21:00:17 #startmeeting EPEL (2022-01-12) 21:00:17 Meeting started Wed Jan 12 21:00:17 2022 UTC. 21:00:17 This meeting is logged and archived in a public location. 21:00:17 The chair is tdawson. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 21:00:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 21:00:17 The meeting name has been set to 'epel_(2022-01-12)' 21:00:17 #meetingname epel 21:00:18 #chair nirik tdawson pgreco carlwgeorge michel dcavalca 21:00:18 #topic aloha 21:00:18 The meeting name has been set to 'epel' 21:00:18 Current chairs: carlwgeorge dcavalca michel nirik pgreco tdawson 21:00:22 .hi 21:00:23 pgreco: pgreco 'Pablo Sebastian Greco' 21:00:30 Hi pgreco 21:00:37 .hi 21:00:38 carlwgeorge: carlwgeorge 'Carl George' 21:00:45 Hi carlwgeorge 21:01:17 morning 21:01:24 Hi nirik 21:01:51 .hi 21:01:52 rcallicotte: rcallicotte 'Robby Callicotte' 21:02:04 Hi rcallicotte 21:02:15 .hi salimma 21:02:16 michel: michel 'None' 21:02:26 Hi michel / sallimma 21:02:27 Wait what 21:02:34 hmm 21:02:37 .hello salimma 21:02:38 michel: salimma 'Michel Alexandre Salim' 21:03:00 Oh yeah I guess there is a Michel in FAS 21:05:20 #topic EPEL Issues 21:05:20 https://pagure.io/epel/issues 21:05:20 https://pagure.io/epel/issues?tags=meeting&status=Open 21:05:40 There isn't any issues marked for meeting ... but ... 21:05:49 .epel 77 21:05:50 tdawson: Issue #77: systemd-rpm-macros included in systemd on EPEL 7 (and maybe 8, too) - epel - Pagure.io - https://pagure.io/epel/issue/77 21:05:56 :( 21:06:03 Sorry :( 21:06:10 Sounds like it still isn't going well. 21:06:21 I always seem to remember about it on Wednestay mornings... 21:06:53 moving on... 21:06:55 :) I tend to consider Wedensdays my EPEL days, cuz that's when I remember to do the things I commited to last week. 21:07:13 #topic Old Business 21:07:26 Same, haga 21:07:29 *haha 21:08:19 The only old business I have is my "Will It Install / Build" ... I'm very close on my will it Build, but having everything fail on fedpkg-minimal makes the output rather large. 21:09:17 We should discuss fedpkg-minimal, either here or when we get to epel9 21:09:59 I'm ok treating it just like any other package, we request it, we wait a couple weeks if nobody does antying, then open a releng ticket. 21:10:05 I have a build failure too in epel9, but seems to be more stream related than epel related 21:11:06 tdawson: yeah, but arguably we requested it months ago 21:11:11 I know carlwgeorge volunteered to be added as an admin for fedpkg-minimal, but since it's rather critical to the epel builds, I'm thinking adding the fed-package-sig to it would be good too. 21:11:15 which one should we keep open? https://bugzilla.redhat.com/show_bug.cgi?id=2007051 or https://bugzilla.redhat.com/show_bug.cgi?id=2039901 21:11:24 carlwgeorge: Keep the older one open. 21:11:33 Yeah, the older one 21:11:45 And, as michel said, we requested it a month ago, I think it falls under a stalled package. 21:11:46 tdawson: epel-packagers-sig right? 21:11:59 alright i closed the newer one 21:12:01 michel Ya ... my fingers weren't cooperating with typing. 21:12:09 If Carl gets admin he can do it for us , but in case they only give commit, better ask for it 21:12:41 My old business: https://pagure.io/epel/ebranch 21:13:21 It works, but haven't got round to teaching it to do recursive deps and state tracking yet. But people can give it a spin. Docs is still very minimal too, but 100% unit tests 21:13:23 I'd give them a chance to respond from today... but if they don't in a bit sure. 21:13:32 neat 21:13:48 So we'll have fedpkg escalated in 2 weeks tops I guess 21:14:02 i would say a week from today 21:14:12 Oh right, today is the ping 21:15:49 Sorry, was reading through ebranch ... very cool 21:16:13 michel I'll see about giving it a spin this week. 21:16:36 Anything else for Old Business? 