20:01:34 #startmeeting tc 20:01:35 Meeting started Tue Aug 9 20:01:34 2016 UTC and is due to finish in 60 minutes. The chair is flaper87. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:01:35 Hello everyone! 20:01:36 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:01:37 As usual, you can find today's agenda here: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda 20:01:40 Let's start with a few "easy" ones 20:01:40 The meeting name has been set to 'tc' 20:01:41 #topic Clarify that cycle-trailing projects follow the milestone deadlines 20:01:43 #link https://review.openstack.org/348501 20:01:44 o/ 20:01:47 ~o/ 20:01:49 AJaeger: I got this 20:01:59 #topic Refresh I18n ATC list 20:02:00 #link https://review.openstack.org/351480 20:02:06 flaper87: you can remove me from the ping list 20:02:15 notmorgan: yup 20:02:16 flaper87: as of today, I'm no longer part of the TC. 20:02:25 notmorgan: thanks again for your service :) 20:02:32 notmorgan: thanks for your service 20:02:38 * flaper87 will update the governance repo 20:02:53 ok, this one has majority already 20:02:55 notmorgan: sorry to see you go, thanks for the support and service 20:02:58 any reason why this shouldn't be merged ? 20:03:08 flaper87: looks good to me 20:03:30 we could try wait for consensus instead, just for giggles, but otherwise, I am happy 20:03:52 now, onto the one I skipped after my failed w/ my copy/paste to start the meeting 20:03:52 +1 20:03:53 johnthetubaguy: 9 is pretty darn close :) 20:03:58 #topic Clarify that cycle-trailing projects follow the milestone deadlines 20:03:59 #link https://review.openstack.org/348501 20:04:15 This one has 8 votes already 20:04:29 that's majority already. 20:04:35 Any reason why this shouldn't be merged? 20:04:42 does anyone have any other questions about that one? 20:04:53 can't think of one myself 20:05:11 * flaper87 is good with it 20:05:45 ok, sold... I guess 20:05:59 Now, onto more heated topics 20:06:11 #topic Replacing Morgan's seat 20:06:13 #link http://lists.openstack.org/pipermail/openstack-tc/2016-August/001233.html, including various proposals 20:06:15 #link https://review.openstack.org/351104 (flaper87) 20:06:17 #link https://review.openstack.org/351295 (dhellmann) 20:06:19 #link https://review.openstack.org/351427 (anteaya) 20:06:23 There are several proposals. The first one proposes using the next closest election to elect all the missing seats. If 1 or more vacants would open before the election, those would be added to the number of seats up for election. 20:06:25 The second proposal is to adopt the same model as the foundation and it proposes three different scenarios: 1) Vacancy openned less than 4 weeks before the election and the seat will be contested, then nothing is done till the election. 2) Vacancy openned less than 4 weeks before the election and the seat will not be contested, then candidates that don't win a seat would be asked if they want to 20:06:26 serve 3) Vacancy opens more than 4 weeks before the election, the TC consults the last results. 20:06:28 The third proposal is to have by-elections as soon as the vacancies open, regardless what time in the cycle we're at. 20:06:30 #link https://en.wikipedia.org/wiki/By-election 20:06:32 I'm fine with either the first or the third option, to be honest. The 2 first points of the second option are fine. The third one is the one that doesn't sit well with me. The main reason is that if it happens too late in the cycle, then I believe we'd be pulling in someone in the TC that could loose to others anyway. 20:06:34 The way the community and its members evolve in the cycle is really hard to predict, which is why I'd prefer to leave it up for elections either to the closest election or a by-election. 20:06:36 * flaper87 hopes that summarizes the three proposals 20:06:53 (fairly summarizes would be better) 20:07:09 flaper87: the third proposal is not by-election as soon as the vacany is open 20:07:23 anteaya: mind clarifying? 20:07:25 anteaya: sorry about that 20:07:41 the third proposal is by-election, timing and details to be determined based on circumstances 20:08:02 anteaya: thanks! Sorry for not summarizing it well :) 20:08:03 it's reasonable to expect that we would want to do it relatively quickly, but it's not required 20:08:12 dhellmann: right 20:08:13 anteaya: but always holding a by-election? 20:08:18 mugsie: yes 20:08:24 OK. 20:08:26 by-election could be held with other election 20:08:47 or if vacancy occurs at christmas the electon could be held when it makes sense to do so 20:08:51 anteaya: other election such as a PTL election? Or along with the already-scheduled TC election? 