2025-03-26 18:00:57 <@tdawson:fedora.im> !startmeeting EPEL (2025-03-26) 2025-03-26 18:00:59 <@meetbot:fedora.im> Meeting started at 2025-03-26 18:00:57 UTC 2025-03-26 18:00:59 <@meetbot:fedora.im> The Meeting name is 'EPEL (2025-03-26)' 2025-03-26 18:01:06 <@tdawson:fedora.im> !topic aloha 2025-03-26 18:01:06 <@tdawson:fedora.im> !meetingname epel 2025-03-26 18:01:07 <@meetbot:fedora.im> The Meeting Name is now epel 2025-03-26 18:01:10 <@dherrera:fedora.im> !hi 2025-03-26 18:01:12 <@zodbot:fedora.im> Diego Herrera (dherrera) - he / him / his 2025-03-26 18:01:13 <@carlwgeorge:fedora.im> !hi 2025-03-26 18:01:14 <@zodbot:fedora.im> Carl George (carlwgeorge) - he / him / his 2025-03-26 18:01:43 <@tdawson:fedora.im> Hi Diego Herrera and Carl George 2025-03-26 18:02:11 <@nirik:matrix.scrye.com> morning 2025-03-26 18:02:20 <@tdawson:fedora.im> Morning nirik 2025-03-26 18:03:37 <@conan_kudo:matrix.org> !hi 2025-03-26 18:03:39 <@zodbot:fedora.im> Neal Gompa (ngompa) - he / him / his 2025-03-26 18:03:52 <@tdawson:fedora.im> Hi Conan Kudo 2025-03-26 18:03:55 <@salimma:fedora.im> !hi 2025-03-26 18:03:57 <@zodbot:fedora.im> Michel Lind (salimma) - he / him / his 2025-03-26 18:04:08 <@tdawson:fedora.im> Hi Michel Lind UTC-6 2025-03-26 18:04:12 <@salimma:fedora.im> I guess the matrix.org interop issue is fixed / better now 2025-03-26 18:04:30 <@nhanlon:beeper.com> !hi 2025-03-26 18:04:32 <@zodbot:fedora.im> Neil Hanlon (neil) - he / him / his 2025-03-26 18:04:39 <@tdawson:fedora.im> Hi Neil Hanlon 2025-03-26 18:04:42 <@nhanlon:beeper.com> morning / afternoon, everyone 2025-03-26 18:04:46 <@conan_kudo:matrix.org> looks like it 2025-03-26 18:04:48 <@rcallicotte:fedora.im> !hi 2025-03-26 18:04:49 <@zodbot:fedora.im> Robby Callicotte (rcallicotte) - he / him / his 2025-03-26 18:05:02 <@tdawson:fedora.im> Hi Robby Callicotte 2025-03-26 18:05:16 <@tdawson:fedora.im> !link https://pagure.io/epel/issues?tags=meeting&status=Open 2025-03-26 18:05:16 <@tdawson:fedora.im> !topic EPEL Issues https://pagure.io/epel/issues 2025-03-26 18:05:42 <@tdawson:fedora.im> We have two open issues ... I think one will be quick ... or hope it will be quick. 2025-03-26 18:06:01 <@tdawson:fedora.im> !epel 323 2025-03-26 18:06:02 <@zodbot:fedora.im> 2025-03-26 18:06:02 <@zodbot:fedora.im> **epel #323** (https://pagure.io/epel/issue/323):**Add selinux-policy-epel to the exception list of packages that have epel in their name** 2025-03-26 18:06:02 <@zodbot:fedora.im> ● **Last Updated:** 6 minutes ago 2025-03-26 18:06:02 <@zodbot:fedora.im> ● **Opened:** 2 weeks ago by tdawson 2025-03-26 18:06:02 <@zodbot:fedora.im> ● **Assignee:** Not Assigned 2025-03-26 18:06:12 <@conan_kudo:matrix.org> I have something for the meeting in AOB I guess afterward 2025-03-26 18:06:33 <@tdawson:fedora.im> OK 2025-03-26 18:07:00 <@jonathanspw:fedora.im> !hi 2025-03-26 18:07:03 <@zodbot:fedora.im> Jonathan Wright (jonathanspw) 2025-03-26 18:07:03 <@tdawson:fedora.im> In short, selinux-policy-epel is a package that will have selinux-policies, for packages that are only in epel, not RHEL. 2025-03-26 18:07:10 <@tdawson:fedora.im> Hi Jonathan Wright 2025-03-26 18:07:54 <@tdawson:fedora.im> It's not a "these binaries / subpackages are missing" package, and thus needs an exception from us. 2025-03-26 18:08:44 <@carlwgeorge:fedora.im> is anyone else having trouble loading the issue? 