20:00:20 <ying_zuo> #startmeeting horizon 20:00:21 <openstack> Meeting started Wed Sep 20 20:00:20 2017 UTC and is due to finish in 60 minutes. The chair is ying_zuo. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:00:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:00:25 <openstack> The meeting name has been set to 'horizon' 20:00:32 <ying_zuo> Hi everyone :) 20:00:37 <lucasxu> o/ 20:00:46 <robcresswell> o/ 20:00:46 <e0ne> hi 20:00:50 <gary-smith> o/ 20:01:32 <ying_zuo> o/ 20:01:40 <ying_zuo> #topic Notices 20:01:41 <david-lyle> o/ 20:01:48 <ying_zuo> #info PTG 20:02:05 <ying_zuo> Thanks everyone for attending the PTG. I am glad most of the active developers were able to make it. It’s such a great opportunity to meet the team and be able to discuss various topics face to face. 20:02:13 <ying_zuo> #link PTG discussion: https://etherpad.openstack.org/p/horizon-ptg-queens 20:02:19 <ying_zuo> #link Our team photos: https://plus.google.com/101186576626172577894/posts/LQnT6pXnRix 20:02:27 <ying_zuo> #link Photos for all the OpenStack team at the PTG: https://www.flickr.com/photos/152419717@N06/albums/72157689092104285/with/37136316176/ 20:03:24 <ying_zuo> #topic Bugs Review 20:03:40 <ying_zuo> We talked about ways to reduce the backlog during PTG. One of them is to review some bug reports during the weekly meeting. Thanks gary-smith for helping out with the bug triage. 20:04:00 <gary-smith> You're welcome! 20:04:07 <ying_zuo> :) 20:04:15 <e0ne> ying_zuo: I started to verify old in-progress bugs. some of them moved to new or fix commited 20:04:43 <ying_zuo> great! thanks e0ne 20:05:04 <gary-smith> e0ne: Thanks for the help. I noticed the bug count dropping between my triaging. I knew someone else was working on that too 20:05:05 <e0ne> there are dozens of in-progress bugs started 2+ years ago 20:05:26 <e0ne> gary-smith, ying_zuo: you're welcome 20:05:58 <ying_zuo> e0ne: why am I not surprised ;) 20:06:12 <e0ne> ying_zuo: :) 20:06:22 <ying_zuo> Today we will discuss a few unassigned and high priority tickets. Mainly to get some attention to those tickets and some pointers to start would be helpful. 20:06:24 <e0ne> there are only 2 ways to cut bugs list... 20:06:25 <robcresswell> If you find bugs have been solved elsewhere, just mark them Invalid. There'll already be a bug report elsewhere 20:06:46 <e0ne> robcresswell: ok, will do 20:06:56 <robcresswell> thanks 20:07:11 <ying_zuo> #link https://bugs.launchpad.net/horizon/+bug/1434241 20:07:12 <openstack> Launchpad bug 1434241 in OpenStack Dashboard (Horizon) " Internal Server Error while updating project quota when using Keystone LDAP backend" [High,New] 20:07:27 <david-lyle> a note that they have been otherwise fixed is helpful on the bug as well 20:08:11 <e0ne> usually, I add something like 'can't reproduce it on master' 20:08:25 <ying_zuo> This one requires keystone LDAP backend. It was opened in 2015. Would be good if someone can help confirming if it’s still valid. 20:08:32 <e0ne> note: only if I tried to reproduce it 20:08:57 <robcresswell> heh, yeah, leave a comment saying "why" its marked invalid 20:09:44 <ying_zuo> +1 20:09:48 <e0ne> +1 20:10:44 <ying_zuo> #link https://bugs.launchpad.net/horizon/+bug/1436239 20:10:44 <openstack> Launchpad bug 1436239 in OpenStack Dashboard (Horizon) "modifying users in non-default domain project fails" [High,Confirmed] 20:10:44 <ying_zuo> #link https://bugs.launchpad.net/horizon/+bug/1563005 20:10:46 <openstack> Launchpad bug 1563005 in OpenStack Dashboard (Horizon) "pseudo translation tool creates uncompilable PO files" [High,Confirmed] 20:10:50 <ying_zuo> same for these two 20:10:52 <david-lyle> there is an ldap plugin for devstack, hopefully it's still functional 20:11:42 <ying_zuo> hopefully 20:12:28 <e0ne> IMO, #1563005 is invalid now. we don't have 'pseudo' now 20:12:42 <david-lyle> I'll take 1436239 20:13:16 * david-lyle assigns himself 20:13:39 <ying_zuo> thanks david-lyle 20:14:03 <ying_zuo> does anyone know about the pseudo translation tool ? 