20:02:24 #startmeeting tc 20:02:25 Meeting started Tue Jan 12 20:02:24 2016 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:02:26 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:02:28 Hi everyone! 20:02:29 The meeting name has been set to 'tc' 20:02:32 i was talking to mestery, i blame him 20:02:34 Our agenda for today: 20:02:39 #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee 20:02:42 * edleafe lurks suspiciously in the corner... 20:02:59 would like to potentially add some more rubber stamp business 20:03:10 https://review.openstack.org/#/c/243348/ 20:03:12 thingee: which ones ? 20:03:13 * mestery watches edleafe from the corner of his eyes 20:03:20 and https://review.openstack.org/#/c/236712/ 20:03:34 thingee: fine by me.. objections ? 20:03:48 well 20:03:55 isn't there a one week thing for all items? 20:04:09 things have had a lifespan more than a week 20:04:15 cross project ruberstamps are not really TC resolutions 20:04:23 * rockyg peeks out from under a stone 20:04:31 ttx: I am obviously confused :) 20:04:32 it's more finally rubberstamping consensus 20:04:42 well, i thought they were 20:04:49 lifeless: we don't really vote on them -- we acknowledge existing consensus 20:05:08 Everyone, these items have existed more than a week. 20:05:09 thingee : the second one appears to be in conflict with the first :-) 20:05:14 hence the name, rubberstamps 20:05:27 thingee: not as a TC meeting agenda item though, so I see what they mean 20:05:34 ah true 20:05:41 thingee: the TC meeting protocol specifies a date when things have to be on the agenda 20:05:47 I'm fine delaying them until properly added to agenda. 20:05:50 thingee: to let participants have time to read them 20:05:55 lifeless: oh I see 20:06:01 * thingee thinks it would be good if people in the tc are watching the repo regardless of agenda item 20:06:10 (or rather, I'm only fine adding them to agenda now if nobody objects, and I would count this as objection) 20:06:15 thingee: Good idea 20:06:15 lifeless: ok, that's a fair point 20:06:49 yeah it's a bit late to look into those 20:06:54 ttx: I'd say add to the agenda 20:06:56 separately, I want to drill into this 'not a resolution' thing 20:07:05 because at least jeblair and I thought differently 20:07:11 but we can do that late 20:07:12 later 20:07:21 cross project specs are things that PTLs need to agree to 20:07:40 ttx: lets defer to open discussion ? 20:07:45 since it's difficult to have voting rules on that, we use the TC to check that consensus is there 20:08:16 but we just ack consensus, otherwise our vote is just like someone else's vote. At least that's how we have done it until now 20:08:23 ok 20:08:25 fine with me. There's an agenda item in the cross-project meeting to talk about getting team consensus. 20:08:27 I don't see a lot of votes on https://review.openstack.org/#/c/236712/ either, fwiw 20:08:31 #topic Rubberstamp cross-project spec: Adds spec for Service Catalog standardization 20:08:36 #link https://review.openstack.org/#/c/181393/ 20:08:58 we have 7 20:09:18 so I'll consider it ready for merging unless someone objects 20:09:22 dhellmann: I don't see objections currently since it existed oct 18 ;) 20:09:40 I had one concern 20:09:42 we've been grinding at this for a long time, hopefully what's there is good for folks 20:09:49 which for some reason was stuck in draft 20:09:55 there's a paragraph that claims to contain a list 20:09:59 but there is no list 20:10:06 lifeless: which bit? 20:10:08 so its a little confusing at best 20:10:12 for the service catalog, we've whittled for a while, and I think merging now is the right thing to do, with edits as we go 20:10:20 sdague: line 117 20:10:25 since I dislike sentence frags as much as anyone :) 20:10:30 sdague: the para above which also says 'list of' has an actual list with it 20:11:04 sdague: I don't want to hold it up 20:11:17 sdague: but I'm only +1 while confused, for what thats worth 20:11:21 lifeless: so the list in the paragraph above is not the list of things after it 20:11:57 It's saying "we'd like to get to having a list of things we need to change" and "here are some guiding principles" 20:12:03 ok, lets approve, but maybe queue a clarification 20:12:24 sdague: ok, so yeah - that didn't come through clearly. perhaps a post merge tweak? 20:12:30 lifeless: right, it's saying we'll make a list as a task. but I can see the confusion point, sure 20:12:31 because at this point, this document is mostly a statement of priciples and direction. The details have to come out as they come out 20:12:43 ok approved 20:12:53 lifeless: yeh, I'm so english blind on this spec at this point, if you have better words, please propose 20:12:57 post merge tweaks welcome 20:13:01 ttx: thanks! 