22:02:10 #startmeeting EPEL 22:02:10 Meeting started Fri Feb 5 22:02:10 2021 UTC. 22:02:10 This meeting is logged and archived in a public location. 22:02:10 The chair is tdawson. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:02:10 Useful Commands: #action #agreed #halp #info #idea #link #topic. 22:02:10 The meeting name has been set to 'epel' 22:02:31 ah. we keep messing up the Server meeting name so I wasn't sure ;) 22:02:45 Overall Plan: 22:02:46 - epel-next is an epel branch that is built against CentOS Stream. It 22:02:48 only has the packages that would be incompatible with released RHEL 22:02:49 builds, or if an EPEL maintainer is updating a package that will only 22:02:51 be released to regular EPEL at the next RHEL release. 22:02:52 - We plan on creating epel9-next when CentOS 9 Stream has a public 22:02:54 repository. We plan on using the EPEL Packaging SIG to populate it 22:02:55 early with common packages, although any EPEL package maintainer can 22:02:57 add their packages whenever they want. 22:02:58 - We plan on creating normal EPEL9 after the RHEL 9.0 GA release, just 22:03:00 like normal. No sooner. All epel9-next packages will be rebuilt on 22:03:01 EPEL9. They will not be "tagged over". 22:03:03 ugggg 22:03:04 sorry 22:03:08 no topic :) 22:03:19 * carlwgeorge waves 22:03:42 #chair nirik tdawson bstinson pgreco carlwgeorge michel_slm 22:03:42 Current chairs: bstinson carlwgeorge michel_slm nirik pgreco tdawson 22:03:51 #topic aloha 22:03:58 .hello salimma 22:03:59 michel_slm: salimma 'Michel Alexandre Salim' 22:04:04 Hi carlwgeorge 22:04:06 Hi michel_slm 22:04:11 Hi smooge 22:04:36 smooge: Did you want me to put you back on that list, or is your comming back just for a few weeks? 22:06:01 I have no idea.. lets just keep me as 'former pope smoogen' 22:06:13 lol 22:06:23 Ah ... I got my normal keyboard back ... I'm sorry for that spam above, I was trying a cut and paste, and it obviously went wrong. 22:06:44 I stay in my keep for the most part but come down to the Holy See just to see how you are keeping up things :P 22:06:48 Emeritus Chair 22:06:59 * nirik is sorta here, I guess 22:07:00 :) 22:07:03 Hi nirik 22:07:12 #topic Old Business 22:07:13 EPEL-Packaging-SIG 22:07:35 michel_slm: any change of status? 22:07:51 I didn't get the template done :( It was a busy week this week. 22:08:12 howdy! not yet, been busy with other things but I can look into enabling branch requests for collaborators after next Wednesday 22:08:16 same here 22:09:00 no more conferences after this weekend, and no more F34 deadlines after next Tuesday 22:09:11 i'm gonna sound like a broken record when i make excuses about not making progress on epel next 22:09:22 Ha! 22:09:36 sorry, hi guys, I"m only about 5% here 22:09:36 carlwgeorge: How is epel8-next coming along? 22:09:41 Hi pgreco 22:10:18 * michel_slm waves at pgreco 22:10:36 pgreco: heard at Dojo today that you're one of those working on alternate LTS kernels :D 22:10:37 no change from last week, fedpkg needs work, along with fedora-infra ansible playbooks. 22:10:44 hey, dealing with fires at work when my weekend should have started already 22:10:50 :( 22:10:53 then we can move forward with implementing epel8-next in production 22:11:08 michel_slm: yeah, I try to keep the 1 or 2 latest lts for centos 22:11:13 ok can I help any 22:11:27 I branch them out of fedora and maintain them for a year or 2 22:12:02 smooge: nothing to worry about, just thing that I'd rather not have to do today ;) 22:12:09 carlwgeorge, what in the playbooks need to look at 22:12:46 smooge: do you still have the link to the hackmd we were working off of? 22:12:58 ..... maybe 22:13:01 https://hackmd.io/@carlwgeorge/ry-qCXDYv last three links 22:13:17 distgit and repo2json roles 22:13:35 ok will look at on Monday 22:13:46 or this weekend while I am 'compiling' 22:14:45 yeah i plan to tinker with fedpkg this