13:00:07 #startmeeting rpm_packaging 13:00:08 Meeting started Thu Mar 31 13:00:07 2016 UTC and is due to finish in 60 minutes. The chair is dirk. Information about MeetBot at http://wiki.debian.org/MeetBot. 13:00:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 13:00:11 hi 13:00:12 The meeting name has been set to 'rpm_packaging' 13:00:49 #topic agenda 13:01:01 everyone, please update https://etherpad.openstack.org/p/openstack-rpm-packaging 13:05:18 o/ 13:05:30 can we start, anything still left? 13:05:51 #topic pypi tarballs or sdisted from git tarball as source 13:05:59 IgorYozhikov: care to explain? 13:06:08 yes 13:07:08 want to clarify is there any concerns about tarball source and how to "cook" them. 13:07:56 or this is in responsibility of package maintainer 13:08:25 at present moment in our templates Source: pointing to pypi tarball 13:08:49 yep 13:08:58 which should be fine for things that are "released", right? 13:09:11 is there a particular package you have in mind where this is a problem? 13:09:21 or is it just for the "package from git" case? 13:09:28 we could point to tarballs.openstack.org 13:10:05 dirk, packages built from git code instead of pypi prebuilt tarballs 13:11:38 with pypi we already have "cooked" tarball 13:11:53 ah, ok 13:11:57 in case of git - tarball should be created from particular tag 13:12:17 and only after that passed into rpmuild | mock |etc 13:12:18 so are you tracking daily commits to the git branches? 13:12:27 or are you creating the tarball yourself from a git tag and sdist it? 13:12:38 dirk, 2nd 13:12:52 own "kitchen" 13:13:27 building sdit-ed tarballs 13:13:40 sdist-ed 13:13:59 and why do you prefer that over pypi? 13:14:14 I mean, if its a requirement we need to add a renderspec variable for it and abstract it away :) 13:14:34 generally I think pypi is preferable because its the canonical source for python packages 13:15:04 yes. and also upstream uses the tarballs from pypi 13:15:37 dirk, understood | clarified. 13:16:32 next topic? 13:17:37 not sure what we agreed on 13:17:45 do we need to create a renderspec filter ? 13:17:52 or will you just switch to pypi ? 13:18:33 dirk, pypi looks fine and in case of necessity 13:19:06 ok, thanks 13:19:16 #agreed upstream spec files will link pypi tarballs 13:19:17 we can switch to another source - like locally prebuilt tarball 13:19:45 #topic design summit session 13:20:33 r 13:20:36 dirk, what news - room, schedule ? 13:20:39 just to give out the update, we have 1/2 fishbowl session 13:20:45 shared with the deb-packaging project 13:20:55 on thursday 11:50 13:21:03 so zigo will be there, right? 13:21:08 and a single work room slot on thursday 2:20 13:21:17 I guess/hope so 13:21:27 PTL is Monty ,so at least he'll likely be there :) 13:21:40 I haven't reached out to him yet regarding topic/slot 13:21:53 do we have things we want to discuss in the fishbowl slot ? 13:22:07 I'd send it as a topic RFQ then to monty to figure out what we're gonna do about it 13:22:24 overall I think the decision from TC to do a shared fishbowl with deb and rpm makes a lot of sense 13:23:35 We could discuss approach which is already used om rpm-side and propose it to deb 13:23:57 templates + auto-fill will make life easier 13:24:19 o/ 13:24:52 yep, shared of experiences is probably good topic 13:25:01 also common gating plans is probably an item to sync on 13:25:09 gating/check gates 13:25:19 And as far as I know deb project working on CI on infra side 13:25:41 yeah 13:25:47 our ci engineers participated in review and meetings helping zigo 13:26:20 o/ 13:26:56 dirk, i really very hope that our ci will be ready before summit and we could also collect some results/data and discuss it too 13:27:12 nice 13:27:16 yep, would be great 13:27:22 working on it 13:27:35 we have mixed experiences with the external zuul setup (its not very stable) still looking into fixing those issues.. 13:27:51 I can,as promised, collect some data/proposal about CI stuff 13:27:57 and share it 13:28:13 yep, anything else, any questions? 13:28:14 I have approve for summit btrip 13:28:55 and believe that can try to drive one-two topics :) 13:29:12 I should be there, and will also draft our plan for CI for RPM packaging 13:30:11 dirk, think that all these more | less enough for fishbowl 13:30:19 yep 13:30:22 worksession: 13:30:28 I'll see if I can invite openstack infra folks to help us 13:30:38 for the work session I guess we don't necessarily need a topic list 13:30:38 (or answer our questions) 13:30:39 yet 13:30:50 for the fishbowl I need to fill in the slot description 13:32:46 next topic? 13:33:17 #topic openstack salt packages 13:33:39 jasondotstar: ping? 13:34:09 dirk: I already pinged him in #openstack-rpm-packaging without response... 13:34:28 jasondotstar recently joined #openstack-rpm-packaging and was interested in getting/contributing openstack-salt rpm packages 13:34:37 salt? salt | saltstack 13:34:42 yes 13:35:50 he wanted to invite people over for the rpm packaging meeting but apparently we miscommunicated on the time slot.. 13:35:56 anyway, I guess we can postpone the topic 13:35:59 dirk, salt* -> rpm-packaging/openstack/* right? 13:36:29 IgorYozhikov: I think it was about packaging the git repo git.openstack.org/openstack/openstack-salt/ 13:36:37 I haven#t looked at what that would require yet 13:36:53 would have been my first question. reportedly there are deb packages but people raised interest in having rpm packages as well 13:37:00 u i c 13:37:27 in case anyone has spec files or wants to write some, contributions are welcome :) 13:38:10 will update old stuff tomorrow, a lot of work during these days :) 13:38:26 any last topic? 13:38:48 othwise I'd end the meeting. I have to rush over to something else but will do a round of reviews later tonight 13:39:31 dirk, i'm fine with finish earlier 13:40:09 nothing from my side. 13:40:22 #endmeeting