20:13:06 heh me too on english-blind 20:13:07 sdague: I can do that 20:13:16 #topic Refactor 4 opens as in Project team guide 20:13:22 #link https://review.openstack.org/264127 20:13:34 This change is about bringing the version of the 4 opens in the governance up to date with the one in the project team guide 20:13:58 we have 7 now, will approve unless someone objects 20:14:24 (tl;dr being it is mostly placing "principles" bits in the "open" they actually belong in) 20:14:37 do we have our stance on open core written anywhere else? 20:14:53 I don't think so 20:14:58 not that I can think of 20:15:04 I mean, we need to retire the wiki version obviously 20:15:16 and there is the subsequent change to remove it from PTG 20:15:18 So I'm fine with this tweak 20:15:28 I still want to do something about the open core discussion 20:15:28 uhm 20:15:33 so - yeah 20:15:37 jeblair: but otherwise not that I know of 20:15:38 sorry, just read this 20:15:43 i'm worried about approving this now because that effectively means we have repealed our open core stance 20:15:47 and the open core thing is vexxing 20:15:49 yea 20:15:59 that's the main reason I wanted to add this to the governance repo in the first place 20:16:00 jeblair: why ? 20:16:01 jeblair: wait, what ? 20:16:09 read the first line 20:16:16 the end lines that I removed are a duplicate stance 20:16:25 OH 20:16:27 which do not bring additional info 20:16:27 Am I reading the right diff ? 20:16:32 yeh, expand the code 20:16:38 expand the view so you can see the whole file 20:16:38 lifeless: no *they* are reading it wrong. 20:16:38 lifeless: it's outside the diff 20:16:44 sorry - I was reading out of context - those lines at the bottom are lines that describe "Open Community" 20:16:47 ah that's different; let me put a comment there 20:16:53 yeah. thanks 20:17:03 * mordred is bad at reading 20:17:08 me too 20:17:28 mordred , jeblair : don't feel bad. I commented about an addition that was actually a removal yesterday. Diffs are hard. 20:17:41 dhellmann: maybe we could replace them with something more trendy and hipster 20:17:50 dhellmann: like emoji or something 20:17:51 which is why I want to remove the duplication in the first place :) 20:17:59 blockchain emoji 20:18:02 mordred: needs more social 20:18:04 sdague: YES! 20:18:06 alright approving now 20:18:08 heh 20:18:15 sdague: we should definitely replace diffs with something based on blockchain 20:18:20 is the mission in the foundation docs? 20:18:20 #topic Add interop and endusers to OpenStack Mission 20:18:23 blockchain is the new docker 20:18:25 #link https://review.openstack.org/264135 20:18:32 So... this resolution is the first step in the likely long road to updating our mission statement 20:18:39 This was introduced by mordred in Tokyo during the joint meeting with the board 20:18:41 mordred: lol 20:18:47 The goal being to mention "interoperability" and "end users" in the stated goals 20:19:05 During that meeting we concluded that the mission statement was sufficiently central to OpenStack that it would be preferable that both the board and the TC agree to the change 20:19:19 The first step is to propose a wording, and this is what some of us came up with 20:19:27 We tried to be additive, and to minimize the change 20:19:41 yeh, seems good to me 20:19:56 especially end users 20:20:02 alright, ready to approve unless there are questions 20:20:11 lgtm 20:20:32 annegentle, mordred, ttx: thank you for that! :) 20:20:33 good to have lots of votes on this one 20:20:42 approved 20:21:08 russellb, mordred: could you bring that to the board, or should I ? 20:21:17 now we all should look for 1-2 tasks we can do to advocate for users :) 20:21:32 ttx: i can ask for it to be on the next board meeting agenda 20:21:37 annegentle: now we can't pretend we didn't know :) 20:21:41 ttx: yes, russelb is good at doing things 20:21:49 I votes for him. 20:21:51 ttx: heh 20:21:52 s/russelb/russellb/ 20:21:53 and voted too 20:22:03 #topic Brainstorming about what the TC would like to see on the "upstream dev" track in Austin 20:22:05 thanks russellb 20:22:06 thanks :) 20:22:09 #link https://etherpad.openstack.org/p/austin-upstream-dev-track-ideas 20:22:16 so weird to see "in Austin" (just have to say) 20:22:18 So.. I hope we'll just have good submissions through the normal CFP process 20:22:27 But there may still be things we really want to see covered there 20:22:30 annegentle: they keep it weird there 20:22:36 and for which nudging some people to make a proposal could help in getting the right content there 20:22:52 let's see what we have 20:23:02 jeblair: so I hear 20:23:11 jeblair: not as weird as portland I also hear 20:23:22 ttx: I added "How do you even doc?" 20:25:05 annegentle: I think portland favors strange rather than weird, since they don't like bandwagoning 20:25:11 ha! 20:25:18 frankly it all feels like things people would naturally submit... is there any must-have in there for which having a volunteer is absolutely necessary ? 