20:08:59 annegentle_: either 20:09:04 whatever makes sense 20:09:09 anteaya: check, thanks for clarifying. 20:09:17 annegentle_: thanks for asking 20:09:28 there are times when I wish we could CIVS a set of competing specs 20:09:34 my preference is dhellmann's proposal, which aligns with how the board does it more or less. It feels to me like 5 month old results are still pretty fresh in the community given our current turn over rate. 20:09:57 i agree with sdague and dhellmann 20:09:59 so I am currently preferring option 2, because it saves lots of people a lot of time, and still seems to respect the will of the voters 20:10:02 sdague: 5 months old results is exactly what doesn't sit well with me 20:10:13 flaper87: no? 20:10:20 the board doesn't actually specify "consult last results" 20:10:26 Things change, people come/go, opinions evolve, etc 20:10:30 it's just up to the individual directors to appoint, and in practice, that's how they've done it so far 20:10:31 russellb: that's why I said "more or less" 20:10:33 but it's not coded that way 20:10:39 I'm ok with the 5 month old results. Not ok with an election to happen and assume voters will be around. 20:10:40 the flexibility is nice 20:10:47 its a pretty short period of time. 20:10:50 I feel / sense that the 4 week window one is best for a bunch of groups; the electorate, the election officials not having to do a by-election on-demand, and the TC members themselves feeling like there's a process for when you just can't finish your term. 20:10:57 thingee: that is why the timing is not set in the patch 20:11:00 russellb: to be fair, I'd also be fine with more flexibility in there as well 20:11:01 flaper87: we elect people for a full year; I doubt that someone who was voted for last election isn't desireable 5 months hence 20:11:03 * annegentle_ still needs to vote 20:11:09 russellb : I would also be ok with a proposal that said it that way, but some other folks seemed to want the details written down 20:11:13 thingee: fwiw, my proposal says we could elect the open sits on the next closest election. Not saying it's better, though 20:11:15 thingee: so the timing can be selected to ensure voters are around 20:11:20 thingee: good point, we don't have the best turn out anyways, a by-election is likely much worse 20:11:39 edleafe: well, if that person won already, there's not much to do. If that person didn't win, then there is 20:11:41 johnthetubaguy / thingee ++ 20:11:45 anteaya: that's hard to figure though 20:11:48 oh, and I still am confused on one point... just a sec. 20:12:01 anteaya: we already have problems with usual scheduled elections 20:12:06 thingee: we can look at email and gerrit activity 20:12:11 flaper87: but "won" is a relative term in CIVS 20:12:15 thingee: yes we do, no argument there 20:12:24 so do we need the wording from https://review.openstack.org/351104 so that the four-week proposal works (https://review.openstack.org/351295)? They are independent, right? 20:12:25 edleafe: yeah yeah, made it into the N seats up for election 20:12:27 and picking voting times then becomes another extra effort 20:12:28 :P 20:12:38 but I think we are futher ahead to have more elections, when warrented, not fewer 20:12:41 sdague: right 20:12:54 anteaya : flaper87's proposal and mine are mutually exclusive. We don't want both of them. 20:12:57 flaper87, dhellmann I think the compromise between the three proposals is my Aug 8, 6:43am comment. with threshold = 4 weeks. 20:13:01 dhellmann: ok 20:13:05 going down a list of previous voted people is less effort imo 20:13:13 oops, meant that for annegentle_ , tab-complete-fail 20:13:15 * thingee is lazy 20:13:24 thingee: yes it is less effort 20:13:27 dhellmann: happens all the time right anteaya ? :) 20:13:35 no question about how much effort it is 20:13:40 I'm talking about value 20:13:46 and doing right by the electorate 20:13:58 given the historical drift of candidates on the TC list (who ranked where cycle after cycle), I think that a special election is largely going to give us the same results as just pulling from the last election, except with a ton of extra work 20:14:00 annegentle_: all the time 20:14:08 vacancies are filled with either appointment or re-vote. Picking from previous results is just one way to handle an appointed replacement 20:14:17 Looking at the list now of the next people in-line, are these people still active and participating? 20:14:27 amrith : I don't think civs results in ties, does it? 20:14:31 would anyone have a problem with them? 20:14:36 dhellmann: it can 20:14:41 * flaper87 doesn't want special elections but a simple election on the closest one 20:14:44 dhellmann, let me check. I think I looked it up and found that it could. 