18:01:19 <smooge> #startmeeting EPEL (2018-09-12)
18:01:19 <smooge> #meetingname epel
18:01:19 <zodbot> Meeting started Wed Sep 12 18:01:19 2018 UTC.
18:01:19 <zodbot> This meeting is logged and archived in a public location.
18:01:19 <zodbot> The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:01:19 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:01:19 <zodbot> The meeting name has been set to 'epel_(2018-09-12)'
18:01:19 <zodbot> The meeting name has been set to 'epel'
18:01:19 <smooge> #topic Chair and Introductions
18:01:19 <smooge> #chair avij bstinson Evolution nirik smooge
18:01:19 <zodbot> Current chairs: Evolution avij bstinson nirik smooge
18:01:25 <nirik> morning
18:01:39 <smooge> morning
18:01:44 <smooge> day
18:01:47 <bstinson> hi all
18:02:04 <avij> hi, for once I'm home at this time
18:02:13 <smooge> yay! we have quormets
18:02:43 <smooge> which leads to the next question
18:02:46 <smooge> #topic Move meetings back to weekly?
18:02:46 <smooge> #info different time/day?
18:03:18 <smooge> So I moved it to biweekly because we had a lot of empty meetings during the spring/summer.. but it became even worse by doing that
18:03:37 <smooge> I would like to find a better time/place and agenda format for future meetings
18:04:06 <nirik> yeah, dunno... seems like a lot of time we don't have much, but bi-weekly is always hard to recall which week is on
18:04:27 <avij> as for myself, I was travelling on many of those days and it wasn't always possible to connect to IRC from wherever I was
18:04:52 <smooge> so I think we should go back to weekly, but I would like to make sure we have a good time for avij to make. Should I set up a whenisgood? or something similar (what do people do these days?)
18:05:31 <smooge> I typed in the wrong name last time and ended up at  'hookup' website :/
18:05:31 <bstinson> i'm ok with a whenisgood
18:05:51 <bstinson> ...but not the hookup one
18:06:01 <Evolution> same.
18:06:08 <avij> this timeslot is generally OK. I try to be flexible.
18:06:33 <Evolution> the timeslot is fine for me. my availabilty is more problematic in terms of 'days' not hours.
18:06:54 * nirik is fine with this... but willing to move if others want to
18:07:38 <bstinson> this time is also good for me
18:08:01 <avij> and either weekly or biweekly, I don't have a strong preference regarding those
18:09:56 <smooge> ok sorry was mesmerized by whenisgood options
18:10:02 <smooge> I will send out an invite shortly
18:10:12 <smooge> #topic Emergency Business
18:10:12 <smooge> #info This is where items needing to be discussed right away go
18:10:21 <smooge> Do we have any options needing to be discussed this week?
18:11:06 <avij> nothing significant seems to have broken recently, as far as I'm aware
18:11:12 <nirik> I don't know of anything urgent
18:11:57 <smooge> http://whenisgood.net/zgt7jmc
18:12:17 <smooge> avij, thank you for answering various people's questions. The python2 this morning seemed to take a lot of time
18:12:31 <smooge> #topic EL-6 end of life
18:12:31 <smooge> #info EL-6 will be reaching its end of life in late 2020.
18:12:31 <smooge> #topic EL-7.6 beta
18:12:31 <smooge> #info time to update/remove packages
18:13:11 <smooge> Those should have been info's.. urg
18:13:57 <nirik> yeah. I haven't looked at 7.6beta much... but I can try to.
18:14:13 <smooge> So with EL-7.6 we have a lot of updates in packages again. I was wondering if I could do some sort of rebuild against it
18:14:15 <nirik> There's a ticket also with existing possible overlap packages...
18:15:14 <nirik> https://pagure.io/releng/issue/7802
18:15:36 <nirik> I really wish we had a list of those limited arch packages... makes it really difficult to fix overlaps.
18:16:26 <smooge> I am wondering if our doing limited arch is actually helping any.
18:17:03 <smooge> because the ones I found were usually done with 7.0 and never updated even though the current package is different
18:17:05 <nirik> I know if I had to do it again I wouldn't. ;) but the horse is out of the barn now.
18:17:18 <smooge> so I will add that to our EPIC
18:18:43 <smooge> nirik, I will get a list of all limited arch ones I can find.. it might not be all of them but it will hopefully get us some starting point
18:18:52 <smooge> I will add tht to that ticket
18:19:00 <nirik> yeah... we should try and maintain a list
18:19:34 <smooge> you know.. in some sort of program that could list product definitions
18:19:40 <nirik> some of those listed there are not right either... like ansible and the pidgin packages I think we added because rhel removed some api...
