20:01:23 #startmeeting tc 20:01:23 * edleafe lurks 20:01:24 o/ 20:01:24 Meeting started Tue Mar 22 20:01:23 2016 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:01:25 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:01:28 The meeting name has been set to 'tc' 20:01:30 o/ 20:01:32 * cdent lurks 20:01:38 o/ 20:01:38 Hi everyone! 20:01:43 * docaedo lurks 20:01:44 Our agenda for today: 20:01:49 #link https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee 20:01:57 o/ 20:01:58 #topic Appointing missing PTLs for Newton 20:02:10 We have a number of project teams that failed to present candidates for PTL election 20:02:18 For those we have to pick a PTL (or make the team unofficial) 20:02:26 There was a ML thread covering it: 20:02:30 #link http://lists.openstack.org/pipermail/openstack-dev/2016-March/089908.html 20:02:38 * EC2-API 20:02:48 For this one we have a late candidate (the current PTL), Alex Levine 20:02:50 I'm here if you need to ask me something 20:03:02 This team is new (been created February 2), so I'm ready to cut them some slack. It's their first election 20:03:13 +1 for some slack. 20:03:29 ++ 20:03:31 wfm 20:03:39 we might need to be clearer with new teamls that there will be reelections with everyone else 20:03:46 yeah 20:03:53 even if that mleans doing an election one month after they are created 20:04:05 ah that might work 20:04:05 +1 for slack 20:04:24 i hope that once the election info is getting auto-published to governance.o.o soon we'll be able to link from the ptg to current election schedules as well 20:04:25 #link http://lists.openstack.org/pipermail/openstack-dev/2016-March/089939.html late ec2-api ptl self-nomination 20:04:31 Any objection to selecting alexandrelevine as EC2API ptl for Newton ? 20:04:34 o/ 20:04:46 none here 20:05:00 none 20:05:04 I'll formally propose it in a change anyway 20:05:19 so we'll have a normal vote on that by next week. I'll assume it's OK if we reach 7 votes though 20:05:31 since we discussed it today 20:05:51 #action ttx to put up a change proposing alexandrelevine as EC2API ptl 20:05:55 * Stable Branch Maintenance 20:05:59 o/ 20:06:01 This one is more of a governance corner case, with the current PTL candidating for another PTLship 20:06:08 and a volunteer replacement blocked by election officializing duties 20:06:18 I think in this case the governance is working as intended by giving us the ability to fix the corner case 20:06:28 I'm +1 on giving tonyb the role 20:06:31 ++ 20:06:33 ++ 20:06:33 +1 20:06:36 +1 20:06:36 wfm 20:06:38 unless mriedem has second thought on being Nova PTL 20:06:41 ++ 20:06:44 +1 20:06:45 it's not TOO LATE 20:06:46 mriedem: now's your chance 20:06:49 oh god 20:06:52 run 20:06:53 mriedem: run 20:06:56 * mriedem makes shifty eyes 20:06:56 run while you can 20:06:57 haha 20:07:02 lol 20:07:13 the current ptl is the only nova candidate? 20:07:13 * mordred hands mriedem a box of chickens and one mostly asleep llama 20:07:19 ha 20:07:27 i mean, thanks for taking on such an important role! you'll do great :) 20:07:27 jeblair: yes 20:07:28 jeblair: Yes, that's the case 20:07:30 #action ttx to put up a change proposing tonyb as stable PTL 20:07:37 he'd pretty much have to abdicate at this point 20:07:42 mordred: Ooooo, fancy gifts! 20:07:47 * Winstackers 20:07:56 This is another relatively young team, added this cycle... so also their first election 20:08:08 We have a late volunteer, the original PTL for the team. The team seems active enough. 20:08:10 so that's going to happen, leaving tonyb as the only potential candidate we know of for stable ptl; so yeah, i think the corner case is not troubling and tonyb sounds good 20:08:12 #link http://lists.openstack.org/pipermail/openstack-dev/2016-March/090111.html late winstackers ptl self-nomination 20:08:17 hello, i'm here, if there's any questions 20:08:44 and I'm sorry for being late 20:08:52 +1 on that one, too,. 