15:59:40 <david-lyle> #startmeeting Horizon
15:59:40 <openstack> Meeting started Tue Jul 29 15:59:40 2014 UTC and is due to finish in 60 minutes.  The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:59:41 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:59:43 <openstack> The meeting name has been set to 'horizon'
16:00:11 <jomara> howdy
16:00:14 <gary-smith> hi
16:00:20 <david-lyle> Hello everyone
16:00:23 <crobertsrh> Hello
16:00:25 <jrist> o/
16:00:26 <jgravel_> hello
16:00:27 <pawels> hello
16:00:41 <lcheng> hello
16:00:46 <amotoki> hi
16:00:55 <tsufiev> hi
16:01:45 <david-lyle> J-2 was tagged last week on Wed, we finished 12 blueprints and closed 103 bugs
16:02:00 <david-lyle> Thanks everyone for your hard work and many reviews to make that happen
16:02:06 <tmazur> hello o/
16:02:11 <tzumainn> hiya!
16:02:24 <jtomasek> hey
16:02:35 <david-lyle> Right after J-2 we merged the rest of the code for the Sahara blueprint as well. \o/
16:02:43 <SergeyLukjanov> yay!
16:03:14 <david-lyle> Thanks to crobertsrh for staying on top of those changes for so long and making sure they were always in a good state
16:03:31 <crobertsrh> Thanks for reviewing the piles of code everyone :)
16:03:35 <jrist> amazing
16:03:48 <jrist> I have to admit, those reviews were ... unpleasantly repetitive
16:03:51 <akrivoka> hi everyone
16:03:54 <absubram_> hi
16:04:15 <lblanchard> hi all
16:04:19 <david-lyle> We also merged the update to bootstrap 3 this week
16:04:47 <david-lyle> there are several bugs related to that, but last week we decided those would be more easily addressed as bug fixes
16:05:06 <david-lyle> than trying to keep that large patch in sync with master
16:05:18 <david-lyle> so let's tackle some of those defects quickly
16:05:36 <david-lyle> one related to that is on the agenda later I believe
16:05:56 <david-lyle> That leaves us with https://launchpad.net/horizon/+milestone/juno-3
16:06:14 <david-lyle> Which I need to massively prune down in the next day or two
16:06:45 <david-lyle> Given past performance and review capacity, I would expect maybe 20 blueprints to land
16:07:35 <david-lyle> I will implement a feature proposal freeze probably a little before the rest of OpenStack so that we can concentrate on merging what we have in the pipeline already
16:08:01 <pawels> when that would be ?
16:08:25 <jpich> Makes sense to me
16:09:13 <david-lyle> I haven't determined the exact date yet, the common one is August 21, so maybe a week before that.
16:09:25 <david-lyle> Let's say 2 weeks from today.
16:10:00 <david-lyle> So August 12
16:10:30 <pawels> I am asking because we are hardly working on https://blueprints.launchpad.net/horizon/+spec/glance-metadata-definitions-base-admin-ui which is partially in glance and in horizon, I am expecting this need to be reviewed first in Glance before going into horizon ?
16:10:42 <david-lyle> without the feature freeze the review load gets too noisy and makes it tough to land anything
16:11:03 <david-lyle> pawels: without support in Glance, we won't merge in Horizon
16:11:32 <amotoki> horizon has many dependencies on other projects.
16:11:38 <pawels> david-lyle: ok, we are starting some review on Glance side this week
16:12:05 <david-lyle> so glance will have to support it, glanceclient will have to support it, glanceclient will have to be released, an openstack/requirements change will have to merge and finally changes to Horizon can merge, that seems very tight for scheduling
16:12:26 <pawels> that's why I am rising it ;)
16:12:32 <david-lyle> pawels: just looking at it realistically
16:13:35 <david-lyle> FFE are a possibility for certain features, but those are the exception
16:13:38 <amotoki> Re client requirements change, how about requirement freeze? does it occur during RC1 phase? I can't remember it well.
