#meeting:fedoraproject.org log

<@jistone:fedora.im>

19:30:53

!startmeeting FESCO (2024-01-22) !meetingname fesco Chairs: @conan_kudo:matrix.org, @ngompa:fedora.im, @nirik:matrix.scrye.com, @humaton:fedora.im, @zbyszek:fedora.im, @sgallagh:fedora.im, @jistone:fedora.im, @dcantrell:fedora.im, @mhayden:fedora.im, @tstellar:fedora.im !topic Init Process

<@meetbot:fedora.im>

19:30:54

Meeting started at 2024-01-22 19:30:53 UTC

<@meetbot:fedora.im>

19:30:54

The Meeting name is 'FESCO (2024-01-22) !meetingname fesco Chairs: @conan_kudo:matrix.org, @ngompa:fedora.im, @nirik:matrix.scrye.com, @humaton:fedora.im, @zbyszek:fedora.im, @sgallagh:fedora.im, @jistone:fedora.im, @dcantrell:fedora.im, @mhayden:fedora.im, @tstellar:fedora.im !topic Init Process'

<@salimma:fedora.im>

19:30:59

!hi

<@zodbot:fedora.im>

19:31:00

Michel Lind (salimma) - he / him / his

<@jistone:fedora.im>

19:31:06

!hi

<@zodbot:fedora.im>

19:31:07

Josh Stone (jistone) - he / him / his

<@nirik:matrix.scrye.com>

19:31:37

I am poking at our ipa cluster outage, so ping me if you need me (but hopefully you don't need me)

<@sgallagh:fedora.im>

19:31:44

!hi

<@zodbot:fedora.im>

19:31:45

Stephen Gallagher (sgallagh) - he / him / his

<@mhayden:fedora.im>

19:31:48

.hi

<@dcantrell:fedora.im>

19:31:49

!hi

<@zodbot:fedora.im>

19:31:50

David Cantrell (dcantrell) - he / him / his

<@mhayden:fedora.im>

19:31:55

!hi

<@zodbot:fedora.im>

19:31:56

Major Hayden (mhayden) - he / him / his

<@mhayden:fedora.im>

19:32:03

👋

<@humaton:fedora.im>

19:32:20

!hi

<@zodbot:fedora.im>

19:32:21

Tomáš Hrčka (humaton) - he / him / his

<@tstellar:fedora.im>

19:32:38

!hi

<@zodbot:fedora.im>

19:32:39

Tom Stellard (tstellar)

<@jistone:fedora.im>

19:33:16

procedural note -- Stephen informed me that nonresponsive-maintainers aren't usually discussed in the meeting, but there's a disconnect in policy. If there's no objection, I think we can skip those two topics in the agenda and talk about that in the open floor.

<@zbyszek:fedora.im>

19:34:10

!hi

<@zodbot:fedora.im>

19:34:11

Zbigniew Jędrzejewski-Szmek (zbyszek)

<@sgallagh:fedora.im>

19:34:14

Right. tl;dr: the policy there disagrees with our general ticket policy. In practice, we've been following the FESCo policy for these rather than the stated one.

<@mhayden:fedora.im>

19:34:53

it's my policy to do what Stephen Gallagher says 99% of the time

<@sgallagh:fedora.im>

19:35:14

mhayden: That is a *terrible* policy. Just to be clear.

<@zbyszek:fedora.im>

19:35:20

That's the policy we all follow.

<@conan_kudo:matrix.org>

19:35:55

!hi

<@zodbot:fedora.im>

19:35:57

Neal Gompa (ngompa) - he / him / his

<@sgallagh:fedora.im>

19:36:01

Anyway, I agree with skipping those topics for now

<@sgallagh:fedora.im>

19:36:16

And I guess that means mhayden and zbyszek do too ✌️

<@zbyszek:fedora.im>

19:36:23

Yes.

