20:00:03 #startmeeting EPEL (2022-08-03) 20:00:03 Meeting started Wed Aug 3 20:00:03 2022 UTC. 20:00:03 This meeting is logged and archived in a public location. 20:00:03 The chair is tdawson. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 20:00:03 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:00:03 The meeting name has been set to 'epel_(2022-08-03)' 20:00:05 #meetingname epel 20:00:05 The meeting name has been set to 'epel' 20:00:06 #chair nirik tdawson pgreco carlwgeorge salimma dcavalca dherrera 20:00:06 Current chairs: carlwgeorge dcavalca dherrera nirik pgreco salimma tdawson 20:00:08 #topic aloha 20:00:13 .hi 20:00:14 pgreco: pgreco 'Pablo Sebastian Greco' 20:00:20 .hi 20:00:21 jonathanspw: jonathanspw 'Jonathan Wright' 20:00:21 Hi pgreco 20:00:29 morning 20:00:30 Hi jonathanspw 20:00:41 Morning nirik 20:00:57 .hi 20:00:59 dherrera: dherrera 'Diego Herrera' 20:01:08 Hi dherrera 20:01:23 before we start carlwgeorge, jonathanspw, thanks for btop!, looks beautiful 20:01:36 It's quite nice ain't it :) 20:01:51 .hello dcavalca 20:01:52 davide: dcavalca 'Davide Cavalca' 20:01:59 hello 20:02:13 .hello gotmax23 20:02:14 (who is a bit preoccupied) 20:02:14 gotmax[m]: gotmax23 'Maxwell G' 20:02:19 Hi davide 20:02:27 Hello smooge 20:02:35 jonathanspw: I left it running for a few hours, and got sad when I had to turn it off to reboot the laptop :D 20:02:40 Hello preoccupied gotmax[m] 20:03:03 Hi tdawson 20:04:05 .hi 20:04:06 rcallicotte: rcallicotte 'Robby Callicotte' 20:04:31 Hi rcallicotte 20:05:12 * tdawson just installed btop right now to see what pgreco was talking about. 20:05:36 #topic EPEL Issues https://pagure.io/epel/issues 20:05:37 https://pagure.io/epel/issues?tags=meeting&status=Open 20:05:47 pgreco: i love it. i only discovered it when i packaged it. I'm still learning how to customize it to my likings 20:06:33 The only open issue is on retiring EPEL packages. 20:06:40 tdawson: I saw some comments a few days ago and installed it out of FOMO, best decision ever 20:07:13 Last weeks discussion led us to talk about trying something like orphaning ... but I didn't see an email about it. 20:07:39 I was about to send it 20:08:05 gotmax[m]: OK 20:08:11 And then I realize that for our purposes, the rerirement process would just be glorified retirement 20:08:21 With a longer time limit 20:08:23 Are we ok waiting on this until gotmax[m] has sent his email, and then discuss it on the email? 20:08:41 I'm ok with that 20:08:47 yeah 20:08:55 yep 20:09:06 Fine with me 20:09:22 OK ... then moving on with Old Business 20:09:33 #topic Old Business 20:10:10 so I was wondering if we should support RHEL 4 20:10:13 We had talked about getting the epel-packaging-sig setup as the bugzilla default. 20:10:16 *laughs* 20:10:17 sorry too old of business? 20:10:21 /kickban smooge 20:10:30 .fire smooge 20:10:30 adamw fires smooge 20:10:35 I'm not the only one with centos 4 machines still kicking? 20:10:57 Sorry ... I can't stop laughing ... give me a minute. 20:11:00 lol 20:11:05 Someone has to tell me the story of .fire sometime 20:12:19 OK ... other than epel4 ... nirik ... it looked like you were able to verify that epel-packaging-sig could be set to be the bugzilla default. 20:12:40 gotmax confirmed that. 20:12:52 Yup :) 20:12:56 I fixed all the ones I processed, but there may be older ones that still need checking 20:13:18 .hi 20:13:19 salimma: salimma 'Michel Alexandre Salim' 20:13:19 Thanks for fixing that! 20:13:20 Oh ... ok. Then I'll have to go back and see if some of mine were fixed. 20:13:24 sorry I'm late, recovering from a meeting 20:13:30 Hi salimma 20:14:29 nirik: So, sounds like we just go through our past requests ... can we do it ourselves, or do we need to open a releng ticket? 20:14:38 * gotmax waves to salimma 20:14:49 you should be able to I think, but if not file a releng ticket. 20:15:32 OK. 20:15:41 I thought only main admins could change the assignee? 