18:19:47 <nirik> heh
18:21:21 <avij> one thing regarding 7.6: I hope that if RH adopts some EPEL package again, I wish they would do it properly, ie. bump the release so that it'd be newer than what is in EPEL
18:21:24 <smooge> bstinson, do you think I could rebuild EPEL against CentOS in CBS? I want to get an idea about any build breakages we currently have
18:21:34 <nirik> avij: that would sure be nice. :)
18:21:59 <smooge> avij, yeah. most of the time they seem to pull it in 6-9 months ago.. so they might bump it from then
18:22:03 <avij> when 7.6 is still in beta, any possible conflicting issues with EPEL could still be fixed before GA
18:22:07 <pgreco> smooge, I've been building epel, almost from scratch for centos/armhfp
18:22:23 <smooge> cool
18:22:32 <pgreco> so I have some idea of that
18:22:38 <bstinson> we can likely rebuild it, but doing it in CBS is probably not the best place at this moment
18:22:46 <smooge> bstinson, ah ok
18:23:27 <smooge> I would do it in copr.. but it isn't the best place either :P
18:23:47 <smooge> I guess I need to ask for that home build system
18:24:08 <bstinson> yeah the thing would be to not make it look like a fork that's shippable
18:24:21 <smooge> ah got it.
18:26:04 <smooge> pgreco, have you run into any 'this is really broke' in the current EPEL using a newer CentOS?
18:26:22 <pgreco> some, yes
18:26:33 <pgreco> some I had to force older versions to actually build
18:27:00 <pgreco> I.E. kstars was failing due to kf5 upgrade
18:27:09 <pgreco> but rdieter fixed it today :D
18:27:15 <smooge> oh that would explain the .whoowns
18:27:25 <pgreco> yeap, exactly
18:28:30 <smooge> ok well I would be interested in getting a list as I would like to look at doing an update/rebuild as things are upgraded more in EL7
18:28:34 <smooge> thanks
18:28:39 <smooge> #topic EPEL/EPIC proposal
18:28:39 <smooge> #info http://smoogespace.blogspot.com/2018/05/blue-sky-discussion-epel-next-or-epic.html
18:28:39 <smooge> #info where do I file a change request?
18:28:39 <smooge> #info dealing with centos-extras conflicts? EPEL-supplemental?
18:28:40 <smooge> #info next
18:29:03 <pgreco> yes, I'm looking at my fixes to see what applies here
18:29:11 <pgreco> and what is armhfp dependant
18:29:54 <smooge> Alright I have a couple of feedbacks I am going to incorporate and then I am going to file a change request in the Fedora system for FESCO? Council to look over.
18:31:18 <nirik> was there a concrete list of what was planned? last I saw there were still questions? or most of those got figured out?
18:31:34 <Evolution> smooge: we could do a working day for EPEL at devconf.
18:31:53 <smooge> ok will put that in as my please send me and nirik to Europa
18:31:56 <Evolution> smooge: also, mattdm had a list of ideas he wanted to see with epel-next/epic
18:32:24 <Evolution> so may be worth asking him (I told him to talk to you)
18:32:53 <smooge> I will do so. Several of them were interesting, but would require some major storage
18:33:19 <smooge> or at least the ones we talked about earlier this summer
18:34:00 <smooge> Europe not Europa.. While I would love to go see Jupiter.. I don't think Red Hat will fund that
18:34:21 <nirik> why not? those moons look lovely!
18:34:36 <bstinson> FLOCK 2019
18:34:57 <smooge> nirik, the robot which runs the ship always seems to get a little cranky
18:35:23 <nirik> I'm afraid I can't do that dave
18:36:04 <smooge> ok I need to iron out definitive steps of what will be needed and I will ahve a couple of days without internet where I can possibly work on it due to Florence
18:36:53 <smooge> #topic Next Meeting
18:36:53 <smooge> #info http://whenisgood.net/zgt7jmc
18:36:53 <smooge> #info Next Scheduled for 2018-09-19
18:36:53 <smooge> #info Need agenda items
18:37:23 <smooge> Please mark the when is good if possible so next week we can have a new time if any marked
18:37:49 <smooge> It may be the same battime and bat-channel
18:38:10 <smooge> Are there any agenda items I need to put up for next week?
18:38:18 <pgreco> for next meeting (or open floor), I'd like to talk about my rebuild of epel for armhfp
18:38:38 <pgreco> interest/support/feedback
18:38:46 <nirik> hum... smooge: that whenisgood only has 3 hours wed and 3 hours friday... is that intended?
18:38:51 <smooge> OK would you prefer in 2 minutes or next week
18:38:55 <nirik> pgreco: cool. :)
18:38:56 <smooge> nirik, no..