<@jistone:fedora.im>

19:36:53

I think *everyone* has chimed in, wow! (with nirik on the side)

<@jistone:fedora.im>

19:37:11

!topic #3145 Delay mass rebuild for Fedora 40 until f40-build-side-81394 side-tag is merged. !fesco 3145

<@sgallagh:fedora.im>

19:37:33

So, this happened, even without a finalized vote.

<@sgallagh:fedora.im>

19:37:40

So I say we just close it and move on

<@zbyszek:fedora.im>

19:37:48

Yeah.

<@humaton:fedora.im>

19:37:49

yup

<@jistone:fedora.im>

19:38:01

(wait, do commands have to be separate lines? I might have messed up init too)

<@zbyszek:fedora.im>

19:38:11

(Unless somebody has a time machine to back and fix things.)

<@zbyszek:fedora.im>

19:38:33

!meetingname fesco

<@zbyszek:fedora.im>

19:39:04

Hmm, dunno. It doesn't seem to have any effect.

<@sgallagh:fedora.im>

19:39:05

Separate lines but not necessarily separate "sends"

<@jistone:fedora.im>

19:39:15

!fesco 3145

<@zodbot:fedora.im>

19:39:19

**fesco #3145** (https://pagure.io/fesco/issue/3145):**Delay mass rebuild for Fedora 40 until f40-build-side-81394 side-tag is merged.** ● **Opened:** 6 days ago by codonell ● **Last Updated:** 18 hours ago ● **Assignee:** Not Assigned

<@jistone:fedora.im>

19:39:27

I was looking for that here

<@jistone:fedora.im>

19:39:49

anyway, I think there's still some question about the actual Change, no?

<@sgallagh:fedora.im>

19:40:21

Is there a question or just the binutils bug they found?

<@conan_kudo:matrix.org>

19:40:23

the Change will be performatively accepted

<@conan_kudo:matrix.org>

19:40:30

we can't back it out, realistically

<@jistone:fedora.im>

19:40:40

sure

<@jistone:fedora.im>

19:41:03

so let's do our dance

<@sgallagh:fedora.im>

19:41:11

💃

<@jistone:fedora.im>

19:42:21

do we just need to ping amoloney to open that change ticket?

<@zbyszek:fedora.im>

19:43:49

I think we should do a formal approval vote here so its clear that the change was accepted.

<@sgallagh:fedora.im>

19:44:02

Works for me

<@sgallagh:fedora.im>

19:44:26

Proposal: F40 toolchain updates are approved

<@zbyszek:fedora.im>

19:44:32

<@conan_kudo:matrix.org>

19:44:37

<@jistone:fedora.im>

19:44:42

<@humaton:fedora.im>

19:44:44

<@dcantrell:fedora.im>

19:44:51

<@tstellar:fedora.im>

19:45:03

<@mhayden:fedora.im>

19:45:10

<@mhayden:fedora.im>

19:45:45

we've... compiled... a good vote here

<@jistone:fedora.im>

19:46:08

(Stephen is implicit +1 as proposer?)

<@sgallagh:fedora.im>

19:46:15

Nice attempt at a dad joke, but you really didn't link it to the conversation all that well ;-)

<@sgallagh:fedora.im>

19:46:36

Yes, implicit +1 unless explicitly stated otherwise

<@jistone:fedora.im>

19:46:58

!agreed APPROVED (+8, 0, -0)

<@zbyszek:fedora.im>

19:47:02

!action Aoife Moloney to create the Tracking ticket for Changes/GNUToolchainF40

<@amoloney:fedora.im>

19:47:21

ack

<@zbyszek:fedora.im>

19:48:10

Aoife Moloney: It's my first meeting on matrix with a working bot. I'm testing out the functionality. :---]

<@amoloney:fedora.im>

19:48:30

its good to see it back up! I think it was down earlier for the QA meeting

<@jistone:fedora.im>

19:48:35

!topic #3146 What to do with webkit2gtk-4.0 removal change proposal?

