21:02:05 #startmeeting project 21:02:05 Meeting started Tue Apr 23 21:02:05 2013 UTC. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:06 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:08 The meeting name has been set to 'project' 21:02:22 We'll skip per-project status this week, in favor of generic advice and open discussion 21:02:28 #link http://wiki.openstack.org/Meetings/ProjectMeeting 21:02:34 should take us less than the hour 21:02:41 #topic Grizzly documentation release status 21:02:47 annegentle: still targeting April 30 ? 21:02:48 I only accept generic advice anyway 21:03:01 Yep, planning to release/publish 4/30 21:03:15 #link https://wiki.openstack.org/wiki/Documentation/Release 21:03:16 What's included in the docs release exactly ? 21:03:17 that's the process 21:03:39 it has to do with URLs, comment threads, builds, um. all the things 21:03:46 everything under docs.o.o ? 21:03:46 #info Doc released planned for 4/30/13 21:04:14 make sure Google indexes correctly, etc. Yep, everything under docs.openstack.org/grizzly anyway 21:04:18 o/ 21:04:24 annegentle: Any specific issue that needs urgent help/attention from others ? 21:04:25 Also I'm working on "who wrote Grizzly docs?" data analysis. 21:04:52 yes. Specifically install docs. 21:05:00 There are two doc patches in review for the Basic Install document: 21:05:10 #link ubuntu https://review.openstack.org/#/c/26954/ 21:05:25 #link fedora https://review.openstack.org/#/c/26919/ 21:05:31 #help Ubuntu folk please help on https://review.openstack.org/#/c/26954/ 21:05:43 #help Fedora folk please help on https://review.openstack.org/#/c/26919/ 21:06:00 Networking team seems to have docs on their radar, thank you to them. 21:06:07 jgriffith: you have a docs Q or one for ttx? 21:06:23 annegentle: no, sorry, just showing up late :) 21:06:25 annegentle: no, I think he was just indicating his presence 21:06:33 jgriffith: good to see you! 21:06:37 :) 21:06:40 ttx: that's it for docs. 21:06:53 annegentle: ok 21:07:07 Questions on doc release ? 21:07:35 #topic Havana Release schedule 21:07:46 During the Design Summit we had, like always, a session on release schedule 21:07:53 The proposed one looks like this: http://ubuntuone.com/p/mhD/ 21:08:07 Release on October 17, Feature freeze on Tuesday, September 3 21:08:20 The main issue raised about that timing was that Monday, September 2 was US Labor day 21:08:36 To reduce the issue I propose we have Feature freeze on the Wednesday (September 4th) and havana-3 tagging on the Friday (September 6) 21:08:44 Does that sound good ? 21:08:49 ttx: +1 on Wed 21:09:00 +1 on Wed 21:09:14 +1 on wednesday 21:09:18 +1 21:09:34 Unless there are objections I'll publish it tomorrow and adjust milestones for projects that are using them 21:09:35 I thought Thurs. was your "magic" release day? 21:09:53 annegentle: It's the "thank god there is friday to catch up in case of problems" day 21:10:19 annegentle: I can publish on Saturday morning if shit happens 21:10:25 s/in case of/due to/ 21:11:06 #action ttx to publish havana schedule with Feature freeze on the Wednesday (September 4th) and havana-3 tagging on the Friday (September 6) 21:11:17 Questions about the release schedule ? 21:12:10 #topic Havana cycle plans 21:12:26 For the next week(s) the main goal will be to turn the summit sessions into a reasonable set of actionable blueprints 21:12:37 The rationale behind tracking work being done as blueprints is to communicate plans to the wider community -- not everyone has time to follow each project closely 21:12:50 With blueprints we know what should be worked on, who should be working on it and when it's planned to land 21:12:56 So the plan is: 21:13:10 1- File blueprints for stuff that people volunteered to work on. A blueprint needs to cover a clear feature that can be "completed" in the Havana timeframe. "Improve database performance" is not really a good one. 21:13:19 (For complex stuff, file multiple