16:00:09 <david-lyle> #startmeeting Horizon 16:00:10 <openstack> Meeting started Tue Sep 9 16:00:09 2014 UTC and is due to finish in 60 minutes. The chair is david-lyle. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:00:11 <tzumainn> hiya! 16:00:12 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 16:00:14 <openstack> The meeting name has been set to 'horizon' 16:00:19 <david-lyle> Hello everyone! 16:00:20 <lcheng_> hello 16:00:22 <jgravel> hello 16:00:31 <amotoki> hi 16:00:31 <crobertsrh> hello 16:01:40 <david-lyle> ok, I filled out a healthy agenda for today 16:01:58 <tsufiev> hi 16:02:05 <david-lyle> https://wiki.openstack.org/wiki/Meetings/Horizon 16:02:46 <jomara_> howdy 16:02:48 <david-lyle> #topic Release Timeline 16:02:55 <jomara_> "healthy" 16:03:15 <david-lyle> So we are currently working on RC1 16:04:05 <david-lyle> that will run until we have closed all things that are critical for RC1 16:04:07 <gugl> hi 16:04:28 <david-lyle> https://launchpad.net/horizon/+milestone/juno-rc1 16:04:51 <david-lyle> so we have some BPs that are FFEs and we'll talk about that in a few minutes 16:05:16 <david-lyle> The remainder are bugs 16:05:55 <pawels> hello 16:06:03 <david-lyle> I/we need to decide what is critical to make it in the Juno Release 16:06:08 <david-lyle> of the bugs 16:06:57 <david-lyle> likely 1-3 weeks away 16:07:21 <david-lyle> after that, we'll open a new rc if we have a critical issue that has to make Juno 16:07:57 <david-lyle> Once RC1 is complete, it will become a branch and master should open up for Kilo 16:08:15 <david-lyle> in the meantime, no bps that aren't FFEs should be approved to merge 16:08:40 <david-lyle> I or another core may -2 your bp related changes to prevent it merging 16:08:56 <david-lyle> unfortunately there's not another mechanism to accomplish this 16:08:58 <lsmola> hello 16:09:22 <david-lyle> Any questions on the timeline, or any bugs for RC1 that anyone would like to talk about? 16:09:44 <ericpeterson> 1363515 and 1366270 seem like the only two that could possibly be included, which are not yet prioritized 16:09:50 <tsufiev> what does mean FFE? 16:10:01 <david-lyle> Feature Freeze Excpetion 16:10:06 <david-lyle> without the typo 16:10:08 <david-lyle> :) 16:10:22 <amotoki> Although I haven't filed, I would like to propose a patch to support Neutron L3-HA (high availability feature). 16:10:26 <tsufiev> ah, ok) 16:10:30 <amotoki> It adds just one attribute. 16:10:46 <amotoki> Neutron side patch is still FEE but it is a community topic. Can I do this as a bug? 16:10:53 <david-lyle> amotoki: sure 16:11:06 <david-lyle> mark it as at least a high 16:11:29 <amotoki> david-lyle: thanks. I will file it and make sure to keep the status up-to-date. 16:11:29 <crobertsrh> We have one small bp from the Sahara side: https://blueprints.launchpad.net/horizon/+spec/data-processing-swift-trust-auth 16:12:00 <crobertsrh> It turns out to be a very small change horizon-wise, but I'm holding on to it until the Sahara backend stuff is ready. 16:12:17 <tsufiev> fix for double login is not critical issue, but it has been bothering everyone for a long while 16:12:30 <tsufiev> 1308918 16:12:32 <david-lyle> tsufiev: I absolutely agree on that 16:12:33 <gary-smith> tsufiev +1 16:12:36 <TravT> +1 tsufiev 16:12:46 <pawels> +1 tsufiev 16:12:53 <david-lyle> crobertsrh: if it's that small, make it a bug 16:13:13 <david-lyle> and ping me with it when you have it so I can prioritize it 16:13:30 <amotoki> crobertsrh: one concern is that it depends on Sahara progress. Can you take care of it? 16:13:38 <crobertsrh> Ok...thanks 16:13:44 <david-lyle> so I will need to release a new version of django_openstack_auth 16:13:52 <crobertsrh> Yes, I'm all over the sahara-side of things. 