12:00:23 #startmeeting Horizon 12:00:24 Meeting started Wed Apr 15 12:00:23 2015 UTC and is due to finish in 60 minutes. The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot. 12:00:25 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 12:00:27 The meeting name has been set to 'horizon' 12:00:32 o/ 12:00:39 hi! 12:01:07 hey 12:01:13 hey 12:01:21 hello 12:02:08 hello 12:02:29 #topic General 12:02:51 Yesterday we tagged RC1 12:02:57 \o/ 12:02:59 congrats! 12:03:14 hooray! 12:03:19 |o| 12:03:27 yay 12:03:32 ....\o/ 12:03:40 Thanks to everyone for your hard work getting to this point 12:04:17 We will have an RC2, but nothing critical is looming for it, but we need it to import final translations 12:04:45 david-lyle: not to be a party pooper, but I'm really quite concerned about this: https://bugs.launchpad.net/horizon/+bug/1444150 12:04:47 Launchpad bug 1444150 in OpenStack Dashboard (Horizon) "New Launch Instance dialog is mostly untranslatable" [Undecided,New] 12:04:56 I was hope that somebody else would tell me it all works fine 12:05:03 but that hasn't happened yet. 12:05:15 I also have two low risk, high gain bugs that we can cherry pick the fixes to it https://bugs.launchpad.net/horizon/+bugs?field.tag=kilo-rc-potential 12:05:21 doug-fish: looking 12:06:36 doug-fish: Does it introduce new strings? 12:06:42 I can't say why it's failing - the code looks right 12:06:44 yeah, lots 12:07:02 and almost none of them are in the po files I generate 12:07:10 I had another bug for kilo-rc-potential: https://bugs.launchpad.net/bugs/1444421 12:07:19 Launchpad bug 1444421 in OpenStack Dashboard (Horizon) "Launch instance fails with nova network" [Undecided,New] 12:07:19 I think it is better to avoid introduce new strings at this stage. the dialog is regarded as experimental in this reelase 12:07:19 but "Availability Zone" must be translated already 12:07:45 so we don't need to tackle it in this release. 12:07:46 david-lyle: seems its not in the right file to be found - not in djangojs.po maybe. 12:08:09 I thought this was our default option for launch instance now 12:08:24 doug-fish: no 12:08:28 ah ok 12:08:42 I chose to leave it as experimental 12:08:47 If it's experimental I feel much better about it. 12:08:53 david-lyle: nice choice! 12:09:11 a lot of that has to do with the fact of how late it landed and how little real world testing it has 12:09:28 we can't just dump that in operators' laps 12:09:30 doug-fish it's unknown why the js translations aren't working? are they just not working for launch instance or all js? 12:09:32 agreed. Is there a setting to enable/disable it? 12:09:40 and hope for the best 12:09:51 doug-fish: yes and it's disabled by default 12:10:11 mattfarina: I know some js is working (like network topology) and some isn't (like launch instance) 12:10:59 the translation mechanism is the same as other JS files, so not sure what would be failing, but the way the JS is forced to use those translated strings is horrible 12:11:15 and fixing that is one of the top priorities in Liberty 12:11:30 yes, that sounds good to me. 12:11:32 mrunge: I tagged yours as such 12:11:41 that shouldn't be happending 12:11:47 david-lyle, thanks, I did as well... 12:12:01 no idea, how launchpad will handle that 12:13:09 mrunge: looks right 12:13:17 I assigned myself 12:13:21 will look today 12:13:41 david-lyle, strange thing is, I remember it worked 12:13:52 not sure, what changed there 12:14:02 I wouldn't have thought a lot 12:14:19 yes, same here. 12:15:05 So as your doing your testing, if you come across a bug you think may be critical to the release, please either tag it as kilo-rc-potential, or ping me 12:15:34 We have a couple weeks left to fix these problems 12:15:56 With RC1 cut, master is now open for liberty 12:16:16 so many of the items that just missed Kilo will start landing 12:16:32 I have not prioritized items for L-1 12:17:02 if they are assigned it's either self-assigned or carry-over 12:17:28 today or tomorrow I will open up a specs repo for Horizon 12:17:53 and we'll move the specifications to that, so we have a better voting record than launchpad 12:18:12 awesome! 12:18:16 I email the dev list when I have that set up 12:18:21 great news! 12:18:50 when we move to using that, self-assigning of blueprints to milestones will be disabled in launchpad 12:19:19 the Horizon Drivers will be the only ones that can set that 12:19:38 since the spec will be the notification mechanism 12:20:00 It will reduce a lot of the milestone planning noise and maintenance that I have to do 12:20:27 we will still have trivial feature adds that won't require a full spec 12:20:45 do we want to handle that with a trivial spec, or just a blueprint? 