blueprints and create dependencies between them) 21:13:31 (People should set the Assignee, the "Milestone target" to when they think it should be landed, and propose the "Series goal" to havana) 21:13:54 2- You PTLs review the proposed blueprints, accept the "Series goal" (you can use the page at https://blueprints.launchpad.net/YOURPROJECT/havana/+setgoals to do that) and set a priority 21:14:09 Priority reflects how important a given blueprint is to your objectives for this cycle. "Essential" means "cannot ship without" and should be used carefully 21:14:22 Blueprints that are accepted in series show up on https://blueprints.launchpad.net/YOURPROJECT/havana -- that's what we'll refine over the next weeks 21:14:48 Questions on this process ? 21:15:24 \o 21:15:33 jd__: you'll have to read the meeting logs :) ^ 21:15:46 gabrielhurley: shoot 21:15:55 ttx: you usually only open the next milestone shortly before the current one closes, but that makes it a little trickier to "lay out" plans for when blueprints will happen 21:16:07 e.g. I can target things to H1, or generally to Havana 21:16:14 but can't say what should be H2 or H3 21:16:17 gabrielhurley: no no, you cvan target stuff to h2/h3 21:16:29 oh, did you open those? 21:16:35 they weren't there when I looked the other day 21:16:38 gabrielhurley: yes, I created them today 21:16:48 heh. you pulled my favorite trick! 21:16:54 ttx: flying under the radar 21:16:54 I should've refreshed ;-) 21:17:03 thanks 21:17:12 Using my whole new https://github.com/ttx/openstack-releasing/blob/master/create_milestones.py script! 21:18:06 gabrielhurley: ideally the whole plan (h1/h2/h3) should be laid out -- but I know that h3 is generally a moving target 21:18:25 Other questions ? 21:19:08 Next week we'll start reviewing those havana plans. let me know if you need help with this step. 21:19:12 #topic Havana Design Summit feedback 21:19:19 ttx: sure thanks, sorry my IRC bouncer got killed at some point 21:19:26 While it's still hot in your memory, any suggestion for improvement on the Design Summit part ? 21:19:53 The seperate section of the conf hall helped a lot IMO 21:20:13 Overlap of sessions was an issue, but I think that's going to be the way it goes 21:20:27 I found it a difficult to know what was going on in unconference sessions on the fly 21:20:29 yeah, I was a bit skeptical about it when Lauren mentioned it... but it was the only thing that kept us sane 21:20:46 better wifi in the rooms needing Etherpads 21:20:47 markwash: +1 21:21:02 we talked a bit about reducing the overlap with the conference part 21:21:04 annegentle: ethernet was available to presenters, afaik 21:21:06 we could have used whiteboards in the rooms, too, for drawing diagrams 21:21:15 markwash: I think we'll have to put the unconference stuff online to improve visibility 21:21:18 wifi improved after day 1 21:21:18 dhellmann: +1 21:21:21 i.e. giving us a couple of days where we can do design summit sessions without the conference in parallel 21:21:23 dhellmann: some rooms had some 21:21:28 to reduce distraction, let us focus 21:21:36 ttx: ok, maybe I just wasn't in those rooms. did we have to ask for them? 21:21:37 * markmc thinks that's worth trying to figure out 21:21:39 I definitely asked for one in every room during the sroom steup thing 21:21:42 unconferences seemed to be generally more productive than regular sessions, IMO 21:21:43 dhellmann: I wanted a white board and took one from another room 21:21:52 markmc: +1 21:21:57 dhellmann: theoretically there should have been one in each 21:22:01 * dhellmann shakes fist at annegentle 21:22:12 * annegentle slinks away with markers 21:22:18 the two big rooms in the back of the dev lounge with round tables were AWESOME for impromptu unconferencing 21:22:22 * jgriffith now knows why not whiteboard in 110 21:22:45 * markmc spotted a whiteboard in the wild 21:22:49 someone even wrote on it 21:23:00 markmc: we can track who stole it then :) 21:23:10 ha! 21:23:16 dolphm: wow I never found those! I needed that. 21:23:25 it would be nice if the dev area wasn't locked down during the keynotes. I know they want us all to attend, but with limited time for hallway track / openspace meetings... 