21:57:23 <tdawson> #startmeeting EPEL (2020-11-13)
21:57:23 <zodbot> Meeting started Fri Nov 13 21:57:23 2020 UTC.
21:57:23 <zodbot> This meeting is logged and archived in a public location.
21:57:23 <zodbot> The chair is tdawson. Information about MeetBot at http://wiki.debian.org/MeetBot.
21:57:23 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
21:57:23 <zodbot> The meeting name has been set to 'epel_(2020-11-13)'
21:57:25 <tdawson> #meetingname epel
21:57:25 <zodbot> The meeting name has been set to 'epel'
21:57:26 <tdawson> #chair nirik tdawson bstinson pgreco carlwgeorge
21:57:26 <zodbot> Current chairs: bstinson carlwgeorge nirik pgreco tdawson
21:57:26 <michel_slm> <Eighth_Doctor "Michel Alexandre Salim: well, I "> the actual Kokomo, not the Beach Boys song :)
21:57:27 <tdawson> #topic aloha
21:57:33 <michel_slm> .hello salimma
21:57:34 <zodbot> michel_slm: salimma 'Michel Alexandre Salim' <michel@michel-slm.name>
21:57:50 <pgreco> hi everybody (officially this time) ;)
21:58:06 <tdawson> Hi pgreco michel_slm carlwgeorge Eighth_Doctor
21:58:42 <carlwgeorge> howdy all
21:58:49 <carlwgeorge> *yall
21:58:58 <Eighth_Doctor> .hello ngompa
21:58:59 <zodbot> Eighth_Doctor: ngompa 'Neal Gompa' <ngompa13@gmail.com>
21:59:09 * Eighth_Doctor doesn't do meetings very often from here :)
21:59:19 <Eighth_Doctor> I'm too lazy to walk across the hall :P
21:59:42 <michel_slm> how's everyone?
22:00:04 <tdawson> I'm doing pretty good ... other than forgetting about DST
22:00:04 <Eighth_Doctor> weary :)
22:00:19 <Eighth_Doctor> every single meeting this week has been disrupted by DST
22:00:28 <carlwgeorge> ready for this week to be over, and next week too
22:00:40 <pgreco> tdawson: happy that you're back, I was a bit tense running the meetings :)
22:00:53 <Eighth_Doctor> my Raspberry Pi 400 kit just shipped though :)
22:01:08 <michel_slm> is that 100 Pi 4 in a Beowulf cluster? ;)
22:01:10 <pgreco> Eighth_Doctor: don't toy with me, hehe, I want one!
22:01:12 <bstinson> hey all
22:01:26 <michel_slm> hello! wow we have a full house today
22:01:28 <tdawson> 400 ?  Was I gone that long?
22:01:29 <michel_slm> wait, Nirik not here yet
22:01:38 <tdawson> Hi bstinson
22:01:56 <michel_slm> they missed an opportunity of making it 404
22:02:02 <Eighth_Doctor> they really did
22:02:16 <Eighth_Doctor> Michel Alexandre Salim: I don't _ever_ want to see RPi Beowulf clusters
22:02:20 <Eighth_Doctor> that would be terrifying
22:03:38 <tdawson> Well, while we wait for nirik, let's get to the first topic
22:03:53 <tdawson> #topic UTC vs DST for meetings
22:04:00 <Eighth_Doctor> DST
22:04:17 <Eighth_Doctor> the problem I'm having is that I'm suffering due to all the meetings being pinned to UTC
22:04:43 <Eighth_Doctor> my schedule is so full now that I have no room left
22:04:51 <pgreco> I'm ok with DST, unlike some people, I know when I'm outvoted :)
22:05:04 <tdawson> At this time slot, my schedule is fine either way.  But I'd prefer DST.
22:05:38 <carlwgeorge> doesn't matter to me, my friday afternoons are typically wide open thankfully
22:05:39 <tdawson> And to ble clear, when we are saying DST, that means this meeting started pretty close to DST (Troy time)
22:05:47 <Eighth_Doctor> haha
22:05:57 <michel_slm> I don't care either way as long as the calendar is accurate
22:05:58 <Eighth_Doctor> carlwgeorge has seen my calendar before, so he knows exactly how full it is
22:06:25 <michel_slm> if Neal's meetings are mostly UTC though, doesn't keeping this in UTC help?
