22:01:29 <david-lyle> #startmeeting Horizon 22:01:30 <openstack> Meeting started Tue Dec 3 22:01:29 2013 UTC and is due to finish in 60 minutes. The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:01:31 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 22:01:33 <openstack> The meeting name has been set to 'horizon' 22:01:56 <julim> hi there 22:02:01 <lsmola> hello 22:02:02 <casanch1> hello 22:02:05 <mrunge> hey 22:02:05 <devlaps> o/ 22:02:07 <jtomasek> hi 22:02:08 <bdehamer> hello 22:02:11 <lcheng> hello 22:02:13 <david-lyle> Hello everyone 22:02:28 <david-lyle> So the freeze for Icehouse-1 is today 22:02:41 <david-lyle> there are still a few items pending review 22:02:53 <jcoufal> o/ 22:03:02 <david-lyle> https://launchpad.net/horizon/+milestone/icehouse-1 22:03:25 <david-lyle> What doesn't merge in the next few hours will just get moved to i-2 22:03:39 <david-lyle> I do think a couple of the items are close though 22:04:16 <orion66> hello everyone 22:04:19 <jomara> howdy 22:04:21 <lsmola> david-lyle, darn, the inline-edit patch got some selenium error, thought it passed like 3 hours ago :-) 22:04:36 <lsmola> david-lyle, it is destined to fail :-) 22:04:46 <david-lyle> Aside, from that, I spent time this week trying to cobble together a plan for icehouse as a whole 22:05:11 <MaxV> o/ 22:05:30 <david-lyle> if anyone feels there's a misalignment or something I overlooked, please change it or contact me to change it 22:06:02 <david-lyle> I may have also assigned a couple of owners that weren't, again I can correct if something is incorrect 22:06:17 <ekarlso> yo gang, is there any clue on when the notifications stuff that have been talked so much about will land ? 22:06:18 <david-lyle> lsmola, I saw that, no worries 22:06:51 <david-lyle> ekarlso, I added the real-time work shadower started back in for i-2 22:07:10 <david-lyle> landing, not sure, that was mostly a proof of concept 22:07:12 <ekarlso> will it land this time during the cycle ? 22:07:16 <ekarlso> sigh 22:07:51 <david-lyle> ekarlso: it needs eyes and time using 22:08:05 <ekarlso> is there any recent patchsets ? 22:08:18 <lsmola> ekarlso, yeah i guess shadower would appreciate some feedback 22:08:25 <david-lyle> ekarlso: not that I am aware of 22:08:27 <ekarlso> shadower: ? ;) 22:08:30 <lsmola> ekarlso, there weren't many reviews 22:08:42 <casanch1> what about the bug fixes that still didn't make it to master? 22:08:43 <ekarlso> it would be a great addition to have :( 22:09:00 <lsmola> ekarlso, yeah, it's the guy assigned to that 22:09:16 <david-lyle> casanch1: 66 bug fixes landed in i-1 22:09:22 <mrunge> casanch1, that can't happen 22:09:25 <ekarlso> shadower: any recent work on it ? 22:09:33 <mrunge> casanch1, or misunderstood you? 22:09:34 <lsmola> ekarlso, or is there some newer patch? 22:09:42 <ekarlso> lsmola: thats 22:09:43 <david-lyle> patches are coming in at about 4.9 per day 22:09:51 <ekarlso> what I am wondering about ;) 22:09:55 <mrunge> casanch1, you mean a patch landed in havana. but not in icehouse? 22:10:18 <casanch1> yes, I mean the bugs that are still under review 22:10:24 <david-lyle> by coming in, I mean up for review 22:10:25 <casanch1> waiting for approval 22:10:51 <casanch1> but if it's ok to postpone them to i-2, there's no problem 22:10:52 <lsmola> ekarlso, no recent work, as far as I know 22:11:30 <david-lyle> casanch1: anything that didn't land in i-1 that was targeted to i-1 automatically gets targeted to i-2 22:12:08 <casanch1> oh ok 22:12:51 <david-lyle> I understand review time lag can be frustrating, I feel that too 22:13:16 <david-lyle> The best thing to speed things up is provide reviews 22:13:36 <david-lyle> :) 22:13:54 <jpich> Yes, please help review other patches, it really helps :) 22:14:36 <david-lyle> #topic Agenda 22:14:48 <MaxV> I'd like to review all futur patches about Angular 22:14:49 <david-lyle> https://wiki.openstack.org/wiki/Meetings/Horizon 