18:00:17 #startmeeting EPEL 18:00:17 Meeting started Wed Apr 20 18:00:17 2016 UTC. The chair is nirik. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:17 Useful Commands: #action #agreed #halp #info #idea #link #topic. 18:00:17 The meeting name has been set to 'epel' 18:00:17 #meetingname EPEL 18:00:17 The meeting name has been set to 'epel' 18:00:17 #chair smooge nirik Evolution bstinson avij 18:00:18 #topic agenda gathering / quorum detection 18:00:18 Current chairs: Evolution avij bstinson nirik smooge 18:00:37 hello 18:00:53 hey avij 18:01:08 not sure if we have people around today, smooge is likely not around. 18:01:31 hi folks, i know Evolution won't be here today 18:01:55 anyone have agenda items? 18:02:30 perhaps that torque issue, https://bugzilla.redhat.com/show_bug.cgi?id=1321154 ? 18:02:50 yeah, I need to re-read that... 18:03:47 are there any clear proposals for a way forward for that? 18:04:01 #topic torque issue - https://bugzilla.redhat.com/show_bug.cgi?id=1321154 18:04:14 hello nucleo 18:05:09 unfortunately I'm not familiar with that package myself 18:06:18 me either. It sounds like from reading the bug the maintainer could make it work by default with non numa by adjusting the default config... 18:07:24 if so they should make a new version with that so the default keeps working. 18:07:56 I can ask that on the bug I guess. 18:10:11 ok, I added that to the bug. 18:10:16 Not sure what else we can really do here. 18:11:02 #topic macros 18:11:14 more macros are in now... tibbs has been knocking them out. ;) 18:11:28 \o/ 18:11:29 The python ones are on the way. 18:11:36 ? 18:11:53 That and %make_build were the last two I had requests for. 18:12:15 rsc: ? 18:12:23 nirik: no meeting protocol in effect for this meeting? 18:12:29 nope. ;) 18:12:31 ask away 18:12:33 nirik: Oh :) 18:12:53 * nirik finds that protocol silly, but I guess it's useful when you have lots of people. 18:13:02 How do the macro changes impact non-EPEL builds, say building a spec file without mock, but directly with rpmbuild on RHEL/CentOS 5 or 6? 18:13:20 Same as if you don't have the other necessary macro packages installed. 18:13:24 if you have epel-rpm-macros installed I imagine it would work fine 18:13:36 I mean, rpmbuild never worked as expected if you didn't install redhat-rpm-config. 18:13:43 This is no different. 18:14:18 Yes, but redhat-rpm-config is in RHEL while epel-rpm-macros is in EPEL. Thus rebuilding of an EPEL package on naked RHEL is no longer working if it depends on macros of that epel-rpm-macros package? 18:14:27 So is there a risk of breaks? 18:14:48 sure, but why would anyone do that? 18:15:06 I don't think we ever made some promise never to rebuild epel packages without other epel packages. 18:15:12 Ah. 18:15:30 Last sentence explains it, because there are enough crazy people out there rebuilding parts of EPEL locally. 18:15:42 (without mock or epel-rpm-macros) 18:15:50 Thanks for the clarification :) 18:15:56 well, they will need to mimic the buildroot from koji if they want to be sure to have everything work 18:16:10 but thats really always been the case. IMHO 18:16:14 Which is as easy as yum groupinstall buildsys-build. 18:16:30 And, yes, this has always been the case. Of course, if you change anything then someone's workflow breaks. 18:16:56 The bigger mess is that fact that I can't define %license on EPEL6 if the scl macros package is installed, because then the scl macros break. 18:16:58 it does make us pay attention on the CentOS side (we rebuild quite a bit in CBS) but that's on us 18:17:28 bstinson: isn't a mock RPM package in a CentOS repo, too? I thought... 18:18:07 rsc: well, we ship a version from EPEL for repoclosure over some of our client tools 18:18:28 Oki. Sorry folks, didn't want to hijack that. 18:18:41 no worries 18:19:00 * nirik still finds it fun that mock is in epel and not rhel, but thats a topic for another day. 18:19:30 #topic Open Floor 18:19:34 anything else from anyone? 18:19:52 nirik: we (CentOS) too have opinions on the mock subject :) 18:19:54 * nirik has been idling considering the idea of a FAD or bug fix weekend or something later in the year. 18:20:00 +1 18:20:16 +1, but virtually or reallife? 18:20:27 whichever would work best. ;) 18:20:44 virtual is easier/cheaper, but real life might get more done depending on if we have targeted stuff. 18:21:26 just a thought at this point... 18:22:33 I guess I could toss a thread up on the list and see how much interest there might be 18:23:00 real life might be good if we are doing something like adding more arches. 18:23:14 bug cleanup/fixing things might be better virtual 18:23:55 anyhow, if no one has anything else, will close in a minute or two 18:24:07 bstinson: what kind of opinions? 18:26:27 dgilmore: CentOS needs a mock to ship for SIG developers (who may or may not have EPEL enabled) 18:26:56 currently we're working on tracking EPEL for that and shipping in CentOS-Extras 18:28:45 not to go too far on a tangent, but we're also looking at submitting some vanilla configs to see if they can be included in mock upstream 18:29:35 mock has always been in a weird place. ;) not in rhel, used all over the place, builds against centos by default, etc etc. 18:30:36 bstinson: okay 18:31:12 bstinson: I see no problem adding vanilla centos configs upstream 18:33:00 we've seeded those to a few people for testing. patch to mock incoming soon-ish :) 18:34:28 cool.. 18:34:36 Thanks for coming everyone. 18:34:39 #endmeeting