16:14:14 <david-lyle> amotoki: good question, that sounds correct
16:14:16 <pawels> david-lyle:: we should have like code complete this week of all the parts glance/glanceclient/horizon
16:14:48 <jpich> The requirements freeze doesn't really apply to the openstack clients, since we control them and they're supposed to be backwards compatible
16:15:09 <jpich> or let's say it's much easier to get a requirement freeze exception for them
16:15:48 <amotoki> jpich: yeah.. but it requires more efforts to us :-(
16:15:49 <david-lyle> jpich, true, but depending on how the feature is supported (optional) we may have to require a newer version of the client
16:16:46 <david-lyle> that should have been (optional or not optional)
16:17:58 <david-lyle> Anyway on J-3, expect pruning, we only have 5 weeks left
16:18:15 <tzumainn> david-lyle, want a dashboard link?
16:18:24 <david-lyle> tzumainn:
16:18:26 <david-lyle> sure
16:18:31 <jpich> Right, just saying that when I asked for a requirement exception for a client requirement last time, people thought it was silly to even have to ask :) http://lists.openstack.org/pipermail/openstack-dev/2014-March/031178.html
16:18:42 <tzumainn> http://goo.gl/RG6HQt
16:18:46 <tzumainn> it'd need to be updated after the pruning
16:19:22 <david-lyle> I also merged a horizon dashboard to the gerrit-dash-creator for general horizon reviews, minus tuskar-ui
16:19:48 <david-lyle> https://review.openstack.org/#/dashboard/?foreach=%28project%3Aopenstack%2Fhorizon+OR+project%3Aopenstack%2Fdjango_openstack_auth%29+status%3Aopen+NOT+label%3ACode-Review%3C%3D-2&title=Horizon+Review+Dashboard&My+Patches=owner%3Aself+status%3Aopen&Needs+Approval%2C+not+me=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D-1+label%3AVerified%3E%3D1%252cjenkins+NOT+owner%3Aself+label%3ACode-Review%3E%3D2+NOT+label%3ACode-R
16:19:48 <david-lyle> eview-1+NOT+label%3ACode-Review%3E%3D1%252cself&Needs+Approval=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D-1+label%3AVerified%3E%3D1%252cjenkins+NOT+owner%3Aself+label%3ACode-Review%3E%3D2+NOT+label%3ACode-Review-1&Not+Reviewed+by+Me=NOT+label%3ACode-Review%3C%3D-1%252cself+NOT+label%3ACode-Review%3E%3D1%252cself+limit%3A50&Needs+Reverify=label%3AVerified%3C%3D-1%252cjenkins+branch%3Amaster+NOT+label%3ACode-Review
16:19:48 <david-lyle> %3C%3D-1+NOT+label%3AWorkflow%3C%3D-1&5+Days+Without+Feedback=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D-1+label%3AVerified%3E%3D1%252cjenkins+NOT+owner%3Aself+NOT+label%3ACode-Review%3C%3D2+age%3A5d&No+Negative+Feedback=NOT+label%3AWorkflow%3E%3D1+NOT+label%3AWorkflow%3C%3D-1+label%3AVerified%3E%3D1%252cjenkins+NOT+owner%3Aself+NOT+label%3ACode-Review%3C%3D-1+NOT+label%3ACode-Review%3E%3D2+limit%3A50&With+Negati
16:19:50 <david-lyle> ve+Feedback=NOT+label%3AWorkflow%3C%3D-1+NOT+label%3AVerified%3C%3D-1%252cjenkins+NOT+owner%3Aself+label%3ACode-Review-1+limit%3A20&Work+In+Progress=NOT+label%3AWorkflow%3E%3D1+NOT+owner%3Aself+label%3AWorkflow%3C%3D-1
16:19:54 <jpich> wow
16:19:56 <david-lyle> wow, that was a mistake
16:20:05 * david-lyle facepalm
16:20:08 <jpich> :-)
16:20:13 <amotoki> IRC cuts off longer link :-(
16:20:25 <jrist> aieee
16:20:35 <david-lyle> you can generate it from the repo :)
16:20:48 <david-lyle> I'll leave it as an exercise for the reader
16:20:57 <akrivoka> lol
16:21:23 <tzumainn> http://goo.gl/ldHJEX would be the shortened version of that, I think
16:21:26 <amotoki> david-lyle: I pasted the similar one to https://wiki.openstack.org/wiki/Horizon/Reviews about a month ago. I will update it.
