20:00:09 <tdawson> #startmeeting EPEL (2022-09-14)
20:00:09 <zodbot> Meeting started Wed Sep 14 20:00:09 2022 UTC.
20:00:09 <zodbot> This meeting is logged and archived in a public location.
20:00:09 <zodbot> The chair is tdawson. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
20:00:09 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
20:00:09 <zodbot> The meeting name has been set to 'epel_(2022-09-14)'
20:00:10 <tdawson> #meetingname epel
20:00:10 <zodbot> The meeting name has been set to 'epel'
20:00:12 <tdawson> #chair nirik tdawson pgreco carlwgeorge salimma dcavalca dherrera gotmax[m]
20:00:12 <zodbot> Current chairs: carlwgeorge dcavalca dherrera gotmax[m] nirik pgreco salimma tdawson
20:00:13 <tdawson> #topic aloha
20:00:17 <rcallicotte> .hi
20:00:18 <zodbot> rcallicotte: rcallicotte 'Robby Callicotte' <rcallicotte@mailbox.org>
20:00:23 <tdawson> Hi rcallicotte
20:00:28 <gotmax[m]> .hello gotmax23
20:00:29 <zodbot> gotmax[m]: gotmax23 'Maxwell G' <gotmax@e.email>
20:00:36 <tdawson> Hi gotmax[m]
20:00:39 <nirik> morning
20:00:48 <neil> o/ hiya
20:00:49 <tdawson> Good morning nirik
20:00:50 <aekoroglu> good evening
20:01:04 <tdawson> evening aekoroglu
20:01:10 * gotmax[m] welcomes aekoroglu
20:01:14 <tdawson> Hi neil
20:01:20 * rcallicotte waves
20:01:30 <tdawson> neil: Are you the neil commonly known as many names?
20:01:56 <tdawson> many different names I mean
20:01:56 <neil> i am the one from rocky linux :)
20:02:04 <neil> idk what y'all call me ;)
20:02:15 <tdawson> Ahh ... then you are different.
20:02:18 <neil> (Not Eighth_Doctor, if that's what you mean)
20:02:26 <tdawson> Yep, that's what I meant
20:02:33 <rcallicotte> Thats what I was thinking too :)
20:02:33 <tdawson> neil: Well, good to see you.
20:02:38 <carlwgeorge> .hi
20:02:39 <zodbot> carlwgeorge: carlwgeorge 'Carl George' <carl@redhat.com>
20:02:55 <rsc> .hello robert
20:02:56 <zodbot> rsc: robert 'Robert Scheck' <redhat@linuxnetz.de>
20:03:04 <tdawson> neil: Is there a specific topic you wanted to talk about?  Or just here in general?
20:03:08 <tdawson> Hi carlwgeorge
20:03:14 <tdawson> Hi rsc
20:03:31 <Eighth_Doctor> .hello ngompa
20:03:31 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
20:04:01 <tdawson> Hi Eighth_Doctor
20:04:11 <neil> tdawson, mostly just hanging out, want to see how things go as I've been wanting to dip my toes into helping out epel for a while
20:04:21 <tdawson> Looks like ya'll spell your name different
20:04:55 <tdawson> neil: Cool ... and welcome.
20:05:18 <Eighth_Doctor> Indeed
20:05:33 <neil> one of us is right, we'll never know which :)
20:05:50 <jonathanspw> .hi
20:05:51 <zodbot> jonathanspw: jonathanspw 'Jonathan Wright' <jonathan@almalinux.org>
20:05:54 <Eighth_Doctor> Mine 😁
20:05:58 <rcallicotte> lol
20:06:00 <neil> tdawson, thanks :)
20:06:03 <aekoroglu> :)
20:06:22 <neil> Eighth_Doctor: tomato tomato
20:06:31 <nirik> nice turnout today. Are we discussing something big? ;)
20:06:52 <rsc> Isn't EPEL big enough? :)
20:07:00 <tdawson> aekoroglu: Was there something you wanted to talk about today?  Or are you just here in general as well?
20:07:40 <tdawson> Well, I'd better get things started.
