Meeting started at 2024-01-31 16:32:04 UTC

The Meeting name is 'fedora_coreos_meeting'

!hi

!hi

Michael Armijo (marmijo)

Dusty Mabe (dustymabe) - he / him / his

!hi

Yasmin Valim de Souza (ydesouza)

!hi

No Fedora Accounts users have the @jdoss:beeper.com Matrix Account defined

!hi

None (jlebon)

!hi

Timothée Ravier (siosm) - he / him / his

!hello jdoss

Joe Doss (jdoss)

travier: FYI: I added a ticket to the meeting checklist

Hello FCOS folks

!hello spresti

Steven Presti (spresti)

!hi

No Fedora Accounts users have the @fifofonix:matrix.org Matrix Account defined

!hello fifofonix

Fifo Phonics (fifofonix)

.hi

https://discussion.fedoraproject.org/t/fedora-coreos-community-meeting-minutes-2024-01-24/102641

We have actions!

Action items spresti to investigate if there are actions we need to take to pre-test golang 1.22 before it hits rawhide dustymabe to open an issue for further investigating `Enable IPv4 Address Conflict Detection` jlebon to open an issue for F40 communications fifofonix to coordinate with travier a time to collaborate and understand the problems more fifofonix to coordinate with travier a time to collaborate and understand the problems related to fwupd more

Action items * spresti to investigate if there are actions we need to take to pre-test golang 1.22 before it hits rawhide * dustymabe to open an issue for further investigating `Enable IPv4 Address Conflict Detection` * jlebon to open an issue for F40 communications * fifofonix to coordinate with travier a time to collaborate and understand the problems more * fifofonix to coordinate with travier a time to collaborate and understand the problems related to fwupd more

Any updates on thoses?

Any updates on those?

Action items: - spresti to investigate if there are actions we need to take to pre-test golang 1.22 before it hits rawhide - dustymabe to open an issue for further investigating `Enable IPv4 Address Conflict Detection` - jlebon to open an issue for F40 communications - fifofonix to coordinate with travier a time to collaborate and understand the problems more - fifofonix to coordinate with travier a time to collaborate and understand the problems related to fwupd more

.hi

fifofonix: travier does that mean we need to find another volunteer - or does it mean fifofonix just needs more help than anticipated?

I think one of those is more difficult that the other and we likely need more people looking at it

Oh goodness can we edit our info's I added an unintentional '\'

Oh goodness can we edit our info's I added an unintentional '\\'

Not sure, maybe give it a try?

So we need an additional volunteer for the fwupd issue

Let's start with what we have on the agenda

Eh, looks like no, It edited fine but I dont think the bot picked it up

Should we resume where we left off last week?

130. Boost 1.83 upgrade

yeah. one sec - let me update the description real quick

130 - Boost 1.83 upgrade

ok - updated

Boost 1.83 upgrade

130 - Boost 1.83 upgrade

Should not impact us

Should not impact us either

I think this is happening progressively already?

I believe so, we updated a bunch of specs in the past.

we had a ticket last cycle I think for making sure all our packages were compliant

i'd be interested if there is a current list of non-compliant packages to see if any of ours are on it

https://github.com/coreos/fedora-coreos-tracker/issues/1497

love the checkbox list there ✔️

Should not impact us (and we're doing osbuild)

i wish 133 went into more details on what the misalignment with osbuild was

I've heard that kiwi is much easier to work with than osbuild

and understands comps, etc.

but I don't have first hand experience with kiwi so I can not confirm

I gotta bounce early folks 👋

travier: interesting, thanks.

134: there's a bunch of ostree services in there. we should turn those on upstream though

134 is a bit vague. Looks like changes will happen on a per-package basis and not system wide

it really should be "try upstream first, and then resort to packaging level", but i can appreciate that being way more work

I can create a issue in ostree to track this and see which ones we can add. It looks like we would not want to enable the private netrwork etc.

Is there actually a service in ostree?

travier: no long-running ones if that's what you mean

ah, it's the onshot units

ah, it's the oneshot units

I worked in kiwi in the past, to create ova image for suse. It was much easier to work than OsBuild

it wouldn't be hard to test the impact of this by opening a PR and seeing what upstream CI says

Renata Ravanelli: Oh, I'll need your help then for another secret project!

Renata Ravanelli: Oh, I'll need your help then for another secret project! 🍎

It accepts scripts and everything inside the image itself. The part that is not good, is that you need to add package by package

OK, should we open an issue to track hardeding our units?

OK, should we open an issue to track hardening our units?

SGTM!

this should not impact us

well, it will, but there is nothing specific to do

on to the smaller ones

agree

I don't think we need to do anything here

travier: fyi https://github.com/ravanelli/kiwi/tree/master/SLES/SLES12SP4

We don't ship kubernetes in FCOS

i mean... ideally we'd have a story there and collaborate way more with the new maintainers 😢

interesting that it's a change request for rawhide

we don't have GHC/Haskell packages in FCOS AFAIK

interesting that the k8s change is a change request for rawhide

SKIPIT

I think it's just a wording/process error

hum, don't we already have something to track that?

yes

https://github.com/coreos/fedora-coreos-tracker/issues/1629

we knew it was coming but they hadn't submitted the proposal yet

some of those dep chain changes might impact us (i.e. we might need to name packages explicitly

some of those dep chain changes might impact us (i.e. we might need to name packages explicitly)

this one is using kiwi to build the cloud images

https://pagure.io/fedora-kiwi-descriptions/pull-request/9

It's interesting for the UKI work to support that in FCOS but not something we need to act on now

This is a name change only.

OK, I think we're good. dustymabe can I let you update the comment with the updates? (later)

yep.. i've been writing down notes as we've been discussing these - thanks for driving travier

So, this was different before as we "preset-all" but it has been fixed in Fedora to match what we do

Should we do anything? Should we disable/mask systemd-homed given that we likely don't support it?

or just do nothin and move on

or just do nothing and move on

Fresh systemd-homed discussion: https://discussion.fedoraproject.org/t/homed-during-installation/103876

hmm

I don't know if I have a strong opinion. I guess I'd be interested to know if it doesn't work at all for us.

are there any fundamental reasons why homed wouldn't work?

heh :)

if it doesn't then masking it might be the right thing to do. but we'd probably want to doc somewhere on our docs site why

apparently, according to user reports, it does not work well but I haven't tried it

As it's not a priority / main use case for FCOS, I'd say we leave it as is until someone complains

or until someone investigates

agreed. was essentially typing the same thing :)

yeah. i mean if it doesn't work in Fedora then it won't work in FCOS and we shouldn't work to make it work I don't think (at least not a high priority)

Sorry all, I have to drop early :/

OK, let's leave it as is. If someone is interested, they can investigate support

dustymabe: should we talk about the last item? We only have 5 min so maybe next time?

I should have ordered it before the previous one

yeah, SGTM - in the meantime add comments to the ticket if you want

Will close in 5 minutes

I've started working / looking at minikube (for teaching and works well as a learning tools). I'm looking at making an FCOS variant of it. It's currently based on buildroot and while it's minimal for sure, it's super custom to maintain.

I've started working / looking at minikube (for teaching and works well as a learning tool). I'm looking at making an FCOS variant of it. It's currently based on buildroot and while it's minimal for sure, it's super custom to maintain.

Alright, if we have nothing else.