21:02:26 #startmeeting crossproject 21:02:27 Meeting started Tue Mar 31 21:02:26 2015 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:28 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:31 The meeting name has been set to 'crossproject' 21:02:32 irc://irc.freenode.net:6667/#link https://wiki.openstack.org/wiki/Meetings/Solum#Agenda_for_2015-03-31_2100_UTC Our Agneda 21:02:37 adrian_otto: don't steal my meeting 21:02:41 lol 21:02:43 Meeting theft! 21:02:47 Our agenda for today: 21:02:52 #link http://wiki.openstack.org/Meetings/CrossProjectMeeting 21:03:05 uh, wait? 21:03:16 adrian_otto: wait what 21:03:16 adrian_otto: wrong channel? 21:03:16 sorry guys 21:03:22 adrian_otto, hi adrian :) 21:03:28 #topic PTL election season 21:03:28 ttx, o/ 21:03:39 Adri2000: you can stay here 21:03:46 adrian_otto: I mean you 21:03:53 Our meeting is more interesting 21:04:00 The PTL election season is coming up: 21:04:07 #link https://wiki.openstack.org/wiki/PTL_Elections_April_2015 21:04:11 that time of year again! 21:04:18 If you don't plan to stand for reelection, it's more than time to encourage people in your teams to step up 21:04:36 Self-nomination period starts end of week 21:04:42 Feel free to reach out to the election officials if you have any questions 21:04:48 as before, we'll also want the governance repo tagged so that our tools aren't hitting a master branch moving target 21:05:06 6am UTC is a weird time to start 21:05:06 fungi: sure I'll be around 21:05:08 Just sayin' 21:05:22 damn round planet 21:05:47 Why not just midnight? 21:05:49 mikal: it's more that i can generate the electorate rolls first thing when i wake up, and 0600 utc is well before i roll out of bed 21:05:49 Not that it matters 21:05:51 mikal: must carry some US-centricity 21:05:53 mikal, i vote 0623UTC ... :P 21:06:05 mikal, my vote doesn't matter though 21:06:11 and for whatever reason previous elections did 0600 utc 21:06:15 ok, no question on that topic ? 21:06:18 fungi: ahhh, fair enough 21:06:41 nothing starts at 6 utc 21:06:49 midnight in US mountain time areas that don't do DST? 21:06:56 it was the utc equivalent of pacific time 21:07:07 0600 utc is merely the cut-off for contributions counting you into the electorate 21:07:09 for midnight on the thursday 21:07:20 and for the self-nomination period 21:07:33 not that anyone cares anyway 21:07:36 #topic Design Summit content 21:07:47 OK, time to talk about the Design Summit in Vancouver 21:08:01 In theory the content will be set by the newly-elected PTLs, but nothing prevents to start the discussion early 21:08:04 anteaya: PDT is UTC-7:00 surely? 21:08:16 eglynn: we have moved topics 21:08:21 ttx: Ack, starting early is well advised! 21:08:21 As far as slot allocation is concerned, we don't have that many new projects after all, so we should be able to give everyone what they requested 21:08:40 Final slot allocation and proposed slot layout should be available around April 10, once we know the final list of projects to be considered 21:08:49 anteaya: ... yet the timezones stay the same 21:09:06 eglynn: I don't think anybody cares what hour the nomination week starts 21:09:30 Questions on slot allocation ? 21:09:33 o/ 21:09:36 do we have some idea if everybody wants fishbowl or conference rooms? 21:09:46 bknudson: let me see 21:09:47 or are we getting a mix 21:10:06 everyone getting what their PTL asked for when I polled them 21:10:17 generally a mix 21:10:21 ttx: How about cross project ones? 21:10:24 great 21:10:33 ttx, good news! 21:10:34 cross-project is all fishbowl 21:10:54 All remaining slots ? or do we have min, max on that? 21:11:04 Currently 133 fishbowls requested, 173 workrooms and 29 half-days 21:11:10 Thanks 21:11:35 ttx: what's the story on for scheduling the Friday afternoon? 21:11:37 nikhil_k: once the allocation is confirmed we may have scraps to share for the needy 21:11:50 That would be great 21:11:55 Friday afternoon is contributors meetups 21:12:03 like Friday nmorning 21:12:33 is there a feedback session on fri aft? 21:12:36 ttx: any projects who requested half day contrib meetup getting allocated the afternoon slot? 21:12:42 I'd like some info on how the economy is going. 21:12:43 bknudson: yes, last slot 21:12:53 bknudson: we invited the same speaker 21:12:58 wonderful! 