20:08:59 +1 20:09:09 So again I don't think that is uncovering a deeper issue -- that team sounds mostly functional to me 20:09:21 no worries claudiub 20:09:24 ++ 20:09:24 so +1 on claudiub from me 20:09:50 ++ 20:09:52 +1 for claudiub from me as well 20:09:54 thanks folks, I'll definetely be more careful next time. :) 20:09:55 #action ttx to put up a change proposing claudiub as winstackers ptl 20:10:12 I'll approve those 3 changes if they reach 7 votes, so that it's more formal 20:10:32 ttx: I have posted several reviews to hopefully help the election to be more known when applying a new project http://lists.openstack.org/pipermail/openstack-dev/2016-March/090091.html 20:10:53 thingee: yes, thank you for those 20:10:58 we'll cover them next week likely 20:11:06 #topic Cross-project workshops at the Newton Design Summit in Austin 20:11:22 It's time to start thinking about how we plan to organize the cross-project workshop day at the Austin Design Summit 20:11:32 Happens all day Tuesday, with the Ops day separated and happening on the Monday 20:11:45 no more conflicts 20:11:47 good, thank you 20:11:50 As a result it will be the only thing running on Design Summit grounds on the Tuesday 20:11:55 \o/ 20:11:59 which gives us a lot of flexibility on the number of parallel rooms we want to run (or the shape of the rooms). 20:12:10 In Tokyo we did 3 parallel rooms (21 * 40-min slots) 20:12:20 Should we do that again ? More parallel rooms ? Less parallel rooms ? Longer sessions ? 20:12:50 it might be worth waiting to see what the content looks like that gets proposed 20:12:58 I want to have a session on the new separated event 20:13:15 sdague: that sounds fine, we can decide quite late how many rooms we actually want 20:13:16 i don't think we want too much parallelism; we were already missing people we expected with 3 tracks :) 20:13:27 on the Tuesday, I had figured you would do that on the Friday 20:13:28 I think last time we barely had enough solid content for 3 tracks 20:13:38 jeblair, ++ 20:13:40 jeblair: we can have times where we have only one and times when we have 3 parallel 20:13:56 ttx, sdague: ++ 20:14:10 so it looks like we should start the topic collection process 20:14:16 and then decide how to schedule 20:14:19 I don't think there should be any competition for you new event session 20:14:20 sounds good to me 20:14:23 so I would mostly just let the suggestions role in on the topics, and we can be flexible based on it 20:14:53 topic collection -- what should we use ? 20:14:55 and not have conflicts unless needed 20:14:56 to clarify, this is like nova/cinder and nova/neutron sessions on tuesday? 20:15:09 b/c we have cross-project stuff to cover 20:15:14 good old etherpad ? Google form and then copy to etherpad ? 20:15:22 mriedem: no 20:15:32 ttx: your old app didn't work last time? 20:15:36 mriedem: ideally those are affecting most projects, rather than 2 or 20:15:38 mriedem: more like "quotas everywhere" "logging everywhere" right? 20:15:40 good old summit app? :) 20:15:46 I vote we start in etherpad if we are going to end up there 20:15:47 oh, good old summit app 20:15:50 -2 for summit app 20:15:51 annegentle: ok, real cross-project 20:15:58 that was a giant pain to sync data 20:16:07 and stuff got lost 20:16:25 (not that etherpad doesn't sometimes lose data on its own, but yeah) 20:16:26 sdague: sync from what to what? 20:16:28 sdague: oh, I didn't know stuff got lost 20:16:32 s/cross-project/cross-openstack/ 20:16:40 yeah maybe etherpad for data collection and then we can move to other more structured docs to vote / select 20:16:45 because there were some things in the etherpad and some in the summit app 20:16:54 it was a giant mess 20:16:57 sdague: i think that's cause we switched halfway through 20:17:01 only one tool please 20:17:16 ttx: was going to post something to the list and later bring in the cross-project meeting next week. 20:17:16 i can agree with that 20:17:33 so... etherpad ? 20:17:37 etherpad seems fine tho 20:17:40 ttx: ++ etherpad 20:17:45 etherpad 20:17:57 consistent with what most others are using 20:18:20 etherpad in previous editions remained mostly readable until we started to vote and people spammed it 20:18:22 what's wrong with the summit app for this? 20:18:32 jeblair: can't vote there ? 20:18:44 so you need to copy the submissions to discuss them ? 