Meeting started at 2026-05-19 16:01:11 UTC

The Meeting name is 'ELN SIG 19 May '26'

The Meeting Name is now eln

!hi

Stephen Gallagher: Stephen Gallagher (sgallagh) - he / him / his

!hi

Troy Dawson: Troy Dawson (tdawson)

!hi

!hi

Davide Cavalca: Davide Cavalca (dcavalca) - he / him / his

jligon: Jeff Ligon (jligon)

ok let's get started

[jligon](https://matrix.to/#/@jligon:matrix.org) any updates?

Yes!

Tenants mr was merged. And now working on compose Mr

After compose Mr is done will finish base-images Mr

I’m now the maintainer for the Fedora-eln bootc image

As the Mrs merge I’ll update the ticket

can we get some co-maintainers so that we're not blocked entirely on you once this all lands?

Sure! Any volunteers?

let me get back to you on that

but in the meantime, if you could update the ticket so that we have some documentation as to where all the necessary changes are, that would help

If you want to walk me through it in detail, I can try to help.

Happy to do both

great, thank you for the update, and for all the work on this

looking forward to having images sometime soon

[Davide Cavalca](https://matrix.to/#/@davide:cavalca.name) you're a comaintainer, there's a patch for ppc64le, do you plan on merging that, coming up with some other fix, or leaving it disabled?

lemme ask upstream

full disclosure this week is a mess at work, so it might take me a bit, but we'll get this sorted out either way

that's fine, it's not urgent, just didn't want to let it fall through the cracks

any updates here?

guess not

this is a headsup that we're planning on proceeding with the z15 baseline bump for s390x in ELN soon, while the POWER10 bump is going to take more time as our infrastructure isn't ready for it yet

but since I don't think any of you are hiding a mainframe in your basement...

this will give us a headstart on finding any compile issues with the newer baseline code generation, so there's a (hopefully slim?) possibility of build failures as a result

ack

and I do plan on starting an ELN mass rebuild real soon (hopefully after that lands if we can get it quickly) in order to preempt issues that won't otherwise materialize until the F45 mass rebuild in July

Landing the baseline generation shouldn't be too hard; unless I'm mistaken, it's just a one-line change in the macros

isn't it gcc's spec file that sets that?

yselkowitz: Please ping me before doing the mass rebuild, it might affect things I am working on.

ack

[Stephen Gallagher](https://matrix.to/#/@sgallagh:fedora.im) [nirik](https://matrix.to/#/@nirik:matrix.scrye.com) I keep pinging on the ticket but I'm not sure how to get this over the finish line

Last time we did this, it was a patch to redhat-rpm-config, I think

oh yeah, I go pulled into other things.

oh yeah, I got pulled into other things.

ok looks like we need both

so, we need to archive all the commits into an archive branch?

https://src.fedoraproject.org/rpms/redhat-rpm-config/blob/rawhide/f/macros#_167

[nirik](https://matrix.to/#/@nirik:matrix.scrye.com) go ahead

this process is per branch, right?

!hi

Michel Lind ☘ UTC+1: Michel Lind (salimma) - he / him / his

sorry, things are chaotic one day before layoffs

So, I am not sure. I guess it is.

but we have no git filter there, so have to do it all remotely

and I am not fully sure how to archive all the existing commits... would have to look into how thats done

[Stephen Gallagher](https://matrix.to/#/@sgallagh:fedora.im) you did this for c10s, it is a per branch action, right?

we can't just archive the old ref, we need all the commits, because the hashes will change right?

I'm not sure, to be entirely honest. I was only *importing* a single branch

Right, the effect of the command I gave you is that it essentially amends the ancient commit and then reapplies all the patches that come after

but you were importing it into an existing repo, right? did you filter-repo before or after merging from rawhide?

Changing all the hashes

I did it locally on my laptop and then pushed the branch to CS GitLab

yeah. I can try and figure it out... just hard to find dedicated time for it with all the irons I have in the fires

or if someone else wants to prepare them, I suppose I could just put them in place (I already saved a copy of all of them)

[Stephen Gallagher](https://matrix.to/#/@sgallagh:fedora.im) can you help figure out the mechanics? you are the closest to this based on the c10s work.

I'm not sure how much this will confuse pagure. Hopefully it will just re-read

Yeah, I'll see what I can do. It's largely paged out of my memory after three years though 😅

basically we need to archive the current rawhide branch, and then rewrite it with filter-repo

we don't care about existing stable branches

As I noted in the ticket, this is also going to be necessary before we can import these packages to Forgejo, whenever we start that converstion

As I noted in the ticket, this is also going to be necessary before we can import these packages to Forgejo, whenever we start that conversion

and for pagure, the final output needs to be a bare repo. Or... I guess changes force pushed to the repo?

The bare repo might be more practical, actually

Otherwise we need to force-push every branch that ever existed on the remote

hold on a second though, for ELN we only care about rawhide

IIRC, the `filter-repo` command does update all refs/branches, but we'd still need to PUSH them all

I'm not sure how easy/possible it is to disable the hook in src that prevents force push

oh it's NOT a branch-specific operation?

I don't think so, but it's been a while. I'll take the afternoon to refamiliarize myself with it

`filter-repo` != `filter-branch` FWIW

Fixing this JUST for RHEL 11 isn't hard; we can do what we did last time. I'm trying to preempt the issues with the inevitable forgejo migration too

well, fixing this in rawhide takes care of 11, 12, etc.

And make it possible for people to have forks of dist-git in other forges if they wanted to

sure, makes sense. I'm all for fixing it at the src.

haha

moving along

any conflicts next week?

No conflicts for me. I should be here.

let's wrap it up then, thank you all for your contributions