21:16:37 big fan of the click library btw 21:16:59 I hate argparser so yeah 21:17:03 Click ftw 21:17:06 :) 21:17:31 #topic EPEL-7 21:18:05 Although I have heard several things dealing with EPEL7 this week, I think all of them were resolved and as far as I saw, nothing needed for this meeting. 21:18:16 Or did I miss anything? 21:19:24 #topic EPEL-8 21:20:00 I have in my notes openssl3 for epel8 ... is that still something we need to bring up here? 21:20:13 I have a bug I filed for updating gflags in c8s, it's blocking some packages 21:20:37 I thought I pushed openssl3 for epel8 already? 21:21:01 Ya ... that's what I was thinking... ok, I'm taking it out of my notes. 21:21:09 https://src.fedoraproject.org/rpms/openssl3/ 21:21:18 I'll add the SIG so we can comaintain 21:21:32 Which, by the way, incase I didn't say it ... good job. 21:21:50 Thanks! We happen to need it internally too 21:22:42 Anything else for epel8? 21:22:51 make sure to retire the rawhide branch as "epel only", or else it will get new fedora branches from here going forward 21:23:04 carlwgeorge: Oh yeah, thanks 21:23:17 Q: if I do that can I still branch for f35 21:23:26 In the unlikely situation we realize suddenly we need it 21:23:50 sure 21:24:00 Oh good 21:24:21 carlwgeorge: Is that something we need to do for -epel packages? 21:24:30 Oh ... I guess I could just check ... just a sec ... 21:25:21 I wonder if we can make fedpkg request-repo do --epel-only 21:25:30 tdawson: yup 21:25:32 Oh, hi all. 21:25:40 And also teach it to request branches immediately and save us time 21:25:56 it doesn't hurt anything to forget, just makes clutter 21:25:59 rsc_: hullo 21:26:05 Youch ... let's put that in the doucumentation. Looks like alot of mine aren't retired. 21:26:12 Hi rsc 21:26:18 retired on rawhide I mean. 21:27:16 They're still wired 21:27:21 😉 21:27:25 I'll show myself out 21:27:31 :) 21:27:41 OK ... sorry for the pauses, just writting myself notes. 21:27:46 Anything else for epel8 ? 21:28:17 #topic EPEL-9 21:28:38 Django is blocked on libraqm, which is now blocking imagemagick too 21:28:56 I forgot to ping the maintainers again last week, so I just did - we can escalate in a week 21:29:11 (ah sorry, this blocks pillow which blocks Django) 21:29:28 so. many. packages. 21:30:06 Yeah, and right now I'm guiltily using Koji to find out what breaks 21:30:17 Actually, ebranch can help now, duh 21:30:46 Thus far my biggest dependency chain requests has gone down 4 levels ... though I'm also not trying to do fedpkg. :) 21:31:02 wowsers! 21:31:49 Yeah, mine is hyperkitty - Django - pillow - raqm 21:31:54 I have some requests where I haven't dug in because it looks daunting. ;( Can try ebranch on em 21:32:23 nirik @nirik:libera.chat: I'll let you know when it can do recursive deps. Haven't had time to code all week 21:32:26 That was for sassc, and when I was creating bugs for the epel9 branches, I saw that the same requests were still sitting there for epel8. so I'm just grabbing both sets ... just have one more week to wait for those. 21:32:57 But our perf reviews are due tomorrow, and after I write the ebranch presentation so I can get it approved I will have double motivation to write the features I promise 😅 21:33:12 tdawson: Yikes 21:34:22 For epel9, do we have anything beyond packages being requested and fedpkg-minimal (which I guess is also a package being requested) ? 21:34:24 side note: does anyone know how to make python-bugzilla (the command line tool) ignore the 20 bug default limit? I saw something was added, but I don't see the option anywhere 21:34:45 Oh, yes, one thing 21:34:58 Does anyone know how Rust is packaged in c9s 21:35:31 I'm hoping we can package Rust crates like we do for Fedora - automatic BRs etc, and not having to vendor the dependencies 21:36:30 nirik @nirik:libera.chat: I don't, but I find that bugzilla tree view also sadly is limited to 20 bugd 21:36:35 * nirik has no idea off hand on rust 21:36:47 so the BRs contain all possible crates? 21:36:57 I believe rust itself, and any rust based pacakges in RHEL9 have things bundled, but I don't think that stops anyone from doing un-bundled rust crates. 