22:02:19 #startmeeting horizon 22:02:20 Meeting started Tue Mar 5 22:02:19 2013 UTC. The chair is gabrielhurley. Information about MeetBot at http://wiki.debian.org/MeetBot. 22:02:21 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 22:02:23 The meeting name has been set to 'horizon' 22:02:27 #topic General 22:02:35 Hi everyone! 22:02:41 Things are going very well with the RC 22:03:00 I triaged the bugs this morning and there wasn't anything unexpected. 22:03:14 What's on the current list is looking good. 22:03:24 #topic Blueprints and Bugs 22:03:44 the one remaining blueprint is cleared for merging, so I'm gonna give that a final review and merge it unless anyone objects. 22:04:04 I would like to talk about this bug for a moment: https://bugs.launchpad.net/horizon/+bug/1060426 22:04:06 Launchpad bug 1060426 in horizon "Login screen doesn't show reason for being logged out" [High,Confirmed] 22:04:20 This bug leads to more bug reports than anything else in Horizon 22:04:27 It confuses people constantly 22:04:34 And what I'm wondering is this: 22:05:18 is it better to 1) show the "reason" (e.g. "Unauthorized, please try logging in again."). Or is it better to *not* log people out and let them see the unauthorized messages in the dashboard for the API requests that can't be completed. 22:05:33 the latter being option 2, since I fail at numbering. 22:06:09 From a faux-security perspective it feels more right to log them out, but I think the better UX is to let them be logged in and unable to act... 22:06:11 thoughts? 22:06:17 yeah, i was thinking about this when fixing another bug 22:06:18 A clear error message in the dashboard would make more sense IMO 22:06:28 Or be more user friendly 22:06:38 i think a dashboard message plus a link to logout/login would be better 22:06:47 As a caveat, fixing this by way of *not* logging them out is a lot easier to implement. 22:07:00 Cool. I think that's the direction I'll take it then. 22:07:10 is lin-hua's comment there correct? 22:07:14 I think that an error message will be great too 22:07:28 i have seen permission denied messages on the login page before 22:07:50 that should be investigated as a related (simple) fix 22:08:15 which I'll do. since the ticket is assigned to me. 22:08:18 lol. 22:08:32 haha 22:08:33 any other bugs folks want to talk about? 22:08:57 i think the comment is saying it's already fixed 22:09:13 Oh, just as a general note: I'm planning on doing a complete Django 1.5 compatibility pass as soon as the H cycle starts. 22:09:15 but the change you suggested would be better 22:10:58 i just have 22:10:59 https://bugs.launchpad.net/horizon/+bug/1072848 22:11:01 Launchpad bug 1072848 in horizon "Switching projects sends user "home" inappropriately" [Low,Confirmed] 22:11:17 i made some changes in openstack_auth to support a fix 22:11:25 is it possible to cut a 1.0.7 release of that? 22:14:39 oh, yes 22:14:41 will do 22:15:06 great 22:15:09 #topic open discussion 22:15:10 the horizon fix is one line 22:15:15 sweet 22:15:36 gabrielhurley, jogo suggested to add tenant deletion on Keystone directly. How this impact my current work on the blueprint? https://blueprints.launchpad.net/horizon/+spec/tenant-deletion 22:15:56 hard to say 22:16:27 The Horizon BP is basically a stop-gap for keystone doing something about it 22:16:36 and getting that into keystone may be a bit controversial 22:16:46 so I'd proceed as planned but stay in the loop on that issue 22:17:20 gabrielhurley, Yeah I agree... 22:17:29 Is there an existing bug or blueprint in keystone for this? 22:17:41 gabrielhurley, Currently I'm having some blockers, but overall progress is good 22:18:15 vkmc: cool. do you know if there's a keystone BP for it? I'm not aware of one, but that doesn't mean it doesn't exist. 22:18:54 gabrielhurley: vkmc: bug 1129037 is one, but it's sort of a dupe... i swear we had another 22:18:55 Launchpad bug 1129037 in keystone "keystone tenant-delete" [Undecided,New] https://launchpad.net/bugs/1129037 22:19:11 yeah. there oughta be one if there isn't 22:19:16 then we could link the horizon and keysotne BPs 22:19:34 gabrielhurley, Saw the one shared by dolphm 22:19:37 dolphm, Thanks 22:20:46 kspear: openstack_auth 1.0.7 is published to PyPI 22:21:01 last time it was discussed it was determined to be an orchestration issue, although there's probably some way for keystone to help with that, such as publishing a list of deleted projects, etc 22:21:17 gabrielhurley: that was quick! thanks 22:21:17 yeah. that's why I say that it's gonna be a while to get it into keystone 22:21:25 so I think Horizon should go ahead with it 22:21:29 s/it/our plan 22:22:04 vkmc: make sense? 22:22:53 gabrielhurley, It does, yeah 22:23:12 gabrielhurley, And it's a relief to hear that :) 22:23:14 cool 22:23:27 other topics from folks? 22:23:49 do we have a date for rc1? 22:25:35 I heard March 12th 22:25:46 gabrielhurley: sorry, accidentally replied off list so i'm hitting you twice 22:25:47 so, next week, roughly 22:25:54 dolphm: no worries 22:26:14 good to know 22:26:25 kspear: yup. good question. 22:28:29 cool. with that I think we'll call it a day. 22:28:42 Great job everyone, same time next week! 22:28:52 #endmeeting