16:14:02 <david-lyle> there are a few changes built up and a couple more to get in. then I'll release 16:14:12 <david-lyle> region stickiness is the other bug 16:14:35 <amotoki> crobertsrh: thanks 16:14:51 <david-lyle> #topic FFEs 16:15:10 <david-lyle> 2 have landed 16:15:21 <david-lyle> amotoki: is IPv6 complete now? 16:15:30 <david-lyle> or do we need a client release from neutron? 16:15:33 <amotoki> david-lyle: yes. done 16:15:42 <david-lyle> no client changes required? 16:15:48 <amotoki> david-lyle: client release is not required. 16:15:54 <david-lyle> perfect 16:16:34 <amotoki> IPv6 neutron support is just for CLI option. 16:16:46 <amotoki> *in neutronclient 16:16:54 <david-lyle> thanks amotoki absubram 16:17:15 <amotoki> I just updated the status. 16:17:21 <david-lyle> amotoki: thanks 16:17:24 <pawels> to FFE "Widget to expose the metadata catalog from glance" (https://review.openstack.org/#/c/104956) we have glance part just completed including python-glanceclinet, reviews are welcome ;) 16:17:39 <david-lyle> just getting to that, thanks pawels 16:18:00 <pawels> we have also python glance client release 16:18:08 <david-lyle> these patches can merge now once approved by reviewers 16:18:11 <TravT> yes, we got the global requirements merged in this weekend for the glanceclient change 16:18:25 <TravT> these widgets use data from glance. 16:18:38 <david-lyle> TravT: we sync'd that is horizon's requirements yesterday as well 16:18:57 <TravT> yep. sample data will be added automatically by devstack as soon as a merge goes through 16:18:57 <david-lyle> so nothings blocking merging other than final reviews/suggestions 16:19:12 <TravT> right... 16:19:23 <david-lyle> I would like to have those merged by sometime Fri 16:19:23 <TravT> FYI, this is the devstack patch currently in zuul gate: https://review.openstack.org/#/c/110955/ 16:19:39 <pawels> many dependecies, but looks like it has happen finally ;) 16:19:49 <david-lyle> thanks TravT and pawels for all your efforts on the glance side to see this through 16:20:03 <TravT> :) 16:20:11 <pawels> ;) 16:20:16 <david-lyle> that leaves one FFE proposal that went to the mailing list 16:20:31 <david-lyle> https://blueprints.launchpad.net/horizon/+spec/cinder-qos-specs 16:20:59 <david-lyle> lcheng__: has signed up as a sponsor to provide reviews, are there any other cores willing to sign up? 16:21:18 <rhagarty> that consists of 4 patches, first of which has landed. Lin has been reviewing... 16:21:57 <david-lyle> I'm willing to sign up, but it makes it an HP heavy patch, want to give someone else an opportunity to sponsor or object 16:23:05 <david-lyle> I'll reach out to some other cores offline 16:23:11 <amotoki> i think it can review but my review bandwidth is my concern... 16:23:25 <rhagarty> david-lyle: thanks 16:24:07 <amotoki> david-lyle: when is the deadline for FFE merge? 16:24:14 <gary-smith> amotoki: thanks for your help 16:24:17 <david-lyle> amotoki: understood, I'll review as well and if we there are no issues we can merge 16:24:43 <david-lyle> amotoki: Ideally, Friday, but that one could probably go until early next week 16:25:17 <david-lyle> #topic django_openstack_auth 16:25:49 <david-lyle> I mentioned this above, but I'll release this once the two issues also mentioned above are fixed 16:26:03 <david-lyle> any other items seen as required for the release? 16:26:41 <amotoki> /FYI/ I created the release milestone for openstack-auth and targeted bugs to the past releases. 