20:15:53 oh yeah, there's a bug/ticket on that 20:16:31 so, perhaps we could gather a full list and do em all at once? but sadly I am not sure how to do it via the api, it's tedious clicking for me. 20:16:40 That would be good to be fixed ... cuz often the main person is no longer responsive. 20:16:48 There's a way to do it via the API 20:17:10 I know how to do it with the API, one sec 20:17:15 https://pagure.io/fedora-sig-onboard/blob/main/f/fedora_sig_onboard/onboarder.py#_167 20:17:21 this is the magic incantation 20:17:25 right. branching and building something is only the first step. Maintaining is important. 20:17:36 it does work for groups as long as you assign the correct string (so @groupname) 20:17:54 You don't need to specify both in the API request if you only want to change the EPEL. 20:18:01 correct 20:18:19 * nirik nods 20:18:19 if we want to start doing this for EPEL Packagers SIG... should we discuss bugs assigned to it during these meetings? 20:18:40 I'd rather use an email list for bugs tbh 20:18:48 so it can be filtered somewhat easily 20:18:59 I think that's what the golang and rust sigs do? 20:19:32 you can do that, yeah. create a user for the mailing list in Bugzilla so bugs assigned there get sent to the list 20:19:53 thats already the case. Thats a prereq for a group. ;) 20:20:05 once we figure all this out, someone please write it down so the next SIG doesn't have to reinvent the wheel :) 20:20:15 I might have done that for lua (copying the Python setup) but I don't quite remember, but Miro or Fabio probably knows. oh - there we are then 20:20:21 as nirik said 20:20:29 How does one go about joining the EPEL SIG officially? Docs just say bring it up at meeting or on ML :) 20:21:01 I think we've done formal votes on tickets before, but I'm fine just doing it in the meeting 20:21:04 https://lists.fedoraproject.org/admin/lists/epel-packagers-sig.lists.fedoraproject.org/ 20:21:20 nirik++ 20:21:20 davide: Karma for kevin changed to 17 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 20:21:30 Correct, both are fine. In the meeting, or open an issue. 20:21:50 Oh, hi! 20:22:42 Hi rsc 20:23:28 Sudden quiet as rsc joins. :) 20:24:14 OK, sounds like this has been cleared up .. 20:24:30 I don't have any more old business 20:24:44 #topic EPEL-7 20:24:46 CentOS 7 will go EOL on 2024-06-30 20:25:02 Any epel7 stuff ... or old business that I forgot ? 20:25:09 Speaking of old business... 20:25:12 .hello ngompa 20:25:13 Eighth_Doctor: ngompa 'Neal Gompa' 20:25:15 HI 20:25:17 HELLLO 20:25:22 Hi Eighth_Doctor 20:25:26 sorry I'm late :) 20:25:33 IWAYEDFUGUSDYFGUYSDGFUYSGDUYFGUSDF 20:25:39 I was busy being mad at Outlook today 20:25:53 Eighth_Doctor: eww ... that's easy to do 20:26:12 HELLO 20:26:13 work was forced to switch from GWS to M365 five weeks ago 20:26:15 SOMEONE REPSOEIDN 20:26:16 POXSIUER9UDGBNHFGIHDBS 20:26:18 LUJBDCIDS 20:26:20 FGEUEUUSORR 20:26:23 I'm not coping well... 20:26:28 There's been a lot of Matrix/IRC spam lately... 20:26:34 HELP 20:26:38 ARE YOU REAL 20:26:41 * nirik sighs 20:26:43 gotmax (He/Him): that's IRC side 20:26:45 sigh. that's new to me 20:26:55 P 20:26:56 'KIBNH 20:26:59 gotmax[m]: I noticed. Was that the old business you were talking about? 20:27:01 most of the spam I see these days is from IRC side thing 20:27:15 PLYGXESDGB 20:27:27 I was saying EPEL 7 is old business 20:27:32 Ha!! 20:27:44 everyone on Matrix, just ignore the user, that seems to work 20:27:48 oh, that works too 20:27:50 nirik++ 20:28:03 I got a spam email from an Outlook.com user recently, and I replied to them and CC'd abuse@outlook.com. Hopefully, that shuts that person up... 20:28:09 WHATS MATIX 20:28:14 Anyways... 20:28:24 Moving on ... 20:28:29 it doesn't work, I've noticed :( 20:28:31 #topic EPEL-8 20:28:32 CentOS Stream 8 goes EOL in 2024-05-31 20:28:33 * Eighth_Doctor grumbles 20:29:09 * gotmax[m] needs to open a new `distribution` bug due to c8s brokenness :( 20:29:10 Anything for epel8? 20:29:28 gotmax[m]: What's broke with c8s? 20:29:39 ansible-core 20:29:53 Or is it really just CentOS Stream, and not really an epel thing? 