22:06:33 <michel_slm> so meetings stay at the same relative offsets from each other
22:06:49 <Eighth_Doctor> Michel Alexandre Salim: no, most of my meetings are US/ET
22:06:52 <michel_slm> (mind you, I do want to keep my lunch time non-illusory so 1pm PT is a bad time for this meeting)
22:07:04 <Eighth_Doctor> my community meetings are a mix of US/ET and UTC
22:07:21 <Eighth_Doctor> all the UTC meetings are hell right now, because we're shifting them anyway
22:08:08 <carlwgeorge> logically it seems to make the most sense to put as many meetings on utc as possible so they don't move, and petition the daylight savings ones to switch too
22:08:13 <Eighth_Doctor> at least one community meeting now sits in lunch now
22:08:38 <Eighth_Doctor> so now I don't get to eat because of back to back meetings every other Tuesday from 9am to 4pm
22:09:21 <tdawson> So, what is the proper way to say the time of this meeting.  Does GMT shift with time zones, or should I just base it off US/ET ?
22:09:25 <Eighth_Doctor> carlwgeorge: I can't make work change to UTC, we operate officially in US/ET
22:09:31 <Eighth_Doctor> tdawson: base on US/ET
22:09:34 <pgreco> seems like sticking to US time
22:09:35 <Eighth_Doctor> GMT does not shift
22:09:53 <Eighth_Doctor> carlwgeorge: even our servers operate in US/ET
22:10:08 <Eighth_Doctor> (even the ones in the west coast... don't ask...)
22:10:27 <Eighth_Doctor> (and some of the European ones too...)
22:10:41 <carlwgeorge> "we're a global company, but operate on the timezone of our headquarters"
22:10:50 <carlwgeorge> not sure if that's true for datto, just poking fun :D
22:10:53 <michel_slm> that's the Google and FB problem too
22:10:56 <pgreco> servers are either in UTC or wrong :)
22:11:04 <michel_slm> carlwgeorge: does RH use UTC?
22:11:06 <tdawson> OK, so it sounds like we want the meeting at 1700 US/ET ... correct?
22:11:13 <carlwgeorge> RH doesn't do anything consistently
22:11:17 <pgreco> yeap,
22:11:31 <Eighth_Doctor> yup
22:11:31 * michel_slm swears my next job will be at a company that doesn't use monorepo and has sane server times
22:11:33 <michel_slm> carlwgeorge: oh no
22:11:42 <pgreco> carlwgeorge: I always say that about Arg, we're only consistent in our inconsistency
22:11:51 <Eighth_Doctor> carlwgeorge: I can promise you every single major tech company does not use UTC
22:11:53 <michel_slm> 5pm ET sounds good, yeah
22:12:02 <carlwgeorge> i would say a lot of rhel stuff standardizes on ET due to the offices in boston and raleigh
22:12:06 <tdawson> I don't really care that it's Red Hat's time zone, it's the farthest east U.S. timezone, so it's closest to Europe
22:12:10 <michel_slm> pgreco: hail Eris, goddess of discord
22:12:23 <Eighth_Doctor> 😆
22:13:06 <Eighth_Doctor> Michel Alexandre Salim: I don't have a monorepo, but we do use either ET or UTC depending on the infra/product stack
22:13:49 <Eighth_Doctor> fun fact, some of Google's APIs sends you crap in US/PT even when you don't want it to, and doesn't have an offset declared in those cases
22:14:11 <tdawson> #info EPEL Steering Committee Meetings will be at 5:00pm Eastern Standard Time
22:14:31 <tdawson> #topic Old Business
22:15:04 <tdawson> I was gone for a couple of weeks.  Did we have any items that we need to discuss?
22:15:12 * Eighth_Doctor shrugs
22:15:13 <tdawson> Or should I call out the old ones I had before?