12:21:22 i'd suggest trivial spec 12:21:37 otherwise people will just declare their blueprint trival and bypass that process 12:21:45 +1 doug-fish 12:21:46 I know that's what I'll do! 12:21:56 that would be more consistent 12:22:16 ok, we can handle both in the specs repo 12:23:38 For the summit, I believe we have 3 fishbowl sessions, 7 working sessions and 2 half-day meetup sessions (like friday last time) 12:24:17 #link https://etherpad.openstack.org/p/horizon-liberty-summit is for topic suggestions 12:24:46 we'll need to start voting that down in the next couple of weeks 12:25:13 I also want to point out #link https://wiki.openstack.org/wiki/Design_Summit/Planning 12:25:44 if you are interested in cross-project topics or other services 12:26:37 I think that's all I had for general items 12:26:42 any questions? 12:27:24 david-lyle, are we still waiting here https://review.openstack.org/#/c/163028/ for new version of openstack_auth? 12:27:49 amotoki, has indicated final translations are targeted for April 23 12:28:26 tsufiev: I release 1.2.0 a couple of days ago 12:28:59 we now need to move g-r to require >= 1.2.0 before we can merge that change 12:29:10 david-lyle, ah, got it, thanks 12:29:30 1.2.0 brought Django 1.7 support which was needed for the release of Kilo 12:29:47 but did not require moving the minimum version of d-o-a 12:29:58 I have a tiny patch currently not passing tests for django-1.8 :D 12:30:28 mrunge: Django 1.8 support needs to be a K-1 item 12:30:36 no more waiting around on our parts 12:30:38 l-1 item? 12:30:44 but yes! 12:30:45 err, L-1 12:30:55 where am I? 12:31:10 since diffs are quite small, we even might be able to backport that 12:31:21 mrunge: sure 12:32:03 and Django will probably release again before we ship Liberty and ruin everything :P 12:32:23 I think 1.8 is a LTS release 12:32:26 david-lyle, 1.8 was released a few days ago 12:32:37 doug-fish: yes it is 12:32:40 mrunge: I know 12:32:41 and yes, it's a lts release 12:32:54 we can stay on it forever! 12:32:55 (almost) 12:32:56 but seems to be a 6 month cadence now 12:33:23 we can also look at dropping 1.4 support once we have 1.8 support 12:33:57 Something I should not about the 1.2.0 d-o-a release, it no longer supports python 2.6 12:34:09 as Django 1.7 does not support 2.6 12:34:54 #topic Open Discussion 12:35:55 Other high priority items are plugin strategy 12:36:57 now that the release pressure has subsided we need to consider plugins from two perspectives, new items in OpenStack and some items already in Horizon 12:37:05 Speaking of plugins... https://review.openstack.org/#/c/128133/ - still not merged as part of Kilo -> plugins are not being added to the proper PanelGroup 12:38:37 that has RC potential 12:39:03 okay, just had to make sure, nobody forgets about it :) 12:39:19 I tagged it 12:39:41 I need to look at that one more closely 12:40:50 mrunge: I see very few patches on tuskar-ui, do you know if that's still a point of interest 12:40:59 and does it need a release for kilo? 12:41:12 asking you with your red hat on 12:41:45 david-lyle, I must admit, I don't know 12:41:51 akrivoka, can you comment? 12:42:08 david-lyle: it is definitely still of interest 12:42:41 ok, is it ready to release? 12:43:12 generally at this time of the cycle someone asks me to release it 12:43:12 david-lyle: let me double check and get back to you soon 12:43:18 akrivoka: thanks! 12:43:27 david-lyle: thank you! 12:47:25 anyone have another item for today? 12:48:03 I will be posting release notes soon to the wiki page for the release 12:49:09 I'm contemplating making the command decision not to merge and rst version in the horizon tree this time 12:49:20 i looked and no other service seems to do that 12:49:38 and I missed the release cut-off 12:49:45 and issues with that 12:50:00 I don't quite follow what you are not going to do ... 12:50:02 s/and/any/ 12:50:32 I found release notes useful in the past 12:50:43 typically release notes for say Kilo are posted to an openstack wiki page and in the source tree 12:50:50 and I must admit, I never looked at them in the wiki 12:51:05 david-lyle: we have one patch we need to land in tuskar-ui before we're ready to release 12:51:14 david-lyle: will let you know when we're ready 12:51:15 ... try to find something in the wiki, that is awful 12:51:19 on the wiki they are combined with all the other services 12:51:41 akrivoka: thank you, once that's merged, I can release 12:52:02 david-lyle: thanks! 12:52:22 mrunge: maybe I can try to slip it into RC-2 12:52:54 just wasn't sure anyone used them 12:53:00 david-lyle, I didn't even knew you were supposed to prepare two different formats 12:53:29 david-lyle, if I'm the only one using them, I'd say: wiki is fine 12:53:36 mrunge: not a big deal 12:54:05 I'll see where I get with them in the next few days 12:54:13 cool 12:54:20 I'll need reviews either way 12:54:28 I always miss items 12:54:40 I just wanted to mention, release notes are really read by humans :D 12:56:44 mrunge: as long as someone finds them useful and I'm not just writing to myself, I'm ok with taking the effort 12:58:31 I'll call it for today. Thanks everyone for your efforts on Kilo, please take the time to test the RC so we don't ship anything major that could have been addressed. 12:58:38 Have a great week! 12:58:42 #endmeeting