21:38:23 yeah the problem isn't how it's packaged, it's just the availability of crates as rpms 21:38:39 ah 21:38:58 tdawson: so I guess for epel we should try doing it Fedora style 21:39:03 (back on keyboard, phew) 21:39:14 rust packaging is painful in fedora even with hundreds (a guess) of crates already packaged, and with epel9 we're starting from scratch essentially 21:39:40 salimma: If that's the way you, or others want to, go for it. I'm staying clear of rust packages as much as possible. 21:40:32 so I just heard back from the other Rust SIG members - there are rust crates in epel9-next, so it works 21:40:44 (this predates the decision to build epel9 against c9s, likely) 21:41:09 tdawson: fair enough, I stay away from Golang and Node for the same reason ;) 21:41:15 salimma: That's right. I remember seing those when I was clearing out my packages for epel9-next 21:41:42 one language requiring rebuild for everything is enough for me. and even the kernel might go Rust this year 21:42:36 Ya ... I think if the kernel goes rust, I would think they'd bundle that, keep it contained within itself ... but, that's just my opinion. 21:42:51 Anything else for epel9 before we move on? 21:43:16 just a comment 21:43:20 the kernel likely won't require additional crates, that'd be insane ;) 21:43:38 I'm trying to build htop for epel9 and I'm hitting something that looks like this error https://bugzilla.redhat.com/show_bug.cgi?id=1869030 21:44:05 works on x86_64 (fedora 35 host) and fails in aarch64 (centos 8 stream host) 21:44:16 if anybody has any chance to replicate, please ping me 21:45:03 pgreco: Will do 21:45:18 #topic EPEL-Packaging-SIG 21:45:51 I know we've already talked about a couple packaging-sig stuff, but it's on the agenda. Does anyone have any other packaging-sig items? 21:46:37 i've got one 21:47:40 at the upcoming centos dojo, i was thinking of putting together an epel hackfest. the target audience would be fedora packagers that are not familiar with epel but want to get involved and would like some hand holding. 21:48:20 i was thinking it would be ideal to have 2-4 epel packaging sig members around to answer questions and mentor 21:49:13 Depending on what time of the day it is, I could be there. 21:49:51 * nirik notes that he found the bugzilla command line thing... '--field=limit=1000' 21:50:02 not sure yet, the cfp is technically closed but rich mentioned the scheduling would work better if he had one more session so i suggested this idea 21:50:04 yeah, same here, depending... 21:50:20 carlwgeorge +1 for this idea. 21:51:13 i'm worried we'll get a lot of non-packagers that don't understand they need to start in fedora, so i'm sure some of it will be guiding people over there 21:51:37 Don't know until you try it. 21:52:02 i'll send the proposal and let yall know what day it ends up getting scheduled for, and if people can come they can come, if not oh well i'll do it myself :D 21:52:09 Could be something we could try at other events as well. Possibly Flock/Nest and places like that. 21:52:27 Wee how things go. 21:52:35 /Wee/See/ 21:52:57 carlwgeorge: ooh, that's nice 21:53:31 Since we're close to time, I'm going to change to Open Floor .... 21:53:33 #topic General Issues / Open Floor 21:53:34 if we discover people need help, this is a good time to give them that guidance too :D 21:53:46 if nothing else i'm sure me and Eighth_Doctor will be there 21:55:28 Anything else before we close? 21:55:45 I'll be there too 21:56:06 though like Seattle weather, I might be called off at random times 21:56:21 Ha! 21:56:45 I will be there (even tho I'm still new-ish) 21:57:05 sounds like we'll be in good shape 21:57:45 Anything else before we close? 21:58:50 Thank you everyone for being here this week, and for all you do for the EPEL community. Ya'll are great. 21:59:04 I'll talk to ya'll next week, if not sooner. 21:59:07 thanks for running things tdawson 21:59:17 #endmeeting