22:01:05 <gabrielhurley> #startmeeting horizon
22:01:06 <openstack> Meeting started Tue Sep  3 22:01:05 2013 UTC and is due to finish in 60 minutes.  The chair is gabrielhurley. Information about MeetBot at http://wiki.debian.org/MeetBot.
22:01:07 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
22:01:09 <openstack> The meeting name has been set to 'horizon'
22:01:11 <gabrielhurley> #topic overview
22:01:19 <gabrielhurley> Boy oh boy has it been a busy week!
22:01:22 <gabrielhurley> Hi folks!
22:01:26 <lsmola_> hello
22:01:27 <jpich> Hello
22:01:28 <david-lyle> Hello
22:01:30 <lblanchard> hey everyone
22:01:42 <jcoufal> hey there
22:01:46 <absubram_> hi
22:01:52 <gabrielhurley> So, first off, AMAZING job getting all that code done and all those reviews in. We all really kicked ass the last two weeks.
22:02:00 <amotoki_> hi, long time no see
22:02:28 <gabrielhurley> We actually got a "leaves me speechless" from ttx for having completed all that
22:02:43 <gabrielhurley> everyone should feel really positive
22:02:53 <jcoufal> the highest title possible :)
22:03:03 <gabrielhurley> So if you haven't heard, we're at the end of the H3 milestone. ;-)
22:03:11 <gabrielhurley> The deadline for merges is EOD tomorrow
22:03:25 <gabrielhurley> What's left in https://launchpad.net/horizon/+milestone/havana-3 is basically what I'd like to see
22:03:46 <gabrielhurley> there's a handful of bugfix reviews (two of which i posted today) that need an additional +2
22:03:51 <gabrielhurley> and one blueprint left
22:03:58 <gabrielhurley> https://blueprints.launchpad.net/horizon/+spec/horizon-cisco-n1k
22:04:12 <gabrielhurley> We'll chat a little more on that in the blueprints topic
22:04:17 <absubram_> ok thanks
22:04:19 <gabrielhurley> before that, a couple more general bits of business
22:05:10 <gabrielhurley> Some of you may have heard that the Technical Committee is going to move to having formal documents like the TC charter stored in github/gerrit so that changes can be officially voted on and tracked in a better way.
22:05:20 <gabrielhurley> Connected to that, the review for the initial set of documents it here: https://review.openstack.org/#/c/44489/2
22:05:30 <gabrielhurley> This isn't just general interest, though
22:05:41 <gabrielhurley> I specifically wanted to point out https://review.openstack.org/#/c/44489/2/reference/extra-atcs
22:06:05 <gabrielhurley> the Horizon team has the only two "extra ATCs" in OpenStack, 'cuz we're just that cool. :-)
22:06:24 <lblanchard> :)
22:06:35 <jcoufal> :) horizon is the best!
22:06:52 <jcoufal> go Horizon, we need more!
22:07:04 <lsmola_> :-)
22:07:05 <gabrielhurley> On a different but related note, I know you have all been waiting for the design summit session proposals to open. Well, I'm gonna steal thierry's thunder and leak the news that it's now live: http://summit.openstack.org/
22:07:19 <gabrielhurley> that'll go out to the ML sometime today or tomorrow
22:07:35 <gabrielhurley> the PTLs and such had early notification to make sure it was working and kick the tires
22:07:47 <gabrielhurley> so basically, go ahead and start proposing
22:08:10 <gabrielhurley> propose as much as you'd like. If I see things that are better suited to other topics, or similar proposals that can be combined I'll start contacting the porposers
22:08:19 <ttx> gabrielhurley: jcoufal already noticed it
22:08:25 <gabrielhurley> I saw ;-)
22:08:33 <gabrielhurley> It wasn't super-secret
22:08:37 <gabrielhurley> just wasn't announced
22:08:53 <ttx> I got lazy and did the beta opening on the public port
22:09:09 <jcoufal> I am not a cheater, I was just lucky :)
22:09:14 <gabrielhurley> ttx: tsk tsk. security would have a field day with you.
22:09:16 <gabrielhurley> ;-)
22:09:34 <gabrielhurley> anyhow, back to main business
22:09:45 <gabrielhurley> So, the Feature Freeze is EOD tomorrow. String Freeze coincides with that.