<@jistone:fedora.im>

19:48:38

!fesco 3146

<@zodbot:fedora.im>

19:48:39

**fesco #3146** (https://pagure.io/fesco/issue/3146):**What to do with webkit2gtk-4.0 removal change proposal?** ● **Opened:** 4 days ago by catanzaro ● **Last Updated:** an hour ago ● **Assignee:** Not Assigned

<@sgallagh:fedora.im>

19:49:00

I'm +1 on zbyszek 's proposal in the ticket

<@tstellar:fedora.im>

19:49:31

Same +1

<@jistone:fedora.im>

19:50:03

I don't have background on it, but isn't it a security risk to keep that around?

<@sgallagh:fedora.im>

19:51:08

A maintainer has stepped up (for now). If they don't keep it well-maintained, the non-responsive maintainer policy exists

<@sgallagh:fedora.im>

19:51:23

(My opinion, YMMV, etc.)

<@jistone:fedora.im>

19:51:30

<@zbyszek:fedora.im>

19:51:47

It probably is a security risk, but if people want to keep an old package, it's their responsibility.

<@jistone:fedora.im>

19:52:10

then I'm +1

<@conan_kudo:matrix.org>

19:52:26

meh +1

<@dcantrell:fedora.im>

19:52:49

<@conan_kudo:matrix.org>

19:52:50

people are going to naturally not want to use this because you can't mix libsoup2 and libsoup3 anyway

<@conan_kudo:matrix.org>

19:53:04

so hopefully its usage will remain minimal

<@sgallagh:fedora.im>

19:53:12

mmm... soup

<@humaton:fedora.im>

19:53:55

<@jistone:fedora.im>

19:54:56

!agreed APPROVED (+7, 0, -0)

<@jistone:fedora.im>

19:55:46

!action catanzaro to update the Change text to describe the status quo

<@jistone:fedora.im>

19:56:11

!topic Next week's chair

<@zbyszek:fedora.im>

19:56:39

I can do it.

<@jistone:fedora.im>

19:57:03

!action zbyszek will chair next meeting

<@jistone:fedora.im>

19:57:11

!topic Open Floor

<@zbyszek:fedora.im>

19:57:30

We had the topic of the nrm tickets…

<@sgallagh:fedora.im>

19:57:34

Right

<@sgallagh:fedora.im>

19:57:52

#link https://docs.fedoraproject.org/en-US/fesco/Policy_for_nonresponsive_package_maintainers/

<@sgallagh:fedora.im>

19:58:15

The basic disconnect is with section 2 of Weed 1

<@sgallagh:fedora.im>

19:58:21

The basic disconnect is with section 2 of Week 1

<@nirik:matrix.scrye.com>

19:58:39

so, completely distracted, but did we talk about the mass rebuild restarting again?

<@sgallagh:fedora.im>

19:58:42

In practice, we've just been handling these tickets like other FESCo tickets.

<@sgallagh:fedora.im>

19:58:52

No, we can add that to the Open Floor topics

<@nirik:matrix.scrye.com>

19:58:58

<@jistone:fedora.im>

19:59:27

we talked about the previous delay, and approving the toolchain Change

<@jistone:fedora.im>

20:00:10

but yes, on NRM I added it to the agenda because of the "Otherwise, FESCo will discuss the issue during a meeting."

<@humaton:fedora.im>

20:00:21

I was just going to mention the mass rebuild

<@nirik:matrix.scrye.com>

20:00:28

yeah, but if we restart it tomorrow... thats almost a week late. Do we want to push the schedule back?