20:25:58 * dhellmann hopes coming up with suggestions wasn't the same as volunteering 20:26:06 mordred: thank you for noticing the difference 20:26:09 dhellmann: ++ 20:26:19 we do usually find session leaders 20:26:22 it is not 20:26:26 yeh, I was hoping this was more of a pull than a push honestly at this stage 20:26:43 dhellmann: though some suggestions clearly point to someone ;) 20:26:46 annegentle : these are presentation ideas, though, right? so we need speakers to write talks. 20:27:09 ttx: I'll work on obfuscating those 20:27:09 dhellmann: yea and they're being recorded so nearly workshoppy if we make the most of videos 20:27:16 annegentle : ++ 20:27:16 like we should probably rope dansmith into doing a developing for upgradability talk (like the cp session from last time) 20:27:49 sdague: yes we should 20:27:58 sdague : added as 4.6 20:28:03 nudge nudge 20:28:26 I made up workshoppy of course but yes, these need to be polished and re-usable 20:29:53 any large development process change that we really need to communicate ? 20:30:01 I think the last one was reno 20:30:24 explain release tags? 20:30:33 (I even stumbled on it myself) 20:30:34 the CFP-closes-two-months-before-event is going to bite us 20:30:39 I've put it on the etherpad 20:30:48 but I think socialising how to work with DLM's is a thing to do 20:30:57 annegentle : let's talk offline about what you mean by that, it may be covered by our presentation in tokyo 20:31:08 dhellmann: ok cool 20:31:17 dhellmann: which no dev attended 20:31:18 if I'm wrong, we can add it to the list 20:31:28 ttx: true, but we can share the link to the video 20:31:34 also, cats 20:31:41 lolcats? 20:31:44 we're not expecting all of them to come to these, are we? 20:31:46 just cats 20:32:07 dhellmann: no but we are expecting all the attendees to be 20:32:14 k 20:33:11 sdague: what would be the next step with this brainstorming ? keep it live and have it trigger volunteers ? 20:33:26 maybe start a -dev thread on it 20:33:31 yeh, it's probably worth sharing as a -dev thread 20:33:34 I should do that 20:33:42 just to clarify, the anticipated audience are mostly not the sort of people who will be in design sessions 100% of the time (currenty active contributors), but rather people with some development background who are attending general conference talks (maybe also new or less active contributors)? 20:33:44 explaining what the track is about and all 20:33:46 ttx: I did this one weird trick with grep through months of IRC logs, to see what questions newcomers had, do we have any similar analysis of what devs are asking? 20:33:54 fungi: no 20:34:03 fungi : no, these will not run concurrently with design sessions 20:34:10 ohhhh 20:34:14 that helps a bunch 20:34:16 fungi: it's running on the Monday, when the design summit is not started yet, and is meant for openstack upstream devs 20:34:18 ttx: looking for a data set that's not us, since we're not the intended audience, ya know? 20:34:24 * thingee can volunteer to fill speaker slots 20:34:32 if only I'd pushed through that dev survey 20:34:40 I keep stalling on it :/ 20:34:41 annegentle: actually, I think the target audience is us 20:34:50 yes, all of us 20:34:59 sdague: ok, for leveling up your dev skillz? that sort of thing? 20:35:11 annegentle: and learn what you need to learn 20:35:16 I feel like this is going to be way more useful if each of us came up with 1 "you know, I really wish I understood X more" 20:35:22 sdague: yeah 20:35:54 sdague: ++ 20:35:59 do we have any data on which contributor pages or videos get the most views already? 20:36:13 #action ttx to explain on -dev what the track is about and point people to the etherpad 20:36:18 submitted anonymously, since we all know all that stuff already, of course 20:36:20 honestly, an overview of the neutron resource model, and what you need to put a thing on the network would be great. 20:36:39 edleafe: or we are big enough to admit that we don't know everything 20:37:16 sdague: :) 20:37:40 annegentle: I think that would have to come from jimmy wouldn't it? 20:37:46 ttx: I'll chase the data angle 20:37:59 talk about new features in some lib that all projects should probably take advantage of 20:38:00 anteaya: I'll start with a request at the foundation 20:38:01 etc 20:38:04 anteaya: who's jimmy? 