20:07:47 <tdawson> #topic End Of Life (EOL)
20:07:48 <tdawson> RHEL 7 will go EOL on 2024-06-30
20:07:50 <tdawson> CentOS Stream 8 goes EOL in 2024-05-31
20:07:51 <tdawson> CentOS Stream 9 goes EOL in 2027-05-31
20:08:00 <tdawson> And ... with that out of the way :)
20:08:08 <tdawson> #topic EPEL Issues https://pagure.io/epel/issues
20:08:10 <tdawson> https://pagure.io/epel/issues?tags=meeting&status=Open
20:09:02 <gotmax[m]> Should we tackle mine?
20:09:05 <tdawson> We have 4 issues marked as meeting.  Unless the modularity one needs something specific this week (#198) I'm going to skip it for a few weeks while we get input.
20:09:06 <gotmax[m]> :)
20:09:29 <tdawson> gotmax[m]: Actually, yep, I was thinking we could start with that.
20:09:34 <tdawson> .epel 200
20:09:35 <zodbot> tdawson: Issue #200: Investigate whether EPEL packagers can get access to the BR repo - epel - Pagure.io - https://pagure.io/epel/issue/200
20:10:04 <tdawson> gotmax[m]: Did you want to do a quick summary?
20:10:09 <gotmax[m]> I guess
20:10:30 <gotmax[m]> Currently, the EPEL koji buildroot repository is unusable
20:10:45 <nirik> unusable remotely you mean?
20:10:50 <gotmax[m]> Yes
20:10:55 <nirik> it works nicely in the buildsystem. ;)
20:10:55 <gotmax[m]> I should've clarified :)
20:11:29 <gotmax[m]> So it's impossible to do local mockbuilds against the koji buildroot
20:11:33 <gotmax[m]> There are several reasons this is a problem:
20:11:42 <gotmax[m]> - You can't access buildroot overrides
20:12:00 <gotmax[m]> - You can't access the modules that are merged into the buildroot repository
20:12:18 <gotmax[m]> "There is no way to replicate this setup locally, and it's a lot easier to just run --enablerepo=local instead of mucking around with mock configs and hoping you enable the right streams."
20:12:34 <gotmax[m]> - You can't do local builds against side tags either
20:12:53 <gotmax[m]> - It's confusing for users
20:12:59 <gotmax[m]> s/users/packagers/
20:13:07 <rcallicotte> so you are left to just kicking off scratch builds?
20:13:19 <gotmax[m]> If you want the buildroot content, yes
20:13:21 <nirik> So, I've not asked anyone, but I very much suspect this is going to be a very firm no. It would basically mean blessing Fedora Infrastructure as a RHEL distributor.
20:13:40 <aekoroglu> tdawson: theoretically I work for Intel but practically for Fedora and I'm here to help by all means possible :)
20:13:53 <carlwgeorge> this is one of those "if we could have we would have already" type situations i suspect
20:14:13 <gotmax[m]> I suggested gating it to packagers or just members of the epel-packager-sig
20:14:19 <gotmax[m]> I'm not asking for Fedora infra to run a publicly available RHEL mirror
20:14:19 <Eighth_Doctor> Another way to fix some of the issues is to expose a tag repo for packagers where overrides live
20:14:20 <tdawson> Outside of the modules issue, how many times does it matter if you are building off real RHEL vs Alma or Rocky?
20:14:35 <jonathanspw> Rarely in my experience
20:14:36 <tdawson> aekoroglu: Cool ... welcome.
20:15:04 <Eighth_Doctor> And side tags have repos, we just don't have a good way to expose them
20:15:05 <gotmax[m]> tdawson: The issue isn't whether you're building against real RHEL. It's that the buildroot content is merged with the private RHEL mirror.
20:15:16 <nirik> giving free rhel to all packagers/sig members is still distributing it. ;(
20:15:49 <gotmax[m]> Conan Kudo: That still wouldn't solve the modules problem, but I guess it's an improvement
20:16:04 <tdawson> I wish smooge were here, cuz I think he is the last one who asked Red Hat for these permissions, and I was/am curious what the full answer was.
20:16:14 <rcallicotte> so this is more of a legal/policy issue than a technical one, eh?
20:16:16 <gotmax[m]> nirik: Yeah, I guess that's true.
