20:00:08 #startmeeting EPEL (2023-05-17) 20:00:08 Meeting started Wed May 17 20:00:08 2023 UTC. 20:00:08 This meeting is logged and archived in a public location. 20:00:08 The chair is tdawson. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions. 20:00:08 Useful Commands: #action #agreed #halp #info #idea #link #topic. 20:00:08 The meeting name has been set to 'epel_(2023-05-17)' 20:00:09 #meetingname epel 20:00:09 The meeting name has been set to 'epel' 20:00:11 #chair nirik tdawson pgreco carlwgeorge salimma dcavalca dherrera gotmax23 smooge 20:00:11 Current chairs: carlwgeorge dcavalca dherrera gotmax23 nirik pgreco salimma smooge tdawson 20:00:12 #topic aloha 20:00:24 .hello dcavalca 20:00:25 davide: dcavalca 'Davide Cavalca' 20:00:27 .hi 20:00:28 carlwgeorge: carlwgeorge 'Carl George' 20:00:30 morning. sort of here, sort of prepping for outage in an hour. 20:00:41 Hello davide 20:00:43 Hi carlwgeorge 20:00:47 Morning nirik 20:01:16 .hi 20:01:17 neil: neil 'Neil Hanlon' 20:01:21 g'day folks 20:01:38 .hi 20:01:39 gotmax: Sorry, but user 'gotmax' does not exist 20:01:45 neil: what was the site for that ranked choice web app you recommended? 20:01:46 .hello gotmax23 20:01:47 gotmax: gotmax23 'Maxwell G' 20:01:58 Hi neil 20:02:02 #link https://rankedchoices.com/ 20:02:03 Hello gotmax 20:02:30 .hi 20:02:31 dherrera: dherrera 'Diego Herrera' 20:02:33 carlwgeorge: Do you want to setup a ballot, or do you want met do? 20:02:37 Hi dherrera 20:02:53 hello 20:03:00 Hello jcpunk 20:03:03 either way, i was getting ready to set it up so you could focus on running the first part of the meeting 20:03:12 carlwgeorge: That would be great, thanks. 20:03:19 .hi 20:03:20 dminer: dminer 'Dalton Miner' 20:03:29 I was going to set it up before, then couldn't find the link. 20:03:32 Hi dminer 20:03:52 carlwgeorge: i'm actually thinking it might be a neat feature to add to nuancier... (ranked choice, that is) 20:05:07 #topic End Of Life (EOL) 20:05:08 RHEL 7 / epel-7 will go EOL on 2024-06-30 20:05:09 CentOS Stream 8 / epel-8-next goes EOL in 2024-05-31 20:05:11 CentOS Stream 9 / epel-9-next goes EOL in 2027-05-31 20:06:08 Usually the agenda does EPEL Issues first, before Old Business, but this time, I want enough time for Old Business. 20:06:10 hello 20:06:15 Hello smooge 20:06:17 .hello ngompa 20:06:18 Eighth_Doctor: ngompa 'Neal Gompa' 20:06:22 hey folks 20:06:23 Hello Eighth_Doctor 20:06:30 hey Eighth_Doctor 20:06:33 #topic Old Business 20:06:33 👋 20:06:50 * Eighth_Doctor is finally recovering from being overheated now that his new air conditioning system is installed 20:06:50 we have rhel 9.2 and 8.8 released here in the last few weeks too for anyone living under a rock. 20:07:09 Yep 20:07:38 So, for Old Business, we can start with the dist-tag discussion for epel 10 20:07:42 being overheated sucks, glad you got the cool air finally Eighth_Doctor 20:07:50 #link https://discussion.fedoraproject.org/t/epel-10-bikeshedding-dist-tag/80510 20:07:50 :) 20:08:01 * gotmax skims the discussion 20:08:25 carlwgeorge: Do you have the survey setup, or do you want met to go into the overview? 20:08:40 close, go ahead and review the options if you like 20:09:00 There are three choices Option 1, 2 and 3. 20:10:00 Option 1 - Each build would have the dist tag of .el10_X with X being the RHEL release it was built against. (ex: .el10_2) 20:11:02 i have the ballot ready, but i'll wait to drop the link till you finish the overview 20:11:18 Option2 - All builds built against CentOS Stream (the latest RHEL) would get .el10 and any built on anything else would get a ~X (exL .el10~1) 20:11:35 * neil finally got around to commenting on the thread.. just now 20:11:39 I possibly didn't explain that one very good, but hopefully you get the idea. 20:11:58 makes sense to me, the pros and cons of each are detailed in the discussion thread 20:12:55 Option3 - Similar to Option 2. The leading branch would get .el10, but package built on anything else would get an _X (ex: .el10_2) 20:13:29 The discussion has it worded much better, but now we have it in the meeting notes. 