16:21:46 <david-lyle> amotoki: ok, I missed that
16:22:14 <david-lyle> with the creator repo, you can also modify to your taste
16:22:52 <david-lyle> so the agenda for today: https://wiki.openstack.org/wiki/Meetings/Horizon
16:23:10 <david-lyle> #topic Current breakages, for general awareness (jpich)
16:23:26 <jpich> Yo! For Your General Awareness:
16:23:41 <jpich> The pep8 job has been broken for about a month, you can't trust that your code is correct by running ./run_tests.sh -p so please be careful. pawels and rdopieralski are working on fixing this (thank you so much!), thanks to amotoki as well for helping to pinpoint where some of the issues are coming from
16:23:54 <jpich> https://bugs.launchpad.net/horizon/+bug/1347472 for the gory details and keeping track of the upcoming patches *refreshes page*
16:24:03 <pawels> together with radomir we are working on the issue
16:24:26 <jpich> The 2nd point is that there's some new intermittent gate failures popping around because of horizon, 2 seem related to the recent scss patch. Not sure if it's because of errors that got in due to lack of pep8ing, but something we'll want to keep an eye on and fix
16:24:34 <david-lyle> thanks pawels and rdopieralski
16:24:36 <jpich> I've been tagging those with "gate" when I see them, https://bugs.launchpad.net/horizon/+bugs?field.tag=gate
16:24:48 <jpich> <end of announcements/>
16:25:11 <david-lyle> oops, I think there's an additional one, that may or may not be a horizon bug
16:25:23 <david-lyle> https://bugs.launchpad.net/horizon/+bug/1349774
16:25:24 <jpich> Thank pawels and rdopieralski, please let us know if you need help or when you need reviews for these ones :)
16:25:36 <pawels> sure, thanks !
16:25:42 <tzumainn> does that bug really mention 2500 pep8 failures?
16:26:11 <jpich> david-lyle: It's in the list. derekh proposed a revert of the giant bootstrap patch as it introduced the issue, I'm not sure what's the best course of action
16:26:30 <david-lyle> jpich, see it now
16:26:32 <david-lyle> thanks
16:26:47 <pawels> tzumainn: if you remove the line "select = H236" you will get that number of errrors
16:26:51 <jpich> I'm a bit anxious huge patches are getting in without being pep8'd, but maybe we can catch the fixes in time with bug 1347472... It'll have to be soon though because from what I understand this breaks the TripleO gate
16:26:54 <david-lyle> not sure either, I don't think a revert is the right course of action
16:27:26 <amotoki> http://paste.openstack.org/show/88908/ show error count per flake8 check.
16:27:29 <jpich> tzumainn: There's about 700 new errors, and from what I understand a couple of 1000s we missed because of new rules that were introduced in a new version of hacking. These ones we'll add to the ignore list for later, we just want to fix what used to work for now
16:27:45 <david-lyle> parsing _stylesheets.html shouldn't be pep8 related
16:28:39 <amotoki> H904 and H307 were introduced in hacking 0.9 and hacking 0.9 came after the bug was introduced :-(
16:28:50 <david-lyle> what is H904
16:28:50 <david-lyle> ?
16:29:01 <amotoki> H904 is a hacking rule number.
16:29:11 <david-lyle> I understand that, what's the rule
16:29:12 <david-lyle> ?
