#meeting-1:fedoraproject.org log

<@alexsaezm:fedora.im>

18:00:03

!startmeeting Go SIG meeting

<@meetbot:fedora.im>

18:00:04

Meeting started at 2025-09-08 18:00:03 UTC

<@meetbot:fedora.im>

18:00:04

The Meeting name is 'Go SIG meeting'

<@alexsaezm:fedora.im>

18:00:07

!topic Roll Call

<@alexsaezm:fedora.im>

18:00:13

Hi everyone! As always we'll wait ~5 minutes for everyone to join.

<@mikelolasagasti:matrix.org>

18:01:47

!hi

<@zodbot:fedora.im>

18:01:49

Mikel Olasagasti Uranga (mikelo2)

<@alexsaezm:fedora.im>

18:04:21

<@mikelolasagasti:matrix.org>

18:04:43

<@alexsaezm:fedora.im>

18:05:39

Looks like today is gonna be you and me

<@mikelolasagasti:matrix.org>

18:06:21

I don't have much for today

<@alexsaezm:fedora.im>

18:07:04

Me neither. I was checking the issues, and I only see one updated recently, but it was updated by gotmax23 so maybe we should skip that one.

<@alexsaezm:fedora.im>

18:07:12

!topic Open floor

<@alexsaezm:fedora.im>

18:08:07

For the record, I performed a mass rebuild on COPR using massprebuild with the DWARF5 by default reverted, and everything seems fine (there were a bunch of packages breaking that I need to change, but not a large amount).

<@alexsaezm:fedora.im>

18:08:31

And, I released Go 1.25.1 already, but some gating tests failed in bodhi that I need to check

<@mikelolasagasti:matrix.org>

18:08:53

are package fails related to DWARF5 or it's simply that dependencies might be missing already?

<@alexsaezm:fedora.im>

18:09:44

I need to check them one by one, I didn't do it yet (today was a bank holiday here, and I went AFK the whole day).

<@alexsaezm:fedora.im>

18:10:02

!link https://copr.fedorainfracloud.org/coprs/alexsaezm/golang-nodwarf5/builds/

<@buckaroogeek:fedora.im>

18:10:19

!hi

<@zodbot:fedora.im>

18:10:20

Brad Smith (buckaroogeek)

<@alexsaezm:fedora.im>

18:11:00

I can already say some of them failed because of missing dependencies.

<@gotmax:matrix.org>

18:11:48

lurking...

<@alexsaezm:fedora.im>

18:13:25

<@alexsaezm:fedora.im>

18:13:43

Not much to lurk today 🙂

<@alexsaezm:fedora.im>

18:14:43

btw, around 200 packages failed.

<@mikelolasagasti:matrix.org>

18:14:56

not that bad

<@alexsaezm:fedora.im>

18:15:02

Now that I think about it, I should use the comparison feature of mass prebuild... but that's is going to take forever...

<@mikelolasagasti:matrix.org>

18:15:07

but you rebuilt all packages, not only binary ones, right?

<@alexsaezm:fedora.im>

18:15:10

Now that I think about it, I should use the comparison feature of mass prebuild... but that's going to take forever...

<@alexsaezm:fedora.im>

18:15:17

all yeah

<@alexsaezm:fedora.im>

18:15:30

I don't really know how to do that distinction with the tool.

<@mikelolasagasti:matrix.org>

18:15:46

luckily the number of packages is getting lower

<@alexsaezm:fedora.im>

18:15:58

1,710 in total

<@mikelolasagasti:matrix.org>

18:15:58

I've been porting a few packages to GVT for example

<@mikelolasagasti:matrix.org>

18:16:10

so slowly we're getting more leaves that are orphaned and later retired

<@mikelolasagasti:matrix.org>

18:16:29

for whatever DWARF5 related it may be interesting just to focus on packages that include a binary

<@alexsaezm:fedora.im>

18:17:35

The way the tool works is by pulling all the dependencies of a given package (in our case, go-rpm-macros and golang to cover all of them). I am not sure if I have a workaround to trigger only binaries.

<@alexsaezm:fedora.im>

18:18:06

I mean I can write the list manually...

<@alexsaezm:fedora.im>

18:18:20

Instead of building everything, a given list

<@mikelolasagasti:matrix.org>

18:19:29

there should be one, as it's tracked in package-data

<@mikelolasagasti:matrix.org>

18:19:45

https://gitlab.com/fedora/sigs/go/package-data/-/blob/main/binaries?ref_type=heads

<@alexsaezm:fedora.im>

18:21:05

got it, I'll try a mass prebuild with that. It's way shorter than 1700 packages.

<@alexsaezm:fedora.im>

18:21:30

in fact.. With that length... I might be able to do two mass prebuilds, the one with and the one without the change.

<@alexsaezm:fedora.im>

18:22:07

which is the selling point of the tool imho

<@mikelolasagasti:matrix.org>

18:23:00

it's true that we're reducing packages, but are slowly getting new ones also, but still binary only packages with GVT will make this easier for you also

<@alexsaezm:fedora.im>

18:23:46

a lot easier I think

<@alexsaezm:fedora.im>

18:26:48

Is there anything else to say, or can we call it a day?

<@mikelolasagasti:matrix.org>

18:27:14

did you update go-rpm-macros for EPEL9 & EPEL10?

<@mikelolasagasti:matrix.org>

18:27:46

I built go-vendor-tools for EPEL10.0, so packages can be built for current EPEL "stable"

<@alexsaezm:fedora.im>

18:28:05

No, not yet. I need to do update it this week ASAP for CentOS

<@mikelolasagasti:matrix.org>

18:29:26

and based on Maxwell's feeback, this is the first PR that works integrating packit with Go Vendor Tools: https://src.fedoraproject.org/fork/mikelo2/rpms/google-guest-agent/blob/c6d7e14023aa4115e86aaea3f10596c58e1bc81a/f/.packit.yaml

<@mikelolasagasti:matrix.org>

18:29:43

I think I'll start pushing that for my packages and sending PRs to others

<@alexsaezm:fedora.im>

18:30:54

!link https://issues.redhat.com/browse/RHEL-112845

<@alexsaezm:fedora.im>

18:30:57

!link https://issues.redhat.com/browse/RHEL-104256

<@alexsaezm:fedora.im>

18:31:05

just in case someone wants to follow the updates

<@mikelolasagasti:matrix.org>

18:31:34

!link https://src.fedoraproject.org/fork/mikelo2/rpms/google-guest-agent/blob/c6d7e14023aa4115e86aaea3f10596c58e1bc81a/f/.packit.yaml

<@alexsaezm:fedora.im>

18:31:35

But they are on the plate for this week without exceptions.

<@mikelolasagasti:matrix.org>

18:33:37

nothing else from my side I think

<@buckaroogeek:fedora.im>

18:35:18

nothing from me

<@alexsaezm:fedora.im>

18:36:58

Awesome. Then, we can call it. I'll restart the mass prebuild with the binaries, and I will work on the CentOS updates first thing tomorrow. Many thanks, everyone!

<@alexsaezm:fedora.im>

18:37:12

!endmeeting