20:14:45 dhellmann: but we have rules for tie breaking 20:14:56 dhellmann: we just have never had to invoke them 20:15:00 k 20:15:20 thingee: that's the vetting process. If they aren't active, or would upset affiliation, move on to the next in line 20:15:24 thingee: yeah I wondered if we had polled the people next on the list yet... 20:15:32 sdague : I think you're probably right. 20:15:36 thingee: they mostly all seem active to me, and I wouldn't have a problem going down that list in order 20:15:36 thingee: or if they are simply no longer interested 20:15:46 edleafe: +1 20:15:47 #link https://wiki.openstack.org/wiki/Governance/TieBreaking 20:15:48 anyone have a link handy to the last TC election results? 20:15:54 yeh, one sec 20:15:58 http://civs.cs.cornell.edu/cgi-bin/results.pl?id=E_fef5cc22eb3dc27a 20:16:07 dtroyer would be the fix ask 20:16:15 just to be clear, did we rule out the not filling the seat until its next up for election approach? 20:16:30 don't think anyone would be upset with dtroyer :) 20:16:34 johnthetubaguy : we did not explicitly decide that. 20:16:36 then david-lyle, sdake, anteaya, cdent (in that order) 20:16:37 dhellmann, anteaya found the link I was loking for; thanks anteaya 20:16:39 thingee: truth 20:16:39 johnthetubaguy: that would leave the seat vacant until next north american spring 20:16:46 anteaya: yes 20:16:52 johnthetubaguy: my seat would be up in spring. 20:16:54 amrith: welcome 20:17:03 long time w/ an unfilled seat. 20:17:06 johnthetubaguy: I didn't see that proposal up for discussion, no 20:17:07 All 4 algorithms end up with same result dtroyer 20:17:26 dims: this time 20:17:28 johnthetubaguy : you raise a good point, and I think we should at least agree that we want to fill the seat if there's some minimum amount of time left. 20:17:37 johnthetubaguy : my proposed minimum is 1 month 20:17:37 anteaya : y 20:17:57 dhellmann: i think your review specified 4wks, i would go with that over 1 month 20:18:05 dhellmann: since months vary in length 20:18:05 notmorgan : tomato, tomato 20:18:09 and the rules we are creating are for filling the seat in an uncontesed way even if someone has an issue with the candidate 20:18:10 dhellmann: the only reason I think we must do that is because we don't have a consensus model for reaching agreement 20:18:11 dhellmann: precision :P 20:18:16 that being filling up the seat 20:18:21 dhellmann: yeah, I think that was a good minimum 20:18:29 otherwise we could live with an empty seat until the next election 20:18:36 yeh, 4 weeks seems like a very reasonable time frame 20:18:40 either works for me 4 weeks or 1 month 20:18:43 dhellmann: yeah, that seems a good minimum, 4 weeks 20:18:50 #link http://civs.cs.cornell.edu/cgi-bin/results.pl?id=E_fef5cc22eb3dc27a 20:18:56 flaper87 : that's a separate discussion, but sure 20:19:06 flaper87: to be fair, I could have also just disappeared and it would be the net effect of an empty seat until next election it was up 20:19:18 notmorgan : we would have noticed 20:19:24 notmorgan: sure, but we would have noticed 20:19:25 dhellmann: what was an exercise in findability for me was "when do we hold TC elections" 20:19:27 are we at least agreed on the 4 week question? 20:19:28 notmorgan: well then we would have had the disappearing leader issue again 20:19:44 +1 on 4 weeks dhellmann 20:19:48 dhellmann: right, but there was no handling of that here, and has the same result. 20:19:49 answer: 5 weeks before the Summit 20:19:50 annegentle_: it is in the charter 20:19:54 http://releases.openstack.org/newton/schedule.html 20:20:00 dhellmann: yeh, seems like 20:20:14 dhellmann: I'm not happy with it, tbh. 5 months old results don't sit well with be but most people seem to be happy with 4 weeks 20:20:18 so, I'll live by it 20:20:19 anyway. sorry. I'm going back afk and getting food. /me ducks out of this convo 20:20:27 flaper87: yeah, one question at a time: so you're ok with 4 weeks? 20:21:03 dhellmann: no, I meant to say I'm not ok with 4 weeks. Perhaps we could make the threshold bigger: 8 weeks ? 20:21:23 flaper87 : so you're saying if I resign 8 weeks before the end of my term, you would not replace me until the next election? 20:21:31 dhellmann: yes 20:21:39 that's not the direction I expected you to go 20:21:40 annegentle_: The election is held 3 weeks prior 20:21:42 that seems long for me 20:21:42 to each design summit, with nominations due 4 weeks prior to the summit and 20:21:44 too long imho 20:21:44 elections held open for no less than five business days. 20:21:48 why would you be willing to wait so long? 20:21:53 anteaya: oh I'd better find the other source I found. 