<@sgallagh:fedora.im>

20:01:08

nirik: I'm in favor of that, as there isn't much time until Branching and this cycle's branching has outsized impact

<@sgallagh:fedora.im>

20:01:23

I'd like there to be time to fix things up

<@nirik:matrix.scrye.com>

20:01:29

yep

<@humaton:fedora.im>

20:01:32

yes

<@sgallagh:fedora.im>

20:01:57

It's no secret that we're branching CentOS Stream 10 from ELN at the same point

<@conan_kudo:matrix.org>

20:02:16

I've personally avoided touching things as I've seen failed mass builds

<@conan_kudo:matrix.org>

20:02:26

so I would appreciate more time

<@humaton:fedora.im>

20:02:53

How does this actually work?

<@humaton:fedora.im>

20:03:02

should releng create ticket for fesco or something else?

<@sgallagh:fedora.im>

20:03:03

In what sense?

<@humaton:fedora.im>

20:03:27

Who and how can announce we are moving things one week later

<@conan_kudo:matrix.org>

20:04:14

it just usually happens

<@conan_kudo:matrix.org>

20:04:23

Ben did this before, I imagine Aoife Moloney would now

<@humaton:fedora.im>

20:04:34

well that was between RCs

<@conan_kudo:matrix.org>

20:04:37

she will need to adjust the schedule files and publish those

<@sgallagh:fedora.im>

20:04:49

Aoife Moloney would be responsible for updating the schedule, but I think FESCo has the authority to declare a slip

<@conan_kudo:matrix.org>

20:04:55

yes

<@zbyszek:fedora.im>

20:05:02

The policy for NRM tickets differs from normal policy in two regards: it it enough to have just one +1 vote, instead of +3, and the vote takes effect more quickly, i.e. after 3 days instead of 7. This was done intentionally, because we wanted to add the new owner as quickly as possible to the package so that they can fix the package. I'd be in favour of keeping those lowered limits. In practice, very few FESCo members vote on tickets, so having the lower limit is practical. So it'd be mostly a question of finding better wording so that tickets are not added to the meeting agenda.

<@conan_kudo:matrix.org>

20:05:02

and I think we'd pretty much agree on it

<@conan_kudo:matrix.org>

20:05:38

now, the question is, is this a slip that we'd shift the schedule or shorten another part of the schedule to compensate?

<@humaton:fedora.im>

20:06:08

I think shift there is a bunch of things people need to do

<@conan_kudo:matrix.org>

20:06:11

historically the first slip means we cut some time elsewhere

<@humaton:fedora.im>

20:06:22

and as mentioned this is important branching

<@conan_kudo:matrix.org>

20:06:23

but I don't like that approach and we don't have to do that

<@sgallagh:fedora.im>

20:06:27

zbyszek: I think we inadvertently switched topics. Let's come back to that

<@conan_kudo:matrix.org>

20:06:44

jednorozec: I'm intentionally ignoring the "important branching", as it's irrelevant for Fedora

<@conan_kudo:matrix.org>

20:07:09

CentOS Stream 10 doesn't exist yet and won't exist yet for some time, they can shift that or not, that's their own choice

<@humaton:fedora.im>

20:07:44

sure but anyhow the week slip is in schedule between the RC's its not counting in we will start rebuild later

<@conan_kudo:matrix.org>

20:07:50

we could choose to push it back by a month, and I suspect they would still disconnect from Fedora in mid Feb

<@humaton:fedora.im>

20:07:57

there might be some broken stuff between the RC's

<@humaton:fedora.im>

20:08:09

and if we just slip now we will get less time later

<@sgallagh:fedora.im>

20:08:31

Conan Kudo: For technical reasons I don't want to get into, we actually have to disconnect at the Branching event, whenever it happens.

<@conan_kudo:matrix.org>

20:08:45

to be clear, I'm in favor of pushing the whole schedule back by up to 3 weeks

<@jistone:fedora.im>

20:09:36

Stephen Gallagher: that schedule dependency is RH's risk alone

<@conan_kudo:matrix.org>

20:09:38

there are people here who don't know what we've done before, so I'm outlining what we've done in the past

<@sgallagh:fedora.im>

20:09:50

I suppose I should clarify.

