18:00:21 <nirik> #startmeeting EPEL
18:00:21 <zodbot> Meeting started Wed Apr 27 18:00:21 2016 UTC.  The chair is nirik. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:21 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:00:21 <zodbot> The meeting name has been set to 'epel'
18:00:21 <nirik> #meetingname EPEL
18:00:21 <nirik> #chair smooge nirik Evolution bstinson avij
18:00:21 <nirik> #topic agenda gathering / quorum detection
18:00:21 <zodbot> The meeting name has been set to 'epel'
18:00:21 <zodbot> Current chairs: Evolution avij bstinson nirik smooge
18:00:26 <nirik> who all is around?
18:00:29 <nirik> for an epel meeting
18:00:31 <Evolution> present and-or accounted for
18:00:41 <avij> greetings
18:01:26 <nirik> Who has agenda items?
18:01:45 <bstinson> hi all
18:01:47 <Evolution> I'd like to bring up the altarch builds, mostly for notice purposes.
18:02:11 <nirik> ok.
18:02:31 * nirik was not sure if we were going to see a smooge today... he's still busy with moving...
18:03:20 <nirik> yselkowitz: you had some agenda item?
18:03:26 <nirik> or we can do it in open floor I guess...
18:03:39 <yselkowitz> nirik, I'm together with Evolution
18:03:46 <nirik> ah. ok. ;)
18:04:06 <nirik> #topic alternate arch builds
18:04:38 <Evolution> so, we've had quite a bit of user/community demand for armhfp and aarch64 epel packages
18:04:55 <Evolution> but getting epel set up to build these is currently slow-going.
18:05:24 <nirik> yeah.
18:05:34 <nirik> there are not at all enough hours in the day. ;(
18:05:35 <Evolution> right now we're looking at cycling the current epel srpms through the offline buildsystem but I'd like to not conflict with epel when/if it finally becomes a thing for altarch
18:06:17 <Evolution> doing this via our offline builders means that we won't pollute the koji community builder and impact the sigs,
18:06:32 <Evolution> however long-term this could be an issue for anyone using priorities or repo protection.
18:06:34 <nirik> ok. Will those results be available somewhere tho?
18:07:05 <Evolution> yes. I'm not certain yet if we'll do it as a one-off unsigned repo... "use at your own risk" sort, or something more official.
18:07:33 <Evolution> but we will make them consumable in some fashion.
18:07:46 <Evolution> if there are ideas on how best to bridge this once epel exists, I'm all ears.
18:07:57 <Evolution> yselkowitz: did you have anything to add here?
18:08:36 <yselkowitz> my point was going to be wrt pushing patches necessary for building on aarch64 in as provenpackager
18:08:38 <nirik> well, I think epel would do bring up just like we did for ppc64le...
18:08:53 <nirik> rebuild everything scratch and import the aarch64 or whatever builds.
18:10:01 <yselkowitz> as for actually getting the srpms to build
18:10:24 <yselkowitz> the rhelsa team has been working on getting this working
18:10:30 <yselkowitz> https://bugzilla.redhat.com/showdependencytree.cgi?id=1285484&maxdepth=1&hide_resolved=0
18:10:42 <nirik> cool. ;)
18:10:52 <yselkowitz> many of the patches have been committed, but many still have not
18:11:15 <yselkowitz> in order to centos to be able to rebuild epel for aarch64, these patches have to go in now
18:11:26 <nirik> IMHO, i'd like to have our better builders up and running before we add aarch64 to epel... but hopefully thats not too far out
18:11:51 * nirik doesn't have any objection to pushing them in... as long as they don't break primary should be fine
18:12:08 <yselkowitz> a few will require coordination with maintainers (namely ghc)
18:12:24 <yselkowitz> otherwise they are mostly simple fixes from newer versions of fedora
18:12:44 <nirik> good ol ghc. ;)
18:12:50 <yselkowitz> yep
18:12:52 <Evolution> yselkowitz: to your knowledge are any of them version bumps that would be problematic for epel?
18:13:24 <yselkowitz> ghc is not a version bump but it does add some patches from f21
18:13:44 <yselkowitz> but iiuc if you sneeze at ghc in the wrong direction it breaks :-)
18:13:48 <Evolution> I don't see that being any more of a problam than ghc already is.
18:13:51 <yselkowitz> so that will take some care
18:13:57 <Evolution> er, problem. spelling ftw
18:14:04 * nirik loves that ghc changes ABI based on how many cpus were used to build it (or it used to anyhow)
18:14:23 <yselkowitz> otherwise, the changes are intentionally conservative, I don't anticipate any issues
18:15:30 <nirik> sounds good.
18:15:59 <avij> it's nice that someone is taking care of this, thanks
18:17:14 <nirik> I might be able to commit some cycles to adding it once we get our better builders going...
18:17:32 <yselkowitz> so is that a consensus for me to proceed with provenpackager-ing these patches?
18:17:32 <Evolution> yselkowitz: is there anything you need from us?
18:17:44 <Evolution> +1 here. I want aarch64!
18:17:49 <bstinson> +1 from me
18:18:04 <nirik> +1
18:18:15 <avij> those changes would need to be done sooner or later anyway, so go ahead.
18:18:27 <nirik> yep. go forth and git push. ;)
18:18:42 <yselkowitz> and fedpkg update too
18:19:03 <nirik> yep
18:19:09 <yselkowitz> ok, thank you
18:19:09 <nirik> ok, anything else on this for now?
18:19:20 <Evolution> that covers my end.
18:19:26 <nirik> cool.
18:19:27 <yselkowitz> probably follow-up next week wrt anything delicate
18:19:33 <nirik> sounds good
18:19:56 <nirik> #info yselkowitz will push needed aarch64 patches into epel packages and update them in prep for mass building.
18:20:03 <nirik> #topic Open Floor
18:20:09 <nirik> anyone have anything for open floor?
18:20:24 <avij> nope
18:20:32 <Evolution> nope
18:20:41 <bstinson> none here
18:20:54 <nirik> There was a nginx thread on the list...
18:21:06 <nirik> which I think we already discussed a few weeks ago...
18:21:22 <Evolution> the version bump ?
18:21:33 <avij> yes, 1.10.x is a better target
18:21:40 <nirik> looks like they plan to go to 1.10.x instead of 1.8.x (1.10.x just came out and has a longer support life and some features that are depreciated in the older ones)
18:21:44 <nirik> yeah
18:21:50 <nirik> I agree, just thought I would mention it
18:22:05 <Evolution> wfm
18:22:08 <avij> yea
18:22:31 <nirik> ok, if nothing else will close out in a minute...
18:23:45 <nirik> Thanks for coming everyone. Go and use the extra 36 minutes scheduled for this meeting for good. ;)
18:23:48 <nirik> #endmeeting