeh ... Okay, I'm going to comment that we approved the things in the ticket, which includes comment about the other cucumber packages max and I linked to.
<@james:fedora.im>
16:39:14
Then they can open a PR for something more general, if they want.
<@james:fedora.im>
16:40:34
Okay, now the fun one for the last 20m.
<@salimma:fedora.im>
16:40:39
sure, let's do that. so unblock on this one package, but ask them to submit a guideline proposal for the rest
<@gotmax:matrix.org>
16:40:43
On to Forge migration?
<@james:fedora.im>
16:40:51
!topic FPC#1527 - Move to Forge https://pagure.io/packaging-committee/issue/1527
<@decathorpe:fedora.im>
16:42:26
sorry, my brains are already leaking out, no energy for bikeshedding here. I'm going to go and get dinner
<@salimma:fedora.im>
16:42:53
I can't even get dinner for 20 mins and I skipped lunch because errands, so I'm still here :P
<@decathorpe:fedora.im>
16:43:21
anything that let's us set up permission structure as needed is fine with me
<@james:fedora.im>
16:43:26
I _really_ don't care what colour this bikeshed is ... so is someone other than Fabio Valentini prepared to do the migration?
<@carlwgeorge:fedora.im>
16:43:36
ryan seems to be under the impression that having one org will be too messy, but didn't explain why
<@decathorpe:fedora.im>
16:43:57
it also sounds as if individual people *can* be added to repos individually
<@gotmax:matrix.org>
16:44:06
I talked to Ryan about this yesterday and there is a proposal that involves a separate packager-tools org where any packager can create repositories (as long as they are in scope)
<@decathorpe:fedora.im>
16:44:06
so we might not need complicated thing after all
<@james:fedora.im>
16:44:19
I do like simple things.
<@gotmax:matrix.org>
16:44:21
I'm wondering if having one packaging repo would be better anyways, though
<@gotmax:matrix.org>
16:44:34
Any packager can create repos for tools there and manage their permissions
<@carlwgeorge:fedora.im>
16:44:40
i still think a single packaging org is a good idea
<@tibbs:fedora.im>
16:44:41
Can we just make "packaging", put "packaging/guidelines" in it, and then worry about the rest of the stuff as it comes?
<@gotmax:matrix.org>
16:44:51
There's an FPC team and a packaging/committee repo for the guidelines
<@salimma:fedora.im>
16:45:14
yeah let's start with that first?
<@gotmax:matrix.org>
16:45:17
There's an FPC team and a packaging/committee (or packaging/guuidelines) repo for the guidelines
<@james:fedora.im>
16:45:18
Yeh, I think so ... In fact I was sure that was the leading option when we last spoke about it on matrix.
<@tibbs:fedora.im>
16:45:20
If it turns out that we just can't manage the permissions properly then the tools can live elsewhere.
<@decathorpe:fedora.im>
16:45:25
yeah. and assuming "proceed with my original suggestion and add people to additional repositories on a case-by-case basis" actually works that would be fine with me.
<@gotmax:matrix.org>
16:46:13
Does the packaging committee want to be admins for the entire packaging org?
<@gotmax:matrix.org>
16:46:42
The other thing Ryan and I were talking about is no matter how it's organized, each org needs *someone* (or some group) to be the org admin
<@salimma:fedora.im>
16:48:59
when everything is quiet I never know if it's a Matrix issue
<@decathorpe:fedora.im>
16:49:16
ha
<@gotmax:matrix.org>
16:49:39
I support a single packaging org
<@gotmax:matrix.org>
16:49:55
The Forge people just need to know what teams to create so they can actually set it up
<@james:fedora.im>
16:50:07
Yeh, it looks like everyone is shrugging and agrees with Fabio Valentini ... so let's just let him do what he proposed, and press the migrate buttons before the next meeting?
<@james:fedora.im>
16:50:16
Yeh, it looks like everyone is shrugging and/or agrees with Fabio Valentini ... so let's just let him do what he proposed, and press the migrate buttons before the next meeting?
<@decathorpe:fedora.im>
16:50:18
wooo more work for me
<@gotmax:matrix.org>
16:50:38
I don't mind dealing with the migration
<@decathorpe:fedora.im>
16:51:25
you were right
<@conan_kudo:matrix.org>
16:51:31
single packaging org plz :)
<@gotmax:matrix.org>
16:51:31
The first step would probably be to create an FPC FAS group because I don't think that exists at all right now?