21:13:00 ttx, was going to ask about that :P 21:13:01 bknudson: the economy is not getting better 21:13:01 I ask because Friday afternoon == graveyard 21:13:04 can we talk economy in it? 21:13:16 ttx, you typed it out before i could ask. 21:13:34 eglynn: most projects requested full day 21:13:49 a-ha, k 21:13:50 bknudson: the Canadian economy? straight down 21:14:04 eglynn: we might be able to schedule all the single half-day ones in the mornings 21:14:19 what economy are you talking about? 21:14:27 eglynn: I share your concern about friday afternoon being a graveyard :) 21:14:27 ttx: excellent, that would be wise I think 21:14:28 OK, next subtopic... session proposals 21:14:37 notmyname: you missed a historic moment in Paris 21:14:47 As far as session proposals go, we let everyone use their own tooling (mostly etherpads) 21:14:49 oh yeah. I heard about it 21:14:57 It would be great to reference those in: 21:15:00 #link https://wiki.openstack.org/wiki/Design_Summit/Planning 21:15:01 I wouldn't say I missed it, Bon 21:15:03 *bob 21:15:11 so if you already have a design summit suggestion etherpad, please add it there 21:15:20 so that people can find them 21:15:25 One team asked that we make an ODSREG instance (the system we used for session proposals pre-kilo) available to them for session suggestion collection 21:15:35 ODSREG code lives at https://git.openstack.org/cgit/openstack-infra/odsreg/ but I guess for those I'll set up team-dedicated instances of the tool 21:15:45 It's not totally impossible we'll use ODSREG for suggesting cross-project sessions as well 21:15:55 Most other projects seemed happy with etherpads 21:16:19 ++ to etherpads 21:16:31 +1 21:16:37 you can also use Google forms, whatever works best for you 21:16:38 googledocs sheets for the win 21:16:47 ... much easier to tally votes 21:17:13 eglynn: you do votes ? What kind of a dictatorship do you live in ? 21:17:31 hah 21:17:42 ttx: we vote, but North Korea stylee ;) 21:17:55 ok.. questions on session proposals ? 21:18:01 lol 21:18:16 * morganfainberg already setup an etherpad for keystone 21:18:20 ttx: more info on the dates to come later? 21:18:22 * morganfainberg is happy w/ etherpads. 21:18:24 morganfainberg: please add link 21:18:27 already done 21:18:32 nikhil_k: dates ? 21:18:47 * morganfainberg also made sure to be the weird project that didn't follow the naming convention for the etherpad :P 21:18:48 Until which people can do proposals 21:18:52 [not on purpose] 21:19:02 ttx: people like to have meetings before slots can be allocated 21:19:12 nikhil_k: you can communicate your own. there will be a deadline to fill the schedule, but otherwise you're on your own 21:19:39 which brings us to our next subtopic 21:19:45 and that deadline to fill the schedule should be close to the main event like the previous time right? 21:19:51 yes 21:19:53 As far as scheduling goes, we won't use ODSREG this time, mostly because the sched.org constraints changed (we share the sched with the main conference now) 21:19:54 thanks 21:20:05 So I created a new tool that lets you directly edit a few things on sched in the slots that are allocated to your teams, I'll send you link to a demo soon 21:20:15 Code currently lives at https://github.com/ttx/summitsched for the curious 21:20:20 ttx: Cool! 21:20:29 it's basically a sched proxy 21:20:48 since shced has a pretty simple idea of ACLs 21:21:00 (stores it in a binary) 21:21:15 We still have a decision to make though. Cheddar enforces a number of things, like making sure all sessions on a track are prefixed by the track name, for clarity 21:21:22 * morganfainberg wonders how many ways we can break sched this time :P 21:21:32 It also enforces boring titles and descriptions for "work sessions" and the "contributors meetups", in order to make them as unattractive as possible for the general attendance 21:21:41 You can see the results at: http://testingsiteforfutureevents2015.sched.org/ 21:22:10 nice! 21:22:15 ttx, i want to go to the "Ops: cake is amazing" session 21:22:20 with "food not provided in the title" ;) 21:22:23 So basically you can give fishbowl sessions a title, but work sessions and contributors meetups have set titles 21:22:43 since they have such limited room you don't want everyone to show up because tyou mentioned "docker" in title 21:22:49 One question I still had is whether to allow custom descriptions in work sessions. Currently you only can customize a link, which looks like this: 21:22:55 http://sched.co/2wp8 21:23:02 That may be a bit impractical for design summit attendees (having to click the link to see what the work session is actually about) 21:23:10 It's a balance between making it unattractive to conference attendees and making it a pain to design summit attendees 21:23:16 So maybe boring title is enough and we should allow custom descriptions for work sessions ? What would be your take ? 