20:21:56 annegentle_: #link http://git.openstack.org/cgit/openstack/governance/tree/reference/charter.rst 20:22:02 dhellmann: 8 weeks before the next election, or the end of your term? 20:22:11 annegentle_: under Election for TC seats 20:22:19 edleafe : either, really 20:22:22 dhellmann: the reason is that I don't feel ok with using a 5months old results to pick someone that will fill the remaining 7 months of the person leaving 20:22:25 annegentle_: the only doc that trumps the charter is the foundation by-laws 20:22:38 edleafe : the question is how long do we want to allow an empty seat 20:23:00 so, I guess the question we should figure out, is whether anyone objects besides flaper87 ? Because we're +6 on dhellmann's proposal right now 20:23:16 dhellmann: exactly. The next election keeps that vacancy short 20:24:00 edleafe : I think I'm not being clear, so I'm just going to stop offering alternatives. Let's pick one of the two that are written up. 20:24:09 sorry, 3 20:24:17 dhellmann: :) 20:24:33 I think 8 weeks is much too long 20:24:33 dhellmann: agreed, we don't need to spend from now until the next election discussing how to fill an empty seat :) 20:24:44 Ok, there are 7 votes on dhellmann review 20:24:51 I guess we should go w/ that one 20:24:54 Honestly, I don't think running a new by-election gains enough to justify the effort of all those nominations, voting, etc, etc 20:24:54 sdague: I disagree 20:25:18 flaper87: this is a sort of important one, so I'd like to get most everyone on board one way or the other if we can 20:25:28 in the past we have had a tc election where some of the seats were for a one-year term and some for a half-year term, when bootstrapping the current committee size 20:25:37 yeah charter changes aren't simple majority 20:25:38 flaper87: only if you can support it. We can work towards consensus here. 20:25:43 dhellmann: would be good to know if the others are abstain / -1 / neutral ... 20:25:54 dhellmann: I'm with your proposal. I'm not happy with the 4 weeks but I'm willing to live by it 20:25:55 fungi : yeah, that's more or less what I based my proposal on. 20:25:56 so there is precedent for a seat in an election being for a partial term 20:25:57 dhellmann: right, I'd like to have everyone on the TC register here 20:26:14 only thing the charter says is, "After the initial election, the elections for the Technical Committee shall be held at least every six months." 20:26:16 * flaper87 is not trying to rush the discussion but rather trying to reach consensus 20:26:16 I don't want to move forward on simple majority 20:26:20 #link https://www.openstack.org/legal/technical-committee-member-policy/ 20:26:21 +1 for getting everyones vote on this 20:26:23 flaper87 : ok. I'm willing to discuss changing that, maybe as a separate thing next week? 20:26:28 sdague: that is not what I said, fwiw 20:26:30 maybe quick poll of whoever hasn't voted yet? 20:26:33 dhellmann: sounds perfect 20:26:34 yeah I'm also fine with waiting 20:26:39 flaper87 : I picked what seemed like a pragmatic number but I'm not wedded to it 20:26:52 Amendments to this Technical Committee charter shall be proposed in a special 20:26:54 motion, which needs to be approved by the affirmative vote of at least 20:26:56 two-thirds of the total number of TC members (rounded up: in a 13-member 20:26:58 russellb: assuming they're around 20:26:58 committee that means a minimum of 9 approvers). 20:27:06 sdague fwiw, I am still available for TC seat filling if necessary -but sounds like dtroyer using the 4 algos gets the seat. 20:27:06 mtreinish: yup. 20:27:08 ttyl ) 20:27:19 #link http://git.openstack.org/cgit/openstack/governance/tree/reference/charter.rst under Ammendment 20:27:23 :) i mean to say 20:27:46 sdake ack :) 20:28:00 sdake: thanks, good to know 20:28:01 so ttx is out, dhellmann hasn't voted, who is the missing vote on there? 20:28:04 charter changes need 9 votes 20:28:23 notmorgan hasn't voted, heh 20:28:28 russellb: right :) 20:28:35 we have 12 votes 20:28:36 i am not a member of the TC, I do not get to vote 20:28:36 #link https://review.openstack.org/#/c/351295/ 20:28:39 ttx did vote +1 at one point 20:28:43 and one is dhellmann's on his own patch 20:28:43 i officially stepped down effective today 20:28:44 just not the current patch :) 20:28:47 which counts like a vote 20:28:48 :P 20:28:54 mestery: ? 20:29:00 notmorgan: i know, just joking ... were couting up the votes (or missing votes) 20:29:02 email was sent to secretary@o.o confirming this. 20:29:08 mestery is probably out, at openstack silicon valley 20:29:19 flaper87 : I went ahead and registered my vote 20:29:25 russellb: he's tweeting from there anyway :) 20:29:27 dhellmann: thanks 20:29:29 that helps 20:29:42 ok, we're missing ttx and mestery 20:29:49 if we hold it open for 2 more days, think both of them will have a chance to weigh in? 20:29:52 10 is more than 2/3 of 12, right? 