21:03:16 #startmeeting project 21:03:16 Hi, but I am taking a quick bio break, back in a sec 21:03:17 Meeting started Tue Apr 22 21:03:16 2014 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:03:18 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:03:21 The meeting name has been set to 'project' 21:03:33 We'll abuse this meeting today to discuss Juno Design Summit scheduling: 21:03:38 #link http://wiki.openstack.org/Meetings/ProjectMeeting 21:03:47 #topic Design Summit scheduling 21:03:57 First, do you have questions on the review/scheduling interface ? 21:04:10 The first step is to review sessions and set their status to Preapproved or Refused 21:04:32 Once that's done, you can use the scheduling screen to drag and drop sessions to available time slots 21:04:48 One thing to keep in mind is that you can pre-approve more sessions than you have available slots, and then merge multiple sessions into the same schedule slot. 21:04:48 ttx: no questions, but is there a wiki page about it? if not I might create one 21:05:02 zaneb: it has Help screens! 21:05:06 would have been good to know some of that stuff earlier :) 21:05:13 i'm not sure if it's a new feature or not, but i found it surprising this cycle that you could drag multiple proposals into the *same* timeslot to merge them 21:05:19 one question, what's the typical over-subscription ratio? 21:05:30 ttx: i only picked up on it from your email 21:05:31 zaneb: feel free to document your experience :) 21:05:32 (i.e. proposals to available slots on track) 21:05:47 ttx: The merging of proposals is quite awesome! 21:05:53 fwiw, ironic is subsribed 18:4 21:05:54 eglynn_: here is the secret URL: http://summit.openstack.org/cfp/topicstatus 21:06:15 devananda: dang 21:06:21 ironic wins 21:06:29 . 21:06:32 ttx, eglynn_: oooh. but rejected sessions are not shown on that 21:06:42 ttx: nice :) 21:06:42 followed closely by Sahara and Swift 21:06:45 zaneb: i believe rejected sessions vanish 21:06:54 dolphm: correct 21:06:59 zaneb: yes, only works for so long 21:07:18 it's more of a "reminder of the problem" type graph 21:07:21 eglynn_: heat was 2:1 oversubscribed fwiw 21:07:24 remainder? 21:07:51 Do you have sessions you would like to push to another topic ? 21:08:16 Remember you will all have a project pod where you can discuss overflow sessions 21:08:29 i've proposed a few sessions to other tracks (eg, for ceilometer and nova integration) 21:08:30 This will be available all 4 days of the summit 21:08:34 ttx: what's the deadline for scheduling? 21:08:36 \o/ 21:08:48 mikal: aim for end of week? 21:08:58 given taht ironic would like to graduate, and that integration is now a pre-requirement, it'd be great to have those discussions some time :) 21:09:07 it's just a round table with a flipchart, but it will be a designated spot (with your program name on it) and there will be a map to help locate it 21:09:08 I was going to discuss it my meeting at the end of the weke, but have just realized I'll be on a plane during that meeting 21:09:21 mikal: goal is to have a basic proposal up at end of week 21:09:30 but then we can tweak it until the last minute 21:09:30 ttx: ok 21:09:51 ttx: I plan to put tripleos things in on monday 21:09:56 mikal: if we have "most" in by the end of week we'll be fine 21:09:59 ttx: what tools will be available in the overflow pods? whiteboard? projector? also, how many ppl can effectively congregate there? 21:10:00 ttx: we're waiting on core team reviews of the proposals 21:10:02 ttx: is that ok ? 21:10:09 lifeless: that's good. Just won't be around to assist in case stuff falls apart 21:10:22 * ttx takes the off week off 21:10:27 ttx: long as we have our 6 slots we're fine :> 21:10:34 so what's the PTL "tribal knowledge" on the sense or o'wise of merging unrelated sessions? 21:10:39 (... just to shoehorn things into the available slots) 21:10:43 I have an etherpad with the neutron schedule, working on pushing those into the tool now. 21:10:53 eglynn_: its doable but not with contentious sessions 21:11:01 eglynn_: try to have things that will attract the same audience 21:11:02 lifeless: fair point 21:11:02 eglynn_: well, previously we didn't have project pods, so the dynamic is slightly different this time 21:11:13 eglynn_: otherwise you get lots of folk moving in and out 21:11:16 eglynn_: markmcclain did that with neutron a few times, you might want to ask him how well it went 21:11:23 ttx: For trove, we're planning on going through the proposals tomorrow. I should have the schedule set by Friday. 