22:01:50 #startmeeting horizon 22:01:51 Meeting started Tue Mar 26 22:01:50 2013 UTC. The chair is gabrielhurley. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:01:52 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 22:01:55 The meeting name has been set to 'horizon' 22:02:03 #topic general updates 22:02:17 So, we've got an RC2 on our hands 22:02:19 that's okay 22:02:29 hooray! 22:02:31 thankfully most of the fixes are already in progress 22:02:59 this new bug that came up during the previous meeting is confusing the heck out of me though... I'll talk about that in a minute 22:03:20 anyhow, the goal is to get RC2 back down to zero by EOD Thursday 22:03:31 so keep an eye on the reviews 22:04:08 Should RC2 bug be submitted to master first or to the milestones proposed branch directly? 22:04:16 I was just typing that :-) 22:04:18 yes 22:04:23 to master first 22:04:28 then backport to the milestone-proposed branch 22:04:28 Ok :) 22:04:41 after it's approved on master? 22:04:46 ideally 22:04:55 Cool 22:04:59 the one to proposed should at least not be merged until the one to master is 22:05:20 but if you don't mind re-spinning both PRs if something needs changing you can propose both at the same time 22:06:55 anyhow, otherwise things are generally stable, and the only other topic of note is that anyone who wants to propose summit topics needs to do so this week 22:07:00 summit.openstack.org 22:07:12 #topic bugs and blueprints 22:07:27 so, I'm a bit lost on https://bugs.launchpad.net/quantum/+bug/1160580 22:07:28 Launchpad bug 1160580 in quantum "Allow extensions already implemented in Quantum in Brocade plugin" [Undecided,New] 22:07:55 I'm gonna follow up with the quantum folks, but I *think* we need another setting to control for this quantum plugin's differences. 22:08:01 I'm just not sure what those differences actually are yet. 22:08:19 reading the ticket leads me to believe it's a floating IP issue, but an email thread on the topic sounded more like a router issue. 22:08:21 I'm not sure. 22:08:37 However, if the change is invasive I'm actually not gonna take it for RC2, just FYI. 22:08:50 I'd rather release note that a specific quantum plugin is not compatible than risk breaking everything. 22:09:57 anybody else got bugs they're concerned about or think might deserve to be fixed in RC2? 22:10:57 I'll take the silence as no. feel free to raise them in the next section if needed. 22:11:02 #topic open discussion 22:11:20 This is mostly a quiet week, so I don't have a lot that needs discussing 22:11:30 I'm open to talking about bugs, or about Havana stuff 22:11:33 or summit ideas 22:11:39 whatever y'all have on your minds 22:11:56 I've proposed http://summit.openstack.org/cfp/details/196 - discussion for folks who build/extend Horizon. 22:13:00 I approved that one as soon as I saw it. I usually am the one to open that session, and it's always useful, so I'm happy to see someone else beat me to it. :-) 22:14:18 I'm curious there is a blueprint https://blueprints.launchpad.net/horizon/+spec/global-ajax-communication is there any concrete plan to get there, or would this be a good discussion topic? 22:14:20 I'm helping some internal folks at HP who have been modifying an old version of Horizon for their purposes get on top of the latest and greatest. They're definitely testing the limits of Horizon. I believe one or two of these folks will be at said session so I think it'll be a good opportunity to get food feedback. 22:14:34 *good 22:15:13 cody-somerville: awesome. I've done a lot of hacking on horizon outside of the "OpenStack Dashboard" context and that's lead to some useful things that landed in master, so I'm always interested in seeing where other people find pain points. 22:15:24 david-lyle: yes and no, lemme explain 22:15:43 There is both the case of using Horizon outside of OpenStack dashboard and also highly customizing the OpenStack Dashboard. 22:16:07 That particular blueprint is sort of a placeholder. It was proposed a while back and quickly got drowned in overscoping. 22:16:25 there was a desire to make it a blueprint about all of ajax and real-time everything, and it died on the vine. 22:16:31 However, I opened this session topic: http://summit.openstack.org/cfp/edit/210 22:17:01 which aims to look at "now that we're a lot farther down the road, can we actually make real-time data-driven interactions happen in Horizon?" 