16:29:28 <amotoki> H904 (Wrap long lines in parentheses instead of a backslash)
16:29:43 <david-lyle> good grief
16:29:48 <david-lyle> who cares
16:29:50 <david-lyle> oops
16:29:52 <tzumainn> lol
16:30:14 <jpich> Definitely for the explicit ignore list till some later time
16:30:23 * david-lyle returns inner voice to internal only
16:30:26 <jpich> in a galaxy far far away
16:30:53 <jpich> ah we've reviewed larger patches :-)
16:30:53 <amotoki> of course we need to discuss which rules should be enabled in horizon, especially like H904.
16:31:37 <david-lyle> amotoki: thanks for catching the larger issue, I'm not trying to discount the checks, I think they're valuable, at least most of them are valuable
16:32:16 <pawels> so we will ignore H904 for the begining but we also need to make a long term decision here
16:33:14 <david-lyle> what was the other new check added?
16:33:34 <pawels> H307
16:33:43 <david-lyle> and that is ?
16:33:48 <amotoki> H307: like imports should be grouped together
16:34:16 <david-lyle> I thought we already had that, did it become more stringent?
16:34:41 <amotoki> It checks all imports are groupd into three categories: standard, thirdparty and project.
16:35:23 <amotoki> It check three groups are separated by blank line and each group does not have a blank line inside group.
16:35:47 <amotoki> most horizon codes do not follow this.
16:36:22 <david-lyle> hmm, I thought we already went through that exercise, apparently I'm mistaken
16:36:31 <amotoki> we have several groups: standard, thirdpaty, django, horizon and openstack_dashboard.
16:36:36 <jpich> Nope, it's been up to the human reviewers so far
16:37:04 <amotoki> Previously grouping check is up to reviewers :-(
16:38:17 <jpich> I think it'd be a nice to have, maybe -- but likewise, later, once we're back to where we used to be with pep8 :)
16:38:28 <david-lyle> ok, I was thinking alphabetical I bet https://github.com/openstack/horizon/commit/cea720e793777519dab542fdc4105dd456e000f6
16:38:34 <david-lyle> anyway
16:39:43 <david-lyle> sounds like we're targeting the old and will move forward with supporting the new rules, which I would propose for early Kilo
16:39:55 <jpich> Sounds reasonable to me
16:40:01 <david-lyle> as we don't need extra code thrash at this point in the release
16:40:28 <jpich> Nope, we don't need all the juno-3 patches to have merge conflicts on feature freeze day :-)
16:40:47 <david-lyle> jpich: exactly, we'll have enough fun already
16:41:03 <david-lyle> #topic Open Discussion
16:41:04 <amotoki> totally makes sense.
16:41:10 <nlahouti> Support for cisco DFA in horizon BP is approved but it show up in  http://goo.gl/RG6HQt  under Blueprints (Review). Is that correct? or list needs to be updated?
16:42:14 <david-lyle> nlahouti: that heading means code for blueprint up for review
16:42:24 <david-lyle> not that the blueprint needs to be reviewed
16:42:50 <nlahouti> david-lyle: thx for clarification.
16:42:57 <david-lyle> nlahouti: np
16:43:24 <mxu> Hi david-lyle: Q on your comment "Given past performance and review capacity, I would expect maybe 20 blueprints to land".    I assume approved BPs will be part of the 20, correct?
16:43:29 <amotoki> I think "Approval" here means "need another +2".
16:44:45 <david-lyle> mxu, there are approved blueprints that won't make it for one reason or another, code's not ready, not enough reviews, etc
16:45:50 <david-lyle> mxu, the reality is, I don't honestly know who's going to stick with a patch long enough to get it in, who's not, how many reviewers are going to look at a particular patch, which ones will be overlooked
16:46:11 <david-lyle> so I approve blueprints that are in line with the direction of the project
16:46:38 <david-lyle> if someone is slated to or actively working on a blueprint, I attempt to assign it to a milestone
16:46:56 <david-lyle> that person may come or go at any point depending on many factors
16:47:00 <mxu> david-lyle: thanks for the answer. The code for our BP has been for a long time.  Can core reviewers please review and approve?
16:47:02 <david-lyle> I can't predict that
16:48:21 <david-lyle> mxu, I'm not sure which blueprint you are concerned about
16:48:21 <tsufiev> i have a kind of announcement :)
16:48:33 <tqtran> announce!