20:18:47 right, we'll end up with an etherpad to work the data to completion 20:18:52 I thought folks could leave comments 20:18:55 (the comments suck since I wrote that app) 20:19:11 (and people can tell I hate web dev after 2 seconds of using the app) 20:19:28 ttx: I just assume you hate everything 20:19:36 mordred: I hate CSS and JS more 20:19:43 ha self deprecation is the best deprecation 20:19:53 Well, I hate PHP the most, but it's not using PHP 20:19:57 yeah, i mean, i like the structured data and comments we get with the app. 20:20:00 i'm fine with etherpad 20:20:15 annegentle, I thought spaghetti code depracation was the best depracation 20:20:23 it just seems like the app, when used well, handles this kind of data in the quantity we throw at it for the cross-project sessions the best 20:20:33 o/ :) 20:20:35 :) 20:20:40 apologies for being late 20:20:41 jeblair: there weren't that many submissions last time 20:21:13 my brain was not large enough to hold all 21 sessions in my head, and i had to make my own external index 20:21:14 it felt like an etherpad would have handled it well enough, and yes we did copy everything to etherpad manually to handle the last stage of selection 20:21:15 right, I think we have a pretty known workflow with etherpad, most projects are using it for planning. 20:21:30 sdague: +1 20:21:49 sdague: do you want to start the doc and write the email about it ? 20:21:55 ttx: sure, can do 20:22:05 ttx: that's what i'm trying to get at 20:22:10 why did we copy to the etherpad at the end? 20:22:39 jeblair: because it's just hard to have a discussion with the comments, and even harder to place votes 20:23:01 * david-lyle waits for the suggestion to use gerrit 20:23:07 a line of +1s on an etherpad is easier to read than pages of +1 comments 20:23:14 and then we can reorder etherpad, move things 20:23:29 I feel like we need to move on from this, we decided what we are using, it's what actually works for the sifting phase 20:23:42 sdague: yes sir! 20:23:47 #action sdague to start the doc and write the email about it 20:23:49 i will just try to be smarter in the future 20:23:57 leadership! No need for zing 20:24:01 * rockyg aims a raspberry at david-lyle 20:24:04 #topic assert*upgrade tags cleanup 20:24:16 We started the discussion on that last week 20:24:26 * Remove rolling-upgrade tag from swift/ceilometer (https://review.openstack.org/292334) 20:24:41 This one is about removing the rolling-upgrade tag from swift/ceilometer, since the tag requires the rolling upgrade to actually be tested in the gate 20:25:24 4 yes, wouldn't mind another 3 20:25:40 since it's likely to draw complaints if we fasttrack it 20:25:46 or if we lazyconsensus it 20:26:16 will add votes later, have a dr appt nowish thoough 20:26:27 fwiw, cool with me on ceilometer 20:26:29 lifeless: ack, thanks for attending some 20:26:32 Will be +1 from me 20:27:23 ok, we have 7 (8 counting lifeless), so I'm ready to aprove 20:27:26 or approve 20:27:35 comments, objections ? 20:28:21 ok, approving now 20:28:22 so did you hear from notmyname ttx? 20:28:33 no, cced him on the review 20:28:45 * notmyname is in a meeting with lawyers. I only just saw this right now when I saw the "swift" ping in IRC 20:28:52 ttx: a lot of people don't see that 20:28:59 sorry, I didn't see the cc 20:29:06 ttx: i am added to thousands of reviews; i never see that. 20:29:16 removing the tag doesn't mean swift doesn't support rolling upgrades, it means that swift doesn't test it at the gate 20:29:19 i actually filter them into a bitbucket i get so many 20:29:28 I have strong feelings on this, but I can't reply now. will have to be after my current meeting 20:29:50 I'm find waiting until next week to give you a chance to object 20:29:53 fine* 20:29:58 yeah I think that's fine 20:30:01 no hurry 20:30:36 #info decision delayed until next week to give notmyname a chance to object 20:30:49 * Add testing requirement to assert:supports-upgrade tag (https://review.openstack.org/292918) 20:30:58 This one is to add a requirement to actually test the basic upgrade in the gate 20:31:08 which will also mean removing it from aodh and gnocchi since they have no grenade-like test 20:31:21 I think this one is fair game 20:31:50 yep 20:32:01 one vote short 20:32:34 ttx: tag author gets an honorary vote right? :P 20:32:48 we have gordc +1 and I think adding grenade-like testing to aodh/gnocchi would be an awesome addition 20:33:05 ok, approving now 20:33:22 dansmith: no you get the right to be consulted every time someone complains about the tag being wrongly applied 20:33:28 oh boy 20:33:29 yeah... 20:33:38 mriedem knows something about that 20:33:41 i did see someone trying to justify that tag on their project b/c they run db schema migrations in dsvm setup 20:33:49 which is a bit bogus 20:34:01 approved 20:34:18 #topic Call out GPL style licenses in testing/validation. 