21:11:26 eglynn_: e.g. we're likel merging *ALL* our CI issues into one session 21:11:27 dhellmann: thanks will do 21:11:31 eglynn_: I would make sure you use your "scheduled" slots to reach out to people and identify the interested people 21:11:32 eglynn_: he says "you need to be a strict time keeper" 21:11:34 eglynn_: which should let sdague drop in... 21:11:51 cool 21:11:56 eglynn_: then the project pod can be used to work, hackathon style 21:11:59 eglynn_: merging 2 uncontested sessions with a fairly even split isn't bad. beyond that, i haven't see nit work well 21:12:19 eglynn_: unless the topics are very similar to begin with 21:12:23 Does anyone expect to have unused slots they could give to someone else ? 21:12:28 ttx: some session proposals would be better for project pods than design sessions - is there anything published about project pods that i can direct people to? 21:12:34 * eglynn_ rapidly cools on the idea ... 21:12:54 * devananda is also wondering what the pods will provide 21:12:58 eglynn_: last time, Glance and Horizon had extra slots they gave out. 21:13:11 dolphm: let me check 21:13:24 eglynn_: I'm planning to do a lot of merging, but only of related topics 21:13:34 https://wiki.openstack.org/wiki/Summit has a paragraph about pods 21:13:36 sahara schedule will be hopefully ready by the end of week, no unused slots already 21:13:41 dolphm: ^ 21:13:46 ttx: all looking over-subscribed this time if I read those graphs right 21:13:52 ttx: thanks! 21:13:55 eglynn_: yes 21:14:00 (so alms for the poor looking unlikely) 21:14:08 eglynn_: less lots overall kinda does taht to you 21:14:15 ttx: "round table" -- so not a reasonable place for the whole team to gather, then 21:14:56 devananda: You can steal extra chairs and invade others space ! 21:14:57 devananda: that was my question too - do we have some sense of the actual number of seats / physical space? 21:15:27 * ttx has no idea if this pod thing will work 21:15:49 kgriffs: let me look at the map I was given 21:15:53 ttx: not if you're stealing chairs! 21:16:07 ttx: 'spare slots' ahahahahahahahahahahahahahahahahaha 21:16:12 ttx: the concept of pods makes the 4:1 oversubscription somewhat more bearable. :-) 21:16:41 kgriffs: so.. the map says the roundtablaes are 66", whatever THAT means 21:17:05 They drew 10 chairs around them 21:17:08 9-10 people 21:17:10 Seats 9-10 21:17:17 1.67 m 21:17:26 that should work OK. If things get crazy we can steal more chairs. :p 21:17:34 you can definitely invest the walls that are near to your table 21:17:58 but they are dark and don't support projectors well? 21:18:18 notmyname: well, you saw the photo. haven't seen them in real 21:18:35 * ttx finds a link 21:18:55 ttx: ah, ok 21:19:16 http://www.gwcc.com/event_planners/Interactive_Planning_Guide.aspx > Click on Level 2 in building B 21:19:35 pods will be in rooms B201 B202 B203 B204 21:19:45 if you want to look what the rooms look like 21:20:07 there is a "photos" button 21:20:15 after you select a specific room 21:20:29 they look substantial enough 21:20:31 now you know as much as I do. 21:20:44 I can forward questions if you have any 21:20:59 Do you have scheduling constraints or conflicts you would like to tell us about ? 21:21:11 like a session where Nova folks need to have Neutron guys around 21:21:12 No yet! 21:21:15 * dolphm it doesn't bother me at all that they misspelled "ceiling" AND "height" in this tool in the same label 21:21:32 dolphm: not exactly reassuring 21:21:33 ttx: there was a mail thread during the TC meeting about that. It looks like we can do it Thursday arvo 21:21:50 ttx: it's probably fine for banqets 21:21:58 mestery: you had a qa/neutron session on the neutron track right? 