13:02:03 <dirk> #startmeeting rpm_packaging
13:02:04 <openstack> Meeting started Thu Jul 28 13:02:03 2016 UTC and is due to finish in 60 minutes.  The chair is dirk. Information about MeetBot at http://wiki.debian.org/MeetBot.
13:02:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
13:02:07 <openstack> The meeting name has been set to 'rpm_packaging'
13:02:17 <dirk> toabctl, dirk, aplanas, IgorYozhikov, jruzicka, number80:: ping
13:02:42 <dirk> #topic roll call
13:02:50 <number80> o/
13:04:30 <dirk> please add your meeting agenda topics to
13:04:37 <dirk> #link https://etherpad.openstack.org/p/openstack-rpm-packaging
13:05:25 <jruzicka> o/
13:07:04 <dirk> #topic  Mascot update
13:07:12 <dirk> so far we have submitted the Donkey
13:07:20 <number80> wfm
13:07:37 <dirk> it will be preapred as a comic like logo with and without text, and there might have both slide templates as well as t-shirts for barcelona potentially
13:07:44 <number80> A/bu19
13:08:28 <dirk> #topic packages reviews (https://review.openstack.org/#/q/project:openstack/rpm-packaging+status:open )
13:09:23 <dirk> i had a pass on the reviews yesterday and this morning
13:09:28 <dirk> anything specific we want to talk about ?
13:09:37 <number80> I did a pass yesterday too, I fixed your comments
13:09:42 <dirk> I've seen a couple of ci failures on both SUSE and mirantis fuel side
13:09:57 <dirk> I have the action item to update the pymod2pkg package on the worker
13:10:00 <dirk> that should fix things
13:10:00 <number80> I also left some hints to fix Fuel CI (e.g missing packages and also provided them CentOS builds)
13:10:12 <dirk> I've also fixed the mitaka gating and merged stuff there
13:10:28 <number80> dirk: out of curiosity, could you add me as reviewer for the worker, I'd like to learn how it works
13:10:44 <astsmtl> Ping me if some review cannot pass our CI.
13:11:04 <number80> astsmtl: I think you're already a reviewer in most of them
13:11:14 <astsmtl> And the fault is on our side. :)
13:11:20 <astsmtl> Ok.
13:11:40 <number80> astsmtl: it's fine, openstack keeps adding a lot of deps, so nothing unusual
13:12:08 <number80> check comments, as I added links to packages to solve missing ones on your CI
13:12:17 <number80> after all, we both builds on CentOS
13:12:36 <dirk> astsmtl: https://review.openstack.org/#/c/342313/
13:12:38 <dirk> astsmtl: there are more
13:12:43 <dirk> just look for failures above
13:13:05 <dirk> astsmtl: I can search through things in more detail, but it is not that many reviews right now, should be quick for you to check all of them quickly
13:13:16 <number80> dirk: btw, I found a weird bug in py2name
13:13:42 <number80> clients name like python-cinderclient are transformed in python-python-cinderclient (same for gnocchiclient)
13:13:57 <number80> maybe you've noticed it already
13:14:46 <dirk> number80: yes, tahts what I just said, I need to update pymod2pkg on the jenkins worker
13:15:05 <number80> oh excellent
13:15:09 <number80> dirk++
13:15:16 <dirk> number80: basically thats review I72977c695c9d4c9937e1602dadad5363def87380 and Ic4da525a2efca118570f150cdd1e37f03c7f7376
13:15:23 <dirk> they're missing
13:15:28 <dirk> I've just pasted a review for that
13:16:05 <dirk> we need to update to 0.5.2 release
13:18:15 <number80> wfm
13:20:26 <dirk> argh
13:20:30 <dirk> sorry, real life interrupt
13:20:45 <dirk> #topic https://review.openstack.org/#/q/project:openstack/renderspec+status:open
13:21:15 <dirk> jruzicka: do you want to discuss https://review.openstack.org/#/c/347447/2/doc/source/usage.rst ?
13:21:44 <jruzicka> yeah
13:21:55 <jruzicka> what do you guys think about that?
13:22:09 <jruzicka> Unless there's some horrible bug, it is a non-intrusive change
13:22:24 <number80> We'd like to reduce the churn on maintaining platform-specific stuff in templates
13:22:38 <jruzicka> yes today I seen "include opstnack-macrost" in review
13:22:53 <jruzicka> so this should be hammer to solve these once and for all
13:23:00 <number80> that's error-prone, and it can be automated most of the time
13:23:30 <jruzicka> yes
13:23:46 <number80> one other thing I'd like to be automated is BR/R sorting as package name could be different (so ordering can be different in RDO/suse for instance)
13:23:58 <jruzicka> so unless you see something wrong with that code I say let's merge this and we can start adding
13:23:59 <jruzicka> DLRN/rdoinfo as jpena is in PTO
13:24:06 <jruzicka> gee, clipboard fail
13:24:19 <jruzicka> {% block build_requires %}{% endblock %}
13:24:23 <dirk> jruzicka: I haven't understood it in detail
13:24:35 <number80> dirk: it's just template composition
13:24:36 <jruzicka> motivation is here:
13:24:37 <jruzicka> https://etherpad.openstack.org/p/renderspec_RFE
13:24:43 <astsmtl> jruzicka, I have one question, why you chose block and inheritance, instead of just variable substitution?