20:34:25 #link https://review.openstack.org/293140 20:34:32 lifeless proposed a clarification around GPL licensing and test dependencies. 20:34:41 I think the clarification looks good, but ideally we would not mention static names there 20:34:49 (and just let requirements-core be aware of who is the licensing resident expert of the day and involve them in the discussion) 20:34:59 but I'm fine approvign that and proposing the name removal as a second step 20:35:02 n for "hey, go through all the prompts again?" I'm 20:35:02 pondering --hard, --force, --reset 20:35:02 05:36 < krotscheck> betherly: ^^ Any thoughts? 20:35:05 bah 20:35:10 ttx: we can update it as the names change 20:35:20 ttx: but I see no reason to *enforce* tribal knowledge 20:35:37 including names seems weird 20:35:50 lifeless: in the end it's requirements-core making the call -- they should know who to ask 20:36:04 not sure we need to codify that list in governance basically 20:36:22 names may be a better fit in requirements README or something 20:36:29 if we feel we need to write them down somewhere 20:36:32 yep, sounds good in requirements README 20:36:39 sure 20:36:41 we probably need to update that anyway 20:36:57 since it has some licensing guidelines and that or may not be compatible with the clarified ones 20:37:11 lifeless: if you remove the names, I'll do the requirements README update proposal 20:37:25 k 20:37:37 will do this afternoon 20:37:53 ttx: i did update requirements with this: https://review.openstack.org/279998 20:37:58 OK, I propose to approve that during the week if it reaches 7+ votes in favor 20:38:18 Nice! 20:38:23 I'll just have to add the names then 20:38:44 "if you have doubts over licensing, like for example a GPL test dep, you can ask the following folks" 20:39:09 #action ttx to propose a "licensing experts" addition to requirements README 20:39:17 #topic Add threat analysis for vulnerability:managed 20:39:28 #link https://review.openstack.org/294212 20:39:39 sdake proposed to slightly modify the wording of the vulnerability:managed requirement around third-party securiity review 20:39:52 fungi did approve it, which in my book counts as an approval from the VMT on the new wording 20:40:05 It's still technically a tag definition change, so we'll need a formal vote 20:40:07 yeah, didn't bother me 20:40:15 (i.e. at least 7 votes in favor) 20:40:24 it was expanding an already subjective list of terms 20:40:42 fungi: do you know what sdake has in his mind for adding that ? 20:41:02 oh, commit message is pretty clear 20:41:11 ttx: also there was an ml thread 20:41:12 ttx what do yu mean? 20:41:35 sdake: missed the rationale for the change, but now I read it 20:41:39 so a threat analysis is a thing the security guys do to cut down their workload on an audit 20:41:54 an audit includes auditing dependencies and whatnot for problems 20:42:01 red hat had an audit tema of 40 people 20:42:09 i tried to get red hat to uadit kolla but they told me to pound sand 20:42:14 don't want to call an audit what is really a threat analysis. Makes sense 20:42:18 but it has to be third party 20:42:32 so i can't get an audit 20:42:34 the issue was mainly that some people had a much more specific interpretation of the terms in that description than the author (me) did when writing it 20:42:36 but i can get a TA 20:42:42 alright, we have enough votes, I'll approve now 20:42:53 thanks folks :) 20:42:54 Thanks sdake for clarifying 20:43:05 in the meeting, it was noted that we probably don't actually have audits for existing vuln-managed projects... 20:43:17 jeblair that is correct 20:43:24 jeblair: it's a bit grandfathered in yes 20:43:39 security by precedent :) 20:43:49 wouldn't hurt to do a refresh tour :) 20:43:50 yes, still on the to-do list for the vmt is to evaluate existing covered projects to see if any need to be removed from support or shored up in some way 20:44:08 but seriously -- are we intentionally raising the bar for new projects to try to keep workload manageable? 