20:38:07 i'm assuming a vulnerability management process presentation would fit best in the cross-project themes section? or is there a better section for horizontal effort presentations? 20:38:16 annegentle: jimmy macarthur 20:38:50 fungi : that seems like a good place for it 20:39:04 fungi: sounds good 20:40:00 alright, I'll push that fun page to the rest of the world 20:40:39 sdague: you can reply to that and suggest people reply with "one thing that as an upstream dev I would really like to learn more about" 20:40:45 since that's a brilliant idea 20:40:48 sure 20:41:07 should be a mine of presentations ideas 20:41:29 alright, lets move on but you can continue on the etherpad 20:41:32 hrm... bindep integration in the new developer-facing features that need adoption category or the developer-oriented infra category? 20:41:46 meh, i'll just pick one 20:41:49 skipping next topic since that was not updated 20:42:14 #topic Open discussion 20:42:22 gothicmindfood had something to mention 20:42:31 hey everyone! 20:42:33 fungi: either 20:42:37 * mestery waves at gothicmindfood 20:42:42 otherwise we can reconsider (again) how we approve cross-project specs 20:43:17 I have a #link https://etherpad.openstack.org/p/Leadershiptraining with some info/thoughts gathered on leadership training 20:43:57 * mordred for one supports his new zingtrain overlords 20:44:10 (would be interested in participating) 20:44:14 after asking around for some opinions/thoughts I put together a few ideas, and I'll be sending an email to Mark Collier soon to talk foundation funding 20:44:28 * thingee has a couple of things for opend discussion 20:44:40 I purposely gathered intel/data from folks who had *ahem* diverse opinions on this the last time it was brought up in a meeting 20:45:21 I'm fine attending, if I can be available 20:45:34 and that was awesome. So I welcome more thoughts or volunteers who'd be interested in participating 20:45:37 sounds like they have an interesting approach 20:45:59 ttx: that, and good food. which is the only reason mordred is interested :) 20:46:03 also in favor of sending a team of potentially skeptical scouts to come and report back 20:46:21 mordred: i was thinking you were in the generally skeptical camp last time we talked 20:46:26 yeah - I put that in the etherpad write up, but I want to reiterate: I think the more skeptics that go, the better 20:46:31 mordred: food is sufficient motivation? :) 20:46:36 I am - but also the sandwiches are AMAZING 20:46:40 jeblair: of course it is 20:46:56 and honestly, if I have any amount of skepticism, I'd be a good candidate to give it a try 20:47:00 gothicmindfood: ok, keep us posted on where that goes 20:47:05 ttx: will do. 20:47:13 back on crossproject stuff... 20:47:22 Last time we discussed cross-project spec approval, we said that it required community consensus, and that the most convenient way to acknowledge consensus was to ask the TC members to validate that there was consensus 20:47:27 I'm fine with a or B 20:47:29 I'd like to go 20:47:31 Hence the current system - everyone reviews the spec and when there is consensus, we push for a rubberstamping by the tc 20:47:33 gothicmindfood: are there dates yet? 20:47:47 I've seen bad things in the space, but also good stuff, and good can be amazing 20:47:49 But I think we have an issue if TC members only look at cross-project specs only once they have reached community consensus 20:47:50 (sorry, had to read then type) 20:48:21 gothicmindfood: as a self-identified skeptic, i appreciate the creative solution you are proposing and the work that must have gone into developing it. :) 20:48:21 because then the current system is broken 20:48:29 jeblair: +1 20:50:01 jeblair, lifeless: so if you consider your cross-project specs vote is similar to your resolution votes, then we have a problem 20:50:18 since you are supposed to have already voted +1 on the spec and use your +2 to say "yes, I see consensus" 20:50:20 annegentle: no dates, unless you look at prescheduled options (option A) - the link I just added shows those. Option B would be on OpenStack's timetable 20:50:44 ttx: is this documented ? 20:50:55 ttx: the thing that caught my eye earlier was the idea that we're not actually voting. i think we are voting on them to give them the weight of a tc vote. 20:50:56 ttx: like, is there a manual I missed reading? 