22:17:15 so I think that blueprint would fit in as an aspect of that session 22:17:29 gabrielhurley: I get forbidden with that link 22:17:32 cody-somerville: both are true and valid, absolutely. 22:17:46 nevermind found 22:17:47 it 22:17:56 david-lyle: let's try http://summit.openstack.org/cfp/details/210 22:17:59 ah, I see you got it 22:18:40 ok, that sounds good 22:18:49 gabrielhurley: Sorry, I meant specifically that there is a group at HP that *is* using Horizon in both capacities. 22:19:05 yeah, I figured 22:19:13 the other big item I don't see is Keystone v3 22:19:28 hmmm... a session strictly on that would be good, actually. 22:19:39 I'd like to discuss both v3 API and multi-region/multi-endpoint stuff. 22:19:47 yes 22:19:50 david-lyle: you wanna propose that session? 22:19:53 sure 22:19:58 (it looks bad when I propose everything) 22:19:59 ;-) 22:20:10 good idea on that 22:21:11 I feel like Horizon should just have X slots at the summit where X is the number of "integrated" projects, and each project gets a slot. lol. 22:21:49 "How can we better serve your project?" 22:21:52 :-P 22:22:08 okay. anybody else have thoughts? 22:22:08 One thing that has struck me is that the OpenStack Dashboard in some ways is almost like an example implementation of Horizon. I think it would be useful for this to be reframed. I believe openstack_dashboard should be implemented in such a way that it is highly reusable so that one can do highly customized *openstack* dashboards, not just extended openstack dashboards. 22:22:31 cody-somerville: hmmmm... I see your point 22:22:35 both are very valuable 22:22:42 I wonder if a session on testing would be useful. The unit tests can be quite complex for new contributors to get their head around, and from another angle we could also use improvements on the selenium front 22:23:01 +1 22:23:04 +1 22:23:11 jpich: I'd be willing to lead or co-lead that if you want to propose it. I'm nt sure how well attended it'd be, but it's absolutely useful. 22:23:50 Would it be a tutorial or working session to improve Horizon tests? or hybrid of both? 22:23:54 cody-somerville: I think that topic could be rolled into the BoF session you listed, most likely. Do you think it deserves more? 22:24:07 Cool, I'll propose it and would welcome all help on leading/directing it 22:24:08 cody-somerville: was thinking tutorial for at least a good portion of it 22:24:23 kind of a "tips and tricks" 22:24:29 'cuz I know every kind of error it can throw 22:24:51 Sounds like I'd like to attend it too then ;) 22:25:06 gabrielhurley: Aye. Sounds good to both. 22:25:17 cool 22:25:40 Another question: https://bugs.launchpad.net/horizon/+bug/1157950 -- this says nova added support for fixed_ips quotas in master, folsom, and essex. Does that mean we should aim to backport the Horizon patch on folsom and essex as well? 22:25:42 Launchpad bug 1157950 in horizon "Add support for nova's new Fixed IP quota" [High,In progress] 22:26:13 jpich: Folsom definitely... I'm not clear on supporting essex currently. 22:26:37 I have no idea how much pain would be involved in backporting that fix to essex 22:26:48 if it's trivial it might be worth investigating 22:27:04 I wonder if increasing the Folsom novaclient required version could cause trouble for distros? 22:27:09 honestly I'm surprised the fix got ported all the way back to essex in nova 22:27:32 I guess I'll have a look then. I'm not familiar at all with Essex 22:27:34 jpich: it's a valid question. I'd take it to the mailing list 22:27:47 Ok! 22:27:54 thankfully the backports don't have a hard deadline the same way the RC2 does 22:30:12 gabrielhurley: This might already been answered, when is the deadline for RC2? :) 22:30:21 end of day thursday 22:31:06 Alright folks. I'm gonna call the meeting here. 22:31:13 I'll follow up on quantum-related issues 22:31:23 keep working on the bugs you're each fixing 22:31:24 I have a patch that is ready for review master. how do I get the patch submitted for RC2? 22:31:34 Sorry for the last minute question.. 22:31:37 lcheng_: link to the ticket? 22:31:45 https://bugs.launchpad.net/horizon/+bug/1148372 22:31:46 Launchpad bug 1148372 in horizon "edit flavor deletes the flavor if special char is used" [Medium,In progress] 22:32:08 I targeted that for to the RC2 milestone in launchpad this morning 22:32:23 once the review gets merged to amster we'll port it to the milestone-proposed branch 22:32:27 *master 22:32:39 Got it. 22:32:42 thanks 22:32:46 no problem 22:32:51 alright. have a great week folks! 22:32:54 #endmeeting