16:49:08 <tsufiev> some time ago (1.5 month) there was a proposal in openstack-dev ML: lists.openstack.org/pipermail/openstack-dev/2014-June/037054.html
16:49:26 <mxu> david-lyle:  here is the code patch  https://review.openstack.org/#/c/83863/  and here is the bp: https://blueprints.launchpad.net/horizon/+spec/horizon-cisco-dfa-support
16:50:11 <tsufiev> finally, that proposed framework for building UIs for projects dealing with complex input data (like Heat, Mistral, Murano) got its name - 'Merlin' and wiki at https://wiki.openstack.org/wiki/Merlin
16:51:15 <tsufiev> there is not yet a line of code (active design and planning of PoC), but I think that Horizon community should know about it
16:51:52 <tsufiev> especially, because I hope that Merlin (if things go well) will become a part of Horizon eventually
16:52:06 <jpich> Thanks tsufiev. Is the OpenStack UX team involved as well? ( https://wiki.openstack.org/wiki/UX  ) -- I usually find their inputs extremely valuable in that domain :)
16:53:11 <rbertram> I'm working with tqtran on JavaScript practices. Including instructions to set up editors with jshint & formatting.
16:53:21 <tsufiev> jpich, so far I've spoken only with Jacki Bauer, she has provided some initial feedback
16:53:24 <rbertram> Besides Sublime & Eclipse, are people any other edtiors for JS?
16:53:26 <jrist> jpich++
16:54:01 <jomara> rbertram: the js support inside pycharm is excellent
16:54:05 <david-lyle> tsufiev: I like the idea
16:54:08 <jrist> pycharm or webstorm yeah
16:54:09 <jomara> rbertram: and it supports jshint, etc
16:54:23 <rbertram> jomara: I will look into that!
16:54:37 <david-lyle> Horizon can't really target Murano, Mistral and Solum yet
16:54:56 <tsufiev> david-lyle, so, the most general overview is to provide framework for building complex wizards, for projects where conventional forms & wizards fail in some way
16:55:06 <david-lyle> however, it would be possible to build the UI toolkit into Horizon
16:55:12 <david-lyle> and build from there
16:55:51 <david-lyle> One of my desires is to completely rebuild the Launch Instances interface, because of the same sort of issues
16:56:19 <david-lyle> it's a usability nightmare and doesn't effectively solve the problem, it needs to be built from the bottom up
16:56:27 <david-lyle> s/built/rebuilt/
16:56:30 <jrist> er...
16:56:39 <tsufiev> david-lyle, is it ok for Merlin to come under 'UI' program? My management asked me to clarify that issue
16:56:42 <jomara> tsufiev: ive seen your proposal doc, it looks great (from the perspective of configuring heat stacks)
16:57:08 <jpich> Sounds like a discussion that would be appropriate on list, esp considering the time left
16:57:20 <david-lyle> tsufiev: I think that would make sense, but that may have to go through the TC
16:57:34 <david-lyle> we might want to have more discussion first
16:57:36 <david-lyle> :)
16:57:54 <tsufiev> jomara, it would be great if you suggested some feedback - either in ML or in wiki :). I'll more official letter in a next few days
16:58:04 <jomara> tsufiev: sure, mail it out
16:58:05 <tsufiev> *I'll write
16:58:45 <tsufiev> jpich, sure, I will make a ML announcement
16:58:49 <david-lyle> mxu: my understanding is you change is still not merged in neutron, until it is, I'm not investing a lot of review time in the Horizon part
16:58:54 <jpich> tsufiev: Cool!
16:58:55 <david-lyle> s/you/your/
16:59:00 <lblanchard> tsufiev: thanks for sharing this! I will take a lot from another UXers point of view and give feedback too
16:59:03 <david-lyle> tsufiev: thanks
16:59:38 <david-lyle> We're out of time. Thanks everyone! Have a great week.
16:59:41 <david-lyle> #endmeeting