20:13:35 carlwgeorge: I'll turn it over to you. 20:13:38 any questions before we vote? 20:14:23 i submitted the options in the same order listed here and in the thread, but i'm not 100% certain the order will be preserved so read the text carefully to make sure it's your intended choice 20:14:24 well on option 3, if I compile first against CS10 and get a el10, then do a fix for say 10_1 and then do another against CS10 do I go back to el10 20:14:46 smooge: yes 20:14:49 yes 20:14:55 option 3 is option 2 without the explicit upgrade sorting via tilde 20:14:59 *tilde versioning 20:15:21 I do like that explicit sorting 20:15:26 option 3 is also what rhel does today 20:15:32 I like option 3 personally, but all of them would work and I don't feel super strongly 20:15:55 as I've said before, I like option 2 because it automatically makes the Fedora workflow and disttag modeling match for EPEL 20:16:12 that is, merging downward (Fedora) instead of merging upward (RHEL) 20:16:15 as does option 1 20:16:40 I should clarify, the reason why option 3 works for RHEL is because they build at the lowest and merge upward 20:16:57 when they don't do that, it requires extra release bumps and such 20:17:17 or more often, they only build for the leading branch and don't build for the already released branches due to the extra red tape 20:17:21 .hi 20:17:22 jonathanspw: jonathanspw 'Jonathan Wright' 20:17:30 Hi jonathanspw 20:17:31 carlwgeorge : \ 20:17:33 carlwgeorge: yup that too 20:17:45 hi jonathanspw 20:17:50 sorry I'm late. Opened the channel 10m early then got distracted. 20:18:11 I proposed option 2 because I like symmetry and we have nice rpm features to make it work without pain 20:18:27 been there jonathanspw. "i'll look at this real quick..." 20:18:33 any other questions or clarifications before we vote? 20:18:52 I'm ready 20:19:00 #link https://rankedchoices.com/epel10disttags 20:19:50 q: can anyone vote? e.g., am I allowed? 20:20:12 Is the top choice my preferred one? 20:20:23 our rules for voting are a bit convoluted, but i think the answer is yes 20:20:33 gotmax: yep. top down, and you can remove ones if you just don't want to vote at all for that option 20:21:01 yeah this didn't sort them in the order i submitted them 20:21:23 carlwgeorge: I thought you were just being nice and putting the one you submitted on the bottom. 20:21:41 so it's randomized, because my original suggestion is second for me 20:21:57 If I had to guess, I'd say it probably randomizes the order 20:22:04 ah 20:22:06 beat me to it 🙂 20:22:13 yeah I think it randomizes 20:22:20 if people want i can start over and put "option 1" etc in the text, or just read carefully 20:22:43 the wording is pretty clear IMO 20:22:44 or we can just vote here for our order? 20:22:51 yeah I think it's fine 20:22:53 here I thought carl was just making mine look like the bad option at the end :P 20:22:54 tdawson: do you agree on interpretation of the voting rules that basically anyone attending the meeting can vote? 20:23:12 you're supposed to drag and drop and you can delete items you don't want considered at all 20:23:16 it's actually a nice voting app 20:23:22 * Eighth_Doctor bookmarks for the future 20:23:34 carlwgeorge: Yes 20:23:49 * dherrera bookmarks too 20:24:02 For this one in particular, I think it's good to get a larger view of what people want. 20:24:24 Just make sure you only vote once. 20:25:41 so it seems Eighth_Doctor's tilde suggestion won? 20:26:19 Seems so, at least so far! 20:26:24 wait, how do we see the results? 