21:23:31 there was even a white board in the unconference room at some point :) 21:23:33 I thought the summit 101 session was helpful 21:23:44 About session overlap - heckj wants to try some more dynamic scheduling 21:23:57 define "dynamic" 21:23:59 dhellmann: +1 21:24:03 he agrees to be the point man to arrange sessions around. I know that's a nightmare to handle 21:24:13 oh, scheduling 21:24:17 sched.org ROCKS 21:24:24 dolphm: no more pre-defined topic layout 21:24:30 oh interesting 21:24:35 dolphm: have keystone sessions scattered all 4 days$ 21:24:37 * markmc literally couldn't have coped without sched synced into his calendar 21:24:41 markmc: have you seen guidebook? even better. actually downloads to your mobile device. 21:24:47 ttx: i would definitely like that 21:24:55 but +1 on the calendar sync 21:24:56 ttx: could you elaborate a bit? 21:25:04 give us time to talk about things in between sessions 21:25:16 ttx: heckj: oo I'd like to hear more too 21:25:17 dhellmann, yeah, all I used was the ical feed 21:25:34 jgriffith: Stuff is scheduled based on everyone's constraints, rather then in one pre-allocated block for that topic 21:25:48 ttx: yeah, got it... big +1 here 21:25:51 jgriffith: generally involves someone not sleeping for the last 7 days before the summit starts 21:26:01 I know that. I did it for the first ones 21:26:02 ttx: who sleeps the week before the summit? 21:26:24 that's what the plane is for 21:26:46 i.e. someone receives all the angry emails about doublebooking and tries to resolve all the constraints... then pisses off more people 21:27:08 not my definition of fun, and the main reason why we switched to preallocated slots 21:27:13 ahhh... I guess I didn't envision that extreme of dynamic 21:27:18 but preallocated slots have their drawbacks 21:27:32 like keystone folks not attending security track at conference etc 21:27:35 I interpretted it as just spaced out, still scheduled but not in blocks 21:27:40 block/tracks 21:27:45 * russellb is around now, sorry. 21:28:03 russellb: read scrollback and let me know if you have any question on the process 21:28:27 I just wish there was a way that common/process could get a half day head start 21:28:33 ack 21:28:45 that would allow more PTL and core involvement in those tracks 21:28:49 blocks/tracks means being in the same room for hours/days, not sure if that's okay for most people? (Efficient) 21:28:58 as a voice in favor of preallocated tracks, I found "clustered" sessions to be very helpful. also it helps when figuring out other schedules for the week when planning for the summit 21:29:06 markmcclain: +1 21:29:24 Maybe we can try to see if broader project topics fall out next time around? 21:29:30 ie: bare-metal 21:29:31 notmyname: I liked being in one room for my track, it was easier if intense 21:29:33 notmyname: it certainly simplifies the scheduling constraints, by solving a lot of them ahead of time :) 21:29:49 rather than having one in Nova, Glance, Cinder etc... combined versions 21:30:31 we'll see what can be done and what can be implemented in the already-negociated time and space 21:30:57 Any more feedback on Design Summit ? On the general event ? 