<@sgallagh:fedora.im>

20:10:00

We can disconnect *no later than* the Branching event

<@sgallagh:fedora.im>

20:10:11

But we have a strong incentive not to do it earlier either.

<@nirik:matrix.scrye.com>

20:10:51

I don't think we need to push 3 weeks...

<@conan_kudo:matrix.org>

20:11:11

I would actually prefer to push more back than less

<@conan_kudo:matrix.org>

20:11:31

simply because of how much of a disruption failed mass builds are

<@conan_kudo:matrix.org>

20:11:34

and we've had two now

<@tstellar:fedora.im>

20:11:36

Can we push back 1 week and then re-evaluate next week?

<@sgallagh:fedora.im>

20:11:39

Why do you think we need three weeks?

<@sgallagh:fedora.im>

20:11:58

I'm supportive of a one-week slip, but I'm not clear on why you think we need more than that

<@tstellar:fedora.im>

20:12:03

What are the two failures? I know of just one.

<@amoloney:fedora.im>

20:12:14

I left for less than 10 mins and I had 3 @'s :-D Ok so FESCo owns the schedule, I just manage it. Therefore, I can adjust and announce any changes, provided there is a formal vote (maybe?) from FESCo. I certainly need your sign off on the slip and any changes you wish to make

<@amoloney:fedora.im>

20:12:29

Formal vote might not be necessary

<@jistone:fedora.im>

20:12:36

delayed side-tag merge, and now the broken binutils

<@amoloney:fedora.im>

20:12:42

but definitely a blessing

<@nirik:matrix.scrye.com>

20:12:45

well, the question is... push the entire schedule out a week? or just change branching?

<@amoloney:fedora.im>

20:13:00

https://fedorapeople.org/groups/schedule/f-40/f-40-all-tasks.html

<@amoloney:fedora.im>

20:13:03

for reference

<@humaton:fedora.im>

20:13:27

I think this view is enough to orient in this discussion https://fedorapeople.org/groups/schedule/f-40/f-40-key-tasks.html

<@nirik:matrix.scrye.com>

20:13:51

I suppose we could also just think about it and decide next week?

<@zbyszek:fedora.im>

20:13:51

What is the actual issue that prevents the mass rebuild from commencing?

<@sgallagh:fedora.im>

20:14:04

Proposal: Delay branching by one week, the rest of the schedule is unaffected.

<@sgallagh:fedora.im>

20:14:11

(And we can revisit if things go pear-shaped)

<@nirik:matrix.scrye.com>

20:15:03

zbyszek: you mean 're-commencing' ?

<@tstellar:fedora.im>

20:15:15

I think we should push the beta freeze back a week too.

<@nirik:matrix.scrye.com>

20:15:23

the fixed binutils is in now I guess, so we could restart anytime (except that the person running it is asleep right now)

<@jistone:fedora.im>

20:17:23

Tom Stellard: why do you think the freeze needs to be delayed too?

<@humaton:fedora.im>

20:17:44

I can restart it

<@tstellar:fedora.im>

20:18:35

Because then there would be only 4 days between post-branch freeze and beta freeze that's not really enough time to do any thing.

<@tstellar:fedora.im>

20:19:06

An 2 of those days are weekends.

<@humaton:fedora.im>

20:19:08

That was my point when I said we should move the whole schedule 1 week

<@humaton:fedora.im>

20:19:41

mass rebuild is kind of day 0 from where we count weeks for other key milestones

<@tstellar:fedora.im>

20:20:03

jednorozec: Yeah the only place we really have any slack is between beta freeze and the beta release. I think we either need to move the whole schedule or everything up to the beta freeze.

<@humaton:fedora.im>

20:20:21

<@sgallagh:fedora.im>

20:20:36

Up to or including Beta Freeze?

<@tstellar:fedora.im>

20:20:41

Up to and including the beta freeze I mean.

