11:59:45 <david-lyle_afk> #startmeeting Horizon
11:59:45 <openstack> Meeting started Wed Dec 10 11:59:45 2014 UTC and is due to finish in 60 minutes.  The chair is david-lyle_afk. Information about MeetBot at http://wiki.debian.org/MeetBot.
11:59:46 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
11:59:49 <openstack> The meeting name has been set to 'horizon'
11:59:56 <david-lyle_afk> Who's here to talk Horizon?
12:00:02 <akrivoka> hello \o
12:00:06 <zhenguo> hello
12:00:15 <ygbo> Hello
12:00:28 <tsufiev> hi
12:00:31 <mrunge> good morning!
12:00:31 <rbertram> hi
12:00:35 <Jean-Brice> hi
12:00:41 <neillc> hello
12:00:45 <doug-fish> hi
12:01:43 <david-lyle_afk> Let's get rolling
12:02:27 <david-lyle_afk> Dec 18 is the end of K-1
12:02:39 <david-lyle_afk> #link https://launchpad.net/horizon/+milestone/kilo-1
12:02:53 <david-lyle_afk> We're really in good shape for that
12:03:25 <david-lyle_afk> so nice job everyone, a couple more of those could use code reviews
12:03:50 <david-lyle_afk> And the couple without code may not make it
12:04:17 <david-lyle_afk> Items that don't make it will slip to k-2
12:04:35 <david-lyle_afk> which I have planned out for the most part as well #link https://launchpad.net/horizon/+milestone/kilo-2
12:05:07 <david-lyle_afk> As blueprints are reviewed, there may be some additions
12:05:58 <david-lyle_afk> beyond milestone information, just a reminder that the former release management meeting is not a cross-project issues meeting
12:06:09 <david-lyle_afk> 2100 UTC on Tues
12:06:19 <david-lyle_afk> all are welcome to attend
12:06:26 <tsufiev> david-lyle_afk, https://blueprints.launchpad.net/horizon/+spec/images-integration-tests - didn't we have a single blueprint for all new integration tests?
12:06:50 <tsufiev> I mean https://blueprints.launchpad.net/horizon/+spec/selenium-integration-testing
12:07:02 <david-lyle_afk> although I must confess that a lot of the cross-project topics don't line up with us very much
12:07:33 <jpich> I would encourage people to open bugs for new integration tests and tag them with integration-tests, rather than create blueprints
12:08:27 <david-lyle_afk> tsufiev: I'm not a big fan of catch-all blueprints, mostly because you never know if they're done
12:08:51 <mrunge> +1 david-lyle_afk
12:09:03 <tsufiev> david-lyle_afk, yep, makes sense. I've just seen how many CRs are already here :)
12:09:15 <david-lyle> I think the images one looks ok because it mention the previous bp
12:09:29 <tsufiev> s/here/there/
12:10:46 <david-lyle> any questions about release scheduling? k-3 is not fully fleshed out yet, but a lot will likely be overflow from k-2
12:11:42 <david-lyle> I know a few more bps need review and I will add those to the blueprint review page, as soon as I'm not distracted
12:12:10 <david-lyle> #link https://wiki.openstack.org/wiki/Horizon/Blueprint_Reviews
12:12:29 <david-lyle> most on the first list are scheduled, a couple require further clarification first
12:12:44 <david-lyle> ok, moving on
12:13:08 <david-lyle> ok, actually one more general item regarding the release
12:13:43 <david-lyle> in the team meeting last week, and on the mailing list we announced that we are canceling the repo split
12:14:12 <ygbo> is it a temperary or definit cancel?
12:14:38 <ygbo> s/temperary/temporary/
12:14:43 <david-lyle> ygbo: at this point I'd say permanent
12:14:48 <mrunge> ygbo, permanent
12:14:58 <tsufiev> I guess if everything will be ok with Horizon, there won't be any time safe for split :)
12:15:25 <tsufiev> i.e. calm enough
12:15:36 <mrunge> tsufiev, IMHO, that's not the point
12:15:49 <david-lyle> the direction of the project is away from the django based toolkit, and taking a lot of time and energy to split and maintain it is not really inline with that
12:15:51 <mrunge> and we'd find a time.
