18:00:07 #startmeeting EPEL 18:00:07 Meeting started Fri Aug 21 18:00:07 2015 UTC. The chair is smooge. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:07 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:12 #meetingname EPEL 18:00:12 The meeting name has been set to 'epel' 18:00:26 #chair smooge avij bstinson dgilmore nirik 18:00:26 Current chairs: avij bstinson dgilmore nirik smooge 18:00:40 #topic Meet and Greets 18:00:46 present 18:01:14 This meeting is to cover things which have been going on in EPEL in the last month or so since the last meeting 18:01:35 hi all 18:01:55 hi bstinson and avij 18:02:32 ... waits to see if we can get 1-2 more people 18:03:00 bstinson and avij were either/both of you at FLOCK? 18:03:09 * bstinson 18:03:50 morning 18:04:12 ok we have minimum quorum 18:04:15 nope, I was rock climbing when FLOCK took place (well, most of the time) 18:04:23 cool 18:04:35 #topic Anything from FLOCK 18:04:49 ok bstinson and nirik was there anything EPEL related at FLOCK? 18:05:43 yeah, dgilmore and I and bstinson talked a bit 18:06:07 dgilmore was saying we could get koji to use different external repos for different arches. (which I didn't think would work) 18:06:24 so, we should be able to add i386 in by using centos i386 for that... 18:06:44 cool. 18:06:52 that would be nice 18:07:03 i know Evolution is keen to get going on some of the ARM bits too 18:07:19 yeah, the aarch64? 18:07:21 would that be 'like' the secondary koji's or just the main koji? 18:07:38 smooge: for i386 we could just add it in the main one... 18:07:48 not fully sure how we decided to handle aarch64/ppcle... 18:08:03 yeah, i don't think we closed on that 18:08:36 but if we have the functional tests done for i686, we can replicate where needed 18:08:41 there was talk about changing how we do secondary arches in fedora... 18:08:41 ok cool. I know more koji's make releng unhappy.. but I want to try and make it so that we aren't breaking main koji if say ppcle is good and then dropped :) 18:08:50 just adding them all into one koji... 18:09:24 that works also... 18:09:48 with these 5 koji's combined, we form kojira 18:10:00 but it was just talked about, not sure if it's been formally proposed yet 18:10:49 well something to look forward to 18:11:28 ok anything else from FLOCK? 18:13:16 alright. I only have one other topic at the moment... 18:13:19 * nirik was giving a talk and wasn't at the centos/fedora releng meetup. hope that went well 18:13:51 yeah, that was a pretty good discussion on container build tooling 18:14:25 yeah, sounds like atomic-reactor is the short term way forward 18:17:03 #topic needing updated/rebuilt packages in 6.7 etc 18:17:33 So after EL6.7 came out there seems to have been a couple of updates in libraries and such which needed packages rebuilt 18:17:54 currently we are finding them via user reports and then having the maintainers rebuild one by one. 18:17:56 ImageMagick was the big one I think. 18:18:01 is there a way we could automate this 18:18:05 but most of them are already rebuilt by now. 18:18:54 I filed a number of bugs for the remaining ImageMagick problems. many of them have been rebuilt now and in epel-testing. 18:19:50 I guess I could have saved a bit of time if I tested against RHEL 6.7, but I only have access to CentOS 18:20:50 ok so the current method is ok for the committee? [I just want to know if I am seeing a real problem or the current methods are ok.] 18:21:12 it would be nice if we were mode coordinated on it. 18:21:23 but not sure we could do much more for this time 18:21:54 so I know we do some sort of bump and rebuild in Fedora at certain times. Is that possible with EPEL? 18:22:28 some sort of an advance notification would be nice for the packagers ("your package will break when RHEL x.x gets released") 18:23:07 yeah.. I don't know if we have any way of knowing that beyond testing with the beta 18:23:33 right, I understand there's not much we can do regarding that 18:24:57 yeah :/ 18:24:59 on the other hand, although I'm not a packager myself, I got the impression that all it takes to fix the issue for these packages is a rebuild. this could probably be done by a provenpackager if needed. 18:25:18 so not necessarily by the official package maintainer 18:26:13 I think that would work 18:26:50 ok I think nirik got paged on something so we can end this now 18:26:55 #topic open floor 18:27:05 Anyone have anything for the open floor? 18:27:12 yeah, provenpackager could fix it... 18:27:15 the mass-rebuild tools might be adaptable for that purpose 18:28:11 that was what I was wondering.. but I think dgilmore would need to help answer that. 18:28:23 as far as I can see, the following packages still need to be rebuilt for the new ImageMagick: drawtiming, perl-Image-SubImageFind, psiconv, spectrum 18:29:33 I will be travelling (in Brussels) next week at this time 18:29:36 well, rpmdev-bumpspec and a build/update. ;) 18:29:46 I can try fixing those this weekend. 18:31:04 ok avij will you be back in 2 weeks? 18:31:47 yes; my trip starts tomorrow morning and ends on 31.8. 18:32:14 ok so next meeting in 2 weeks 18:32:22 +1 18:32:41 my cell phone is going crazy.. am going to clal this meeting over 18:32:46 #endmeeting