weekend in between fosdem talks 22:15:06 s/in between/during/ 22:15:34 I failed at waking up much earlier so I'll probably only attend talks that start at/after 14:00 CET 22:16:04 oh crap i didn't even think about changing my sleep schedule to catch more talks 22:16:18 :) 22:16:33 * nirik is kinda online conferenced out... but will watch recordings later of interesting ones. 22:16:46 * carlwgeorge can relate 22:17:08 I find it really depressing to do online conferences 22:17:18 Anything else on epel8-next? 22:17:23 thinking that I should be freezing in Brussels right now.... 22:17:28 nah we can move on 22:17:55 OK ... the other thing from last week was "documentation changes" ... going from the wiki to fedora style docs. 22:18:20 I don't suppose anyone had any time to look at that? I know I didn't. 22:18:39 I looked a long time ago 22:18:56 the biggest issue is always 'which f*ing thing they are using this week' 22:19:19 I rewrote stuff in rst and got told to move it to markdown. I did markdown and now its asciidoc 22:19:32 I thought it had calmed down for at least a year. 22:19:43 that would be when I quit :) 22:20:06 :) 22:20:25 once i'm done with epel8-next stuff i plan to take a stab at converting the epel wiki guidelines into asciidoc and submitting a pr 22:20:36 in any case it comes down to taking a day off and doing it 22:20:43 carlwgeorge: cool. That would be great. 22:20:58 it helps that no one is opposed to that change :D 22:21:11 i think you should to it in TeX 22:21:16 True ... it's really just a matter of getting someone to do the initial changeover. 22:21:27 old school TeX, not even LaTeX? you're mean 22:21:40 There is only the Lion book 22:22:16 :) 22:23:12 OK, moving on 22:23:13 actually ... my Lion book seems to have finally fallen apart completely. it is only half here 22:24:05 The other topic we talked about last week was getting ready for RHEL9 / CentOS 9 Stream 22:24:18 I sent an email out, about 5 minutes before this meeting. 22:24:44 I actually read it 22:24:53 thanks for doing that 22:24:55 WhooHoo :) 22:25:03 thanks for putting that together tdawson 22:25:11 I was hoping to get a summary out before the meeting, so we didn't have to re-hash what we'd already talked about. 22:25:27 Did it sound fairly accurate? 22:25:48 it did. 22:26:36 having it match to stream will make it go a lot faster because public feedback of 'how the f* do you expect this to f* compile?' can be done 22:27:04 Anything in it anyone wants to discuss? Or should we just do it via email for the next week or two *laughs* 22:27:19 what i've read so far looks accurate 22:27:25 I would say it probably needs some reading and digesting. 22:27:40 OK, then I'll put it on the topic for next week. 22:28:09 I think that's it for old business. 22:28:16 #topic EPEL-7 22:29:08 Anything for EPEL 7 this week? 22:29:46 I'll take the silence for a no 22:29:59 #topic EPEL-8 22:30:41 Any EPEL8 issues or things to discuss? 22:31:17 I have one question but not sure if it's EPEL8-specific, I can wait until open floor 22:31:44 Well, I don't have any EPEL8 stuff, so let's move on 22:31:48 I have no issues for EL7, EL8, and EL9 22:31:49 #topic General Issues / Open Floor 22:32:18 michel_slm: what's your question? 22:34:33 Must be a long question 22:34:39 oh 22:34:43 haha, sorry, was in another window 22:35:03 just wondering about the relationship between EPEL and other repos that SIGs might produce 22:35:30 epel (and epel-next) presumably can only depend on repos that ship in RHEL and their CentOS (Stream) equivalents, right? 22:35:43 Correct 22:35:43 but any additional repos can choose to depend on EPEL? 22:35:47 yes. 22:36:08 that must be a record for question:answer length ratio :p 22:36:13 most don't because they can't be assured their consumers do 22:36:23 ah. yeah, was wondering about Hyperscale 22:36:28 so instead they replicate it 22:36:51 this is where epel being in CentOS makes a bit more sense at times. 