22:10:11 <Daisy> any exceptions?
22:10:12 <gabrielhurley> So, be extra-mindful of changing translated strings in any bugfix reviews, etc. that happen in RC1
22:10:33 <gabrielhurley> There isn't really a formal process for "string freeze exceptions"
22:10:48 <gabrielhurley> so far it's just a "pay attention to what's changing and try not to do it"
22:10:57 <Daisy> If translation team run the Horizon in their locale environment, they may find some bugs.
22:11:04 <Daisy> are there a deadline for these bugs reporting?
22:11:21 <gabrielhurley> Daisy: I'd be happy to see bugs/patches from the translation team, because they're the ones being affected by it
22:11:41 <gabrielhurley> We just have to make sure that the changes get sync'd between github and transifex if things do get changed
22:12:03 <Daisy> so any string changing need to change the po files too.
22:12:06 <gabrielhurley> Daisy: when you set up the new translation resources, did you reconnect the auto-updating of the Transifex resources from github?
22:12:18 <Daisy> I use this checkbox: auto sync.
22:12:23 <gabrielhurley> heh. sounds good
22:12:40 <Daisy> only en po files are enough.
22:12:43 <Daisy> no need to change all.
22:12:48 <gabrielhurley> so as long as we run the script to update the base translation files those'll get sync'd over
22:12:50 <gabrielhurley> yep
22:12:54 <gabrielhurley> perfect
22:12:56 <Daisy> yes, I think so.
22:13:34 <gabrielhurley> so like I was saying, be mindful of string changes. If you *do* see any, make sure that the base en po files get updated in the same commit so we can keep everything in sync
22:13:41 <Daisy> "be extra-mindful of changing translated strings in any bugfix reviews, etc. that happen in RC1"
22:13:56 <Daisy> is that mean, only before RC1 allow string changing?
22:14:00 <gabrielhurley> yep
22:14:26 <gabrielhurley> at least "arbitrary" string changing
22:14:26 <Daisy> or translation changing before RC1?
22:14:40 <gabrielhurley> changing strings marked for translation in the codebase
22:14:51 <jpich> Any thought about a FFE for https://blueprints.launchpad.net/horizon/+spec/messages-on-login-page if the solution in the current review doesn't become more complex? (There would be string changes)
22:15:14 <gabrielhurley> tbh I'm not up-to-date on the current solution there
22:15:29 <Daisy> when is RC1?
22:16:30 <gabrielhurley> jpich: glancing at it currently that's probably reasonable for now. let's look to land that in the next 24 hours if we're going to.
22:16:37 <jpich> It uses a short-lived cookie to carry the log out reason. It's ok if it slips, though it would be a nice UX improvement especially when it comes to the new session timeouts (since no reason is given to the user at the moment)
22:16:45 <gabrielhurley> Daisy: RC1 will be cut on Thursday
22:17:00 <gabrielhurley> jpich: agreed
22:17:08 <gabrielhurley> let me put the BP back into H3
22:17:09 <jpich> gabrielhurley: Ok, cheers
22:18:34 <gabrielhurley> I think that's all the general business
22:18:41 <gabrielhurley> #topic bleuprints
22:19:33 <gabrielhurley> So, we just added https://blueprints.launchpad.net/horizon/+spec/messages-on-login-page back to the list
22:19:58 <gabrielhurley> there seems to just be the one question about the whether the trans block in the template is appropriate or not, otherwise I see no reason that can't merge
22:20:29 <gabrielhurley> so the other BP remaining is https://blueprints.launchpad.net/horizon/+spec/horizon-cisco-n1k
22:20:53 <absubram_> yes :)
22:21:31 <gabrielhurley> there have been many many many iterations on that review. it currently has had all the comments addressed.
22:22:16 <absubram_> yes.. hopefully it'll get a +1 from Paul and amotoki on this iteration?
22:22:37 <gabrielhurley> I feel like there may be some impetus to do a little reorganization in I, particularly when we're looking at navigation changes, but for now it seems like adding this in is probably still acceptable given the extreme isolation and narrow scope of the functionality...
22:22:54 <gabrielhurley> anyone else have other thoughts?