2025-03-26 18:09:17 <@tdawson:fedora.im> I guess it sorta is missing binaires, but not in the normal sense, so it would be best to get the exception for it. 2025-03-26 18:09:41 <@tdawson:fedora.im> It loads for me, but I also loaded it before the meeting. 2025-03-26 18:10:31 <@carlwgeorge:fedora.im> weirdly it loads in chromium, but in firefox it's just an hourglass indefinitely. oh well, i'll just assume it's a local issue, carry on. 2025-03-26 18:10:56 <@conan_kudo:matrix.org> (I have the same issue) 2025-03-26 18:11:07 <@rcallicotte:fedora.im> it loads in chrome not firefox for me 2025-03-26 18:11:12 <@salimma:fedora.im> it loads in Firefox for me - I'm on the ESR version though (Debian VM) 2025-03-26 18:11:13 <@tdawson:fedora.im> Are the any questions, or discussions people want to have for this before we vote? 2025-03-26 18:11:33 <@conan_kudo:matrix.org> this package is weird and I think it's asking for trouble, but hey if they want to do this 2025-03-26 18:11:36 <@conan_kudo:matrix.org> let's do it I guess 2025-03-26 18:11:47 <@nirik:matrix.scrye.com> loads fine here. 2025-03-26 18:11:55 <@nirik:matrix.scrye.com> I'm fine with the exception. 2025-03-26 18:11:56 <@salimma:fedora.im> I can't think of a better name either 2025-03-26 18:12:36 <@tdawson:fedora.im> I agree ... on all accounts. 2025-03-26 18:12:40 <@salimma:fedora.im> I think in general we should try and move selinux policies to the individual packages and have less big "omnibus" packages like this, but this is probably not the best place to discuss that 2025-03-26 18:12:56 <@carlwgeorge:fedora.im> as stated in the issue i'm +1 on an exception, as long as it's explicit not implicit 2025-03-26 18:13:16 <@tdawson:fedora.im> Meaning, just this package ? 2025-03-26 18:14:07 <@carlwgeorge:fedora.im> right, originally they just assumed it was fine, but i think it should be a limited explicit list of packages that go outside the documented usage of that suffix 2025-03-26 18:14:09 <@tdawson:fedora.im> Ready for a formal vote then? 2025-03-26 18:14:51 <@conan_kudo:matrix.org> sure 2025-03-26 18:15:10 <@tdawson:fedora.im> All those that want to grant an exception for selinux-policy-epel (and just this package) give a +1 2025-03-26 18:15:18 <@tdawson:fedora.im> +1 2025-03-26 18:15:24 <@jonathanspw:fedora.im> +1 2025-03-26 18:15:25 <@rcallicotte:fedora.im> +1 2025-03-26 18:15:42 <@salimma:fedora.im> +1 2025-03-26 18:15:43 <@conan_kudo:matrix.org> +1 2025-03-26 18:15:58 <@carlwgeorge:fedora.im> +1 2025-03-26 18:16:09 <@nirik:matrix.scrye.com> +1 2025-03-26 18:16:41 <@tdawson:fedora.im> !agreed - selinux-policy-epel has been given an exemption for it's name 7(+1) 0(-1) 0(0) 2025-03-26 18:17:44 <@tdawson:fedora.im> OK, I'll put that in the ticket. Moving on. 2025-03-26 18:18:00 <@tdawson:fedora.im> !epel 324 2025-03-26 18:18:00 <@zodbot:fedora.im> ● **Opened:** a week ago by carlwgeorge 2025-03-26 18:18:00 <@zodbot:fedora.im> **epel #324** (https://pagure.io/epel/issue/324):**EPEL 10 minor version upgrade path** 2025-03-26 18:18:00 <@zodbot:fedora.im> 2025-03-26 18:18:00 <@zodbot:fedora.im> ● **Last Updated:** 6 days ago 2025-03-26 18:18:00 <@zodbot:fedora.im> ● **Assignee:** carlwgeorge 2025-03-26 18:18:20 <@tdawson:fedora.im> Carl George: Did you want to take this one 2025-03-26 18:18:34 <@carlwgeorge:fedora.im> i've been stewing on this one, and do have some details to share that i've learned that will shape the solution 2025-03-26 18:19:14 <@carlwgeorge:fedora.im> first is that the first suggestion i made of epel10m won't work because or mirrormanager logic, the version has to be at the end with nothing following 2025-03-26 18:20:08 <@carlwgeorge:fedora.im> another is that the mirrormanager scanner is additive, so any changes we make to those mappings just change new repo detection, and existing repos stay in place in the database, which will help with any potential name changes 2025-03-26 18:21:21 <@carlwgeorge:fedora.im> i don't have the whole plan pinned down yet, but the thing that seems obvious to me is that we need to start publishing epel 10.1 to pub/epel/10.1, not pub/epel/10 like we're doing now. we'll want to free up that path for a repo redirect and a symlink. 2025-03-26 18:21:39 <@nirik:matrix.scrye.com> Might help to have a written out timeline of what points where and when it changes/example? it's all kinda complex to wrap your brain around... 2025-03-26 18:23:15 <@carlwgeorge:fedora.im> it also occurred to me that with the scanner we currently have a mapping regex that matches both major and major.minor paths, we can change that to use one mapping for major version only (epel9 and earlier) and a different one for major.minor (epel10.x going forward) 2025-03-26 18:25:12 <@carlwgeorge:fedora.im> which would expand to epel-10 on centos and epel-minor-10 on rhel and derivatives 2025-03-26 18:25:12 <@carlwgeorge:fedora.im> the last specific detail is that given the available dnf variables and logic, the best way to handle this is have the mapping set a prefix that includes a string we can conditionalize on releasever_minor. that string is firmly in bikeshed territory, but could look something like: 2025-03-26 18:25:12 <@carlwgeorge:fedora.im> 2025-03-26 18:25:12 <@carlwgeorge:fedora.im> `epel${releasever_minor:+-minor}-$releasever` 2025-03-26 18:25:12 <@carlwgeorge:fedora.im> 2025-03-26 18:25:55 <@carlwgeorge:fedora.im> we could set up the mapping prefix as epel-minor, and then use repo redirects for the primary repo names 2025-03-26 18:26:05 <@tdawson:fedora.im> That's interesting. 2025-03-26 18:26:22 <@carlwgeorge:fedora.im> epel-minor-10 → epel-minor-10.0 2025-03-26 18:26:22 <@carlwgeorge:fedora.im> epel-10 → epel-minor-10.1 2025-03-26 18:26:22 <@carlwgeorge:fedora.im> ``` 2025-03-26 18:26:22 <@carlwgeorge:fedora.im> ``` 2025-03-26 18:26:48 <@carlwgeorge:fedora.im> on the left are the repo redirects, and on the right are the actual repo names that get set by the scanner mapping 2025-03-26 18:27:22 <@tdawson:fedora.im> And just to double check, when you say 'minor', you reall do mean the word, correct? 2025-03-26 18:27:23 <@carlwgeorge:fedora.im> said mapping for reference https://pagure.io/fedora-infra/ansible/blob/main/f/roles/openshift-apps/mirrormanager/templates/scan-primary-mirror.toml#_27-29 2025-03-26 18:27:35 <@smooge:fedora.im> that name confuses my brain 2025-03-26 18:27:50 <@carlwgeorge:fedora.im> in this example, yes the literal word, but that is the bikeshedable string i mentioned 2025-03-26 18:27:51 <@smooge:fedora.im> but naming is hard 2025-03-26 18:27:52 <@tdawson:fedora.im> It's must better than 'm' ... which hurt my brain worse. :) 2025-03-26 18:28:40 <@tdawson:fedora.im> Yep, naming is hard. But I'm following the logic at this point. 2025-03-26 18:28:48 <@conan_kudo:matrix.org> worksforme, I just have one detail to ask for 2025-03-26 18:29:02 <@conan_kudo:matrix.org> can we have this also reflected in https://kojipkgs.fedoraproject.org/compose/updates/ somehow? 