22:36:58 where I think new packages that don't conflict with RHEL/CS should go straight to EPEL, and only overrides and things that don't make sense without the Hyperscale repo should go to that repo 22:37:08 if it was over there in there build system then builds could be tagged into the sigs content 22:37:16 smooge: oh, is there any plan to make EPEL part of CentOS? 22:37:24 hahahahahahahahahhahahahahahahaahahahhahaahahaha 22:37:27 but that has a number of downsides too. 22:37:34 * smooge has been removed from the room 22:37:38 * michel_slm broke smooge 22:37:53 no longer building against rhel. Not being in the same place as fedora where maintainers are already maintaining things, etc. 22:38:16 maybe it'll be clearer once more of the infra get merged 22:38:44 we build epel against centos now, right? I see mock has both epel* and rhelepel* 22:38:47 there is nothing wrong with sigs shipping packages that don't exist anywhere else (i.e. only make sense with other hyperscale sig packages) 22:38:52 no we built it against RHEL 22:39:24 * smooge is given a nice electroshock and happy pills and comes back to play 22:39:31 michel_slm: mock configs use centos for their epel targets, but epel itself does not 22:39:42 *epel's koji mock configs 22:40:01 nirik, I agree on those which is where this breaks down also. 22:40:25 * nirik isn't sure whats actually happening with the 'more infra merged' anymore. ;) 22:40:44 nirik, however since multiple fedora maintainers have told me over the years what I can do with epel packages I don't see the second part as strong as I did 22:40:51 but thats likely beyond this' meetings scope 22:41:36 well, for each "I don't care about epel" there's lots more quietly just building it with their fedora packages. 22:41:39 carlwgeorge: yup. for things that only make sense in hyperscale, those packages shouldn't be in epel 22:42:05 yeah, the ones who care about epel tend to be the one that screams if they can't fast-forward update :) 22:43:04 in any case, for the CentOS Stream SIGs it causes a problem because they usually need to build in/tag against specific items in their koji 22:43:30 when i first heard about the idea of making centos upstream of rhel, my first thought was that rhel would be a subset of centos packages, and centos could have additional packages with community maintainers, obsoleting the need for epel. 22:43:47 now that i'm inside the machine, i realize there are people that will never allow that to happen. 22:44:11 carlwgeorge: huh, interesting idea... 22:44:39 centos stream couldn't even ship a devel repo (well, we did and got yelled at and had to empty it out) 22:45:01 but those were things in rhel... 22:45:24 * smooge asks for another round of treatment 22:45:55 Yep, that's stuff I don't even want to touch. 22:46:26 yeah.. it is things like that make one Emeritus 22:46:34 :) 22:46:52 Anything else before we close the meeting? 22:47:04 I'm all for having an extra 10-15 minutes of weekend. 22:47:28 carlwgeorge: yeah the idea that c9s and rhel are still built in parallel makes legal sense, but requiring them to match 1:1 is a bit sad 22:47:37 weekend! 22:48:14 Thanks everyone for being here. I'll talk to you all next week. 22:48:20 thanks tdawson 22:48:24 minor nit pick, we're trying to refer to stream as cs8 and cs9, despite what branch names would have you think 22:48:30 * smooge goes back to his mountain retreat 22:48:36 counterstrike! 22:48:59 cs9 instead of c9s? ok 22:49:00 i.e. the distro is "centos stream", the version is 8 or 9 (not 8-stream, again despite what mirror directories would have you believe) 22:49:07 carlwgeorge: Good to hear ... I was having a dicussion today on what to call it, and none of us knew. we'll go for cs9 cuz it's easeir to say. 22:49:08 yeah, c8s is a bit too close to k8s haha 22:49:14 yup 22:49:21 #endmeeting