22:15:25 <Eighth_Doctor> I don't think we had much in the way of meeting progress the past two weeks
22:15:27 <michel_slm> I have a sheepish update about fedpkg
22:15:31 <Eighth_Doctor> oh?
22:15:46 <tdawson> michel_slm: Go for it
22:15:50 <michel_slm> so I never checked Pagure, and Ondrej found a tiny error in my PR, fixed it and merged it
22:16:00 <michel_slm> and also did the commit to fix missing package.cfg
22:16:07 <Eighth_Doctor> so wait, does that mean we're set there now?
22:16:14 <michel_slm> so... I think there's no new release yet but the playground issues should be fixed. yeah.
22:16:17 <pgreco> michel_slm: that's a lot
22:16:23 <tdawson> Ya!!
22:16:31 <michel_slm> if someone knows his office I want to send him a beer
22:16:48 <tdawson> Don't need to be sheepish about that.
22:17:19 <Eighth_Doctor> I'm sure tdawson or carlwgeorge can hook up Ondrej with beer
22:17:24 <Eighth_Doctor> or at least a gift card for beer
22:17:36 <carlwgeorge> haha yeah rewards zone points
22:17:55 <carlwgeorge> link me the pr and i'll include it in the note
22:18:10 <michel_slm> https://pagure.io/fedpkg/pull-request/416
22:18:33 <michel_slm> and https://pagure.io/fedpkg/c/781036e7da46dab79f301e350e225a1073db55f6?branch=master
22:19:33 <tdawson> #info fedpkg commit to revert playground changes has been merged.  No build yet, but next build should have it in.
22:20:55 <tdawson> So, as soon as that get's built, playground transition is done.  Ya!
22:21:20 <carlwgeorge> hell yeah
22:21:42 * Eighth_Doctor is happy to celebrate the death of epel8-playground
22:21:48 <tdawson> epel8-next ... carlwgeorge was there any progress on that?  When I left we were still waiting, but I don't remember when it was going to start moving forward.
22:22:35 <carlwgeorge> did some initial tag work experimentation in stg koji
22:22:46 <tdawson> Well, the death of the autobuilds/autobranches for playground ... it's "not dead yet"
22:23:20 <carlwgeorge> now we're waiting on getting cs8 content mirrored (https://pagure.io/releng/issue/9850) and grobisplitted (https://pagure.io/releng/issue/9851)
22:23:51 <carlwgeorge> that will allow us to set up the external repos and actually test it
22:24:28 * Eighth_Doctor wishes we could actually use the modular content properly
22:24:46 <tdawson> Cool.  Things are in the works then.
22:25:31 <tdawson> carlwgeorge: Any help you need?
22:26:03 <carlwgeorge> i don't think so at this point
22:26:26 <carlwgeorge> while working on it with mboddu and nirik, we started talking about the repo file again
22:26:54 <tdawson> Which file is that?
22:27:06 <carlwgeorge> specifically if it's a good idea to ship it in epel-release, disabled by default.  the alternative would be a separate subpackage.
22:27:12 <carlwgeorge> epel-next.repo
22:27:21 <tdawson> The /etc/yum.repos.d/ ... ah, ok, I remember
22:27:51 <tdawson> I'm still of the opinion of a seperate sub-package.
22:27:53 <Eighth_Doctor> well, epel-next would be needed on stream, no?
22:27:55 * mboddu wondering this meeting is surely going on for a long time :)
22:28:05 <Eighth_Doctor> so a subpackage that explicitly requires centos-stream-release would make sense?
22:28:17 <carlwgeorge> i think shipping it disabled in epel-release is cleaner, and consistent with playground's repo file
22:28:41 <carlwgeorge> Eighth_Doctor: no, then it's not usable on rhel betas
22:28:56 <tdawson> mboddu: Ya ... well, it was run on "Troy Time" and first order of business was if that included daylight savings time.
22:29:05 <Eighth_Doctor> welp, then disabled in the main epel-release package makes sense then
22:29:20 <michel_slm> I'm looking forward to testing epel-next with RHEL 9 beta. want to make sure it actually works fine on laptops unlike 8 (bluetooth)
22:30:03 <carlwgeorge> first we got to get epel8-next done before we worry about epel9-next
22:30:15 <pgreco> if we do a separate package, it will have to provide epel-release somehow
22:30:30 <pgreco> because other packages may require it
22:30:54 <carlwgeorge> they would be satisfied by the actual epel-release.  they would be used together.