22:23:29 <david-lyle> I honestly haven't had a chance to look through it yet, will try to tonight
22:24:03 <absubram_> yes.. I definitely have to re-org the unit tests.. I'd really appreciate if we can get this infra in so that I can work on the enhancements in smaller diffs
22:24:30 <absubram_> it is by default not turned on now though..
22:24:39 <absubram_> it would be great if you could david-lyle
22:24:58 <amotoki_> my concern is the navigation structure and i am not sure the policy of adding new dashboard.
22:25:21 <gabrielhurley> the navigation isn't ideal, but we can't really fix that in H and will revisit in I
22:25:28 <amotoki_> i hope other folks chimes in and hear their opinions.
22:25:29 <jcoufal> In I we would like to review information architecture for Horizon in general
22:25:40 <gabrielhurley> as for the policy of adding a new dashboard, there isn't one.
22:25:53 <gabrielhurley> tbh I don't think it's the right solution here, but I'm not sure what it is
22:25:54 <david-lyle> use short names is the policy, I think
22:25:58 <gabrielhurley> s/it is/is
22:26:06 <absubram_> :)
22:26:20 <gabrielhurley> that's why I said earlier that I think we may want to move/reorganize this entire thing in I
22:26:31 <gabrielhurley> it's sort of a canary to see if we end up being okay with it or not
22:26:31 <david-lyle> absolutely
22:26:37 <gabrielhurley> then we can build a policy around that
22:26:40 <absubram_> sure
22:26:49 <jcoufal> agree
22:27:01 <lblanchard> agreed
22:27:06 <amotoki_> agree
22:27:07 <absubram_> reorg I agree.. you;d like to move it itself?
22:27:08 <gabrielhurley> but absubram_ has made a very good faith effort this entire release cycle to get it in and I don't think he should be punished for us being unclear on what we want in this regard. ;-)
22:27:11 <lsmola_> sounds reasonable
22:27:20 <absubram_> haha.. thank you!
22:27:24 <david-lyle> +1
22:27:41 <gabrielhurley> like I said, I don't love it being its own dashboard, but we don't *have* a home for this. we should come up with one at the summit
22:27:58 <absubram_> I'm planning on proposing that already as a topic ;)
22:28:20 <jcoufal> +1
22:28:31 <ftcjeff> +1
22:28:36 <lsmola_> +1
22:28:41 <gabrielhurley> it'll also dovetail with jcoufal's proposal on the future of UX Direction of the dashboard, but probably won't overlap depending on how you phrase things.
22:28:51 <gabrielhurley> so, +1
22:29:16 <Toshi> +1
22:29:25 <gabrielhurley> alright, it sounds like we're generally okay with the functionality, and agree it should have *some* home in the dashboard, so let's go forward with it and try to get it reviewed and merged in the next 24 hours as well.
22:29:41 <absubram_> thank you!
22:29:47 <absubram_> :))
22:29:59 <david-lyle> absubram: thanks for sticking with it
22:30:13 <gabrielhurley> great!
22:30:16 <gabrielhurley> #topic bugs
22:30:22 <jcoufal> I think that the whole IA/navigation enhancements might get it's own session, I expect pretty big discussions around this
22:30:27 <jcoufal> (sry late)
22:30:30 <gabrielhurley> agreed
22:30:44 <gabrielhurley> Real quick I want to call attention to https://review.openstack.org/#/c/44929/ and https://review.openstack.org/#/c/44938/ that need an extra +2 for the H3 cut
22:31:17 <gabrielhurley> and wanted to point out that I've moved a few things that seemed urgent into https://launchpad.net/horizon/+milestone/havana-rc1 with the "high" priority
22:31:42 <gabrielhurley> that's about it for bugs at the moment
22:32:03 <gabrielhurley> one of those two reviews I posted will fix the docs generation, which will make me happy.
22:32:06 <jpich> Yay to working API references again
22:32:08 <gabrielhurley> the docs being all broken made me sad
22:32:18 <david-lyle> https://bugs.launchpad.net/horizon/+bug/1212748 needs a release of openstack-auth to close
22:32:19 <uvirtbot> Launchpad bug 1212748 in horizon "log in for user with first project disabled fails" [High,In progress]
22:32:30 <gabrielhurley> david-lyle: yep. I'm gonna do that today
22:32:42 <david-lyle> ok, thanks!