22:15:04 <jomara> MaxV: +1 22:15:16 <david-lyle> MaxV: please do 22:15:21 <jomara> MaxV: you did an outstanding job with mine 22:15:22 <MaxV> If someone sees one, or make one ping me 22:15:26 <david-lyle> your input is greatly valued 22:15:43 <david-lyle> #topic Review IA proposal 22:16:17 <david-lyle> This was a carry over from last week. I posted a mind map diagram to http://ask-openstackux.rhcloud.com/question/1/openstack-ui-information-architecture/ 22:16:29 <david-lyle> I still need to write the prose 22:17:02 <david-lyle> So, this is just a reminder to provide feedback if there's anything you disagree with 22:17:26 <david-lyle> A lot of what's in there in contingent on the RBAC work which is progressing slowly :) 22:17:42 <david-lyle> but I think bdehamer may be coming to assist on that 22:17:54 <jcoufal> that's great news 22:18:13 <jcoufal> just quickly connected to that 22:18:25 <david-lyle> #topic Jasmine (MaxV) 22:18:41 <jcoufal> ok, I'll leave it to open discussion :) 22:18:41 <david-lyle> MaxV made a proposal on the mailing list 22:18:59 <jomara> +1 to proposal 22:19:00 <jtomasek> +1 22:19:18 <MaxV> you also have a blueprint, if someone wants to check https://blueprints.launchpad.net/horizon/+spec/jasmine-integration 22:19:50 <jtomasek> I still need to have a look, but MaxV, how does it integrate with selenium? Are there any issues? 22:19:55 <MaxV> there also are a few POC https://review.openstack.org/#/q/status:open+project:openstack/horizon+branch:master+topic:jasmine-jshint,n,z 22:20:07 <lsmola> jasmine sounds very good, +1 to that 22:20:16 <MaxV> I made it working through selenium 22:20:41 <MaxV> the error messages are retrieved server side 22:20:54 <MaxV> and fire as a failure of the test 22:21:08 <MaxV> I also wrapped jshint 22:21:34 <jomara> awesome! jshint is very necessary 22:21:40 <jomara> as you can tell by my flagrant lack of semicolons 22:22:01 <MaxV> It logs the first 10 failures 22:22:20 <MaxV> I also wants to make a runserver-test command 22:22:35 <MaxV> which will brings up a testing server 22:22:48 <MaxV> some people are interested? 22:23:17 <MaxV> selenium will be used for the packaging and checking regressions 22:23:47 <MaxV> I'm a little on fire, you can say calm down :) 22:23:55 <david-lyle> MaxV: the test server is for the Jasmine tests only? 22:24:20 <jpich> What do you mean with testing server, how much does it do? 22:24:25 <MaxV> no, It just run a server with debug = true and compress false 22:24:49 <jomara> oh, you mean just django, but with uncompressed assets and debugging 22:24:52 <jomara> it is the entire horizon app 22:24:52 <mrunge> why compress = False? 22:25:00 <david-lyle> debugging 22:25:03 <david-lyle> javascript 22:25:15 <MaxV> runserver-dev must be a better name 22:25:33 <mrunge> well, that wouldn't test, what will be deployed in production 22:25:48 <mrunge> wouldn't catch issues with compressing 22:25:56 <david-lyle> no, but when developing the JavaScript it's useful 22:26:05 <MaxV> no test, my bad, it is a dev server 22:26:16 <mrunge> ah, ok 22:26:19 <david-lyle> gates and checkin verification would still run the way they do now 22:26:24 <MaxV> yes 22:26:36 <MaxV> it is just an entry in run_test.sh 22:26:52 <david-lyle> or a flag to --runserver 22:27:00 <jomara> +50000 to that 22:27:47 <david-lyle> I would like to take some time to look into Jasmine and the patches you've put up for review 22:27:47 <MaxV> the main discussion was about Jasmine 22:27:54 <MaxV> sure 22:28:29 <david-lyle> that said Jasmine sounds reasonable, I would just like to do my homework and let others check it out too 22:28:53 <david-lyle> I'll add it as a topic for the next meeting so we can revisit it 22:29:00 <lsmola> david-lyle, +1 22:29:12 <david-lyle> Any other questions for MaxV or re: Jasmine in general? 