<@decathorpe:fedora.im>
16:51:45
gotmax23: your messages starting with "On to Forge migration?" are just arriving here now
<@decathorpe:fedora.im>
16:51:59
oh man
<@gotmax:matrix.org>
16:52:07
Uh oh
<@decathorpe:fedora.im>
16:52:08
I hate out of order message delivery
<@gotmax:matrix.org>
16:52:23
That message was 10 minutes ago
<@james:fedora.im>
16:52:43
I assume all of max's messages are in order, just they all arrived at once 10m late :(
<@salimma:fedora.im>
16:52:57
this is why I only use my fedora.im account - and I never encourage anyone to move to Matrix anymore :P
<@james:fedora.im>
16:54:00
Other servers aren't perfect, but matrix esp. seems much more prone to delays.
<@tibbs:fedora.im>
16:55:12
So, uh, IRC was nice.
<@conan_kudo:matrix.org>
16:55:26
I was at least not responding because I was trying to figure out lunch
<@conan_kudo:matrix.org>
16:55:36
Max's messages were not late for me :)
<@james:fedora.im>
16:55:55
Anyway ... did you want gotmax to do this, or you do it? Maybe give the job to him while we can't see his replies quickly? ;)
<@decathorpe:fedora.im>
16:56:15
I'm fine with both
<@gotmax23:fedora.im>
16:56:22
I can deal with the migration
<@decathorpe:fedora.im>
16:56:46
whatever route we take, ending up at packaging/guidelines would be nice though.
<@gotmax23:fedora.im>
16:57:35
We need a FAS FPC group first
<@gotmax23:fedora.im>
16:57:49
And then a decision on who wants to admin the packaging org
<@james:fedora.im>
16:58:00
Okay ... I think we are agreed that gotmax23 will paint the shed in the next couple of weeks, and we mostly know what colour he'll do it in?
<@carlwgeorge:fedora.im>
16:58:16
i suspect ryan will want the adming group to be `forge-packaging-owners`, and we can just use that as the fpc group
<@james:fedora.im>
16:58:17
AIUI once we have the FAS group then that can be the admin?
<@gotmax23:fedora.im>
16:58:37
Yes, that will work
<@carlwgeorge:fedora.im>
16:58:51
the other migrations i've seen they prefer to create groups in that format, not use existing groups
<@gotmax23:fedora.im>
16:59:47
All the forge-* groups inherit from other groups
<@gotmax23:fedora.im>
17:00:00
So we'd need to create an FPC group I think
<@gotmax23:fedora.im>
17:00:30
Can we call it `packaging-committee` in FAS and make James the sponsor?
<@gotmax23:fedora.im>
17:00:42
And then the forge-packaging-owners can inherit from that
<@james:fedora.im>
17:00:52
There's already: https://accounts.fedoraproject.org/group/packaging-committee/
<@gotmax23:fedora.im>
17:00:57
Derp
<@tibbs:fedora.im>
17:00:59
It seems as if this should be so simple and yet....
<@carlwgeorge:fedora.im>
17:01:01
i don't think we have group inheritance in fas
<@james:fedora.im>
17:01:26
And both tibbs and I are the sponsors.
<@gotmax23:fedora.im>
17:01:39
Okay, ignore me. I did not realize that already existed...
<@decathorpe:fedora.im>
17:02:03
not in FAS, but in ipsilon
<@decathorpe:fedora.im>
17:02:14
so behind the scenes, groups can contain other groups.
<@gotmax23:fedora.im>
17:02:37
Can someone add me to that and then I'll request the packaging org with that group as the admin?
<@tibbs:fedora.im>
17:02:49
I am also happy to work on this if others are out of bandwidth.
<@gotmax23:fedora.im>
17:03:15
And I guess request a separate team that allows packager members to create new repositories in the org but gives no other permissions
<@james:fedora.im>
17:04:48
Done ... and added Conan Kudo 🤢 too ... and removing very old members.
<@james:fedora.im>
17:06:00
Okay, I'm going to end now ... as we're 5m over.
<@gotmax23:fedora.im>
17:06:17
Sounds good. Thanks for chairing!
<@james:fedora.im>
17:06:44
!endmeeting