20:26:31 https://rankedchoices.com/results#epel10disttags 20:26:31 https://rankedchoices.com/results#epel10disttags 20:26:33 https://rankedchoices.com/results#epel10disttags 20:26:38 first :) 20:26:53 uhh, i don't think we have 17 people attending this meeting 20:27:12 it was obviously Eighth_Doctor padding his option /s 20:27:13 It probably wou;d 20:27:15 voops 20:27:16 I wonder if some people clicked multiple times by accident 20:27:17 exclude the ones taht didn't put a name in? 20:27:31 Ya ... although I said hello to alot of people, I don't think it was that many. 20:27:36 but then it won't calculate the ranked choice properly 20:27:46 re-vote w/ names, i guess is one way 20:27:52 yeah 20:27:59 make name a required field if possible 20:28:01 neil: do you know if i can reset this one or do i have to create it again 20:28:01 can you make it require names? 20:28:01 yeah, I think we need to do that :/ 20:28:03 I think I see the flaw in this app 20:28:08 it can be made a required field 20:28:10 i'm not 100% sure 20:28:56 tdawson: lets put a pin in this while i recreate the poll, move on to the next topic and come back to this one 20:29:30 carlwgeorge: Sounds good. 20:29:45 carlwgeorge: If there is a way to make the name non-optional, that would be good. 20:30:05 i didn't see a setting for that but i'll look again 20:30:43 voter names required is an advanced setting 20:30:48 carlwgeorge: check the box for 'Advanced settings' 20:30:49 :nod: 20:30:53 you can also make it anonymous 20:31:03 and/or make the voters hidden 20:31:11 also it seems doesn't render in the result, i'll substitute numbers 20:31:13 We usually don't vote anonymously in meetings 20:31:23 smells like an XSS 20:31:23 neil: nice, thanks 20:31:29 So, other Old Business. EPEL2RHEL 20:31:31 yeah i don't think this needs to be anonymous 20:31:56 We got our few EPEL2RHEL with the new wording. 20:32:38 RHEL 9.2 and 8.8 are out, so now those are being proc'd right? 20:32:39 That unblocks me for two things. 1) putting in another request so we don't have 4 emails for each bug. 20:32:50 yeah, that would be nice :) 20:33:04 2) Starting to write a script that automatically checks things. 20:33:15 Eighth_Doctor: Correct. 20:33:38 tdawson: what do you mean by "checks things"? 20:33:40 I was able to add mpv to EPEL 9 because of RHEL 9.2's release :) 20:33:46 https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2023-c1fb825f3e 20:33:46 🎉 20:34:05 ðŸĨģ 20:34:09 Ha ... ya, you know, that whole conversation I just had in my head = "check things" ... 20:34:59 "check things" = See if a package is now in RHEL, and if it is, remove it from epel, or at the very least, tell someone. 20:35:25 so you mean a script to mass retire everything? 20:35:48 gotmax: Yes, exactly. You worded it much better than me. 20:36:38 You can take a look at https://git.sr.ht/~gotmax23/fedora-scripts/tree/main/item/go-sig/retire.sh and https://git.sr.ht/~gotmax23/fedora-scripts/tree/main/item/go-sig/close-rawhide-bzs.py that I used for Changes/Mass_Retire_Golang_Leaves 20:36:56 Ooooh ... thanks. 20:37:09 .hi 20:37:11 michel: michel 'None' 20:37:14 Sorry I'm late 20:37:18 Argh 20:37:21 .hello salimma 20:37:22 michel: salimma 'Michel Alexandre Salim' 20:37:29 Hi michel 20:37:35 its ok you can vote 4 times like the rest of us soon 20:37:38 you didn't miss anything 20:37:49 The first script clones the distgit repositories and retires the packages and then calls the second one to close the Bugzillas with a helpful message 20:38:00 One person, four votes! 