20:16:30 <tdawson> rcallicotte: correct
20:16:34 <nirik> side-tag repos are exposed in the same way the main build repo is... but it runs into the same problem.. the packages aren't
20:16:46 <neil> fwiw my experience matches jonathanspw's with respect to the difference being rarely important
20:16:53 <nirik> rcallicotte: it's completely legal
20:17:24 <gotmax[m]> Didn't someone ask at one point for there to be a way for EPEL packagers to get free access to RHEL repositories by using there FAS account?
20:17:35 <gotmax[m]> I think they said no, but this is kind of equivalent to that
20:17:46 <gotmax[m]> *their FAS
20:17:49 <nirik> yes, and that process was for them to request it, and I filled out a form and they got a rhel entitlement
20:18:07 <nirik> now that gone and people are supposed to use the developer subscription
20:18:40 <gotmax[m]> Yeah, I know. It's just not the most friendly/simple process.
20:18:43 <tdawson> I guess it can't hurt if I ask again.  Red Hat is more friendly to EPEL, and I don't think we've asked since that time.
20:19:03 <tdawson> "that time" = Since they changed their stance towards EPEL.
20:19:06 <nirik> I'm not even sure who to ask... I think this is just a non starter.
20:19:10 <carlwgeorge> i can't imagine a different answer than "dev sub"
20:19:25 <nirik> Could we take the repodata and somehow have a way to transform it to work with another repo?
20:19:25 <Eighth_Doctor> Yeah, I don't think this would be workable
20:19:43 <gotmax[m]> Well, the grobisplitter stuff makes that murky
20:19:46 <tdawson> Josh is who I would start with, he usually knows who to point me to.
20:20:21 <smooge> ... wonders why no one has started a meeting... realizes he joined #epel-meeting
20:20:27 <nirik> ha
20:20:29 <Eighth_Doctor> Why not just do grobisplitter on a RHEL clone?
20:20:39 <nirik> well, ask if you like, I think it's not got any chance. ;)
20:20:50 <smooge> what did I miss?
20:20:51 <nirik> someone could! (not it!)
20:21:04 <gotmax[m]> We're talking about
20:21:06 <gotmax[m]> .epel 200
20:21:08 <zodbot> gotmax[m]: Issue #200: Investigate whether EPEL packagers can get access to the BR repo - epel - Pagure.io - https://pagure.io/epel/issue/200
20:21:37 <tdawson> smooge: We're planning on asking if we (at the very least packagers or epel-packging-sig) could get access to the  epel buildroots
20:21:47 <smooge> hahahahhahahahahahahahahahhaahahahahhaahahahhahahaha
20:21:50 <smooge> <breath>
20:21:53 <tdawson> smooge: Were you the last person to ask ... :)
20:21:54 <Eighth_Doctor> We could also give up on epel8
20:21:54 <smooge> hahahahahahahahahahaha
20:21:56 <gotmax[m]> <tdawson> > I wish smooge were here, cuz I think he is the last one who asked Red Hat for these permissions, and I was/am curious what the full answer was.
20:22:23 <carlwgeorge> Eighth_Doctor: even using the fastest clone would mean a multi day delay for packagers to start building
20:22:50 <tdawson> I believe that answers my question.
20:23:00 <rcallicotte> oh boy
20:23:18 <nirik> I might be able to expose the overrides repos (just by adding a bogus target for them)... would that be of us?
20:23:20 <nirik> use?
20:23:31 <gotmax[m]> Eighth_Doctor: That might be worth exploring
20:23:54 <nirik> although many folks are using sidetags at this point... overrides are kind of old hat.
20:24:03 <gotmax[m]> All of the scripts are in fedora-infra/ansible so someone would have to do the work
20:24:24 <gotmax[m]> nirik: I think that's better than nothing
20:24:29 <jonathanspw> nirik: overrides still have their place.  it's much faster than side tags for many tasks at least in my workflows
20:25:01 <nirik> huh, it should be just as fast...
20:25:35 <gotmax[m]> The docs and fedpkg say that you have to run `koji wait-repo` before building in a side tag
20:25:36 <jonathanspw> I always have far fewer delays with overrides it seems but I've actually mostly used side tags lately.  Maybe it's a weird placebo effect.
20:25:55 <jonathanspw> You have to wait-repo and then build the first package in it, so 2 waits instead of 1.