<@sgallagh:fedora.im>

20:20:56

Thanks, that's what I thought. Best to be sure, though

<@jistone:fedora.im>

20:21:57

ok, do we need a vote?

<@sgallagh:fedora.im>

20:22:02

Yes

<@sgallagh:fedora.im>

20:22:12

First, a formal proposal

<@tstellar:fedora.im>

20:23:02

If/when we do push the schedule back we should make sure we enumerate which tasks/changes we are waiting for and make sure the owners of those tasks are aware.

<@jistone:fedora.im>

20:24:05

are we still waiting? I thought we were in the clear

<@tstellar:fedora.im>

20:24:58

I thought we had to re-do at least some of the mass rebuild due to the binutils bug.

<@jistone:fedora.im>

20:25:28

that's the restart

<@jistone:fedora.im>

20:25:51

Proposal: delay the F40 schedule 1 week for everything up to and including Beta Freeze

<@humaton:fedora.im>

20:26:01

<@sgallagh:fedora.im>

20:26:12

<@zbyszek:fedora.im>

20:26:14

Please clarify: the whole beta freeze incl. end?

<@jistone:fedora.im>

20:26:29

not the end IMO

<@sgallagh:fedora.im>

20:26:30

Ah, I was assuming start only

<@sgallagh:fedora.im>

20:26:33

So good to clarify

<@tstellar:fedora.im>

20:26:43

Yeah should be the start of the beta freeze.

<@nirik:matrix.scrye.com>

20:26:54

so I guess we want to decide this right now? ;)

<@zbyszek:fedora.im>

20:27:16

OK. So maybe say "and including the start of the Beta Freeze" explicitly.

<@sgallagh:fedora.im>

20:29:02

nirik: I do, yes. The time until the branch date matters to ELN quite a lot, and I'd rather have the time to adjust our plans than wait and find out it's delayed.

<@jistone:fedora.im>

20:29:47

Proposal: delay the F40 schedule 1 week for everything up to and including the start of the Beta Freeze

<@zbyszek:fedora.im>

20:29:58

<@tstellar:fedora.im>

20:30:20

<@sgallagh:fedora.im>

20:30:29

<@dcantrell:fedora.im>

20:30:32

<@jistone:fedora.im>

20:31:46

!agreed delay F40 1 week through the start of Beta Freeze (+5, 0, -0)

<@jistone:fedora.im>

20:32:03

time's up, shall we consider NRM later?

<@zbyszek:fedora.im>

20:32:10

Yes, please.

<@sgallagh:fedora.im>

20:32:16

Yeah, it's not urgent

<@jistone:fedora.im>

20:32:42

is there a bot command for that?

<@jistone:fedora.im>

20:33:44

!action jistone will file a ticket about NRM voting policy

<@jistone:fedora.im>

20:33:49

!endmeeting

<@jistone:fedora.im>

20:34:27

uh oh, the end is still broken by FAS?

<@sgallagh:fedora.im>

20:35:21

Alas, we are trapped in this meeting for the remainder of our days...

<@amoloney:fedora.im>

20:35:27

I take it I will make the necessary changes to the schedule?

<@sgallagh:fedora.im>

20:35:34

Yes, plese

<@amoloney:fedora.im>

20:35:35

do I need to announce this too or will this come from FESCo?

<@sgallagh:fedora.im>

20:35:41

Yes, please

<@sgallagh:fedora.im>

20:36:15

I think it might be easiest if you send an announcement.

<@sgallagh:fedora.im>

20:36:22

Noting that it was a FESCo decision

<@amoloney:fedora.im>

20:36:35

Certainly, I will link the meeting log in the announcement then too

<@jistone:fedora.im>

20:36:53

(once we can get a meeting log...)

<@amoloney:fedora.im>

20:37:12

👀

<@zbyszek:fedora.im>

20:37:22

!endmeeting