12:16:23 <tsufiev> mrunge, ok
12:17:10 <david-lyle> ok, now moving on
12:17:34 <david-lyle_afk> #topic Next step with 3rd party components & packaging (xstatic / bower)
12:17:55 * david-lyle hates that he forgot to switch nick before starting this
12:18:04 <mrunge> ;-)
12:19:58 <david-lyle> I don't see richard in the room
12:20:10 <david-lyle> and I think the next item is his as well
12:20:42 <david-lyle> hmm, suboptimal
12:21:08 <neillc> He was planning to be here.
12:21:08 <mrunge> david-lyle, I just added this to the agenda
12:21:25 <mrunge> since it was a bit overflow from last week
12:22:17 <david-lyle> ok, I think my summary of the current status of 3rd party packaging is we're stuck with xstatic for the time being
12:22:30 <mrunge> there was another item from Dec 3: cinder rest api...
12:23:04 <david-lyle> mrunge: DuncanT jumped in a bit a the end, but we could talk about it
12:23:23 <rbertram> stuck with xstatoc for Kilo?
12:23:43 <rbertram> does "for the time being" mean "for Kilo"?
12:23:53 <mrunge> what's the issue with xstatic approach?
12:24:00 <david-lyle> rbertram: to make forward progress
12:24:17 <rbertram> ok
12:24:36 <david-lyle> mrunge: other than it being a bit clunky and a repeat of say what bower is doing, I'm not sure
12:24:43 <rbertram> mrunge: I'm not advocating either side, due to lack of experience w/ bower. Just wanting to know what to do.
12:25:38 <david-lyle> pulling tagged bits from a repo would make releasing difficult for distributions
12:25:47 * david-lyle understates
12:26:05 <mrunge> my idea currently was to drop xstatic etc. to replace installs from bower with distro packaging
12:26:28 <mrunge> not sure, how this would work on the gate...
12:27:24 <david-lyle> mrunge: should put the js in an rpm without the xstatic wrappings?
12:27:37 <mrunge> david-lyle, at least, we can
12:27:48 <david-lyle> s/should/would/
12:27:59 <david-lyle> question rather than implication
12:28:12 <mrunge> david-lyle, and I assume, the same will work for .deb
12:28:42 <david-lyle> gates will work with slightly different mechanism
12:28:44 <mrunge> but as far as I understand the gate, every dep is installed via pip install....
12:29:02 <david-lyle> storyboard has gone down that path
12:29:31 <david-lyle> dynamically installs the toolset
12:29:35 <david-lyle> and dependencies
12:31:19 <david-lyle_afk> #topic Open Discussion
12:31:42 <david-lyle> jumping here, if richard makes it, we'll go back
12:32:17 <david-lyle> I will mention the policy proposal here
12:32:31 <david-lyle> #link https://review.openstack.org/#/c/136980/
12:32:35 <david-lyle> this is from Cinder
12:33:05 <david-lyle> I met with them at their mid-cycle in Juno and talked a bit about our policy difficulties
12:33:06 <doug-fish> I was just looking at that - it seems like a good idea at a high level
12:33:28 <david-lyle> it would be nice to push that to the APIs
12:34:00 <david-lyle> the potential short-coming I see are splintered implementation across services
12:34:12 <doug-fish> david-lyle: do you think we can take advantage before all of the APIs have it?
12:34:19 <david-lyle> and the target information is not always general
12:34:28 <david-lyle> doug-fish: certainly
12:34:51 <david-lyle> but it's more service specific code, like pagination, etc
12:34:52 <amotoki> iirc, there was a dicsussion on common policy service in a context of keystone at Paris, thogh I am not aware of the status/progress.
12:35:14 <david-lyle> amotoki: the first step for that would be a policy store
12:35:34 <david-lyle> essentially one true source for policy.json files in the cloud
12:36:09 <amotoki> david-lyle: agree
12:36:53 <doug-fish> is there any blueprint/etherpad/wiki regarding that discussion?
12:37:03 <david-lyle> I think there are grander aspirations to maybe provide something similar to what Cinder is proposing, but not entirely sure that an outside service is best equipped to make those determinations
12:37:09 <david-lyle> doug-fish: keystone wise?
12:37:13 <doug-fish> yeah
12:37:24 <david-lyle> I'm sure there's a spec
12:37:39 <david-lyle> they wouldn't have a bp until the spec passes
12:37:44 <doug-fish> oh sure
12:37:51 <doug-fish> those guys are formal.  :-)
12:39:43 <david-lyle> at quick glance I see 4 policy related specs in keystone-specs, but none seem to be what we're discussing
12:40:32 <david-lyle> #link https://review.openstack.org/#/c/134655/ seems related
12:41:32 <david-lyle> So, I'll provide some feedback on the proposal, but I encourage others to take a look
12:41:38 <david-lyle> other topics
12:41:39 <david-lyle> ?