22:31:35 <pgreco> oh, ok
22:31:46 <Eighth_Doctor> I'm now in the "ship it disabled in epel-release" camp, because the only reason a subpackage makes sense is to tie it to stream explicitly
22:31:47 <carlwgeorge> i.e. you do rhel8 + epel8, or cs8 + epel8 + epel8-next
22:32:06 <carlwgeorge> or rhel8beta + epel8 + epel8-next
22:32:09 <michel_slm> right (re: epel8-next), it's just that it's nice if we pick a way forward that allows testing epel9-next with rhel9 beta as Carl suggested
22:32:54 <carlwgeorge> yes, i think creating epel9 and epel9-next at the same time makes sense
22:33:07 <Eighth_Doctor> well, there's no reason that once the el9 tree exists we wouldn't start making epel9
22:33:28 <Eighth_Doctor> we wouldn't have the same handicap that existed in el8 where the rhel8 beta was completely, utterly useless
22:33:59 <tdawson> Hay hay ... I created that beta ... it wasn't useless to me. :)
22:34:03 <carlwgeorge> shall we vote on the repo file approach we're gonna use?
22:34:15 <carlwgeorge> once we have a decision i can do that pr and have it ready to go
22:34:23 <Eighth_Doctor> tdawson: if you saw my bug report about el8 beta, then you'd know :P
22:34:37 <Eighth_Doctor> but sure, it wasn't totally useless for some folks, just me :)
22:34:39 <tdawson> carlwgeorge: Do you know which way nirik was leaning about the files?
22:35:15 <carlwgeorge> iirc him and mboddu agreed a separate subpackage didn't make sense
22:35:25 <tdawson> Eighth_Doctor: I'm litterally still picking RHEL8 beta bug reports out of my email ... just got another reminder of something yesterday.
22:35:40 <Eighth_Doctor> hah
22:35:42 <mboddu> Yeah, better go with with a disabled file in epel-release package
22:36:00 <carlwgeorge> modularity kinda rocked everyone's world
22:36:45 <tdawson> I'm not feeling strongly about the seperate sub-package, but it's how I would do it.  But if everyone else wan't it disabled in the epel-release, I'm fine with that.
22:36:47 <mboddu> tdawson: Correct if I am wrong, epel7 was created before RHEL 7 was announced, right? As in after rhel7 beta was out?
22:37:16 <pgreco> separate repo disabled would make it easier to use I guess
22:37:36 <pgreco> even --enablerepo=epel8-next for picking and choosing specific updates
22:37:39 <tdawson> mboddu: That I don't know.
22:37:45 <pgreco> from the beta I mean
22:37:56 <carlwgeorge> mboddu: i think so, i think it was called epel7-beta at first
22:38:06 <rsc> mboddu: Yes, EPEL 7 beta was created with RHEL 7 beta or so.
22:38:35 <mboddu> carlwgeorge, rsc : Thanks :)
22:38:46 <mboddu> tdawson: ^^ now we know :)
22:39:20 <tdawson> Votes: subpackage: tdawson pgreco  no-subpackage: nikik mboddu Eigth_Doctor
22:40:01 <pgreco> tdawson: I'm with no supbackage, just separate repo
22:40:05 <carlwgeorge> no vote for me? :P
22:40:09 <pgreco> I may have misspoken
22:40:10 <tdawson> carlwgeorge: michel_slm: I dont' know if I've heard which you prefer on the epel8-next repo files.
22:40:26 <Eighth_Doctor> carlwgeorge: is for no subpackage
22:40:27 <mboddu> tdawson: Whats the advantage of subpackage?
22:40:29 <carlwgeorge> i would prefer disabled in epel-release, no separate supackage
22:40:54 * mboddu scrolling back in case I missed the advantage of a subpackage
22:41:11 <tdawson> mboddu: I just know that I do alot of --enablerepo=* and if ti's installed but disabled, I run into issues.