22:33:12 <gabrielhurley> we won't need to do anything in the hrizon requirements.txt, though, since it's currently a >=1.1.1
22:33:26 <gabrielhurley> so I'll just bump the version on openstack-auth and push a new tag
22:33:33 <david-lyle> sounds good
22:33:45 <gabrielhurley> #topic open discussion
22:33:58 <gabrielhurley> I think that's it for business that must be discussed this week
22:34:08 <gabrielhurley> the next phase (RC1) is all about testing and polishing
22:34:17 <gabrielhurley> so be sure to, you know, *use* the dashboard and stuff. ;-)
22:34:41 <lsmola_> :-)
22:34:44 <gabrielhurley> particularly the new features like keystone domains/groups, heat, and ceilometer
22:34:48 <Daisy> I think the translation team need very detail guideline to "use" dashboard.
22:35:26 <gabrielhurley> Daisy: devstack is the place to start. that gives you a full running openstack. from there I suggest looking at the docs teams' basic users guides
22:35:27 <Daisy> how can we get the guidelines? are there any test cases?
22:35:46 <gabrielhurley> I'm not sure what kind of guidelines/tests you're looking for?
22:35:52 <amotoki_> now i am writing a guideline how to check their own translations  using devstack.
22:36:05 <Daisy> amotoki_: when can it be ready?
22:36:18 <amotoki_> will share this week.
22:36:23 <Daisy> thanks.
22:36:25 <gabrielhurley> amotoki_: great
22:36:33 <amotoki_> now i am busy on review rushes
22:36:38 <Daisy> I think I'm looking for the new features you point out.
22:36:44 <gabrielhurley> hopefully it's not much more than "run devstack, log in to horizon, go to user settings, change language"
22:37:20 <gabrielhurley> from there "use the dashboard" is very nebulous, but describing how to use every feature of every project is the docs teams' purview and they already have substantial books on the subject
22:37:22 <gabrielhurley> there's no short answer
22:37:26 <amotoki_> how to get translations from transifex , update po/mo files and so on.
22:37:27 <Daisy> amotoki_ writes the guide lines to use the translations in their environment, not the test guideline.
22:37:36 <gabrielhurley> ah
22:37:44 <gabrielhurley> gotcha, like pulling down the translations from transifex
22:37:45 <gabrielhurley> that makes sense
22:38:08 <Daisy> I think we need to regard translators as people who don't understand openstack that much.
22:38:44 <Daisy> but they can do as what the guideline writes.
22:39:15 <Daisy> maybe I will write such one for translation team.
22:39:34 <gabrielhurley> No question there. I just don't think writing a separate set of "here's how to use an OpenStack cloud" for translators helps, because the informationshouldn't be different than for anyone else trying to use OpenStack, and if you write something too narrow then everyone looks at the exact same parts of the dashboard and that doesn't encourage getting good coverage.
22:40:22 <Daisy> I will try to look for something from doc team.
22:40:34 <Daisy> If there are existing url, that would be very helpful.
22:40:41 <gabrielhurley> I'm not arguing there's a disconnect and that this is a problem, I just want to make sure we don't duplicate a lot of effort with docs
22:41:10 <gabrielhurley> Daisy: there definitely are things available online, I don't have a URL off the top of my head. ping annegentle on what she thinks would be most helpful.
22:41:20 <Daisy> ok.
22:42:09 <gabrielhurley> any other topics on people's minds?
22:42:15 <absubram_> david-lyle, amotoki_: just wanted to quickly point out.. I have a link in my blueprint on launchpad that points to a short demo on how the new dashboard was used internally when using the cisco n1k plugin..
22:42:16 <Daisy> another question from me: when can you allow us to import the translations to the code repository? Are there any deadline? Are there any pre-conditions?
22:42:17 <amotoki_> gabrielhurley: what do you think about FWaaS and VPNaaS support?
22:42:28 <amotoki_> Should they go to icehouse?
22:42:29 <gabrielhurley> amotoki_: great targets for I1
22:42:30 <absubram_> if it helps with the review and understanding :)
22:42:56 <amotoki_> I think so.
22:43:11 <amotoki_> they are not tested enough from my impression.
22:43:16 <gabrielhurley> Daisy: the final deadline will be the day before we cut the Havana final release. That'll be several weeks at least. I don't have an exact date.