20:26:27 <gotmax[m]> I like side-tags since you can use `fedpkg chain-build`
20:26:40 <nirik> I guess thats it making the initial repo. anyhow, we are sidetracking
20:26:48 <smooge> so first off we are in the weeds
20:26:50 <gotmax[m]> Yes, we are
20:26:52 <smooge> what nirik said
20:26:53 <tdawson> So, I'm going to ask (despite smooges encouragement) ... give me a couple weeks.  But let's assume the answer is going to be "no" and how about we put some of these ideas into the issue.
20:27:35 <tdawson> And with that ... how about we move onto the next issue.
20:28:16 <tdawson> .epel 199
20:28:17 <zodbot> tdawson: Issue #199: ensure EPEL bugzilla assignees point to valid packagers - epel - Pagure.io - https://pagure.io/epel/issue/199
20:28:55 <gotmax[m]> This is somewhat related to the devel@ discussion about non-packagers having permissions on distgit repos
20:28:55 <tdawson> Oh ... I don't know if davida is here ...
20:29:09 <tdawson> I think it is.
20:29:44 <Eighth_Doctor> He's in Ireland for Linux Plumbers right now
20:29:51 <tdawson> Are we ok holding off on this one until dcavalca is here to talk about it?
20:30:33 <nirik> sure
20:30:34 <dherrera> sounds reasonable (hi)
20:30:44 <gotmax[m]> Perhaps we should work with the rest of Fedora on the two related issues.
20:31:09 <gotmax[m]> I don't think it needs to/should be an EPEL specific policy
20:31:13 <nirik> I already cleaned up all the people who were commit or higher on packages without being in packager.
20:31:29 <gotmax[m]> Thanks for that!
20:31:46 <tdawson> Cool.
20:31:55 <nirik> this is related, but different... peope who are assigned on bugs but don't have commit or higher
20:32:09 <nirik> yeah, it could be fedora/epel both
20:32:13 <tdawson> smooge: We were going to skip the modularity issue for a couple weeks, unless you had any input you wanted to say.
20:32:30 <smooge> nope
20:32:35 <rcallicotte> more hahaha? oh ok
20:32:40 <rcallicotte> :)
20:32:44 <tdawson> :)
20:32:48 <smooge> people have been responding on the thread and ticket as asked.. others not responding at all
20:33:36 <tdawson> I'm sorta dreading getting to the retirement issue ... cuz that also deals with permissions, and each time we go there, it ends up being a circlular talk.
20:33:40 <smooge> move along.. nothing to see here
20:33:49 <gotmax[m]> Me too
20:34:21 <gotmax[m]> I think we need to define clearer goals
20:34:30 * nirik nods
20:34:32 <gotmax[m]> What are we trying to do, what do we want to acomplish
20:34:41 <gotmax[m]> before discussing implementation
20:34:46 <smooge> what? clearer goals? THIS IS FEDORA (said in 300 voice)
20:35:01 <gotmax[m]> And perhaps we should do that more on the ML
20:35:20 <tdawson> gotmax[m]: I'm for that.
20:35:24 <smooge> are we still talking about the modularity or a different one?
20:35:49 <tdawson> smooge: Oh, we were subtly talking about
20:35:53 <tdawson> .epel 39
20:35:54 <zodbot> tdawson: Issue #39: Retiring EPEL Packages (End of lifing CEPH in EL7 and process improvement) - epel - Pagure.io - https://pagure.io/epel/issue/39
20:35:59 <gotmax[m]> We're talking about retirement/orphaning
20:36:22 <smooge> ah yeah. I can do some clearer goals on the list
20:37:21 <tdawson> Is there anything anyone wanted to talk about the retirement / orphan stuff, before me move it over to the mailling list?
20:37:43 <tdawson> Nope ... ok, moving on. :)
20:38:03 <tdawson> And ... with the new agena format ... we've moving to ...
20:38:10 <tdawson> #topic General Issues / Open Floor
20:38:32 <gotmax[m]> I'm thinking I might stop building ansible in EPEL 8 Next
20:39:09 <tdawson> gotmax[m]: If it is built on the regular epel8, does it not work on CentOS STream 8?