2025-03-26 18:29:18 <@smooge:fedora.im> BIKESHED epel-rhel-current-release-10 -> epel-rcr-10.0 2025-03-26 18:29:32 <@conan_kudo:matrix.org> when I mirror/download composes, I'd like to be able to suitably automatically sync things properly the same way we push for mirrors 2025-03-26 18:29:48 <@conan_kudo:matrix.org> iirc, symlinks are okay here, so I'd just like some kind of maintained symlink 2025-03-26 18:29:49 <@carlwgeorge:fedora.im> i did originally want to use the string rhel, but troy and bstinson don't like it 😁 2025-03-26 18:29:52 <@smooge:fedora.im> if for no other reason than causng all kinds of naming sins 2025-03-26 18:30:47 <@smooge:fedora.im> Troy Dawson: what was the old naming convention? Prominent North American Enterprise Linux Vendor (PNAELV)? 2025-03-26 18:31:00 <@conan_kudo:matrix.org> lol 2025-03-26 18:31:06 <@conan_kudo:matrix.org> that's an oldie 2025-03-26 18:31:12 <@smooge:fedora.im> what scientific linux and other distros were using during the cease and desist days 2025-03-26 18:31:21 <@tdawson:fedora.im> There's a name I haven't heard for a while ... yes, I believe that was it. 2025-03-26 18:31:41 <@carlwgeorge:fedora.im> we have something similar for the koji local repo, we can work something out and have it as part of the sop for minor versions 2025-03-26 18:31:48 <@smooge:fedora.im> there epel-pnaelv-release 2025-03-26 18:31:51 <@conan_kudo:matrix.org> thanks 2025-03-26 18:32:07 <@nirik:matrix.scrye.com> This is sure looking like a lot of trouble. ;) 2025-03-26 18:32:25 <@conan_kudo:matrix.org> great, I need it for being able to easily pull centos-stream + epel composes in sync automatically, so some stuff for this would be awesome 2025-03-26 18:33:00 <@carlwgeorge:fedora.im> the alternative is we leave things in place as we designed it, and train people to do `dnf -y update redhat-release && dnf -y update` 2025-03-26 18:33:11 <@smooge:fedora.im> epel-cs-release , epel-pnaelv-release 2025-03-26 18:33:15 <@nirik:matrix.scrye.com> yeah, or dnf update twide 2025-03-26 18:33:17 <@nirik:matrix.scrye.com> yeah, or dnf update twice 2025-03-26 18:33:28 <@smooge:fedora.im> epel-everyoneelse-release 2025-03-26 18:33:31 <@jonathanspw:fedora.im> I'm a hard -1 on this. It will be trouble that will follow us for 5 years straight and is poor UX. 2025-03-26 18:33:47 <@nirik:matrix.scrye.com> yeah, I am sure it will bite people... sadly 2025-03-26 18:34:15 <@jonathanspw:fedora.im> It will also break dnf-auto or whatever it's called now. Agree or disagree that's a core feature of yum/dnf and we shouldn't break it. 2025-03-26 18:34:16 <@carlwgeorge:fedora.im> if we resort to separate release packages, we can do all kinds of stuff, but i would definitely like to have a single epel-release package that has the logic in metalink/baseurls to pick the right repos 2025-03-26 18:34:29 <@conan_kudo:matrix.org> it is actually too bad dnf can't pre-update variables based on redhat-release changing in the transaction 2025-03-26 18:34:43 <@conan_kudo:matrix.org> it's not like the data isn't known from rpmmd repodata 2025-03-26 18:34:51 <@nirik:matrix.scrye.com> well, it shouldn't break it, it just means you will upgrade minor on one run and all of the updated epel stuff on the next 2025-03-26 18:34:52 <@jonathanspw:fedora.im> Wouldn't that be neat 2025-03-26 18:34:53 <@conan_kudo:matrix.org> but that's more engineering than I think we have time for 2025-03-26 18:35:00 <@conan_kudo:matrix.org> perhaps for the future 2025-03-26 18:35:07 <@carlwgeorge:fedora.im> where? 