20:29:55 we can wait for those and address questions/concerns next week if there are more 20:29:55 it is 20:30:03 dhellmann: you need 9 votes 20:30:03 but we should be fine 20:30:12 dhellmann: it is but this is a good chance to practice consensus 20:30:15 I just want to make sure people had the ability to voice an opinion 20:30:18 annegentle_ : sure 20:30:31 So, Does anyone have concerns about dhellmann's proposal? 20:30:34 #link https://review.openstack.org/#/c/351295/ 20:30:40 anteaya: the TC is only 12 members as of today, so technically, 8 are needed 20:30:46 Any other questions about the process? 20:30:52 Any other questions about the requirements? 20:30:52 alright, I'm fine leaving it open until next week. we should ask dtroyer to join us for the meeting. 20:31:07 dhellmann: sounds good 20:31:08 fwiw, I'm here 20:31:13 edleafe, normally in the computation of vote margins, it is the size of the body, not the number currently present. so 9 is still required; IMHO 20:31:19 dtroyer : I meant to make sure you were here next week :-) 20:31:21 dtroyer: thanks for being here 20:31:28 amrith: the body is only 12 20:31:37 dhellmann: roger 20:31:38 no, the body is 13, 12 members. 20:31:41 amrith: the wording would indicate 12 memebers 20:31:42 dtroyer: so, question, do you remain interested in serving on the TC? 20:31:50 we'll have to wait till ttx is back anyway. I don't have +W and It'd be great to have his opinion (and mestery's too) 20:31:58 amrith : 10 is also > 2/3 of 13 20:31:58 because if the answer is yes, that simplifies talking with folks down the list 20:32:00 sdague: yes, I would happily fill the remaining term if asked 20:32:12 ok, cool, so we have that data point in public 20:32:22 thanks dtroyer 20:32:33 dtroyer: danke 20:32:41 ok, I move we leave voting open until ttx and mestery express a vote, or next meeting, whichever comes first 20:32:48 and we move to next topic 20:32:49 ++ 20:32:52 sdague: just said that :P 20:33:01 ++ 20:33:12 ++ flaper87 sdague 20:33:15 dhellmann, you are correct. sorry; 9 or > 20:33:19 flaper87 : as chair, will you ping both ttx and mestery in the mean time? 20:33:28 dhellmann: yup, on it 20:33:40 moving on 20:33:42 #topic Community-wide goals 20:33:43 #link https://review.openstack.org/349068 20:33:46 #link https://review.openstack.org/349069 20:33:48 #link https://review.openstack.org/349070 20:33:51 dhellmann: I'll let you summarize/introduce this one 20:34:03 ok 20:34:18 this is the write-up of the idea that we discussed the thursday of the training in ann arbor 20:34:40 the idea is to be a little more active in encouraging the community to take some specific actions each cycle 20:35:00 I've included 2 specific proposals from the list that we discussed as examples 20:35:26 there has been a fair amount of concern about consequences for not following through on goals, but very little discussion of the specific goals proposed 20:35:42 I think both goals are good small scoped ones 20:35:59 sdague: +1 20:36:06 * flaper87 likes what's up there 20:36:08 I'll try to summarize ttx's comments on the consequences by saying that failing to follow through is not an immediate trigger for being dropped, but it may be seen as part of a larger set of signals that a project does not really consider itself part of the community 20:36:20 I also think that starting from a "hey everybody, let's all try to ..." is great. nobody has done it yet. might someone choose to not play ball? sure 20:36:23 his wording on the mailing list is probably better 20:36:25 good to test the process, mostly I've been thinking about how to use something similar to help drive the upgrades conversation 20:36:27 and then yeah, ttx said that well 20:36:28 dhellmann: are the two proposed the "typical" scope you'd expect? 20:36:34 dhellmann: or are you ready for more to be proposed by others? 20:36:57 sdague: yeah and I was thinking of something around api-ref now that I have the numbers and the nav is coming together. 