22:43:30 <gabrielhurley> If it would be helpful to commit interim sets of changes we can do that too
22:43:41 <jcoufal> one more topic here about IE support
22:43:57 <Daisy> ok.
22:44:06 <jcoufal> there is BP around upgrading Bootstrap to v3
22:44:19 <jcoufal> v3 doesn't support IE6 and IE7
22:44:26 <gabrielhurley> amotoki_: I agree on the testing. I want to make sure they get a lot more robust before we land them, and/or that we land them at the beginning of a release cycle instead of the end like LBaaS in G
22:44:32 <jcoufal> so the question to the discussion here is if that's any problem for Horizon
22:44:37 <david-lyle> jcoufal: I think IE9+ would be a good support list
22:44:38 <Daisy> gabrielhurley: so there is also no bug reporting deadlines, right? just before final release?
22:44:49 <gabrielhurley> jcoufal: IE6 and 7 are categorically off the table
22:44:55 <jcoufal> My opinion is that we might be safe stating that we support IE8 or IE9+
22:45:11 <gabrielhurley> I've been delaying the decision on IE8 because the final final extended support drpo date for microsoft in April 2014
22:45:13 <lsmola_> d3 library has ie9+ I think
22:45:14 <jcoufal> gabrielhurley: great, just wanted to assure with more people
22:45:33 <gabrielhurley> after April 2014 I'm calling Horizon IE9+ and that's the end of it
22:45:43 <lsmola_> +1
22:45:44 <jcoufal> Bootstrap is IE8+, so we are safe
22:45:55 <gabrielhurley> Daisy: nope, no deadline for bugs
22:45:56 <jcoufal> +1 for 2014 IE9+
22:46:00 <jcoufal> thanks
22:46:01 <Daisy> ok. thanks.
22:46:05 <lblanchard> +1
22:46:14 <Toshi> +1
22:46:40 <gabrielhurley> In practice that actually means we need to decide between dropping IE8 for Icehouse Final (which will be released in ~March 2014) or in J1... we'll talk about that later though
22:47:14 <david-lyle> other than anything graphical IE8 works currently
22:47:46 <gabrielhurley> yep
22:48:17 <david-lyle> is there a max number of design sessions for Horizon?
22:48:26 <david-lyle> or is it demand based
22:48:37 <gabrielhurley> we do have a limited number of slots
22:48:42 <gabrielhurley> but there's usually a little wiggle room
22:49:09 <gabrielhurley> off the top of my head I don't recall the exact number
22:49:13 <gabrielhurley> it's usually 6 or 7
22:49:20 <gabrielhurley> it's gonna be split across two days this time
22:49:32 <gabrielhurley> so we'll have a couple early in the week and a couple towards the end
22:49:39 <gabrielhurley> I think it's Tuesday and Thursday, offhand
22:49:42 <lblanchard> ah that will be nice to split them...
22:49:45 <gabrielhurley> yeah
22:49:47 <david-lyle> The pagination item obviously effects all services, so hopefully we'll have good turnout from those groups
22:50:02 <gabrielhurley> also, I'm leaving halfway through Friday, so I couldn't schedule any then
22:50:38 <gabrielhurley> david-lyle: yeah. I'll do what I can to not schedule it against any other critical cross-project sessions
22:50:43 <gabrielhurley> this always comes up
22:51:33 <david-lyle> I suppose designated cross-project sessions would be too noisy
22:52:01 <gabrielhurley> I've advocated for a "cross-project" track before and it's never gotten much traction
22:52:12 <Daisy> gabrielhurley: just come up a question: do you know if there is a formal team to test Horizion?
22:52:29 <gabrielhurley> "test Horizon" how?
22:52:29 <Daisy> is QA team responsible for the test ?
22:52:35 <gabrielhurley> the answer is probably "no" regardless
22:52:43 <Daisy> just run and test.
22:52:59 <gabrielhurley> Infra/QA manage the infrastructure that *runs* the tests and gates all the projects. We write our own tests, though.
22:53:09 <Daisy> unit test ?
22:53:41 <ekarlso-> gabrielhurley: how goes the notificitions stuff?
22:53:47 <Daisy> ok. I think I get it. so there is no functional test. Like set up the env, and run to see if there are problems.