2025-03-26 18:35:11 <@tdawson:fedora.im> Question: When we are at epel10.4, and someone is running an EUS RHEL 10.2 ... what will they be pointed to in this plan? 2025-03-26 18:35:20 <@conan_kudo:matrix.org> provides are expressed in primary.xml 2025-03-26 18:35:27 <@conan_kudo:matrix.org> so it can be pre-fetched 2025-03-26 18:35:27 <@smooge:fedora.im> ``` 2025-03-26 18:35:27 <@smooge:fedora.im> epel-pnaelv-10 → epel-10-minor-0 2025-03-26 18:35:27 <@smooge:fedora.im> ``` 2025-03-26 18:35:27 <@smooge:fedora.im> sorry I misread an earlier part 2025-03-26 18:35:27 <@smooge:fedora.im> epel-10 → epel-10-minor-1 2025-03-26 18:35:39 <@carlwgeorge:fedora.im> with eus they get releasever set to say 10.2 and are pinned to the right repo 2025-03-26 18:36:40 <@tdawson:fedora.im> OK ... so with this plan, it's just for the mirror managers, but under it all, we'll have the 10.0, 10.1, 10.2 etc ... correct? 2025-03-26 18:36:49 <@carlwgeorge:fedora.im> yup 2025-03-26 18:36:59 <@tdawson:fedora.im> OK, sounds good to me. 2025-03-26 18:37:07 <@carlwgeorge:fedora.im> we'd also have symlinks on the mirrors for baseurls, which mirrormanager skips 2025-03-26 18:37:13 <@tdawson:fedora.im> Or at least 'try-able' 2025-03-26 18:37:50 <@conan_kudo:matrix.org> worksforme 2025-03-26 18:38:03 <@conan_kudo:matrix.org> as long as we address having some kind of stable URL thing for composes too, I'm good 2025-03-26 18:38:09 <@carlwgeorge:fedora.im> yeah this sounds like a complex undertaking, and if that's where we eventually go i hope it happens early enough that we're not doing an exception to get it into rhel 11 2025-03-26 18:38:29 <@conan_kudo:matrix.org> yeah it's something for el11 2025-03-26 18:38:43 <@conan_kudo:matrix.org> it would be broadly useful for a lot of things, including simpifying how EUS enablement works 2025-03-26 18:38:55 <@nhanlon:beeper.com> ++ 2025-03-26 18:39:19 <@nhanlon:beeper.com> I think what's being done for 10 is an awesome step forward, though. Kudos to Carl for driving this 2025-03-26 18:39:29 <@nhanlon:beeper.com> (and others of course!) 2025-03-26 18:39:40 <@nhanlon:beeper.com> I'd give you a cookie but I think I already did, so... 2025-03-26 18:39:47 <@nirik:matrix.scrye.com> for el11 we might be able to use the action plugin stuff. ;) 2025-03-26 18:39:52 <@tdawson:fedora.im> Yes, thank you very much Carl ... and any others that have helped with this. But definatly thank you Carl. 2025-03-26 18:39:55 <@carlwgeorge:fedora.im> for the bikeshed string using pnaelv could be a funny easter egg 2025-03-26 18:40:05 <@nhanlon:beeper.com> lmao 2025-03-26 18:40:09 <@nhanlon:beeper.com> i'd die, please do it 2025-03-26 18:40:10 <@smooge:fedora.im> april fools is coming up 2025-03-26 18:40:19 <@conan_kudo:matrix.org> I'm all for that :P 2025-03-26 18:40:27 <@carlwgeorge:fedora.im> someone talk me out of this 2025-03-26 18:40:28 <@nhanlon:beeper.com> (not really die, stop celebrating..) 