21:22:14 mikal: yep, that would fall in the space where no neutron is running 21:22:16 dolphm: you should know that GA education is dead last in the US 21:22:46 markmcclain: I think they prefer "first in opportunities to improve" 21:22:49 mtreinish: Yes: http://summit.openstack.org/cfp/details/366 and http://summit.openstack.org/cfp/details/401 21:22:56 mtreinish, jeblair, dhellmann : ideally we would publish our schedules soon, so that the others can plan around it 21:23:16 ttx: ack 21:23:19 ttx: working on it 21:23:21 ttx: What happens if we accidentally press "push to sched" early? :) 21:23:22 I've got one talk that should probably be moved to the cross-project session 21:23:29 mestery: oh they're multiplying, I only knew about the first one 21:23:36 i.e. a neutron QA session in the neutron track could try to avoid times where a QA session is running 21:23:41 notmyname: what is that? 21:23:43 notmyname: I'm pretty sure the cross-project session is full 21:23:43 mtreinish: We're combining those ones for now, as they overlap a bit. 21:23:44 mestery: you can redo it later 21:23:48 sdague: Thanks. 21:23:56 ttx: should have it Thurs. I want to discuss it during the meeting this week before I schedule anything 21:24:00 http://summit.openstack.org/cfp/details/192 <-- glance + swift + keystone 21:24:01 mestery: it will just push the current state of the schedule 21:24:03 sdauge: Not that I accidentally hit "push to sched" or anything. :) 21:24:03 so pushing early isn't and issue 21:24:14 thanks ttx 21:24:25 notmyname: checking how stuff aligns 21:24:36 notmyname: we tried to keep the cross-project sessions to those affecting all (or at least most) projects, not just >1 21:24:53 Remember slot layout lives at https://docs.google.com/spreadsheet/ccc?key=0AmUn0hzC1InKdGNXcWlWX0FIekQxbUtvRVlnVF9IV3c&usp=drive_web#gid=5 21:25:20 notmyname: probably worth seeing if you can do that with a joint swift/glance/cinder session 21:25:24 notmyname: as long as you avoid the first two slots of the swift track, you should have all of them 21:25:35 yeh, I think a session like that we'd have kicked back to project tracks. Like the various neutron+nova things we kicked back out. 21:25:43 if you want cinder as well, that limits choices 21:25:46 well as one of 22 proposed for 8 slots, I'm not sure if it will get chosen anyway in the swift track 21:25:46 ttx: notmyname we could possibly pu tit in Cinder 21:25:59 Oh, an offer to take one slot! 21:26:14 jgriffith: how generous of you :-) 21:26:35 notmyname: I'd of course like to tweak the session a bit :) 21:26:58 jgriffith: Friday after 2:10pm you would have swift/glance and keystone folks available 21:27:03 jgriffith: ya, makes sense 21:27:06 even more generous! ;) 21:27:22 ttx: notmyname we can take it, I have been wanting to propose a glance/cinder 21:27:32 jgriffith: cool, thanks! 21:27:38 ttx: notmyname I think this counts and could be abstracted a bit 21:27:48 jeblair: anything you'd like to try to sneak in release management ? Or can I run with my canned 4 ? 21:28:28 jgriffith: FWIW, Donagh is in Ireland, so keep aware of that when getting in touch with him 21:28:39 ttx: nope, thanks 21:29:09 jeblair: I have a release branch vs. QA vs. infra session filed in release management already 21:29:17 where we can also disucss requirements 21:29:27 ttx: are you the only one who can rehome a talk or can we do it? 21:29:39 notmyname: no you can do it 21:29:53 on the review screen, click in the topic column (where it says Swift) 21:29:55 ttx: i'd like to ask that http://summit.openstack.org/cfp/details/198 be scheduled tuesday morning (not overlapping ironic) 21:30:23 russellb: ^ 21:30:25 ttx: of the cross project sessions taht i dont want to miss, it's probably the most important for ironic 21:30:25 ttx: ah, cool 21:30:27 ttx: yeah, i saw that. ++. i don't see anything overlapping that... 21:30:30 jgriffith: moved. thanks 21:30:42 mtreinish, ttx: do you have any thoughts on this session: http://summit.openstack.org/cfp/details/285 21:30:59 notmyname: np 21:31:21 jeblair: I'm not 100% convinced we need a session for that 21:31:33 jeblair: yeah I'm not sure either, but I know we want to do it 21:31:49 it seems like it's more working through issues as they come up 21:31:50 devananda: it's already scheduled at http://junodesignsummit.sched.org/ -- does it conflict ? 21:32:07 * ttx checks 21:32:30 tuesday 5:30 pm 21:32:30 oh, it does 21:32:36 yep 21:33:03 ttx, mtreinish: that's kind of what i'm thinking. let me know if you see a session we might be able to combine it with; otherwise, maybe hallway track or pod for that one. 21:33:06 ttx: is it worth shuffling the incubated slots instead? 