13:24:49 <dirk> jruzicka: so basically we add this block build_requires and then the distro flavor part will be expadned?
13:24:57 <dirk> jruzicka: thats interesting, we need this e.g. for the copyright header
13:25:01 <number80> yep
13:25:02 <jruzicka> astsmtl, because it's a generic solution to many problems that can come
13:25:04 <jruzicka> as in
13:25:13 <jruzicka> you can define blocks easily in already existing files
13:25:22 <jruzicka> and they can use the macros which main template uses
13:25:40 <jruzicka> also undefined block will remain empty
13:25:51 <jruzicka> so no failures due to unexpected renderspec version
13:26:22 <jruzicka> astsmtl, since you ask that, how do you propose we solve the sorting of requiremenst? :)
13:26:26 <dirk> jruzicka: ok
13:26:44 <dirk> jruzicka: I'll try to concentrate on the review in the next days and comment there. so far I like the idea. thanks for the change
13:26:53 <jruzicka> one more thing
13:27:01 <jruzicka> this only solves certain case of problems
13:27:04 <astsmtl> jruzicka, Dunno.
13:27:20 <jruzicka> to solve "Group: only in suse but not rdo"
13:27:35 <jruzicka> and to sort requirements, easiest way I found is to postprocess
13:28:11 <jruzicka> so some additional logic in renderspec to manipulate text... remote Group and order requires, pretty easy to do.
13:28:46 <jruzicka> If you hate the idea, let me know, otherwise I'll provide reviews ;)
13:28:50 <jruzicka> that's all
13:29:44 <astsmtl> It is more transparent for user if rendered spec is not post-processed.
13:29:56 <astsmtl> But if there is no other good solution, we can do it.
13:30:10 <dirk> jruzicka: well, if the group is realy hurting for RDO, then we need to add a renderspec macro that only expands on suse
13:30:20 <dirk> jruzicka: that should be okay I think
13:30:56 <number80> dirk: it can be post-processing, but that's not high prio
13:31:47 <number80> it can be kept around until tooling can handle it gracefully
13:32:34 <jruzicka> group aside
13:32:39 <jruzicka> (which is low prio indeed)
13:32:46 <dirk> anything else on this topic?
13:32:48 <jruzicka> only other way to do the sorted requires would be
13:32:54 <jruzicka> {{ requires(
13:32:54 <jruzicka> 'foo',
13:32:54 <jruzicka> 'bar',
13:32:54 <jruzicka> 'pbr'
13:32:55 <jruzicka> ) }}
13:33:04 <jruzicka> that would mean drastic change to all templates
13:33:14 <jruzicka> more power, but not sure if worth
13:33:29 <astsmtl> Btw, can we finally merge or reject: https://review.openstack.org/#/c/336151/
13:34:10 <jruzicka> astsmtl, yes, I have that planned for today
13:34:26 <astsmtl> jruzicka, I thought about this way. I think we should evaluate both solutions carefully.
13:35:01 <jruzicka> astsmtl, yup, so I'll propose the less intrusive but less transparent one first
13:35:14 <number80> astsmtl: I am not -2 but still hesitating between -1/+1
13:35:41 <jruzicka> I hoped for someone else's numbers on that
13:35:48 <astsmtl> :D
13:36:08 <number80> just to say that you made good points
13:36:26 <dirk> astsmtl: I am undecided on this still
13:36:33 <dirk> I kind of see value in both approaches
13:36:59 <astsmtl> Ok, lets proceed with the next topic.
13:37:12 <dirk> I'll try to get to it
13:37:18 <number80> *nods*
13:38:49 <dirk> #topic Meeting chair for next week
13:39:00 <dirk> I'm  travelling next week this time
13:39:07 <dirk> any volunteer?
13:39:12 <number80> same for me
13:39:19 <dirk> toabctl might be back from PTO, but I don't know for sure
13:39:26 <jruzicka> I can
13:39:30 <number80> excellent
13:39:33 <astsmtl> Igot might be back too.
13:39:34 <dirk> jruzicka: thanks!
13:39:36 <astsmtl> Igor.
13:39:58 <dirk> #agreed jruzicka / IgorYozhikov  can chair the meeting
13:40:04 <dirk> #topic Open floor
13:40:11 <dirk> anything else?
13:40:40 <number80> nope, just to say thanks to everyone for making it more lively here :)
13:40:48 <number80> keep up with the good work!
13:40:55 <number80> astsmtl++
13:42:11 <astsmtl> :)
13:43:25 * jruzicka is very happy to have some time allocated for rpm-packaging again
13:46:17 <number80> we can close the meeting I think :)
13:46:34 <dirk> agreed
13:46:44 <dirk> thanks to all for the attendance and the increased participating!"
13:46:49 <dirk> keep the reviews flowing!
13:46:51 <dirk> #endmeeting