2025-03-26 18:40:36 <@nhanlon:beeper.com> you've come to the wrong place Carl lol 2025-03-26 18:40:47 <@conan_kudo:matrix.org> look man, we need to get our jollies somewhere 2025-03-26 18:41:19 <@carlwgeorge:fedora.im> well that's enough for this meeting i think, nothing final yet, will keep on planning in that issue 2025-03-26 18:41:34 <@tdawson:fedora.im> OK, thank you. Let's move on. 2025-03-26 18:42:03 <@tdawson:fedora.im> Conan Kudo: You said you had something to bring up ... would that be on open floor or old business? 2025-03-26 18:42:17 <@conan_kudo:matrix.org> it's a ticket in the epel tracker 2025-03-26 18:42:23 <@conan_kudo:matrix.org> two rather 2025-03-26 18:42:30 <@conan_kudo:matrix.org> so I guess old business? 2025-03-26 18:43:00 <@tdawson:fedora.im> Good, cuz that's next. 2025-03-26 18:43:02 <@carlwgeorge:fedora.im> fyi, you should have perms on that repo to add the meeting label, that helps me remember to have things on the agenda 2025-03-26 18:43:08 <@tdawson:fedora.im> !topic Old Business 2025-03-26 18:43:20 <@conan_kudo:matrix.org> oops 😅 2025-03-26 18:43:22 <@tdawson:fedora.im> Go for it Conan Kudo 2025-03-26 18:43:36 <@conan_kudo:matrix.org> okay, the first thing is the pair of fish update tickets 2025-03-26 18:43:44 <@conan_kudo:matrix.org> !epel 319 2025-03-26 18:43:46 <@zodbot:fedora.im> ● **Last Updated:** 5 days ago 2025-03-26 18:43:46 <@zodbot:fedora.im> **epel #319** (https://pagure.io/epel/issue/319):**Proposing incompatible update of fish to version 4.x for EPEL 10 series** 2025-03-26 18:43:46 <@zodbot:fedora.im> 2025-03-26 18:43:46 <@zodbot:fedora.im> ● **Opened:** 3 weeks ago by ngompa 2025-03-26 18:43:46 <@zodbot:fedora.im> ● **Assignee:** Not Assigned 2025-03-26 18:43:49 <@conan_kudo:matrix.org> !epel 320 2025-03-26 18:43:50 <@zodbot:fedora.im> 2025-03-26 18:43:50 <@zodbot:fedora.im> **epel #320** (https://pagure.io/epel/issue/320):**Proposing (semi-)incompatible update of fish to version 3.7.x for EPEL 9** 2025-03-26 18:43:50 <@zodbot:fedora.im> ● **Opened:** 3 weeks ago by salimma 2025-03-26 18:43:50 <@zodbot:fedora.im> ● **Assignee:** Not Assigned 2025-03-26 18:43:50 <@zodbot:fedora.im> ● **Last Updated:** 3 weeks ago 2025-03-26 18:44:18 <@conan_kudo:matrix.org> I've heard no negative feedback for either of these requests in nearly a month 2025-03-26 18:44:36 <@conan_kudo:matrix.org> and Michel Lind UTC-6 and I would like to process these 2025-03-26 18:44:37 <@carlwgeorge:fedora.im> has anyone attempted a backport? 2025-03-26 18:44:52 <@conan_kudo:matrix.org> I did one for fish4 in COPR already 2025-03-26 18:45:03 <@conan_kudo:matrix.org> the fish3 security update is relatively trivial 2025-03-26 18:45:25 <@carlwgeorge:fedora.im> if it were a high severity and the backport had been established as very complex, it would be a slam dunk exception. lacking those, i think it could go either way. 2025-03-26 18:47:00 <@tdawson:fedora.im> Sorry, I just realized I was reading through both of those. 2025-03-26 18:47:02 <@smooge:fedora.im> ah I think this answers my question.. I need to do a major update to vsomeip3 in EPEL-9 so it needs to be a ticket in the system versus just an email to the list 2025-03-26 18:47:20 <@carlwgeorge:fedora.im> technically both 2025-03-26 18:47:55 <@tdawson:fedora.im> I agree that for epel10, if you are going to do the update, do it now, before RHEL 10.0 GA. 2025-03-26 18:48:21 <@smooge:fedora.im> well I have it only in EPEL-10.1 currently because well 2025-03-26 18:48:35 <@smooge:fedora.im> I built it before realizing that it needed a different thing 2025-03-26 18:48:41 <@carlwgeorge:fedora.im> to avoid confusing the two, can we focus on the