16:27:00 <amotoki> we have less open bugs now :-) 16:27:02 <rdopieralski> I'm not sure what will happen with the js libs that didn't get unbundled -- I suppose it's the problem of the packagers 16:27:27 <david-lyle> thanks amotoki 16:27:46 <david-lyle> with less bugs, it must be that much better too :) 16:28:32 <amotoki> rdopieralski: are talking about horizon side? I think openstack-auth has no js libs. 16:28:37 <david-lyle> rdopieralski: my understanding is most of them replace the built in JS anyway 16:28:50 <rdopieralski> no, sorry, I didn't see the context, please ignore me 16:29:01 <rdopieralski> I wasn't paying attention for a moment 16:29:05 <rdopieralski> sorry 16:29:07 <amotoki> np 16:29:11 <david-lyle> no worries 16:29:50 <david-lyle> ok, so I'll revisit the openstack_auth buglist as well and make sure there's nothing else essential 16:30:05 <david-lyle> #topic rdopieralski talks JS :) 16:30:15 <rdopieralski> lol 16:30:47 <david-lyle> As I mentioned, my understanding is that distros just gut the embedded js and have it point to the version they deliver 16:30:47 <rdopieralski> I suppose they will just replace them with whatever version they have in the distro 16:30:56 <david-lyle> exactly 16:31:07 <david-lyle> I think it's no worse than before 16:31:19 <rdopieralski> the problem is, with jquery-ui, we can't be compatible with both versions at the same time 16:31:42 <rbertram> rdopieralski: what about datepicker, where we are switching to xstatic and updating at same time, fixing bugs? 16:32:08 <rdopieralski> rbertram: datepicker is not such a problem, because we are the only ones who use it 16:32:23 <rdopieralski> rbertram: so the distros will package whatever version we use 16:32:23 <ericpeterson> seems like we either re-package / copy all the js libs or we place a hard requirement on particular versions 16:32:56 <rdopieralski> rbertram: with jquery-ui the problem is that they have other software that depends on it 16:33:02 <rdopieralski> rbertram: so they can't just upgrade 16:33:06 <ericpeterson> I think the package / distro people can say xstatic-jquery-ui >= x.y.z 16:33:12 <rbertram> rdopieralski: I see 16:33:42 <ericpeterson> so if the focus is to make distro people happy, then we copy all the js libs 16:33:45 <david-lyle> rdopieralski: do we have a hard requirement on a newer version? 16:34:18 <david-lyle> ericpeterson: they just replace them anyway 16:34:26 <rdopieralski> david-lyle: at the moment we still have an ancient version bundled, which is incompatible with both Debian's and Fedora's versions 16:34:29 <david-lyle> they don't want to support the security issues of multiple versions 16:34:53 <david-lyle> rdopieralski: and moving forward breaks what? 16:35:30 <rdopieralski> david-lyle: the project changed the directory structure (twice), so I have to change the path in the template's <link> tag 16:35:44 <rdopieralski> david-lyle: and the new path depends on the version 16:36:03 <rdopieralski> david-lyle: I suppose that's trivial for the distros to change 16:36:14 <ericpeterson> so the packing teams replace our files.... and then this becomes a problem for us??? that ain't right 16:36:29 <rdopieralski> ericpeterson: we try to cooperate 16:36:55 <david-lyle> back, network kicked me off 16:37:17 <rdopieralski> david-lyle: the project changed the directory structure (twice), so I have to change the path in the template's <link> tag 16:37:22 <rdopieralski> david-lyle: and the new path depends on the version 16:37:25 <rdopieralski> david-lyle: I suppose that's trivial for the distros to change 16:37:50 <david-lyle> I think they change the path anyway to point at the system library 16:38:05 <rdopieralski> then there are some smaller incompatibilities in there, but we won't know them until we switch to new version and start using it 16:38:17 <david-lyle> that scares me more 16:38:36 <rdopieralski> so the current state is the worst of the possible solutions 16:38:49 <david-lyle> the sweet-spot then 16:39:24 <david-lyle> rdopieralski: do you have time to test with the newer version of jquery-ui 16:39:42 <david-lyle> or anyone else, so we can get a handle on what breaks? 