20:37:03 ttx's wording http://lists.openstack.org/pipermail/openstack-dev/2016-August/100670.html 20:37:11 annegentle_ : my sense is that we only want a couple of these each cycle, so I don't think we want more right now unless we reject one of these 20:37:14 mordred: sure - but I am not sure I would be happy top have this morph into "do this or leave" conversation 20:37:22 #link http://lists.openstack.org/pipermail/openstack-dev/2016-August/100670.html 20:37:22 annegentle_: right, we'd probably have done *this* for api-ref this cycle if we had this process 20:37:24 mugsie: totally 20:37:31 we kind of ad-hoced it 20:37:48 as far as scope, we want things to be finished in a cycle, rather than being of indefinite length, and that's going to naturally limit the scope 20:37:52 but, it would be a very reasonable 3rd thing to use this process this next cycle 20:37:53 which is why I think there needs to be more concrete wording about consequences 20:37:53 dhellmann: question: I don't remember if it was mentioned/discussed whether some of this goals would eventually become requirements for entrance to the big tent. I think I read new projects have 1 year to be compliant with some of these goals, is that right? 20:37:54 large initiatives will need to be broken into phases 20:38:00 mugsie: I think I'd rather just see people get excited about a few shared esprit-de-corps types of things - without us holding an axe over anyone's head 20:38:00 dhellmann: ok, makes sense 20:38:17 For example, if all projects manage to migrate to py35, we might want to ask new projects to be py35 compliant 20:38:18 flaper87 : I believe somewhere in that patch I say that all teams need to do the goals. that would apply to new teams, too. 20:38:19 sdague: I was just going to suggest we might want to add it still 20:38:24 mordred: yeah - but the ML has been hinting at it having consequences 20:38:31 flaper87 : if we need to clarify that, we can work on language in another patch? 20:38:35 mugsie: I don't really find it productive to speculate that there would be a goal agreed that would cause problems. 20:38:45 dhellmann: Also, forgive me if I'm behind on the ML discussion: do we see these as themes for a release such as what the product group has discussed? 20:38:50 dhellmann: yup, sounds perfect. I;ve voted already 20:39:13 annegentle_ : a bit more narrow, though ttx did mention that the foundation might use them as "look what we did" type messaging at summits 20:39:15 annegentle_: I think these goals are generally more specific 20:39:17 thingee: well, teams are differning sizes. A team might decide that actully writing docs this cycle is better for users, for example 20:39:17 honestly, I really rather just think of this as information gathering and exposure 20:39:24 mugsie: I tihnk thingee just said what I was thinking - hopefully these are all just putting words to things that are already on people's lists anyway 20:39:27 sdague: +1 20:39:47 mugsie: and I think there's a difference in "we couldn't get to it" and "screw you guys, we're not doing it" 20:39:47 the current 2 are fine, nice small, and common sense 20:39:59 because I think that we state some goals, ask people to work towards them, make sure we captures what's getting done, and it helps us all have a better picture of the state of openstack 20:40:00 mordred: +1 20:40:03 I think this is a very good start for the community-goals effort 20:40:15 flaper87: agreed, it has to start somewhere. 20:40:29 I sorta read it as "replace the giant cross-project spec with smaller goals" -- is that an accurate assessment? 20:40:46 annegentle_: I think specs will still exist as reference 20:40:59 there is no replacing here 20:40:59 annegentle_ : not really. by the time something is ready to be approved as a goal, I would expect the path to finishing it to be very clear already. 20:41:12 dhellmann: so I was kinda expecting us to merge the goals post summit, but I guess this is us going to the summit with some merged ideas to spark the conversation? 20:41:19 for example, one of the other things discussed was moving all projects so they can be deployed behind apache 20:41:20 dhellmann: +1 20:41:21 this feels more about execution support 20:41:27 johnthetubaguy: yeah 20:41:29 figuring out how to do that is going to require work up front 20:41:31 ok that's useful thanks dhellmann thingee 20:42:06 johnthetubaguy : right, we need to land these in time for teams to (a) know about the fact that we expect them to dedicate time to them and (b) have time to do any planning at the summit 20:42:15 johnthetubaguy : in future cycles, we'll be doing this right before the PTG 20:42:18 only other question that I hadn't put on review is related to: if you author the goal are you then responsible/accountable for follow up in supporting teams to get what they need to complete the goal? 20:42:32 because a bunch of the openstack-specs were like "let's do this thing!" and even if they got approved, then actually doing that thing, and figuring out where that stood in all the projects, was pretty nebulous. And if you can't see how far away the end game is, people get demotivated 20:42:57 sdague: yeah exactly the scenario I played through too 20:42:59 it was amazing how much just doing this http://burndown.dague.org/ helped us work through our api-ref on nova 20:43:02 annegentle_ : that's a good question. I hope not. I've written this so that we only want goals to be documented by a tc member, in part to avoid having the governance repo turn into the cross-project specs repo. 20:43:14 dhellmann: ayup, got it. 20:43:15 so, in that line of thinking, would goals that only really affect a few projects be accepted? 