epel9 one first (the linked issue) and then handle the epel10 one next? 2025-03-26 18:49:21 <@smooge:fedora.im> well currently you are focusing on fish and I am wandering into the weeds 2025-03-26 18:49:45 <@conan_kudo:matrix.org> how very fishy 2025-03-26 18:49:56 <@tdawson:fedora.im> Sounds like a good idea. epel9 - issue #320 2025-03-26 18:50:01 <@carlwgeorge:fedora.im> oh so three things, fish in epel9, fish in epel10, and vsomeip3 in epel9 2025-03-26 18:50:20 <@carlwgeorge:fedora.im> into the...seaweeds? 2025-03-26 18:51:12 <@nhanlon:beeper.com> security related: has anyone heard about the atop stuff? 2025-03-26 18:51:32 <@nhanlon:beeper.com> the versions in epel i suspect are too old?... i should go to fedora-security 😐 (oh hi Michel 😉 ) 2025-03-26 18:51:34 <@carlwgeorge:fedora.im> looks like https://github.com/fish-shell/fish-shell/commit/09986f5563e31e2c900a606438f1d60d008f3a14 is the commit fixing the cve in question for fish in epel9 2025-03-26 18:51:35 <@conan_kudo:matrix.org> we have only 9 minutes left, so I'd really like a resolution on the fish stuff 2025-03-26 18:52:03 <@nhanlon:beeper.com> sorry for derailing. I'm +1 on the fish ones, fwiw 2025-03-26 18:52:47 <@carlwgeorge:fedora.im> my answer is contingent on the backportability. if the fix can be backported i'm -1, if it can then i'm +1. not ok voting without the possibility of a backport being explorted. 2025-03-26 18:53:09 <@carlwgeorge:fedora.im> the commits seems small, but that doesn't guarantee it will apply cleanly 2025-03-26 18:53:24 <@carlwgeorge:fedora.im> my answer is contingent on the backportability. if the fix can be backported i'm -1, if it can't then i'm +1. not ok voting without the possibility of a backport being explorted. 2025-03-26 18:55:32 <@tdawson:fedora.im> Sorry, got dropped for a bit there. I'm not a fish person, so when you say there are slight changes, how bad is that? 2025-03-26 18:56:02 <@nhanlon:beeper.com> IMO it's worth bumping to the 3.7.1 now given the rest of the lifetime of epel9 2025-03-26 18:56:04 <@conan_kudo:matrix.org> generally not user-noticeable 2025-03-26 18:56:10 <@salimma:fedora.im> let's do that next? 2025-03-26 18:56:26 <@nhanlon:beeper.com> it will make future backports easier, if needed (probably?) 2025-03-26 18:56:47 <@conan_kudo:matrix.org> fish is overly verbose about their incompatibility guidance, but in practice it doesn't break with users or scripts much, if at all 2025-03-26 18:57:38 <@dherrera:fedora.im> will there even be future backports or maintainiance on the c++ side considering they moved to rust? 2025-03-26 18:57:38 <@salimma:fedora.im> yeah, and the reverse is normally true - if you run an older fish and try to use a newer version of say starship little things break 2025-03-26 18:57:59 <@conan_kudo:matrix.org> having been a fish user for 10+ years across fish2->fish3->fish4, upgrades have been fine, it's holding back that becomes a problem over time 2025-03-26 18:58:05 <@carlwgeorge:fedora.im> without ruling out a future incompat update for a worse cve, just doing the backport for now is how i'm leaning 2025-03-26 18:58:17 <@salimma:fedora.im> upstream does not support older 3.x releases except 3.7 right? 2025-03-26 18:58:39 <@conan_kudo:matrix.org> that's correct 2025-03-26 18:58:39 <@jonathanspw:fedora.im> I have something minor but we're running out of time 2025-03-26 18:59:20 <@tdawson:fedora.im> Jonathan Wright: Ya, it will probrubly need to wait till next week, or on the regular epel channel. 