12:43:04 <david-lyle> There have been more questions regarding a mid-cycle meetup for Horizon.
12:43:26 <david-lyle> My thoughts are that while I feel it would be useful and probably very productive,
12:44:03 <david-lyle> the globally distributed nature of our team makes this a high burden on 2/3s of the team to travel
12:44:31 <david-lyle> I also worry about the timing of a topic target meetup
12:45:18 <david-lyle> I am open to the idea, but I have never got too many we really shoulds, other than from US folks who assume it will be here
12:45:24 <david-lyle> other thoughts on that?
12:46:19 <david-lyle> I think trying to work on some online collaboration times would be beneficial and meet many of the same objectives
12:46:27 <doug-fish> do you know how other projects handle that?
12:46:27 <doug-fish> that= picking a location
12:47:00 <david-lyle> doug-fish: they ask hosts to volunteer, then select based on where the team can make it
12:47:20 <david-lyle> or reverse that order
12:47:55 <david-lyle> then schedule a date and try to get discounted hotel blocks etc
12:47:59 <amotoki> I think a location is not so important. Nova helds the mid-cycle at bay area and neutron has it at salt-lake city on a different schedule.
12:48:10 <amotoki> They have volunteer to offer spaces.
12:48:31 <david-lyle> amotoki: we have volunteers too
12:48:43 <jpich> +1 to figuring out how to better collaborate online with the whole community and relying less on face-to-face meetings
12:49:06 <sambetts> jpich: +1
12:49:07 <david-lyle> but the ones I know of are in the US, and I'm not sure how many folks outside the US would be able to travel
12:49:28 <david-lyle> It feels like a high bar
12:49:44 <doug-fish> agreed.
12:50:13 <zhenguo> jpich: +1
12:51:16 <david-lyle> I know infra just conducted a meetup online, I'll talk to them and get more details on how they did it
12:51:35 <doug-fish> cool
12:51:52 <david-lyle> I'm not sure google hangouts scales enough or works for everyone, I know they went a different route
12:52:27 <david-lyle> other items?
12:52:35 <neillc> I believe google hangouts have a limte for number of participants
12:52:48 <david-lyle> neillc: yeah, I think it's 10
12:53:00 <tsufiev> david-lyle, afaik hangouts has 15 or 20 people limitation.
12:53:10 <david-lyle> ok more than I thought
12:53:14 <tsufiev> WebEx?
12:53:27 <neillc> I think 10 may be right. We've hit it with our team which is < 15
12:53:58 <amotoki> i wonder how the bp review process works? I am sure how to know/share their review status...
12:54:01 <bradjones> 15 if you have a business account otherwise 10
12:54:37 <david-lyle> amotoki: so far I've been putting blueprints in the review state that I thought were ready for review
12:54:37 <tsufiev> we've been using WebEx for meetings with large number of participants (in Mirantis)
12:54:56 <david-lyle> trying to add to wiki, when I remember
12:55:04 <david-lyle> imperfect solution
12:55:35 <david-lyle> thinking I'm going to advocate moving to specs in L
12:56:11 <ygbo> thers is http://bluejeans.com/
12:56:22 <ygbo> I don't know the limit
12:56:37 <ygbo> but read-hat people might know since they use it :-)
12:56:42 <amotoki> david-lyle: thanks. updating the status on wiki sounds good at the moment.
12:56:45 <david-lyle> once they are in the review state and I've publicized them, if I don't get negative feedback and I also agree with the direction of the bp, I schedule it
12:56:46 <ygbo> s/read-hat/red-hat/
12:57:07 <david-lyle> amotoki: sure, I'll make sure I do that
12:57:16 <david-lyle> will save so many clicks
12:57:27 <amotoki> :-)
12:58:22 <david-lyle> I want to draw attention to the rest API patches and Identity panel reworks patches that Richard and Thai have been working on
12:58:33 <david-lyle> make sure to review those
12:58:43 * david-lyle doesn't have links handy
12:59:15 <david-lyle> we'd like to have this in good shape by early k-2 so we can build on it
13:00:12 <david-lyle> times up. Thanks everyone for finding the new meeting time. Have a great week!
13:00:21 <akrivoka> thanks everyone
13:00:23 <david-lyle_afk> #endmeeting