20:14:40 <e0ne> ying_zuo: it's removed from master but still available in older branches 20:14:48 <robcresswell> No, its still available 20:15:02 <robcresswell> It's just in a management command, instead of a script 20:15:24 <amotoki> https://docs.openstack.org/horizon/latest/contributor/topics/translation.html#pseudo-translation-tool 20:15:26 <robcresswell> I can take that anyway. I ported it to the management files, which is what broke it I assume :p 20:15:42 <david-lyle> creating work for yourself :) 20:15:49 <e0ne> :) 20:15:52 <gary-smith> #1655487 has a patch, https://review.openstack.org/461490, that needs some core review love 20:16:08 <ying_zuo> :) thanks robcresswell 20:16:12 <robcresswell> david-lyle: I'm keeping myself employed forever 20:17:43 <ying_zuo> gary-smith: I will take a look at that one 20:18:01 <gary-smith> ying_zuo: Thanks 20:18:20 <ying_zuo> #link https://bugs.launchpad.net/horizon/+bug/1715066 20:18:21 <openstack> Launchpad bug 1715066 in OpenStack Dashboard (Horizon) "Cannot launch instance with Django Launch Instance form with multiple networks for a tenant" [High,Confirmed] 20:20:21 <ying_zuo> I was able to reproduce. There’s no obvious error so someone will need to look into it. 20:20:43 <gary-smith> e0ne: glad to see you taking those MS browser bugs; I have no windowz box anywhere to even reproduce them on 20:20:59 <robcresswell> gary-smith: Very few people do; they tend to sit around for ages 20:21:11 <e0ne> gary-smith: at lease, you can verify than nothing is broken in other browsers 20:21:26 <gary-smith> good point 20:21:27 <e0ne> s/ at lease/at least 20:21:42 <amotoki> reg 1715066, It seems generated HTML has been changed and horizon.list.generate_html JS cannot handle it. 20:21:47 <e0ne> ying_zuo: I can take this one #1715066 20:21:54 <amotoki> one question for queens and django 20:22:15 <amotoki> do we bump the minimum version of django for Queens or keep the minimum >=1.9 ? 20:22:31 <ying_zuo> cool. thanks e0ne 20:22:36 <david-lyle> we're jumping straight to django 4 :) 20:22:58 <david-lyle> wait, wrong framework 20:23:06 <e0ne> david-lyle: you missed django 3.6 for python 3.6 ;) 20:23:07 <robcresswell> bump to 1.11 20:23:22 <david-lyle> I vote to bump to the LTS as well 20:23:33 <robcresswell> everything else is EOL during this cycle anyway iirc 20:23:49 <e0ne> david-lyle: +1 20:23:56 <robcresswell> Djangos lifecycle is about as aggressive as openstacks XD 20:24:08 <amotoki> one thing to note is bug 1715066 exists in Pike, so even when we bump the min to 1.11 we need to explore a solution which works for Pike too 20:24:09 <openstack> bug 1715066 in OpenStack Dashboard (Horizon) "Cannot launch instance with Django Launch Instance form with multiple networks for a tenant" [High,Confirmed] https://launchpad.net/bugs/1715066 20:24:44 <e0ne> #link https://www.djangoproject.com/download/ (Supported Versions section) 20:25:37 <gary-smith> Should bug 1547290 (gen'ing JS docs) be downgraded? It doesn't seem to "prevent a key feature from working properly" 20:25:38 <openstack> bug 1547290 in OpenStack Dashboard (Horizon) "We currently do not generate Javascript API docs" [High,In progress] https://launchpad.net/bugs/1547290 - Assigned to Richard Jones (r1chardj0n3s) 20:26:25 <e0ne> gary-smith: only if we can develop w/o these docs :) 20:26:52 <david-lyle> I think more of a medium 20:27:27 <ying_zuo> medium is good 20:27:48 <gary-smith> ying_zuo: looks like you are editing that one. Can you change to medium? 20:27:57 <ying_zuo> I just did :) 20:27:59 <gary-smith> Thx 20:28:48 <david-lyle> one could argue since we don't have API docs, that it's a bp rather than a bug, but details 20:29:17 <ying_zuo> That's right. It's a feature 20:30:21 <ying_zuo> I guess a bug report will do for now. 20:31:24 <ying_zuo> Anything else we should discuss regarding bug 1715066? 20:31:25 <openstack> bug 1715066 in OpenStack Dashboard (Horizon) "Cannot launch instance with Django Launch Instance form with multiple networks for a tenant" [High,Confirmed] https://launchpad.net/bugs/1715066 - Assigned to Ivan Kolodyazhny (e0ne) 20:32:07 <e0ne> here is a patch: https://review.openstack.org/#/c/289932/ 20:32:26 <e0ne> oops, sorry, a wrong window:( 20:34:30 <ying_zuo> cool. unit test is added :) 20:34:52 <ying_zuo> #link https://bugs.launchpad.net/horizon/+bug/1717542 20:34:53 <openstack> Launchpad bug 1717542 in openstack-secaudit " Possible client side template injection in horizon login screen" [Undecided,New] 20:35:25 <ying_zuo> This one is the last one I want to bring up today. It’s a security issue. Just wondering if someone is aware of this issue or has any thoughts on it. 20:35:56 <david-lyle> that wasn't marked correctly, hmm 20:36:01 <e0ne> ying_zuo: it wasn't me 20:36:04 <david-lyle> it's a public bug now 20:36:48 <gary-smith> sorry about that 20:37:09 <david-lyle> but I think the discussion went something along the lines of since we changed the escape characters, this is a false positive 20:37:41 <david-lyle> gary-smith, the bug filer has to mark it correctly 20:39:14 <david-lyle> which is why automated tool results have to be double checked 20:39:35 <ying_zuo> david-lyle: can you add a comment on the ticket? 