2025-03-26 18:59:21 <@carlwgeorge:fedora.im> i also had a thing, will defer it to the main channel after the meeting 2025-03-26 18:59:22 <@salimma:fedora.im> e.g. with the epel9 build it fails because $(...) is not supported - they added it later as a compatibility option but starship uses it unconditionally 2025-03-26 18:59:29 <@jonathanspw:fedora.im> 2025-03-26 18:59:29 <@jonathanspw:fedora.im> I'm not sure who/where we need to report this. 2025-03-26 18:59:29 <@jonathanspw:fedora.im> 2025-03-26 18:59:29 <@jonathanspw:fedora.im> ``` 2025-03-26 18:59:29 <@jonathanspw:fedora.im> https://pagure.io/releng/fedora-scm-requests/issue/73849 2025-03-26 18:59:29 <@jonathanspw:fedora.im> https://pagure.io/releng/fedora-scm-requests/issue/73848 2025-03-26 18:59:29 <@jonathanspw:fedora.im> https://pagure.io/releng/fedora-scm-requests/issue/73847 2025-03-26 18:59:29 <@jonathanspw:fedora.im> jonathan@jon-office:~/fedora-scm/python-syrupy$ fedpkg request-branch epel10.0 2025-03-26 18:59:29 <@jonathanspw:fedora.im> ``` 2025-03-26 18:59:29 <@jonathanspw:fedora.im> 2025-03-26 18:59:29 <@jonathanspw:fedora.im> For example: 2025-03-26 18:59:29 <@jonathanspw:fedora.im> If I request epel10.0 it open the proper ticket for the 10.0 branch, but it also opens 2 more tickets related to modules. 2025-03-26 18:59:29 <@jonathanspw:fedora.im> I've noticed today when requesting epel10 branches with fedpkg everything works fine. 2025-03-26 18:59:42 <@jonathanspw:fedora.im> I'll c/p that to regular epel channel. 2025-03-26 18:59:46 <@smooge:fedora.im> report it in the main channel please 2025-03-26 19:00:07 <@nirik:matrix.scrye.com> It's already fixed in fedpkg, just not released yet. 2025-03-26 19:00:16 <@jonathanspw:fedora.im> awesome! 2025-03-26 19:00:27 <@conan_kudo:matrix.org> so uhh, I think I have the votes for fish4 in the ticket, I would just like to know what I'm supposed to do 2025-03-26 19:01:45 <@tdawson:fedora.im> fish4 is a +1 for me, I'm still on the fense for fish3 2025-03-26 19:04:10 <@tdawson:fedora.im> Our time is up. Let's mark fish3 in epel9 with a meeting tag for next week, and if people could put comments in the issue for discussion before then, that would be good. 2025-03-26 19:05:27 <@tdawson:fedora.im> fish4 in epel10, we didn't discuss it too much, but it's already got a majority of steering committee meembers giving it a +1 in the ticket. 2025-03-26 19:07:02 <@tdawson:fedora.im> I'm not feeling comfortable saying it passed, because we didn't have a formal vote ... is one more week ok? 2025-03-26 19:07:59 <@tdawson:fedora.im> anyway, we are way over time. It's been a good discussion, I feel bad leaving the last discussion in this state, but I'm going to. 2025-03-26 19:08:03 <@conan_kudo:matrix.org> I guess? I expect to become very busy once fedora final freeze kicks in next week 2025-03-26 19:08:38 <@carlwgeorge:fedora.im> we can keep discussing in the main channel, and do the final vote in the ticket whenever everyone is comfortable 2025-03-26 19:09:05 <@tdawson:fedora.im> That sounds like a good idea. Thank you ... I'm good with that. 2025-03-26 19:09:26 <@tdawson:fedora.im> Thank you all for coming, and thank you for all you do for EPEL and it's community. 2025-03-26 19:09:29 <@tdawson:fedora.im> Talk to you next week. 2025-03-26 19:09:43 <@tdawson:fedora.im> !endmeeting