20:43:20 annegentle_, dhellmann: I don't think it should. it should be up to the team. 20:43:21 annegentle_ : +1 should have sufficient help to the teams 20:43:23 sdague: lol, I read "execution support" entirely wrong 20:43:30 sdague: good thing you elaborated on that 20:43:31 flaper87: heh 20:43:37 sdague: it was so good we stole that for the configuration options: http://45.55.105.55:8082/config-options.html 20:44:04 annegentle_, dhellmann the TC may review who has followed through the goals though and may reach out to projects. 20:44:17 mugsie : the oslo goal that's up now affects fewer projects than I thought when I wrote it up, but we wouldn't be likely to take something as "community wide" if we knew from the outset that it was only 2 projects. 20:44:31 thingee : yes, we need to do that review at the end of each cycle 20:45:02 dhellmann: great, I think that's the only follow-up that needs to happen. leave the goal author out of it 20:45:04 sdague : I may be interested in the code for that for release tracking 20:45:08 dhellmann: and is it worthwhile to propose something that affects one team only? Or is that a 1:1 convo with that team's lead? 20:45:20 annegentle_ : these are for community-wide goals 20:45:28 dhellmann: yeh, sure, I have thoughts about how to make it more generic. We can do that offline. 20:45:41 dhellmann: I think it's going to be enough work just getting a goal agreed. 20:45:44 annegentle_ : if we need a way to make recommendations to individual teams, we need a new process 20:45:45 sdague: markus_z did a bit of that, possibly 20:45:49 thingee: ++ 20:45:58 annegentle_ : I'd be happy to collaborate on designing that 20:46:19 dhellmann: ok, thanks 20:46:30 ok, we have 5 more mins for this topic. Anything else people would like to ask? Other open questions? Huge objections? 20:46:41 dhellmann: would we not just use the team's own process for work, if its team specific? 20:46:49 The review will be open till ttx is back so there's still time to comment on the review 20:46:49 i like it. thanks for driving this, dhellmann 20:46:52 that's all I have, hadn't voted yet as I needed to ask these 20:46:52 I think this process is useful for even fully understanding how far away we are from things, as dhellmann said, we didn't even know how big the oslo gap was (which is smaller than anticipated) until doing that analysis 20:46:57 and the goals seem like obviously good ones 20:46:59 sdague: +1 20:47:00 which is incredibly valuable all on it's own 20:47:02 johnthetubaguy : if the tc is asking a team to do something they're not already prioritizing, we may want to do that a different way 20:47:15 annegentle_: I always look forward to your questions, fwiw 20:47:17 :) 20:47:29 dhellmann: honestly, that feels bad to me, but thats a separate conversation 20:47:32 heh :) 20:47:33 so I love the goals 20:47:39 and I agree with russellb, thanks dhellmann for driving this 20:47:41 johnthetubaguy : right 20:47:49 its totally something I wanted to see happen while I was on the TC, and love this approach 20:48:33 Ok, unless there are other questions/comments, I think we can move on 20:48:36 I am tempted to say what we have before the PTG/summit is a draft that could get torn up, but thats just splitting hairs 20:48:53 johnthetubaguy: yeah I do think timing will matter too 20:49:09 we need something before, else nothing will happen 20:49:17 but its likely to get revised 20:49:18 yeah, we're going to need to work quickly to get goals out of this summit to work on before the ptg in feb 20:49:21 but honestly, we should just do it 20:49:24 johnthetubaguy: yeah, that was one of the goals 20:49:46 we should encourage PTL's to file some too? (ex stevemar to add a goal for keystone v3 api support in all projects) 20:49:58 dims: good Q 20:50:07 dims : as written, this says tc members will write them up 20:50:14 I want *us* to be active in this process. 