22:29:53 <jomara> just as a comment, i think it'll be very nice to have a testing framework people who know angular are familiar with (since we now have angular) 22:30:33 <david-lyle> #topic Client side validation 22:30:53 <david-lyle> last meeting the issue of client side validation was brought up 22:31:08 <casanch1> yes, that was a quick fix I made 22:31:23 <david-lyle> a patch https://bugs.launchpad.net/horizon/+bug/1125232 was proposed that used html5 22:31:26 <uvirtbot> Launchpad bug 1125232 in horizon "Object Upload is validated after object upload" [Medium,In progress] 22:31:56 <david-lyle> there was input that angular might provide a better solution once in 22:32:04 <casanch1> it's pending approval https://review.openstack.org/#/c/57499/ 22:32:09 <david-lyle> now that it's in, is that the case? 22:32:10 <jomara> angular has nice inline form validation options 22:32:23 <jomara> they are very very simple 22:32:40 <david-lyle> or should we accept this approach until the angular method is better laid out 22:32:55 <david-lyle> it's a question of time, I suppose 22:33:00 <jomara> david-lyle: is there a thread about this? i can write up an example if we're interested 22:33:10 <casanch1> the fix is really simple, it's just needed to add one line of code in each form 22:33:18 <mrunge> how long will it take for angular? 22:33:46 <jomara> depends on how many forms you want to set up and how complex the validations are 22:33:48 <MaxV> it seems easy 22:33:59 <jtomasek> david-lyle: this html5 method is very unintrusive, so there is no issue with it for now 22:34:31 <jtomasek> MaxV: jomara: good read: http://django-angular.readthedocs.org/en/latest/angular-form-validation.html 22:34:39 <jomara> oh cool, will check that out 22:34:57 <david-lyle> it felt clean to me, I just knew there were concerns expressed 22:35:11 <mrunge> casanch1, does your "fix" work with older browsers? Internet explorer? 22:35:49 <jtomasek> mrunge: not really, client side validations should be backed up by server side validations still 22:35:50 <david-lyle> IE 8 is as far back as we support 22:36:22 <jtomasek> not sure about this instance though 22:36:22 <casanch1> it doesn't work on old versions of IE, IE10 for sure, that IE9 or 8, not sure 22:36:25 <casanch1> I can check 22:36:28 <mrunge> jtomasek, point taken. let's make it directly right, no quick fixes 22:36:44 <devlaps> jtomasek: +1 22:36:50 <MaxV> angular support IE8 and form validation is easy 22:37:08 <MaxV> but as mentionned jtomasek we have to check on server side too 22:37:23 <casanch1> it is being checked on server side 22:37:25 <mrunge> yes of course! 22:37:27 <casanch1> that is already in 22:37:42 <casanch1> this is for the missing client-side validation 22:38:05 <jtomasek> casanch1: then I think it is good enough 22:38:14 <jtomasek> casanch1: as discussed last time 22:38:26 <casanch1> and again, I completely agree that angular is the best final solution, but this fix will prevent wasting resources while uploading large objects 22:38:41 <casanch1> while we wait for angular... 22:38:49 <MaxV> it is merged 22:39:29 <mrunge> MaxV you mean angular is merged 22:39:37 <david-lyle> ok, casanch1 and MaxV can you two use that bug to collaborate on an angular based solution 22:39:43 <MaxV> yes 22:39:47 <casanch1> oh, didn't know that Angular was already merged to master 22:39:49 <mrunge> not the qickhack casanch1 made 22:40:02 <casanch1> ok, I can rework the fix to use angular 22:40:16 <david-lyle> ok, casanch1, thank you 22:40:34 <casanch1> no problem 22:40:44 <lsmola> MaxV, would be nice to find a way how to generate client side validation from Django validations 22:41:15 <MaxV> good idea 22:41:35 <bdehamer> Yeah, the tricky part is making sure your server-side validations stay in sync with your client-side logic. 