21:31:23 I was really impressed with the crowd handling considering the growth. And I had a darn good time to boot. 21:31:29 To me, best yet. 21:31:35 definitely 21:31:51 annegentle: +1 21:31:51 the ratio of people attending design summit sessions and not participating is starting to get quite high 21:31:58 dhellmann: about general session blocking devs: I think Lauren mentioned that we would not do that anymore 21:32:16 dolphm, you thought so? I actually thought it was often less than previous times 21:32:20 dhellmann: They may actually have stuff running in the main big room all the time 21:32:27 + growing participants from women - Anne thxs for hosting social 21:32:46 dhellmann: with some tickets only allowing entracnce to that "keynote" space 21:32:50 ttx: good 21:32:52 patelna_: happy to play host! 21:32:53 ah, I like that, too 21:33:03 dhellmann: less people potentially wandering in our direction 21:33:05 markmcclain: other than the unconference sessions I attended, it seems like the majority of design session attendees were just there to lurk 21:33:25 patelna_: +1 21:33:34 TTX +1 21:33:37 I didn't mind the lurkers 21:33:49 we'll also mark the design summit session more clearly on the schedule, so that they can't be mistaken for a regular presentation 21:33:52 I was just glad the lurkers didn't participate if they were OT or distracting 21:34:02 something like "Design session: FOO" all over 21:34:02 dolphm: definitely 21:34:14 ttx: awesome, i was really confused when i was told the dev lounge was closed 21:34:18 ttx: could we mark them as separate tracks in sched.org to make it easier to follow the themes with different colors? 21:34:32 too many people in the design summit ...we need to split the days liek what we had before 21:34:42 Sometimes too many lurkers prevent you from having the discussion, that's the limit number we don't want to cross (hence relatively small rooms) 21:34:59 ttx: they can be smaller! 21:35:05 with bouncers 21:35:10 dhellmann: design summit sessions have their own color already 21:35:22 ttx: but they're not color coded by project anymore 21:35:24 ttx: I want separate colors for each project 21:35:36 we had that 21:35:44 or track or whatever (incl. docs, qa, process, etc.) 21:35:51 the problem was it was overcrowded 21:36:03 dhellmann: i'll see what sched lets you do. Having them grouped under a single "design summit" headline is convenient too 21:36:25 maybe we'll go back to separate sched websites :) 21:36:37 ttx: create two seperate schedules again, maybe? 21:36:40 * dhellmann didn't mean to open a can of worms 21:37:05 no no, that's good feedback, keep it coming 21:37:48 #topic Open discussion 21:37:56 Any preference on the order you guys should have on the regular meetings ? 21:38:13 I tend to place "provider" projects (Oslo, Keystone), before "consumer" projects (Horizon, Heat), but if anyone is based in Europe they can go earlier 21:38:22 i like being early in the meeting 21:38:22 I was thinking something like: 21:38:29 Oslo, Keystone, Ceilometer, Swift, Glance, Quantum, Cinder, Nova, Heat, Horizon 21:38:36 oh sure. put me last. ;-) 21:38:41 (that's fine) 21:38:46 gabrielhurley: you have a meeting next :P 21:38:58 I thought I was doing you a FAVOR 21:39:01 lol 21:39:03 Still waiting for my beer 21:39:07 haha 21:39:16 next time you're in california 21:39:51 ttx: fine with me! 21:40:11 ttx: no objections here 21:40:19 It's in the predator order 21:40:30 ha, yeah, it's fine with me 21:40:40 openstack food chain? lol 21:40:41 ttx: you should hold out for the germain-robin 21:40:46 Well it's not set in stone, we can always adjust 21:41:23 Anything else ? Questions about the release cycle, process, blueprints, anything ? 21:41:45 so the proposer should be able to set the milestone target? 21:41:45 are we nearly there yet? 21:41:59 russellb: yes, if he sets himself as assignee, certainly 21:42:06 ok, cool. 21:42:17 can we make "baby steps" the theme for this release? 21:42:22 maybe have a logo and stuff? 21:42:27 * russellb stares at his pile of 175 blueprints 21:42:34 markmc: how is "baby steps" related to "Havana" ? 21:42:43 Some kind of dance thing ? 