20:40:12 <david-lyle> I believe it's on the actual security bug, but will double check if that's open yet 20:40:47 <david-lyle> which is this https://bugs.launchpad.net/horizon/+bug/1567673 20:40:48 <openstack> Launchpad bug 1567673 in OpenStack Dashboard (Horizon) "[OSSA-2016-010] Possible client side template injection in horizon (CVE-2016-4428)" [Critical,Fix released] - Assigned to Tristan Cacqueray (tristan-cacqueray) 20:40:59 <david-lyle> which is now public because fixed 20:41:10 <david-lyle> and OSSA published 20:41:37 <david-lyle> but will add comment 20:42:31 <ying_zuo> thanks david-lyle 20:43:36 <ying_zuo> That's the list of bug reports I have for today 20:43:54 <e0ne> ying_zuo: thanks for bringing this to the meeting 20:44:19 <david-lyle> I marked the issue as incomplete 20:44:35 <david-lyle> asking for verification 20:44:51 <ying_zuo> cool 20:45:05 <ying_zuo> As always, if you have any thoughts of any tickets, please leave a comment there. 20:45:31 <amotoki> i have one question 20:45:37 <ying_zuo> Also, confirming the bugs will be very helpful too 20:45:51 <e0ne> ying_zuo: +1 20:45:55 <ying_zuo> #topic Open Discussion 20:45:58 <amotoki> I see some heat related bugs were marked as Won't Fix 20:46:28 <david-lyle> I think until heat is excised marking them as won't fix is premature 20:46:31 <amotoki> don't we keep them open (with heat tag) and move them to heat-dashboard (once it is split out 20:46:32 <amotoki> ? 20:46:49 <david-lyle> I should think so 20:47:03 <ying_zuo> +1 20:47:37 <amotoki> when I checked bugs targeted to 'next' milestone, I noticed it but I have no good way to recover them... 20:48:15 <robcresswell> ? Just mark them as new again or something 20:48:28 <robcresswell> And target them to Heat instead of Horizon 20:49:20 <david-lyle> I wouldn't retarget them until there is a separate plugin 20:49:33 <david-lyle> but moving to new is correct, IMO 20:49:44 <ying_zuo> maybe just add a tag as heat 20:50:13 <gary-smith> sounds good 20:50:26 <amotoki> I added 'heat' tag to heat related bugs several days ago. 20:50:35 <robcresswell> Either is fine 20:50:50 <robcresswell> When heat decide how they want to manage bugs, we can move them over 20:51:45 <ying_zuo> cool. what's next 20:53:21 <ying_zuo> any other questions or concerns? 20:53:30 <amotoki> i have one more thing to raise. https://bugs.launchpad.net/horizon/+bug/1321865 20:53:31 <openstack> Launchpad bug 1321865 in OpenStack Dashboard (Horizon) "Networks panel can't be actually disabled" [Medium,Confirmed] 20:53:45 <amotoki> it is related to the template design and raises a question on where we should have common templates. 20:54:43 <amotoki> i just would like to know opinions to the team: do we really support disabling some specific project panel? 20:54:54 <david-lyle> we should 20:55:03 <e0ne> 5mins reminder 20:55:16 <robcresswell> All you can really do atm is hide it :/ 20:55:50 <amotoki> the problem is if we hide a project network panel, the admin network panel does not work.... 20:56:14 <amotoki> it depends on templates from the project network panel.... 20:56:24 <david-lyle> ideally, this is a policy configuration 20:56:49 <david-lyle> and only admin type users have roles to see any networking panels 20:57:14 <amotoki> policy configuration on templates? 20:57:26 <david-lyle> on panels 20:57:37 <robcresswell> What I meant was, you can hide the pabel, but still register it 20:57:45 <robcresswell> Right ^^ 20:57:57 <amotoki> ah.... 20:57:58 <david-lyle> amotoki, you've kind of hit two use cases which are different 20:58:07 <robcresswell> So then it doesn't affect admin 20:58:08 <david-lyle> one is remove all network panel support 20:58:24 <david-lyle> the other is hide project/network but not admin/network 20:58:38 <amotoki> yes, my point is the latter. 20:59:07 <david-lyle> ok, in that case I would suggest policy rather than enabled/disabled 20:59:44 <amotoki> ah I see. I think i got your point 20:59:55 <david-lyle> unless you're just wanting to change horizon and not cli/API behavior 21:00:07 <robcresswell> We are about at time :) 21:00:29 <robcresswell> david-lyle: Nothing to stop you altering Horizons copy of policy only 21:00:29 <david-lyle> at that point I think common template location should be above dashboard like usages 21:00:43 <ying_zuo> that's right. we can continue the discussion in #openstack-horizon. 21:01:04 <robcresswell> Thanks everyone! \o 21:01:06 <e0ne> see you next week 21:01:19 <ying_zuo> thanks everyone for joining the meeting! 21:01:25 <ying_zuo> #endmeeting