22:41:24 <michel_slm> tdawson: I don't have enough context to say either way so I'll abstain
22:41:57 <carlwgeorge> i routinely do `--disablerepo \*` but i don't think i've ever done `--enablerepo \*`
22:42:20 <Eighth_Doctor> I don't think I've ever done `--enablerepo=\*` either
22:42:30 <mboddu> How about a subpackage that doesn't require anything? People can just install it if they want to use epel next
22:42:35 <tdawson> carlwgeorge: Looks like 5 to 1.  And I don't have a strong vote for sub-package.
22:42:35 <Eighth_Doctor> that seems kind of crazy since we have debuginfo and source repos also installed by default
22:42:54 <pgreco> Eighth_Doctor: that's what I was thinking about
22:43:09 <tdawson> Ya ... I sorta move the debug and testing out of the way first thing. :)
22:43:15 <pgreco> I think I only use --enablerepo=* for clean
22:43:27 <pgreco> tdawson: then you can move -next out too in that step ;)
22:43:34 <tdawson> Yep
22:43:43 <carlwgeorge> pgreco: `rm -rf /var/cache/dnf/*` does that faster :D
22:43:46 <tdawson> Which is why I don't feel that strongly about it.
22:44:19 <carlwgeorge> we can always change course if we change our minds or find some issue with it, but i'll do the initial pr as a disabled repo in the main package
22:45:04 <pgreco> carlwgeorge: I'm a little worried about running sudo and rm in the same command :)
22:45:22 <tdawson> #info epel8-next will have it's repo config file disabled in epel-release
22:45:33 <carlwgeorge> sudo?  you don't just do everything as root already? :P
22:46:01 <pgreco> always :P
22:46:07 * michel_slm wonders if anyone ever had sudo as their unixname
22:46:24 <tdawson> *laughs*
22:46:24 <carlwgeorge> 99.99% of my sudo commands are just `sudo -i` to get a root shell
22:46:46 <cyberpear> `sudo -i`++
22:46:48 <Eighth_Doctor> hehe
22:46:52 <Eighth_Doctor> indeed, same here
22:47:03 <Eighth_Doctor> I used to do `su -l` and now I do `sudo -i`
22:47:23 <carlwgeorge> very rarely when i need root privs is it for a single command, i almost always want to run a few commands
22:48:00 <tdawson> I just always have a terminal open as root ... labeled very distinctly ROOT
22:48:05 <pgreco> I normally have some passwordless commands that I run (like dnf), and then try to use sudo as much as possible
22:48:22 <tdawson> Anyway ....
22:48:25 <tdawson> #info EPEL-6 is End of Life in 2020-11-30. It will be moved to archives in 2020-12
22:48:27 <tdawson> #info THIS IS NOT A DRILL - 2 weeks left
22:48:46 <carlwgeorge> SOON
22:48:49 <michel_slm> Thanksgiving present
22:48:49 <tdawson> I've sent out my last warning email .... next one will be the real thing.
22:48:59 <michel_slm> (not for sysadmins who still have EL6 I guess)
22:49:12 <pgreco> I can neither confirm nor deny...
22:49:46 <tdawson> #topic EPEL-7
22:49:57 <carlwgeorge> have you met my friend RHEL 6 ELS?
22:50:04 <tdawson> CentOS 7.9 came out ... Ya!!
22:50:25 <Eighth_Doctor> and now I expect nothing to change ever again
22:50:25 <pgreco> yeap, the last one!
22:50:28 <tdawson> carlwgeorge: I think that's your imaginary friend ... there is no such thing. :)
22:50:41 * Eighth_Doctor wards away the evil that is ELS
22:50:54 <rsc> ELS solves the 2020-11-30 thing :)
22:50:59 <carlwgeorge> https://access.redhat.com/support/policy/updates/errata#Extended_Life_Cycle_Support begs to differ
22:51:03 <tdawson> Next week (tuesday I'm hoping) I'm going to do my "will it install" and start filing bugs.
22:51:06 <Eighth_Doctor> ugh noo
22:51:12 <Eighth_Doctor> haha
22:52:00 <pgreco> surely there will be some people to fix those problems, oh wait....