20:39:16 <nirik> There was a request to upgrade Xfce in epel7... :(
20:39:25 <gotmax[m]> ansible-core was broken for two months until carlwgeorge asked them to fix it
20:39:34 <nirik> :(
20:40:02 <tdawson> nirik: So ... they want the latest and greatest desktop ... on the oldest release?
20:40:17 <gotmax[m]> In the same compose where the ansible-core update that was submitted two months ago landed, the python39 module broke
20:40:20 <nirik> tdawson: yes.
20:40:47 <gotmax[m]> This is the forth time one of the Python modules in c8s has broken
20:40:47 <nirik> gotmax[m]: yeah, hard to want to keep things up if others don't do their part
20:41:36 <carlwgeorge> until/unless c8s gets converted to the c9s workflows, this type of stuff will keep happening
20:41:47 <tdawson> I'm really really hoping that when CentOS STream 8 has the same workflow as CentOS STream 9, all those missing / old / broken packages will be cleaned up.
20:42:01 <gotmax[m]> tdawson: Well, I was trying to build the newer version of ansible that corresponds to the newer versions of ansible-core in Stream
20:42:10 <gotmax[m]> But it will still work if I don't do that
20:42:19 <smooge> gotmax[m], I would say stop building it until things improve elsewhere
20:42:20 <tdawson> But that's still a couple months away :(
20:42:49 <gotmax[m]> This has been a success with EPEL 9 Next, but EPEL 8 Next is a headache
20:42:57 <gotmax[m]> And I already have a headache disorder ;)
20:43:16 <smooge> you only have so many mana points and this isn't the boss level
20:43:43 <tdawson> gotmax[m]: Yep, I agree with smooge ... if it keeps breaking due to stream ... don't worry about it and point people to the cause of the problem.
20:44:55 <tdawson> nirik: You didn't say what you told them about xfce
20:45:01 <gotmax[m]> Yeah, that's probably what I'll do
20:45:25 <gotmax[m]> Should I file a c8s `distribution` bug about the broken python39?
20:47:04 <carlwgeorge> that or against the python39 component
20:47:39 <nirik> tdawson: I told em no, but my co-maintainer is asking them more... see https://bugzilla.redhat.com/show_bug.cgi?id=2126508 for discussion
20:48:23 <tdawson> nirik: At least for KDE I have some valid "the libraries in RHEL are too old"
20:48:28 <gotmax[m]> carlwgeorge: I usually file against distribution. I guess I'll do that and CC the python-maint address.
20:48:56 <gotmax[m]> I'm getting a bit tired of filing these bug reports, but I also don't like broken things, so I'll do it
20:49:10 <tdawson> nirik: You should point them to icewm ... that's "lean and mean" ... it makes xfce look bloated.
20:49:22 <smooge> fvwm4life
20:49:44 <nirik> ratpoison is great too (oh wait, I maintain that too!)
20:50:07 <tdawson> *laughs* ... oh ... man ... I threw up a little there.
20:50:20 <tdawson> Sorry ... didn't mean to comment on others choices of desktop. :)
20:50:28 <nirik> it also may be that epel7 is too old for Xfce 4.16.
20:50:47 <smooge> tdawson, if you eat ratpoison, throwing it up is a good option
20:51:00 <smooge> but please see a proper poison control
20:51:09 <nirik> (it's a don't use the mouse/screen for X type thing)
20:51:10 <tdawson> *laughs*
20:51:51 <smooge> anyway.. good luck with that nirik.
20:51:59 <smooge> epel7 is still growing in usage!
20:52:11 * smooge goes back to laughing and laughing
20:52:28 <tdawson> gotmax[m]: Just so you know, the person(s) fixing those bugs on CentOS STream 8 do appreaciate the bugs so they know when they are broken ... but they also really wish their tools would quit messing things up.
20:53:04 <tdawson> Anything else for this week?
20:53:19 <smooge> not from me.
20:53:22 <tdawson> This was shorter than usual.
20:53:40 <smooge> well cutting out all about EL6 EL7, EL8 and EL9 takes 10 minutes off
20:53:51 <tdawson> It does
20:54:10 <tdawson> Thank you all, for all you do for the EPEL community.
20:54:15 <smooge> thank you tdawson
20:54:16 <tdawson> I'll talk to you next week, if not sooner.
20:54:23 <tdawson> #endmeeting