21:23:34 I'd allow a link to an etherpad, add a boring description "see etherpad for me" 21:23:37 "more"* 21:23:39 maybe a very short description? 21:23:47 as a reminder 21:23:51 morganfainberg: that is what we have for work sessions right now 21:24:04 ++ for very short description asalkeld 21:24:11 FWIW it seems like it should be possible to work around some of this by creating an etherpad for “here are our project’s work sessions” and putting descriptions there 21:24:18 generic description looks like http://sched.co/2wp8 21:24:32 possible, but slightly annoying 21:24:34 ttx, i'm a fan of that. it means that it's easy to change the work session topic if suddenly everyone is busy elsewhere for that topic 21:24:37 the work group titles are hopefully going to have the project name ? 21:24:55 that basically means you need to train your teams to click on the etherpads 21:24:56 ttx, i'd rather just keep the boring description and the link. 21:25:11 I think it's pretty nice - at the end working session are for contributors and they know what they plan to discuss 21:25:14 but i'm not going to vote strongly against a short description 21:25:36 so let's do a quick poll 21:26:04 (was just thinking if you are on a phone, short description is nice) 21:26:10 ttx: Error: Can't start another meeting, one is in progress. Use #endmeeting first. 21:26:16 #startvote should work sessions have custom descriptions, or just a boring description and a link ? custom, boring, don't care 21:26:17 Begin voting on: should work sessions have custom descriptions, or just a boring description and a link ? Valid vote options are custom, boring, don, t, care. 21:26:18 Vote using '#vote OPTION'. Only your last vote counts. 21:26:23 #vote boring 21:26:30 #vote custom 21:26:33 #vote boring 21:26:33 #vote custom 21:26:34 #vote boring 21:26:35 #vote boring 21:26:49 #vote custom 21:26:55 actually... if we can just have it default to boring, and allow an override... 21:27:01 #vote custom 21:27:02 i think that solves the use for everyone 21:27:10 morganfainberg: ++ 21:27:12 morganfainberg, +1 21:27:12 hmm, yes. 21:27:17 #endvote 21:27:18 Voted on "should work sessions have custom descriptions, or just a boring description and a link ?" Results are 21:27:19 boring (4): david-lyle_, edleafe, mestery, morganfainberg 21:27:20 custom (4): nikhil_k, eglynn, notmyname, asalkeld 21:27:34 Let's make it custom, but encourage boring 21:27:34 lol 21:27:42 ;) 21:27:43 you can come with your own flavor of boring 21:27:48 i like it! 21:27:48 haha 21:27:49 I wouldn't like to see "boring" used as a means of realtime rescheduling of sessions 21:27:49 A tie ... but it doesn't matter anywya because morganfainberg had a better solution :) 21:28:15 the vote was a rhetorical trick 21:28:16 (hard enough to manage the clashes between multiple tracks in parallel) 21:28:24 "this is a boring work session, don't come here... we're going to be quiet and not doing performance art... we swear... and definitely not docker" 21:28:32 LOL :) 21:28:37 haha 21:28:39 No! You said Docker! ;) 21:28:57 open the floodgates! 21:28:59 put a sign that says docker on a room. 21:29:07 LOL 21:29:08 "10 things you won't believe about Docker" 21:29:16 we just need that nfv docker honeypot session 21:29:19 ttx, this one amazing trick about docker... 21:29:29 "Docker and OpenStack: Better Together!" 21:29:30 * morganfainberg apologizes for derailing 21:29:36 lol 21:29:40 sdague: lol 21:29:53 sdague: in another building 21:29:57 Oh, and Cheddar (the sched proxy) also supports a session showing up in multiple tracks! 