20:30:13 Well, it's affecting the ansible package that I maintain. Not really important, though 20:30:22 Ah ... ok. 20:30:42 what breakage is this? 20:30:42 Anything else epel8 related? 20:30:51 The rebase from python38 to python39 wasn't properly handled on the CentOS side 20:31:12 ah. :( 20:31:13 And the package has been FTBFS for 3+ weeks due to the python39 packages missing from ansible-core's built tag 20:31:30 ouch 20:31:48 oh dear 20:31:49 not nice. ;( 20:32:23 I follow ansible-core bugs in RHEL and I expected it to break when I saw the bug for the update. I asked the maintainer to notify the CentOS people, but it seems that didn't happen... 20:32:46 HOW FEDORA 20:33:05 OOGA BOGA 20:33:08 OBOBOFA 20:33:11 I feel your pain, but I'm going to move on 20:33:21 #topic EPEL-9 20:33:23 CentOS Stream 9 goes EOL in 2027-05-31 20:33:30 OGA BOGA 20:33:43 Yeah, this really isn't that relevant :) 20:33:49 Any epel9 things? 20:34:28 I had one... 20:34:39 nirik: ++ 20:34:47 nirik++ 20:34:47 rcallicotte: Karma for kevin changed to 18 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 20:34:47 me too, but nirik first 20:35:00 nirik++ 20:35:01 nirik: Go for it 20:35:01 where should this be asked to refiled: https://bugzilla.redhat.com/show_bug.cgi?id=2115102 20:35:23 man-db-cron is a fedora thing I think, it's not in the epel9 man-db... but it's part of man-db in fedora? 20:36:22 ??? I think you put the wrong link in ... that's a package review request. 20:36:33 for pylero 20:36:53 oops. 20:37:21 https://bugzilla.redhat.com/show_bug.cgi?id=2115091 20:37:26 the one right next to it. 20:38:27 :) 20:39:06 Well ... it's a missing package ... although it's not a -devel package, I think it would go through the missing package steps. 20:39:09 Well, I can tell you that this is definitely the wrong place. 20:39:21 Probably the epel mailing list or the Fedora package 20:39:45 This would need a -epel package (i.e. it's a missing subpackage)? 20:40:02 Correct 20:40:26 yes, I think so... 20:40:27 I just double checked, and it is build on CentOS Stream 9 and RHEL9, just not shipped. 20:40:44 but likely the reporter isn't a packager... but I haven't checked 20:41:30 It's a fairly low maintenance package if someone wants to take on the -epel package ... but like gotmax[m] said ... I think sending it to the epel-devel mailling list would be a good idea. 20:41:56 And also reporting to RHEL to see if they would provide it 20:42:11 the approach I've taken with these is to parallel path 20:42:28 so request the one you need in RHEL, and make the -epel one to unblock things in the meantime 20:42:28 Yep ... do them both. 20:42:31 if someone would like to update/close/move that, feel free. ;) 20:42:32 and then retire the -epel one once the RHEL one is sorted out 20:42:38 that's obviously more work though 20:42:43 Yeah, I think that's pretty standard 20:43:00 Based on their Bugzilla history, they don't appear to be a packager 20:44:20 I just barely noticed that it's set to "Package Review" as the component ... ya ... I don't think they are a packager. 20:44:51 * gotmax[m] starts writing a response 20:45:11 salimma: You said you had an epel9 thing also ... go for it 20:45:18 yup 20:45:36 was chasing a new dependency to rubberband and got to this - https://bugzilla.redhat.com/show_bug.cgi?id=2112430 20:46:17 it's another package where the test suite still depends on nose, which for good reasons is not in EL9. in general, do we just prefer disabling tests, or helping port the test suite? I haven't had time until now but can try give it a stab today or tomorrow 20:47:01 I think disabling... especially if they are enabled in Fedora to catch things there... 20:47:03 Trying to port the test suite is always best... 20:47:05 porting is a lot more effort. 20:47:18 if you can tho yeah, it's better... 