20:50:57 so I'm fine with getting more consensus from ptls or cross-project spec liaisons (https://review.openstack.org/266072) before approaching tc. At some point though, we need to move on though regardless of everyone being attentive. 20:50:57 thanks gothicmindfood 20:51:08 lifeless: it was discussed in this meeting 20:51:12 jeblair: thanks! Coming from my favorite skeptic, I appreciate that :) 20:51:22 might be documented, let me check 20:51:25 ttx: the TC rotates 20:51:27 ttx: I agree with jeblair. I thought that once we approved these specs, we backed them up. 20:51:50 yah 20:51:52 jeblair: oh, we ARE voting. But before it comes to rubberstamp 20:51:56 so there are multiple layers here 20:52:07 which is not the same as vote +1 at first and +2 later, which is a confusing process but I guess OK. 20:52:11 basically our vote is not a +2, it's the same level as a PTL +1 20:52:20 README.rst in openstack-specs does *not* describe special voting mechanisms or anything 20:52:26 PTLs are at least as important than us in getting the spec approved 20:52:36 ttx: how they get on the agenda, and how we go about actually achieving the consensus we desire are not things i have strong opinions on -- only that our vote means it is a real tc vote. 20:52:40 ttx: probably more, since they have to implement them 20:52:46 dhellmann: right 20:52:48 jeblair : ++ 20:52:55 nor does its CONTRIBUTING.rst 20:53:13 lifeless: yeah, we may have relied on oral tradition there :/ 20:53:20 ttx: it sounds like we need a TC section in the PTG 20:53:31 mostly because the crossproject meeting was just after this one and that's where those were discussed 20:53:33 or an expanded one, if there is already one 20:53:34 dhellmann: that would be nice as well 20:53:46 but also, repos with different rules should have them locally documented IMO 20:53:47 oh well 20:53:54 that doesn't solve the issue 20:53:59 lifeless : true 20:54:10 which is TC members should look at those specs earlier 20:54:15 lifeless : but a "here are your duties" list would be helpful. maybe that should go into the governance repo, though 20:54:18 the same way we ask PTLs to look at them 20:54:25 dhellmann: yes 20:54:37 if you wait for it to be pushed to the TC agenda, we have an issue 20:54:56 since it's not a TC item, it's a crossproject item 20:55:11 the only reason it's pushed as a TC item is a technicality 20:55:23 to get final rubberstamp on it 20:55:34 since we have no other way to "approve" a cross project thing 20:55:59 so 20:56:11 there are two disconnected issues 20:56:19 one is the weird voting setup on the repo 20:56:30 by weird I mean non-standard 20:56:34 whew 20:56:35 right 20:56:42 * jeblair sits back down 20:56:54 separately there is 'are tc members not reading cross project specs enough' 20:57:02 I don't know that we have evidence/concerns on that 20:57:18 since jeblair, me and dhellmann seem to be more on the 'if we're voting, we're voting' angle 20:57:28 but ttx seems worried about the 'are folk reading' angle 20:57:40 lifeless : I'm also worried that folks aren't reading 20:57:47 dhellmann: ack 20:58:31 not just the TC, though 20:58:43 basically we raise those at the TC meeting to give it the final approval push, but lately it felt like that's when TC members start to register objections on them 20:58:52 (or others for that matter) 20:59:03 see https://review.openstack.org/#/c/236712/ for example, which thingee wanted us to consider for approval today but that has a very small number of contributors 20:59:51 meaning the "consensus" we are supposedly rubberstamping is not really there yet, and we are having a technical discussion about the spec rather than a social discussion about the presence of consensus 20:59:52 we're basically in a state where if someone wants to do cross-project work, they already have to be well known and very noisy about it to get anywhere 21:00:06 which means we're not doing enough to grow leaders within the community 21:00:08 dhellmann: yeah 21:00:09 excluding non-TC members from the technical discussion 21:00:25 and we are out of time 21:00:31 and the cross-project meeting starts :) 21:00:34 I think you're getting at some of the fundamental economics of openstack there 21:00:42 we're paid 21:00:43 I would like to say quickly while I have the tc's attention 21:00:44 it make it weird 21:01:15 please attend the cross-project meeting to discuss the cross-project spec liaison which aims on getting consensus from each team for spec for approval 21:01:19 at least improving it 21:01:19 yes 21:01:21 please 21:01:27 I'll be there in a minute 21:01:32 closing here first 21:01:35 Thanks everyone 21:01:39 * harlowja joined twitter recently to be more noisy 21:01:40 #endmeeting