12:01:09 #startmeeting rpm_packaging 12:01:10 Meeting started Thu Aug 24 12:01:09 2017 UTC and is due to finish in 60 minutes. The chair is toabctl. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:01:11 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 12:01:13 The meeting name has been set to 'rpm_packaging' 12:01:17 ping toabctl, dirk, apevec, aplanas, IgorYozhikov, jpena, jruzicka, number80, kaslcrof 12:01:57 anybody arround? 12:02:15 o/ 12:02:22 hey jpena 12:02:48 let's wait some minutes of somebody else is arroung 12:03:01 ok 12:03:05 feel free to add agenda points to https://etherpad.openstack.org/p/openstack-rpm-packaging 12:06:24 let's start 12:06:30 #topic pymod2pkg - handling singlespec 12:06:41 o/ 12:07:01 for https://review.openstack.org/#/c/492931/ I removed the pymod2pkg handling 12:07:54 which does not work (obviously) . but the %python_module macro expands to python-%{**} python3-%{**} so it expects the name without the leading python- 12:08:04 not sure how to handle that with pymod2pkg. any suggestions? 12:08:08 hey number80 12:08:56 I wonder if pymod2pkg should return 2 names instead of one. 12:09:33 let me check the macros 12:09:44 jpena, it is the first line in the changeset 12:09:51 %{?!python_module:%define python_module() python-%{**} python3-%{**}} 12:09:59 ah ok 12:10:29 just removing the python- prefix would not work in all cases I think (eg. for python-cinderclient) 12:13:04 so pymod2pkg could return 2 names 12:14:29 what if we have a new call in pymod2pkg where we return 2 names (for backwards compat)? Then we could have a second macro to get them, like py3pkg (or a better name) 12:14:58 hm. yeah 12:15:27 jpena, so we would need a second mapping for python3 ? 12:15:56 toabctl: yes. We have several instances where the python2 and python3 packages have different names in CentOS 12:16:00 we don't not have to use the python_module macro. we could also expand this ourselfs 12:16:16 jpena, oh. interessting. do you have an example? 12:16:32 yes: PyYAML vs python3-PyYAML 12:16:35 jpena, that would mean that we need to handle that anyway in pymod2pkg, righ? 12:16:39 right 12:16:58 same for pyOpenSSL 12:17:06 you are doing crazy things in CentOS :) 12:17:14 * jpena shrugs 12:17:45 ok. so somebody needs to work on pymod2pkg 12:18:04 any volunteer? 12:18:25 it's just you and me, with number80 hiding :) 12:18:31 heh 12:18:38 I can give it a look 12:18:49 jpena, ok. cool. thx 12:19:01 #action jpena looks into a py3 mapping for pymod2pkg 12:19:22 next topic? 12:19:29 yes 12:19:31 #topic packaging reviews 12:19:58 anything special? 12:20:20 nothing for me 12:20:32 ok. then next 12:20:37 #topic backports merging policy 12:21:14 I was merging backports with just a single +2 (pardon) and wonder if we could change our policy to allow that 12:21:22 jpena, number80 any opinion on that? 12:21:47 did we have a policy? 12:21:52 we do not have many reviewers and from time to time things are slow 12:22:03 jpena, heh. I thought we need 2 +2 :) 12:22:36 jpena, but tbh I think for smaller updates where CI is green we could just merge with a single +2 in general 12:23:41 I'm ok with that for cherry-picks. To make it official and reach the other cores, could you open a thread in openstack-dev? 12:24:19 sur 12:24:20 e 12:24:45 #action toabctl mail to openstack-dev about single +2 for merging things 12:25:03 #topic open floor 12:25:06 anything else? 12:26:07 not for me 12:26:52 ok. then thanks and goodbye 12:26:54 #endmeeting