22:52:14 <pgreco> we can fix those now ;)
22:52:26 <tdawson> Ya, hopefully EPEL7 will remain stable and there will not be a 7.10
22:52:30 <Eighth_Doctor> I'm okay with letting the clock run out on my EPEL6 packages
22:52:49 <Eighth_Doctor> and as for EPEL7, nobody's said anything so far :)
22:53:08 <pgreco> nope, we'll know more next week I guess
22:53:17 <pgreco> do we have a bugzilla group for epel sig?
22:53:36 <michel_slm> epel-sig or epel-packagers-sig?
22:53:40 <michel_slm> the latter, yes
22:53:50 <pgreco> yes, that's what I meant
22:54:07 <tdawson> Oh, time is short ...
22:54:09 <pgreco> so we can be aware of what is going on with those issues
22:54:18 <tdawson> #topic EPEL-8
22:55:06 <tdawson> I wanted to thank smooge and nirik and possiby mboddu ... for archiving EPEL8 off  on our usual archives/epel/8.2
22:55:33 <tdawson> I guess it's a snapshot, not an archive ...
22:55:58 <tdawson> #topic General Issues / Open Floor
22:56:11 <michel_slm> https://pagure.io/fedora-infrastructure/issue/9427
22:56:14 <mboddu> All the archiving (snapshots) are done by smooge mostly
22:56:37 <michel_slm> for epel-packagers-sig, pingou looked at our request that collaborators should be allowed to request branches
22:57:02 <michel_slm> it's marked as medium gain, high effort -- if people could take a look and see if we can help make that happen, that'd be great
22:58:11 <pgreco> does the move to gl affect this?
22:58:20 <Eighth_Doctor> it eliminates that possibility entirely
22:58:26 <michel_slm> GL? ugh gitlab
22:58:33 <Eighth_Doctor> but GitLab move isn't happening anytime soon
22:58:34 <tdawson> michel_slm: Thanks for bring this up, and working on that.  I totally skipped it.
22:58:37 <michel_slm> wait, is that definite?
22:58:46 * Eighth_Doctor shrugs
22:58:57 <Eighth_Doctor> nobody knows, but given that there's literally no plan on how to do literally anything...
22:59:15 <pgreco> ‾\_(ツ)_/‾
22:59:41 <Eighth_Doctor> and pretty much everything we hear is "welp I guess CPE is magically going to become master Gophers and Rubyists _and_ write pkgdb3 for GitLab"
22:59:59 <Eighth_Doctor> 🤦‍♂️
23:00:19 <rsc> With each pkgdb rewrite, some things are left behind ;-(
23:00:35 <Eighth_Doctor> we're now at a point that we have more features in pagure than we did in pkgdb2+cgit
23:00:39 <michel_slm> and on that bright note... it's 6PM
23:00:49 <tdawson> Looking at his first point ... I thought there was a different issue specifically making it so collaborators could be given branches they have access too ...
23:00:51 <rsc> Not sure why we need to migrate the framework every few years, it doesn't get much better each time.
23:00:59 <michel_slm> <Eighth_Doctor "we're now at a point that we hav"> so of course it's time to start over
23:01:07 <Eighth_Doctor> 🤦‍♂️
23:01:17 <tdawson> Thank you everyone for coming.  Especially for such a long meeting.
23:01:27 <michel_slm> <tdawson "Looking at his first point ... I"> yeah, I'm not sure why a new endpoint is needed
23:01:32 <Eighth_Doctor> rsc: because CPE didn't want to add a req to work on the codebases
23:01:33 <carlwgeorge> feels like we've been chatting for 2hrs lol
23:01:33 <michel_slm> but I'm not familiar with pagure internals
23:01:50 <tdawson> I'll talk to ya'll next week ... hopefully for just one hour this time. :)
23:01:55 <pgreco> thanks tdawson for taking the reigns back from me ;)
23:01:55 <Eighth_Doctor> rsc: apparently spending ~$4 million a year is easier than hiring a couple of people
23:02:12 <tdawson> #endmeeting