21:42:46 baby steps + havana == baby with cigar 21:42:47 hmm 21:43:03 baby swaggering with cigar 21:43:05 anyway, I digress 21:43:17 I shall raise a thread about the I naming soon, we have a complex issue coming up 21:43:28 oooh, you tease 21:43:32 heh, still can't find an 'I' near next summit? :) 21:43:33 haha.. that's an understatement 21:43:33 as Chinese don't seem to start any word with I 21:43:41 hehe 21:44:01 there are a few places in the general vicinity, but not in China :) 21:44:22 * ttx doesn't want to create a diplocamtic incident, at least not until he is back home 21:44:29 diplomatic* 21:44:32 there's a manmade lake that starts with 'I', but a man-made tourist area is kinda lame 21:44:58 there are a few towns in Taiwan and Indonesia that start with I 21:45:00 someone write a google maps query to find the absolute closest anything to the summit that starts with the letter I, and we'll skip the vote 21:45:13 dolphm: you're the man 21:45:24 ttx, that sounds like a sure fire diplomatic incident 21:45:30 ttx, throw some flags into the mix too 21:45:44 Inspiration Lake: http://goo.gl/maps/8l79a 21:45:59 the Inspiration release 21:46:04 cheesy, but not bad 21:46:24 not the worst name ever. 21:46:26 especially cheesy once you see that the lake is artificial and the thing hosts a theme park 21:46:39 yeah that's the disappointing part :( 21:46:56 anyway, watch out for the thread, sometimes next week probably. You can do your research in the mean time 21:46:56 we should just call it "Island" 21:47:07 markmc: +1 21:47:10 markwash: that kinda reinforces the "non-interoperable" meme 21:47:18 lol 21:47:19 markwash: yeah, i thought about that ... but that's terrible messaging for a project that wants to promote federation and interop :) 21:47:21 lol 21:47:24 ttx: perfect 21:48:09 Name: "I...don't know what to call this", short "idunno" 21:48:24 ttx: are there any specifics on where the summit will be yet, other than 'hong kong'? 21:48:35 Can that thing be described with a thing that starts with a I : http://www.flags.net/images/smallflags/CHIN0100.GIF 21:48:47 dolphm: still under contract negociation 21:49:08 anything else before we end the meeting ? 21:49:11 oh, since this is an open session 21:49:27 I'm working up some improvements on documenting Oslo processes etc. 21:49:38 Isquare, Hong Kong (a 31 story mall) 21:49:41 one idea we talked about at the summit was incubating APIs having explicit maintainers 21:49:45 we should just skip to K and call it Kowloon (http://en.wikipedia.org/wiki/Kowloon_Walled_City) 21:49:48 * markmc just proposed this: https://review.openstack.org/#/c/27379/1/MAINTAINERS 21:49:51 feedback welcome 21:49:56 ttx, you're in there :) 21:49:57 there's also an Ikea 21:50:06 lol 21:50:29 will they offer corporate sponsorship of the next release? 21:50:52 that could fit the openstack theme of "some assembly required" :-) 21:51:05 markwash: I like it, but that would confuse the hell out of the watchers 21:51:15 (Kowloon) 21:51:52 nod 21:51:54 markmc: looks good 21:52:16 ttx, cool, thanks 21:53:09 ttx: we can compensate for the Kowloon confusion by naming the subsequent release starting with the letter I 21:53:44 then we won't have skipped anything and the OCD in all of us will be satisfied (?) 21:54:07 dolphm: or find some alphabet that skips "I" 21:54:22 ttx: the OpenStack(tm) Alphabet 21:54:32 Isquare sounds like some Apple store 21:54:47 ttx: i guarantee there's a knockoff apple store there 21:54:56 markmc: are we skipping j? 21:55:17 letters with dots over them are clearly unacceptable 21:55:20 i win? http://en.wikipedia.org/wiki/Ice_House_Street 21:56:02 oh, and we like naming things after beer, right? 21:56:09 IPA? 21:56:15 dolphm: not too bad. Still two words where single words are preferred, but probably the best one suggested so far 21:56:21 OpenStack Ice 21:56:22 done 21:56:36 can pretend it's one word 'icehouse' 21:56:57 russellb: +1 21:57:06 we need ice 21:57:07 i like it more than the fake lake 21:57:14 OpenStack I Ching? 21:57:22 Ice may have a trademark conflict ... 21:57:24 russellb: yeah, as I say, probably the less intrusive abuse 21:57:48 http://www.zeroc.com/ice.html 21:57:49 russellb: it's safer than Ikea 21:58:04 maybe not 21:58:11 anyway, time running out, let's close this 21:58:28 #endmeeting