16:39:46 <rdopieralski> I think so 16:40:18 <rdopieralski> https://review.openstack.org/#/c/113184/ <-- here is the patch for the global-requirements 16:40:24 <david-lyle> if the problems are large, we can make release dependent on getting that xstatic patch through 16:40:51 <rdopieralski> I will upload another one for the most recent version, and try to convince zigo to just replace it with an older version in Debian 16:41:19 <rdopieralski> though it feels wrong for me to take the side of Fedora 16:41:19 <zigo> rdopieralski: Another what? 16:41:37 <rdopieralski> zigo: with the version that you have -- you will just have to change one path in the template 16:42:07 <rdopieralski> zigo: would that work for you? 16:42:54 <david-lyle> I'm going to move on in the agenda and pick this back up at the end 16:42:56 <zigo> rdopieralski: 1.10.1 is already what I have in Debian (eg: it points to the files from libjs-jquery-ui 1.10.1+dfsg-1). 16:43:20 <rdopieralski> zigo: yeah, but mrunge wants 1.11 16:43:48 <david-lyle> #topic Allow translators to translate BatchActions https://review.openstack.org/#/c/91338/ (amotoki, dougfish) 16:44:02 <rdopieralski> zigo: let's continue on #openstack-horizon 16:44:38 <amotoki> This is related to translations. It force developers to do more like https://review.openstack.org/#/c/91338/17/openstack_dashboard/dashboards/project/instances/tables.py.... 16:45:13 <amotoki> it is useful for translations but I believe we need a consensus and decide the direction for juno as developer community. 16:46:06 <amotoki> The impact on I18N team seems acceptable per today's mail in i18n list. 16:46:33 <amotoki> I understand concern's from david comments. thougtht? 16:46:35 <david-lyle> amotoki: the impact of handling the additional translations? 16:47:15 <amotoki> david-lyle: yes. it splits the existing words into two or more, so translators needs additional work. 16:47:27 <david-lyle> amotoki: just wanted to make sure I followed 16:47:57 <david-lyle> so my concern is the additional boilerplate 16:49:09 <amotoki> david-lyle: understood completely. I just want to move the thing forward even if it is +1 or -1... 16:50:11 <amotoki> perhaps I think we need to improve django or gettext utils to reduce the complexity :-( 16:51:07 <amotoki> I hope other developers take a look at https://review.openstack.org/#/c/91338/ and leave comments. 16:52:08 <david-lyle> amotoki: If we don't we get anymore feedback with a better suggestion, we can just move forward with the proposal 16:53:03 <amotoki> sounds good as translator.... ok 16:54:04 <david-lyle> #topic Summit Format Changes 16:54:23 <david-lyle> so for the upcoming summit there are some format changes 16:54:42 <david-lyle> most of this has been covered on the email, but I wanted to just summarize here 16:55:05 <david-lyle> First day should be a limited set of cross-project session and incubated projecst 16:55:10 <david-lyle> *projects 16:55:38 <david-lyle> 2 days of individual project sessions, where Horizon should have 4-5 40 minute sessions 16:56:11 <david-lyle> topic proposal for those sessions will be done on etherpad rather than the previous tool 16:56:27 <david-lyle> I will post a link to the etherpad this week 16:57:30 <david-lyle> finally, Horizon will have a half day of unschedule time 16:57:51 <david-lyle> more free-form to tackle larger issues, or issues that arise at the summit 16:58:35 <david-lyle> The last topic about potential changes to the blueprint process will have to take place next week 16:58:46 <david-lyle> #topic Last Minute items? 16:58:56 <david-lyle> we have two minutes :) 16:59:38 <thinrichs1> Hi all—it's time for the Congress meeting 16:59:44 <openstack> thinrichs1: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 16:59:52 <david-lyle> one minute early :) 17:00:03 <david-lyle> til next week Horizon folks 17:00:04 <thinrichs1> Sorry—I thought I saw the end-meeting. 17:00:09 <david-lyle> #endmeeting