21:29:59 I bet those folks are as easy to trap as ... LEMMINGS 21:30:00 * dansmith runs 21:30:13 So it has a main track, but you may also make it appear in other tracks 21:30:19 dansmith: lol 21:30:23 oh thats cool. 21:30:30 example at: 21:30:33 lol 21:30:58 http://sched.co/2vhQ 21:31:07 That's an Ops session, but it also appears in swift 21:31:26 nice. swift + cake == awesome 21:31:26 so that when your contributor go to: http://testingsiteforfutureevents2015.sched.org/type/design+summit/Swift they still see it 21:31:41 That should facilitate multi-project work 21:31:45 * devananda sneaks in, tries to catch up on scrollback 21:32:15 devananda: well that's not sneaky at all 21:32:17 devananda: Summary is: NFV Docker honeypot! 21:32:20 If someone wants to beta-test Cheddar, I can set up access to the demo instance 21:32:33 just send me some email so that I remember to do it 21:33:09 Alright, any question on design summit ? 21:33:54 all good, thanks ttx 21:33:59 #topic Open discussion & announcements 21:34:06 We had 1:1s syncs today in #openstack-relmgr-office, logs at: 21:34:07 ttx: is there a list of the per-project slot allocations anywhere? 21:34:15 #link http://eavesdrop.openstack.org/meetings/ptl_sync/2015/ptl_sync.2015-03-31-08.04.html 21:34:21 ttx: i would have voted boring :) 21:34:23 ttx: for open discussion, question on string freeze and string freeze exceptions 21:34:41 eglynn: not yet, but the google doc I colected answers on is still up 21:34:48 when we need to change a string for bugs at this point, what are we supposed to do to alert i18n teams 21:34:53 ttx: a-ha, cool, ta 21:35:05 ttx: What seem to be the community plan for the smaller release cycle in L? 21:35:21 sdague: so that is a good question 21:35:23 to be community's general plan* 21:35:41 sdague: we generally do a thread on -dev, but I have no idea if that reaches the right people 21:35:55 sdague. we should do a i18nImpact tag? 21:35:58 ttx: i ma have missed it in my skim - is the ops track taking suggestions and/or having project-specific discussions? 21:36:09 yeh, this seems like a very non closed loop, as I never have any idea what the i18n progress is 21:36:10 Actually, I have no evidence that string freeze exceptions get seen by translators 21:36:12 ttx, i use the i18n mailing list. 21:36:24 ttx, got a response from andreas about it. 21:36:28 I know they appreciate us not gratuitously changing strings after FF, but that's about all the feedback I got 21:36:28 morganfainberg: there's a separate mailing list? 21:36:29 ttx, so that seems to work. 21:36:31 sdague, yes. 21:36:39 * mestery didn't know of the separate ML either 21:36:42 devananda: there's an etherpad for ops 21:36:52 * morganfainberg has had to do this a few times in the past. 21:36:55 Rockyg: link isn't on https://wiki.openstack.org/wiki/Design_Summit/Planning yet :) 21:37:05 * devananda adds ironic to ^ 21:37:14 openstack-i18n@lists.openstack.org 21:37:28 ttx: so it seems like it would be good to get i18n leadership to be banging appropriate drums on -dev list around this 21:37:34 * Rockyg is looking 21:37:38 Here's the etherpad for the Ops topics: https://etherpad.openstack.org/p/YVR-ops-meetup 21:37:40 sdague: yes 21:37:42 because otherwise we're just going to fail a lot 21:38:25 ops summit ether pad: #link https://etherpad.openstack.org/p/YVR-ops-meetup 21:38:27 we tend to be good w/ closes-bug, and docimpact (not as good as closes bug), so maybe a tag in the commit message is the right approach... maybe it's just a bad idea as well. 21:38:31 morganfainberg: you seem to know your way around there, care to introduce that topic ? 21:39:04 we just ened to make sure the i18n leadership shows up 21:39:12 ttx, uhm.. what needs to be introduced? or you mean for a session? 21:39:24 morganfainberg: we need to know what to do now 21:39:28 no I mean ttx: so it seems like it would be good to get i18n leadership to be banging appropriate drums on -dev list around this 21:39:30 on -dev 21:39:50 morganfainberg: you seem to know who to reach to 21:40:00 sdague, i mean i'll bug andraes and send a message to the i18n list :) 21:40:05 i18n is a black box to me. To reach it I usually dial fifieldt 21:40:06 you even knew there was a list 21:40:18 morganfainberg: so I think the point is this needs to be on -dev 21:40:25 because few people know that list is a thing 21:40:38 so clearly won't be getting info out to the review teams that need it 21:40:41 sdague, right, the point is getting them in on the convo. 21:40:46 and the current process involved [stringfreeze] on dev list 21:40:55 ttx, oh i didn't know that :P 21:40:59 ttx: ok, news to me 21:41:03 * morganfainberg did something totally different. 