20:50:31 dhellmann : ok collaborate with one of us to file it 20:50:33 need to keep a reasonable limit on the number of these, too 20:50:33 everyone can propose them, but I want us to take the step of documenting them, and choosing them 20:50:36 I guess its just a case of getting a sponsor via the ML 20:50:37 dims: and I see those as different goals but we can elaborate on this later 20:50:41 dhellmann: ++ 20:50:49 right 20:50:51 and just approve a couple/few highest priority ones.. 20:51:00 let's start with us deep in the middle of this 20:51:01 dhellmann: yeah I'd imagine a deadline, then a discussion meeting at regular TC meeting time. 20:51:02 at least to start with 20:51:06 dims : also, I do not expect us to be choosing from a bunch of written up goals each cycle. I expect us to have a list of things coming out of the summit, get it narrowed down, then write up a couple of finalists. 20:51:13 dhellmann: ++ 20:51:17 russellb: I thin kthe priority of some of these goals will be the key of how they will be merged 20:51:18 dhellmann: ++ 20:51:19 dhellmann: oh even better 20:51:26 dhellmann: ++ 20:51:26 dhellmann: ++ thats what I was kinda trying to say 20:51:38 ack dhellmann 20:51:40 ok, 10 mins left, let's move onto Open Discussion! Thanks again, dhellmann 20:51:42 we're the last step in this process 20:51:47 dhellmann: ++ 20:51:49 makes sense 20:51:49 +1 20:51:50 well, aside from implementation :-) 20:51:55 heh 20:52:00 #topic Open discussion 20:52:08 dhellmann: got mestery to vote on the patch 20:52:13 flaper87 : ++ 20:52:13 ttx will vote when he's back 20:52:16 * dims is away on vacation next week 20:52:22 * Agree to at least temporarily abandon stalled items: 20:52:23 #link https://review.openstack.org/338796 20:52:25 #link https://review.openstack.org/314691 20:52:27 #link https://review.openstack.org/295971 20:52:29 #link https://review.openstack.org/295528 20:52:31 * flaper87 is away on vacation next week too 20:52:39 Any comments on those reviews? 20:52:43 flaper87 : those release model items can be dropped 20:52:48 flaper87: I'll abandon mine on that list 20:52:51 Any reason why those shouldn't be abandoned ? 20:52:54 mtreinish: thanks 20:52:57 sdake: ^ 20:52:57 I still need to write up the followup proposal 20:53:00 flaper87 : sorry, the 'type' tag items 20:53:01 I keep forgetting to do that 20:53:03 don't think 338796 is stalled ... the author has been active and waiting for a response from the tc 20:53:05 the release team isn't going to use those for now 20:53:13 mordred: ^ 20:53:23 flaper87 it makes it hard for me to find the original comments 20:53:31 thingee: yeah I'd agree with that assessment too 20:53:43 mestery voted on the tc vacancy review as well 20:53:45 flaper87 but i guess I can unabandon them when I get around to it 20:53:49 sdake: yes 20:53:52 sdake : yes, please 20:53:54 sdake: please :) 20:53:58 sdake: owner:self status:abandoned 20:54:00 sdake: both of them, if you don't mind 20:54:03 it's easy to search for them 20:54:05 thingee: yah - it's on my list 20:54:25 ok, who's out next week? 20:54:25 oh, that reminds me we need the vmt to weigh in on https://review.openstack.org/350597 (which will just be a request to start with a thread on the -dev ml) 20:54:32 flaper87 feel free to abandon them - they will be in my inbox 20:54:36 flaper87: me 20:54:42 spain and burning man 20:54:43 sdake: either you or ttx can do that :( 20:54:48 sdake : flaper87 doesn't have permission to do it 20:54:51 flaper87 i'll get around to em some day when shit isn't on fire ;) 20:54:55 oh ok 20:54:58 got specific links? 20:55:08 sdake: can we not swear please 20:55:11 sorry 20:55:17 some day when things are not on fire 20:55:18 sdake : https://review.openstack.org/#/c/295971/ and https://review.openstack.org/#/c/295528/ 20:55:20 thank you 20:55:23 thanks dhellmann 20:55:23 sdake: https://review.openstack.org/295971 20:55:23 sdake: https://review.openstack.org/295528 20:55:33 so, thingee dims and myself will be out 20:55:35 anyone else? 20:55:40 flaper87: your linking is on fire today :) 20:55:52 good job flaper87 :) 20:55:56 flaper87: next week? 20:55:57 sounds like there will be quorum so, there should be meeting 20:55:59 yeah, thanks again for being chair flaper87 20:55:59 mtreinish: yes 20:56:05 flaper87: yeah thanks 20:56:05 my pleasure :) 20:56:18 yeah - kudos flaper87! 20:56:39 edleafe: annegentle_ danke danke :) 20:57:09 ok, anything else pressing to discuss ? 20:57:11 thanks, flaper87, and enjoy your time off! 20:57:26 otherwise, we can call it a meeting and be happy to have closed it a couple of mins earlier 20:57:36 Or I can do the flaper87 thing and keep talking till time's up 20:57:41 * flaper87 is Italian, he knows how to do that 20:57:44 russellb: thanks :) 20:58:00 thanks flaper87 20:58:10 flaper87 your wish is my command :) 20:58:10 * flaper87 bows 20:58:11 enjoy 20:58:17 ok, let's call it! 20:58:18 thanks everyone 20:58:25 have a great rest of your week 20:58:26 thx flaper87 20:58:28 #endmeeting