22:41:54 <david-lyle> a linked solution would be ideal 22:42:14 <kspear_> jtomasek's link above talks about that 22:43:07 <david-lyle> #topic Follow up on: Updates from I18N team 22:43:20 <jtomasek> lsmola: this is the right solution: http://django-angular.readthedocs.org/en/latest/angular-form-validation.html 22:44:17 <david-lyle> I think there are just two remaining reviews that need to be merged on stable, although if the deadline is today, that seems unlikely 22:44:17 <lsmola> jtomasek, excellent 22:44:26 <david-lyle> thanks jtomasek 22:45:41 <david-lyle> #topic Open Discussion 22:45:43 <jpich> We may be able to backport a translation patch to the milestone-proposed branch if it's proposed within the next couple of days 22:46:13 <david-lyle> jpich ok, that would work 22:46:14 <jpich> but I'm not sure what's the status on Transifex, if the latest strings have been updated or if the i18n has had a chance to translate yet 22:46:32 <jpich> I'll check with amotoki tomorrow, he may be more up to date than I am 22:47:04 <david-lyle> ok thanks jpich 22:47:21 <jpich> sure 22:47:22 * david-lyle jumps topics a little too quickly 22:47:32 <jpich> :-) 22:47:46 <david-lyle> alright, open season 22:48:43 <jcoufal> one quick update 22:48:55 <jcoufal> there is updated navigation proposal 22:48:57 <jcoufal> http://ask-openstackux.rhcloud.com/question/2/openstack-dashboard-navigation-redesign/?answer=99#post-id-99 22:49:39 <jcoufal> so whoever is interested, feel free to give any feedback, everything is welcome 22:50:17 <lsmola> will do 22:50:44 <jcoufal> it's mostly reflecting summit discussions 22:50:54 <devlaps> jcoufal: thanks for the info. definitely interested. 22:51:22 <david-lyle> jcoufal: thanks for putting that together. I think there is more than one bp in there 22:51:38 <david-lyle> the main navigation and header updates is one 22:52:02 <jcoufal> yeah, BP are already there AFAIK 22:52:08 <david-lyle> breadcrumbs would be another 22:52:17 <jcoufal> david-lyle: +1 22:52:39 <jcoufal> i think there was already one about breadcrumbs 22:52:48 <jcoufal> but it might be obsolete by this time (not sure) 22:53:05 <jpich> By the way, 2 of the 4 interns who will participate in the next OPW round will be doing work related to Horizon: Sayali will be working with lsmola on sparklines, and Anna will be working with me on the instance actions 22:53:09 <david-lyle> ok, I thought I read through -all- of them in the last week, but I may have missed one 22:53:19 <jpich> The internship starts on December 10, say hello if you see them around :) 22:53:20 <lsmola> yaaay 22:53:25 <david-lyle> fantastic 22:53:36 <lsmola> jpich, sayali has already started 22:54:09 <mrunge> one quick update for django-1.6 22:54:18 <mrunge> I have a fix for horizon for revuew 22:54:25 <jcoufal> david-lyle: it was pretty old one, maybe that's my mistake 22:54:36 <mrunge> the larger issue is django_openstack_auth 22:54:38 <jpich> lsmola: Cool!! Let her know about these meetings so she can join in and say hello, if it's at a reasonable time where she lives :) 22:55:19 <lsmola> jpich, ok, will do 22:55:26 <david-lyle> mrunge, what's the issue? 22:55:41 <david-lyle> jcoufal: could be a ux bp 22:55:44 <mrunge> david-lyle, the review is here: https://review.openstack.org/#/c/58947/ 22:56:09 <mrunge> david-lyle, and the issue is: somethings in authenticate changed 22:56:21 <mrunge> let me check for the bug I filed 22:56:47 <mrunge> https://bugs.launchpad.net/django-openstack-auth/+bug/1256805 22:56:49 <uvirtbot> Launchpad bug 1256805 in django-openstack-auth "django-1.6: TypeError at /auth/login/" [Undecided,New] 22:57:59 <david-lyle> yeah, openstack_auth seems to break every django roll 22:58:20 <mrunge> once authenticated, one can use horizon with 1.6 (with the proposed fix for horizon) 22:58:35 <mrunge> so, the larger issue is openstack_auth 22:58:54 <mrunge> when we fixed that, we could start gating at 1.6 as well ;-) 22:59:14 <lsmola> mrunge, great 22:59:18 <david-lyle> mrunge: great 23:00:27 <david-lyle> times almost up, any last items 23:01:26 <david-lyle> ok, thanks everyone 23:01:30 <david-lyle> #endmeeting