22:54:37 <gabrielhurley> Daisy: there are some superficial selenium tests, but not nearly as in-depth as they should be
22:54:45 <gabrielhurley> ekarlso-: what notifications stuff?
22:55:15 <ekarlso-> live notifications or what the bp is called again
22:55:18 <amotoki_> Daisy: almost all projects have only unit tests and few functional tests. devstack-tempest and smokestack are the only system tests.
22:55:18 <david-lyle> real-time?
22:55:24 <ekarlso-> yes
22:55:35 <Daisy> thanks, amotoki_
22:55:37 <gabrielhurley> ekarlso-: push to the I release so we can spend a lot of time on it
22:55:50 <gabrielhurley> we had a ton on our plate supporting two new projects and new APIs in H
22:55:51 <ekarlso-> gabrielhurley: that sucks :p
22:56:09 <gabrielhurley> more contributors makes things go faste... ;-)
22:56:23 <lsmola_> :-)
22:56:30 <gabrielhurley> but it's a significant priority for everyone from a "we know we want to get there and that'll be really cool" standpoint
22:56:32 <ekarlso-> gabrielhurley: heh, too bad I got way to much atm :p
22:56:54 <gabrielhurley> there's some POC code available from Tomas Sedovic (sorry if I spelled that wrong off the top of my head)
22:57:04 <gabrielhurley> that'll serve as a good foundation/discussion point in the next cycle
22:57:06 <ekarlso-> gabrielhurley: it always seems the tasty candy gets postponed to the next release though ?
22:57:11 <lsmola_> no it is right
22:57:16 <gabrielhurley> not always
22:57:20 <gabrielhurley> we just do as much as we can
22:57:28 <gabrielhurley> sometimes that's more, sometimes less
22:57:33 <lsmola_> the realtime communiction is almost complete
22:57:34 <gabrielhurley> this one is *huge*, though
22:57:40 <gabrielhurley> and we have to get it right
22:57:50 <gabrielhurley> it's not one we can land late in a cycle and just sort of let sit there
22:57:56 <gabrielhurley> people expect a lot out of this
22:58:08 <gabrielhurley> and getting it wrong will frustrate users and deployers alike
22:58:29 <gabrielhurley> From my current vantage point I don't see it *not* landing in I
22:58:36 <rdxc> are you planing to abstract the message broker or just use RMQ and call it done
22:58:59 <ekarlso-> gabrielhurley: how early you think in I ?
22:59:31 <lsmola_> ekarlso-, giving some feedback to the realtime communication patches would be very good :-)
22:59:31 <gabrielhurley> rdxc: there's an element of using the code in oslo to listen on the message bus, but our browser <--> server channel is websockets
22:59:38 <gabrielhurley> ekarlso-: no idea
22:59:53 <gabrielhurley> ask again after the summit
23:00:05 <ekarlso-> ok :p
23:00:18 <lsmola_> ekarlso-, it all depends of how many people will give feedback
23:00:20 <gabrielhurley> we've all been focused on getting Havana out, with all the good things that *are* in it
23:00:24 <rdxc> ok, we built this architecture out in rails w/ web sockets for HPCS
23:00:31 <gabrielhurley> now we get to start looking forward again today
23:00:32 <rdxc> we learned much from that
23:00:53 <gabrielhurley> rdxc: would lov to hear your learnings. will you be at the summit by any chance?
23:00:57 <rdxc> getting the underlying plumbing is key to scale
23:01:28 <rdxc> no, but I'll make sure @david-lyle is properly armed
23:01:41 <gabrielhurley> fwiw, Nebula's dashboard also relies on websockets (and we've worked towards direct browser to service communication via CORS lately, too), so I've learned a few things myself. :-)
23:01:44 <gabrielhurley> awesome
23:02:08 <gabrielhurley> oh, we're out of time
23:02:13 <gabrielhurley> any last minute urgent items?
23:02:14 <gabrielhurley> last call
23:02:33 <lsmola_> nope
23:02:36 <david-lyle> Thanks!
23:02:37 <gabrielhurley> Thanks everyone. You've all done great! H3 is jam-packed full of goodness.
23:02:40 <lblanchard> thanks all!
23:02:42 <gabrielhurley> talk to you next wek
23:02:43 <amotoki_> thanks!
23:02:44 <gabrielhurley> #endmeeting