21:33:25 sdague: maybe 21:33:27 i will be delegating folks from ironic for the afternoon cross project sessions that we need to sync with, but where I dont need to be present 21:33:48 like switch marconi and ironic 21:33:50 devananda: how about you swap with Marconi slots ? 21:34:02 kgriffs: would that work for you ? 21:34:20 the morning cross proejct slots are not especially pertitnent to ironic 21:34:36 yep, I'm flexible 21:34:38 consistency across APIs would be nice to attend, but i dont think we'll have much input there 21:34:54 devananda: looks like we have a winner 21:35:01 kgriffs: thanks! 21:35:09 devananda: i'll have to edit the slots on the scheduling site 21:35:19 yep. i wont schedule anything yet 21:35:53 #action ttx to move slots around between Marconi and Ironic 21:36:04 will do that tomorrow morning. At this hour I would fumble it 21:36:20 So, perhaps I'm slow. I can see a commenting interface on summit.o.o. Where do I set things to approved / rejected etc? 21:36:43 mikal: you should have a "Review Nova topic" button on the main screen 21:36:53 mikal: you have to click on the "unreviewed" word on the far right 21:36:54 ttx: yep, clicked that 21:37:00 mikal: not the topic name 21:37:01 notmyname: ahhh, ok! 21:37:11 took me a while to find that too 21:37:19 mikal: I've been doing this for years and ttx just tought me about the rehoming today! 21:37:23 mikal: also the "Help" screen (link at bottom) is supposed to contain useful help 21:37:25 Oh, you can't do that in the proposal view screen? 21:37:28 Just that summary list? 21:37:48 mikal: yes, you have to be on the review screen 21:37:55 Ok 21:37:57 I can live with that 21:38:12 first person who tells me the site sucks gets a link to the openstack-infra repo which contains the code for it :) 21:38:25 heh 21:38:28 The site is wonderful! 21:38:29 ttx: it's awesome :) 21:38:50 mikal: I KNOW! It has this little Launchpad feeling! 21:39:37 #topic Open discussion 21:39:44 might as well switch to that 21:40:27 jeblair: looks like you have a duplicate on the SSO stuff 21:40:45 ttx: yes, i'm combining those 2 21:40:56 jeblair: by the way, we are sharing the same pod! 21:41:06 * mtreinish has to drop 21:41:08 I have questions about the Dev/Ops session(s) that were proposed 21:41:12 If we combine two proposals, do we end up with a proposal with two presenters? 21:41:14 were those proposed to every project? 21:41:23 i.e. is that more politically acceptable than accepting one and rejecting the other? 21:41:24 dhellmann: any chance you have space for this in oslo? http://summit.openstack.org/cfp/details/69 21:41:26 mikal: basically, yes 21:41:36 notmyname: I think that was Tom's plan yes 21:41:42 And then I can edit the proposal text to make it make sense? 21:41:43 ah, yes. I see them now 21:41:54 mikal: so.. let me explain 21:41:58 jeblair: only if you promise all of you guys come and help us decide what to do -- I need your input on that one 21:41:59 mikal: you end up with sessions like this http://junodesignsummit.sched.org/event/20465cd62e9054d4043dda156da5070e 21:42:01 notmyname: I believe only projects who requested it, but that was quite a lot 21:42:07 mikal: and all you can do is add some prefix text, which i didn't do there 21:42:14 (yet) 21:42:39 Hmmmm, ok 21:42:48 mikal: if you add more than one session on a time slot, it becomes a combined session with the two descriptions concatenated. You can edit the title of the session and add a small paragraph on top so that the thing doesn't look too ugly 21:42:50 dhellmann: definitely, probably want qa folks there too 21:42:52 And the way you merge is in the scheduling UI? 21:42:58 we have ten interns working this (northern hemisphere) summer on OpenStack through the Google Summer of Code and GNOME Outreach Program for Women programs 21:43:02 mikal: i suppose you could also stomp on the proposed session descriptions to eliminate redundancy, etc 21:43:03 jeblair: true 21:43:05 ten! 21:43:13 jeblair: ok, send it over, and I'll try to make room 21:43:13 annegentle: nice! 21:43:15 mikal: but do that individually before getting to the scheduling screen 21:43:22 mikal: but it's also valid to tweak one so that it includes the other and then reject session 2 21:43:26 jeblair: oops, I can change that one, nevermind 21:43:34 annegentle: that's awesome :) 21:43:37 mikal: if you are comfortable with the proposers 21:43:43 jeblair: hmm, no, I can't change the track 21:43:58 Ok, I will muddle around and do my best, making a lot of enemies in the process and leading to my ultimate downfall. Its a greek trajedy of summit organization really. 