20:38:36 "vote early, vote often" :-) 20:38:42 lol 20:38:46 gotmax: Cool ... that will help trememdously. 20:39:03 tdawson: i'm ready when you're done with whatever topic we're in 20:39:17 carlwgeorge: sounds good. I said all I needed to say. 20:39:27 names are now required and the choices say "option X" to correspond with the descriptions in the thread 20:39:27 You'd want to change it to clone the correct branch (i.e. not rawhide), close the Bugzillas for the correct Product and Version, and change the message to something about how users should refile any bugs with RHEL 20:39:38 also i got rid of the < and > characters that were tripping things up 20:40:11 what is the link? 20:40:16 #link https://rankedchoices.com/epel10disttag 20:41:02 oh neat, secretness 20:41:29 hopefully I voted properly :) 20:41:47 the ballots won't be released until 5:30pm :o 20:42:00 "The ballot you selected will not have the results released until May 17th 2023, 2:30 pm America/Los_Angeles Time" 20:42:09 No ... it says 2:30 pm 20:42:20 :) 20:42:23 "The ballot you selected will not have the results released until May 17th 2023, 5:30 pm America/New_York Time" 20:42:28 it corrects for your timezone 20:42:34 oh one of the advanced options was to set a cut off time, i wasn't trying to delay the results 20:42:37 it's very smart 20:42:51 Well sure ... if you want to live in New York. :) 20:43:09 or Florida ;) 20:43:16 carlwgeorge: Can you at least see the results? 20:43:19 or North Carolina :) 20:44:04 no, the results are also hidden 20:44:13 oh I guess the question is if that setting can be changed 20:44:40 yes 20:44:49 let me see if i can adjust the results visibility 20:45:02 * gotmax hopes he voted on the new one instead of the old one twice 20:45:25 * carlwgeorge sees gotmax in the results 20:46:20 hmm, no way to change settings now it seems. i guess setting a cut off time was a mistake. 20:46:38 well, I guess we'll find out next week ðŸĪŠ 20:46:44 Yep :) 20:46:59 I don't know why, but it's sort of fun not knowing the results. 20:46:59 we find out in less than an hour, actually 20:47:09 * Eighth_Doctor hopes his proposal wins 20:47:10 yeah, makes it feel like a real electino 20:47:15 is there any reason why voting can't here like any other normal proposal? 20:47:19 should I be campaigning? 20:47:38 * neil hands out stickers 20:47:40 wait, i found it 20:47:50 I guess most proposals are yes/no while here there are multiple options 20:47:56 yselkowitz[m]: Because having three choices makes it complicated. 20:48:04 yselkowitz: we couldn't agree enough last time that we resorted to this 20:48:04 and... we either do approval voting or ranked choice, both are a pain to do by hand 20:48:19 ah the time changed! 20:48:26 The ballot you selected will not have the results released until May 17th 2023, 1:50 pm America/Los_Angeles Time 20:48:38 yup! 20:48:39 I see 4:50pm now 20:48:44 two minutes! are we doing exit polls? 20:48:47 so 2 mins? 20:48:59 the other one was our exit polls 20:49:05 if we go by that, I'm winning :P 20:49:10 i just stripped off the cutoff, it weirdly kept logging me out 20:49:31 > option 1: .el10_MINOR in all branches has exceeded the quota and is elected. 20:49:41 #link https://rankedchoices.com/results#epel10disttag 20:49:46 aww 20:49:56 I didn't campaign hard enough 20:50:03 lost by two votes 20:50:05 13 voters sounds more realistic 20:50:29 smooge changed his vote too :P 20:50:32 and with names we know who to tar and feather for voting wrong 20:50:36 :D 20:50:38 woah, the results are not secret 20:50:44 * Eighth_Doctor shrugs 20:50:49 the names I mean 20:50:50 well I tried for symmetry 20:50:54 it's a technical poll, it doesn't need to be anonymous 20:50:55 we are all winners! 