20:47:23 nirik: sad thing is in Fedora... the test suite is run with || : 20:47:28 because of known existing failures :( 20:47:40 Agreed. It depends how difficult it is. 20:47:45 so its really no test at all 20:47:48 salimma: sad trombone emoji 20:47:58 so yeah, I've requested an epel9 branch, if I get a working test suite port before then I'll put it up, if not I'll just use bcond 20:47:58 If someone updates rdflib from 5.0.0 to the current 6.2.0 in fedora then the problem will go away 20:48:10 At a certain point, it becomes an upstream problem. They shouldn't be depending on packages that have been unmaintained for years. 20:48:11 upstream removed the nose dependency 20:48:16 music: oh interesting. upstream changed their ... nice 20:48:18 Well ... yes and no. At least you get the output of the test to know that it's broken ... 20:48:26 In that case, it's our problem :) 20:48:34 so updating rawhide, then branching for epel9, would be the quickest fix. don't touch f36/f35 unless the maintainer wants to 20:48:54 music++ 20:48:54 salimma: Karma for music changed to 2 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 20:49:02 that sounds like the best course of action 20:49:12 I'll prepare a PR for 6.2.0 once I get it to build on both epel9 and rawhide 20:49:18 Am I the only one that noticed that when someone played the "sad trombone" emoji ... that music[m] suddenly appeared? 20:49:34 hehe 20:49:44 ha. really? 20:50:17 salimma: sounds good. And thanks music[m] for the info. 20:50:19 PSA: Matrix reactions don't transfer over the IRC bridge 20:50:38 Yeah, thanks music :) 20:50:55 gotmax[m]: yup, but I try not to spam IRC users :) 20:51:02 anything else for epel9 before we go to open floor? 20:51:09 the really sad thing is that there isn't a sad trombone emoji (yet) 20:51:17 Fair enough 20:51:31 but every time someone types it I hear it.... 20:51:40 #topic General Issues / Open Floor 20:52:10 I gave a talk at SCALE last week that covered some of the work we're doing in EPEL (among other things) 20:52:23 #info nest is starting tomorrow. Everyone should attend. ;) 20:52:28 https://www.youtube.com/watch?v=snHjDfjDdBs if anyone's interested (skip over the first few minutes) 20:52:43 The very last talk is my "State of EPEL" presentation. 20:52:51 nice 20:52:56 this is perfect timing, I just got done with my managerial duties for the year :) 20:53:04 Although I'm giving the presentation, everyone is invited for join in on the Q&A 20:53:22 also a bunch of us will be in Boston for DevConf.US 20:53:30 if you're gonna be in the area stop by and say hi 20:53:30 Ya!! 20:53:37 :) 20:53:38 I've gotten used to devconf being on hopin, this will be weird 20:53:51 hope I can join in person next year (at least for the US one, maybe the CZ one the year after? we'll see) 20:54:07 I've never physically been to a devcon ... so I'm excited for that. 20:54:15 I don't think CZ is happening next year, or at least I haven't seen anything so far 20:54:18 you need to be mentally prepared for devconf.cz 20:54:19 huh? interesting 20:54:22 the cold is unbelievable... 20:54:38 ah 20:54:55 pgreco: I just moved to Chicago, so the next time they have it I think I'll be rady 20:54:58 i think that some places in the Andes in winter are comparable 20:55:18 yeah.. Chicago at -40C is good test 20:55:31 it's less the cold, and more the ice everywhere that always gets me in Brno 20:55:31 count me out... :D 20:55:31 smooge: the C is redundant btw 20:55:32 hehe 20:55:41 haha 20:55:46 yes it was 20:55:46 oh yeah, having to walk in ice is a pain. I lost one laptop that way 20:56:42 I look forward to seeing those that I see there 20:56:48 I think we're close to time 20:56:56 Anything else before I close up? 20:57:08 * gotmax[m] feels the need to clarify that the ansible package in EPEL 8 Next isn't actually broken. It just can't be updated. 20:57:12 nothing from me. 20:57:14 see everyone at next tomorrow! 20:57:20 s/next/nest 20:57:28 yay!! 20:57:36 Thank you all for the great discussions, and for all you do for EPEL and the community. 20:57:46 Talk to you all next week, if I don't see you at Nest. 20:57:49 thanks y'all 20:57:52 tdawson++ 20:57:52 salimma: Karma for tdawson changed to 3 (for the current release cycle): https://badges.fedoraproject.org/tags/cookie/any 20:57:52 I'm sure tdawson will be discussing EPEL Next at Nest... 20:57:57 #endmeeting