21:41:22 well, that is what I replied to people who asked me 21:41:28 so i'll try and chase someone (some i18n folks) down and get them to make some noise on -dev about. 21:41:31 not sure that qualifies as "process" 21:41:39 not guarantees that i'll be successful. 21:41:42 but i'll try 21:41:50 when we get to these parts of the release I think the [all] emails need to go out with "x is frozen, to get an except do Y" 21:42:05 the only way i knew someone read i18n list is because someone responded this time (i think this is the 4th or 5th string freeze i've sent out) 21:42:11 because there has been a ton of confusion with code review where everyone's trying to guess the rules 21:42:38 sdague, ++ i'll chase some people down and get them to send something to -dev. 21:42:40 sdague: but but but difference in culture is good right ? 21:42:44 which means they are so inconsistently applied, it's not clear they are useful at all 21:42:54 * sdague walks away slowly 21:42:55 sdague, at least we have some direction then. 21:43:35 morganfainberg: yep, thanks 21:43:40 morganfainberg: thx! 21:43:51 nikhil_k: you had a question ? 21:44:22 ttx: What seems to be community's general plan for the smaller release cycle in L? 21:44:41 nikhil_k: I don't think we'll do a smaller release cycle in L 21:44:50 but we'll dsicuss release cycle lengths in Vancouver 21:45:06 so not 6 weeks then? 21:45:18 There was some proposal for abandoning older PS 21:45:26 * fungi votes for a 6-minute release cycle 21:45:28 well, the cycle will still be 6 months. You may release more often though 21:45:35 we'll run out of letters. 21:45:43 nikhil_k: don't believe everything you read on MLs 21:45:52 I thought the rhythm matches well with our 6 weeks style 21:45:59 * morganfainberg believes everything on the internet 21:46:05 i really didn't get the impression that the "release more often but less thoroughly" discussion really went anywhere except in a circle back to describing why we have milestones that aren't releases 21:46:06 we could use Liberty, Lemming and Lizard 21:46:06 ttx: I kinda like that idea :) 21:46:17 (of abandoning older PS) 21:46:23 david-lyle, and love. don't forget love 21:46:30 >.> 21:46:35 All you need is Love 21:46:46 All I need is a nap 21:47:23 nikhil_k: I'll invite you to that session :) 21:47:33 OK... Anything else, anyone ? 21:47:59 ttx, does that mean we move to semver? 21:48:05 ttx, releasing more often that is? 21:48:14 or is that just a rabbit hole not worth talking about right now 21:48:16 mikal: http://www.tumblr.com/tagged/i-need-a-nap 21:48:17 where are cross-project session topics being collected? 21:48:20 morganfainberg: rabbit hole 21:48:32 david-lyle: nowhere yet 21:48:33 ttx, sold. happy to merrily go on my way 21:48:37 ttx: heh 21:48:45 david-lyle: the etherpad ended up a bit messy last time 21:48:52 russellb: ^ 21:49:05 we need to get started on that soon 21:49:11 I was wondering if we should not use something less messy, like ODSREG of some google form 21:49:13 one problem was that we started kinda late, IIRC 21:49:19 ttx: that too, yes 21:49:36 submission by completely freeform text is kind of difficult 21:50:00 works well for teams brainstorming, not so much for random individuals posting 21:50:02 how many cross-project tracts concurrently running? 21:50:06 yeah 21:50:11 let me see 21:50:29 is it a cross project day again? 21:50:43 2 sessions in parallel 21:50:47 yes, Tuesday 21:50:50 ok 21:50:52 ok 21:50:53 ODSREG is fine with me 21:51:04 Ops track runs in parallel though 21:51:18 so we'll have to choose where to go 21:51:28 usually the case .. 21:51:46 right 21:51:49 let's just add more days! 21:51:53 * russellb headdesks 21:52:05 i think i might collapse 21:52:23 ok so, cross-project track. we should do somethign about that. 21:52:27 OK, I think we are done for today, unless someone has a last-minute question/topic 21:52:29 ttx: ODSREG++ 21:52:46 russellb: ok, will set up an instance for cross-project 21:52:50 awesome 21:52:54 notmyname asked for one for swift too 21:53:06 notmyname: still interested ? 21:53:11 yes, i think so 21:53:19 I prefer it to etherpads 21:53:33 notmyname: some people use Google forms 21:53:48 notmyname: maybe discuss with team and let me know what you need 21:53:52 ok 21:54:01 would be nice if it was consistent 21:54:02 oh well 21:54:37 ok, let's close this 21:54:41 #endmeeting