18:39:16 <smooge> that is me not knowing what the henry what I needed
18:40:05 <smooge> nirik, please try now :)
18:40:13 <smooge> #topic Open Floor
18:40:24 <nirik> smooge: yeah, that looks better
18:40:40 <smooge> pgreco, we can talk about armhfp now or next week whichever you would like
18:40:51 <pgreco> now works for me
18:41:15 <pgreco> Arrfab had been rebuilding srpms from epel
18:41:26 <smooge> ok... you have the conch
18:41:50 <pgreco> but completely blind, without checking success/failure
18:42:06 <pgreco> I picked up from there and started bootstrapping/fixing
18:42:18 <pgreco> my current FTBFS is about 10 rpms
18:42:33 <pgreco> after removing the "exclusivearch" from the list
18:42:50 <smooge> cool. that is a lot less than I would have guessed
18:43:12 <pgreco> the thing is, some of the "fixes" were reapplying things from fedora
18:43:17 <smooge> it would be interesting to get a list of builds and rebuilds needed in the bootstrapping
18:43:27 <pgreco> others upgrading to a newer version
18:43:38 <pgreco> I can get that in a minute
18:44:51 <pgreco> https://paste.fedoraproject.org/paste/pPkCdCiAt~pdYM9eF8Jn7g
18:45:07 <pgreco> that is what I had to do some sort of bootstrapping
18:45:18 <pgreco> ghc being the hardest, by far....
18:45:48 <nirik> lots of langs... bootstrapping? ie, ghc, rust, mono...
18:46:34 <pgreco> nirik, to build rust, you need rust
18:47:00 <pgreco> same thing for mono
18:47:10 <nirik> yeah. bootstrapping.
18:47:17 <pgreco> it really burns your head, hehe
18:47:22 <nirik> yeah
18:47:49 <smooge> pgreco, was that link the FTBFS or the must rebuild multiple times from ancient fedora
18:47:56 <pgreco> and this is the list of "modified" somehow packages https://paste.fedoraproject.org/paste/eJn~D5sBIgtm06swyLT8yA
18:48:28 <pgreco> smooge, list 1 is "need to build first with changes, then rebuild normally"
18:48:35 <smooge> got it.
18:48:50 <smooge> wonders how many raspberry pis were killed to make texworks compile
18:49:12 <pgreco> second list is "not working, needs fixing", some fixes may be due to armhfp, others for deps
18:49:35 <smooge> got it
18:49:39 <smooge> thanks
18:49:45 <pgreco> hehe, there were some casualties in the process
18:49:50 <pgreco> now, to the point
18:49:58 <smooge> I think this list would be similar for the other OS people want i3865
18:50:12 <pgreco> smooge, hughesjr is working on that
18:50:33 <smooge> sorry I derailed your point
18:50:46 <pgreco> he is a little behind, because he started after, but the idea for him to use my experience in his rebuild
18:50:51 <pgreco> back to point
18:51:00 <pgreco> is there any interest from the maintainers in those fixes/changes
18:51:06 <pgreco> for armhfp/i686
18:51:19 <pgreco> to keep them all together within epel?
18:51:33 <avij> some maintainers may have interest, some may not
18:51:35 <nirik> I would think so... as long as they don't cause problems for x86_64/etc
18:51:52 <pgreco> or should we keep those on a separate and reapply?
18:52:11 <pgreco> most of the changes should be within %ifarch
18:53:50 <smooge> I agree with avij.. it will be mostly a package by package thing. Most of the packages I think you have made changes to I expect would be open to it
18:54:05 <pgreco> if there is interest, I could work with each maintainer (at least the ones who care) to make those changes
18:54:30 <avij> filing bugs with the proposed changes for each package would be a good first step, I'd think
18:54:41 <smooge> I think that would be good. It would be useful for future releases
18:54:59 <pgreco> avij, if there is a general interest to keep armhfp changes in epel, yes
18:55:12 <smooge> nirik, someone is at my front door. can you take over meeting?
18:55:29 <nirik> you could also file PR's for them...
18:55:41 <nirik> make it easy for people to take.
18:55:44 <nirik> smooge: sure...
18:56:00 <nirik> anyone have anything further? or should we call it a meeting?
18:56:40 <avij> some package owners may reply with something similar to "I have zero interest in epel build of cinnamon or it's deps", like in one of the bugs I filed. but I would believe most would be OK to accept patches.
18:56:58 <nirik> yeah
18:57:15 <avij> https://bugzilla.redhat.com/show_bug.cgi?id=1578521 is where that was (needs rebuild)
18:57:54 <pgreco> yeah, banshee has been broken since 7.4
18:58:01 <avij> another thing, very briefly, I tried to get people to enable epel-testing to catch any issues in EPEL packages. https://twitter.com/avij/status/1039855950193152000 and the related forum post.
18:58:03 <pgreco> same idea
18:58:20 <nirik> huh, that package is actually orphaned
18:58:37 <avij> and now I figured I'd make a note about epel-testing on https://wiki.centos.org/AdditionalResources/Repositories as well. I'll do that soon. that's all I had on this.
18:59:02 <nirik> anyhow, if you all just need one off fixes, I'm happy to just go make them as a provenpackager... not long term sustaintable, but might help for some of this.
19:00:11 <pgreco> nirik, I'll prepare some patches and ping you over #epel with them
19:00:23 <pgreco> if you want, and we'll see from there
19:00:37 <nirik> sure happy to help as time permits.
19:00:49 <nirik> ok, if nothing else we are over time... so will close out here...
19:01:02 <nirik> #endmeeting