20:51:01 lol 20:51:09 I feel... defeated 20:51:27 mucho saddo 20:51:42 * Eighth_Doctor blinks 20:51:44 meh 20:51:50 I was ok with either 1 or 2, but I prefered option 2. 20:51:54 oh well, at least the bikeshedding is over 20:51:54 it'll be one of those little annoyances in life 20:52:05 I just realized my second choice is actually the one that is most like Fedora, so I'm glad that's the one that won 20:52:07 I'll try again for EPEL 11 20:52:08 Eighth_Doctor: You might find that you like it in the end. 20:52:12 :) 20:52:14 tbh these are all fine in practice 20:52:24 tdawson: I already don't like it in RHEL 20:52:30 yeah, time to stop bikeshedding 20:52:34 Davide Cavalca: option 3 would be a problem, the other two technically work 20:52:38 davide: agreed, they were all acceptable enough 20:52:52 meh 20:52:59 we're done, and it'll get rolled out for EPEL 10 20:53:05 we'll see how it goes 20:53:18 #info EPEL 10 dist tags voted to be .el10_MINOR in all branches 20:53:54 Ya ... that part of epel 10 is decided. 20:54:03 Eighth_Doctor: for EPEL 11 maybe we can do emoji 20:54:09 oh god no 20:54:11 OK, time is close, moving to Open Floor. 20:54:12 :D 20:54:17 I don't want to think about sorting emoji 20:54:21 #topic General Issues / Open Floor 20:54:23 i'm still hoping for more feedback on the other bikeshedding thread 20:54:38 easy, you just sort them by their code point in hex 20:54:41 #chair gotmax 20:54:41 Current chairs: carlwgeorge dcavalca dherrera gotmax gotmax23 nirik pgreco salimma smooge tdawson 20:54:47 carlwgeorge: Ya, I don't think I've contributed enough to the other one, it's more complicated. 20:54:57 have a link handy? 20:55:03 https://discussion.fedoraproject.org/t/epel-10-bikeshedding-pointing-users-at-the-right-repo/80551 20:55:28 Hmm, I was going to #info my Ansible announcement, but now I don't see it 20:55:30 tldr, if dnf upstream will give us a releasever_minor variable we can do some neat stuff 20:55:59 we can also hack it up with file triggers, but i'm skeptical of that approach 20:56:21 s/dnf/dnf5/ 20:56:30 Oh ... I haven't made an epelrelease variable yet ... for a prototype. 20:56:42 Can someone ack my `Ansible in EPEL 8` epel-announce post before the meeting ends :)? 20:57:00 can someone poke the dnf team and get a sense of whether https://bugzilla.redhat.com/show_bug.cgi?id=1789346 is likely to happen or not? 20:57:07 that will inform what to do here 20:57:26 gotmax: I thought I did ... looking again. 20:57:37 Davide Cavalca: I'll poke them 20:57:44 i poked it back in october, someone else should poke it next so the interest looks higher 20:57:56 thanks 20:58:01 How about changing the component to dnf5? 20:58:16 we can also bring this up in Brno in person for maximum poking 20:58:28 both are good ideas 20:59:05 gotmax: your Ansible in EPEL 8 isn't in the epel-announce hold queu 20:59:19 I thought I acked it eariler? 20:59:48 except... yeah, I don't see it. weird. 20:59:54 I don't see the EPEL 8 one in https://lists.fedoraproject.org/archives/list/epel-announce@lists.fedoraproject.org/2023/5/ 21:00:24 gotmax: Maybe resend it. 21:00:26 fwiw this bug was filed before dnf5 was announced, so the component was correct originally 21:00:41 right 21:00:58 Thank you all for comming. It was nice having such a large crowd today. Thank you for the good discussions, and for all the work ya'll do for EPEL and it's community. 21:01:10 Talk to ya'll next week. 21:01:18 #endmeeting