21:44:10 mikal: I can help you. 21:44:41 dhellmann: it's probably going to be tricky scheduling it; there's a lot of infra/oslo/qa/docs/release overlap. i definitely want it to happen and attend though. 21:44:48 mikal: don't hesitate to ping me when you want to work on it 21:45:10 mikal: beware the ides of may 21:45:13 jeblair: yeah 21:45:21 mikal: and yes, you merge by dragging more than one session to the same time slot. 21:45:36 ttx: thanks, I'm starting to fight with it now, but slowly 21:45:40 notmyname: so true 21:45:56 ttx: how do we move to another topic? 21:46:05 ttx: does the UI support merging >2 proposals? 21:46:12 devananda: yes 21:46:13 jeblair: I know this one! click on the topic on the far left 21:46:20 dolphm: thanks 21:46:30 ttx, notmyname: got it thx 21:46:31 devananda: the session i linked above is 4 proposals merged into one session 21:46:40 notmyname: looks like I should make that a bit more discoverable. 21:46:47 devananda: yes 21:47:32 dhellmann: http://summit.openstack.org/cfp/details/69 moved to oslo. mtreinish: fyi on a session infra and qa folks should try to attend 21:47:46 dhellmann: maybe you should wait until the Docs/RelMgt/Infra/QA room is scheduled to do the Oslo scheduling 21:47:46 jeblair: ack 21:47:50 ttx: IMO having one "edit" screen accessed by any link on that session would be nice 21:47:59 ttx: yeah, good point 21:48:10 dhellmann: that should make choices easier 21:48:44 notmyname: yes. I'll admit not having worked AT ALL on that app this cycle. 21:49:23 seems much more important to spend time on story board 21:49:23 ttx: also, when I was doing open compute project stuff, they were really wanting to use the same system for scheduling sessions at their conference. it's a popular thing! 21:50:17 ttx: I'd really like to let more folk in the project twiddle the layout 21:50:35 ttx: ptl-only takes ptl from an arbitrator to a only-one-that-does situation 21:50:58 lifeless: it's just a convenience gateway, you don't have to make all the calls 21:50:58 lifeless: +1. tbh, I'd love to see the selection/scheduling done in Gerrit 21:51:17 ttx: copying stuff out of it to an etherpad is painful and time consuming 21:51:19 I just need to have one go-to name to complain to if it's not done 21:51:34 ttx: thats fine - have a the PTL as ultimately responsible 21:51:37 lifeless: do the twiddling in google docs or something, and then enter the final outcome into summit.openstack.org ? 21:51:56 dolphm: sure - we're using etherpad. Its the copy-paste-sync thats painful 21:52:00 lifeless: but yeah, I guess the tool could allow for more collaboration in scheduling 21:52:02 because you have an app that doesn't delegate 21:52:14 and we tell people to use that app 21:52:29 lifeless: or we could have some export/import that would let you iterate off-tool with more people 21:52:32 but many? most? projects don't scheule in it 21:52:48 ttx: import-export to something gerrit managed would be sweet 21:52:57 ttx: maybe an export session titles and details links to csv feature? 21:53:00 gerrit for EVERYTHING 21:53:01 :) 21:53:13 ttx: first cut of infra is pushed to sched 21:53:43 (but it takes awhile for sched.org to reflect) 21:53:57 yes, somehow sched takes time to show up stuff 21:54:33 dolphm: funny how it fails to detect a new topic 21:55:02 it takes anywhere from 5 to 60minutes for the new sessions to be recognized as a new topic 21:56:12 jeblair: pushed the tentative RelMgt schedule 21:57:01 anteaya: ^ 21:57:49 OK, we are running out of time 21:58:00 Don't hesitate to ping me this week if you have questions 21:58:12 I'm away next week, so better ask questions this week 21:58:31 Last words ? 21:58:33 thanks 21:58:42 Oh. Vote! 21:58:46 ttx: thanks for the hand holding 21:58:47 :D 21:59:15 #endmeeting