20:44:10 moving on 20:44:11 or what fungi said :) 20:44:12 err 20:44:28 jeblair you mean lowering the bar - ta = bar lowering 20:44:41 I think ultimately the same rules will apply to everyone, it's just a manpower thing 20:44:52 sdake: i think he meant as opposed to holding existing projects to the same criteria we're requiring of new projects 20:44:55 at this stage it's mostly about not adding undue burden 20:45:01 sdake: point taken, but what i really meant was that we have set the bar higher for new projects than for existing ones 20:45:01 fungi oh got it 20:45:12 jeblair ya that sucks 20:45:13 is that intentional or is it just debt we need to pay off on the old projects 20:45:16 but it is what it is 20:45:21 absolutely debt 20:45:24 jeblair: it's not the only one. See activity requirement vs. the Packaging-deb project team 20:45:38 ok, I gotta go, sorry. 20:45:46 lifeless: we are mostly done. Thanks 20:45:53 fungi: cool, good to know. 20:45:58 i think that answers my question 20:46:03 thanks tc for your time on that change - tthat will be super helpful for new projects ;) 20:46:07 ttx: thanks for humoring me and not moving on. 20:46:38 jeblair: we should use slack so that I can edit my stupid "moving on"s 20:46:48 ttx: let's move on. :) 20:46:53 #topic Change mission statement by recent change 20:46:59 I failed at trollbaiting 20:47:00 ttx: you should totally propose that 20:47:12 you guys ain't fun 20:47:19 #link https://review.openstack.org/292610 20:47:25 * mordred hands ttx a box full of "fun" 20:47:27 This change proposes we align the mission in governance with the proposed mission from our recent resolution 20:47:35 This is slightly ahead of the game, since we'd like the board to approve the mission statement change first 20:47:45 As such I propose we defer this one and come back to it once the board decided 20:47:52 Which should hopefully happen next week, russellb ? 20:47:56 hopefully, yes 20:48:00 i asked for it to be on the agenda 20:48:05 well, if not, we can reevaluate our options 20:48:06 if the board can't agree to this, then ... 20:48:14 yes. 20:48:24 it's scheduled for 11:00! 20:48:38 or rather the newton TC members will reevaluate options 20:48:43 who has to approve this is a bit of a gray area anyway, isn't it? 20:48:47 11/00... PST ? 20:48:59 russellb: yes, not defined in governance 20:49:09 pdt i'd imagine 20:49:11 which is why it's simpler if we just all agree 20:49:11 * fungi can't remember if lake tahoe is pdt or mdt 20:49:27 ttx: yep 20:49:34 well let's see what happens next week 20:49:39 ok, so deferring until we hear back from the board meeting 20:49:40 fungi: it's p*t; m*t is the nevada/utah border 20:49:48 #info deferring this one until we hear back from the board meeting 20:50:02 #topic Open discussion 20:50:07 o/ 20:50:18 anteaya: yes? 20:50:30 will there be a tc/board meeting the sunday before summit? 20:50:32 oh, right, lake tahoe is in california. i keep thinking it's further east 20:50:47 anteaya: yes 20:50:48 fungi: most of it. some of it is in nevada. 20:50:58 russellb: do you know when ? Sunday ? 20:51:02 russellb: do we have any details, like time or location yet? 20:51:21 Joint TC/UC/Board April 24, 2:30pm - 5:30pm 20:51:28 ok 20:51:29 thank you 20:51:32 Does the TC have an initiative to drive all the projects towards py3 compatibility? I mean actually running functional tests on py3? 20:51:36 #info Joint TC/UC/Board April 24, 2:30pm - 5:30pm 20:51:37 https://wiki.openstack.org/wiki/Governance/Foundation/24Apr2016BoardMeeting 20:52:01 bswartz: we declared to support that effort, but haven't applied any coercion yet 20:52:10 I had one quick topic for open discussion 20:52:20 Design summit session moderator advice video 20:52:31 location: JW Marriott 20:52:31 With the big tent we have a lot of teams at the Design Summit, and sometimes the people moderating the session there have no idea what they are running into 20:52:41 Some of them might come to Design Summit 101 on Tuesday, but this is mostly geared toward new attendees, not moderators 20:52:52 So there was this idea to host a video panel / hangout / webinar with a few design summit old-timers 20:53:03 that would result in a video that we could send to newish session moderators and PTLs 20:53:14 Who would be interested in participating in that ? 20:53:32 I think sdague could be interested but he had to run 20:53:39 when are you creating the video? 20:53:42 That would likely be setup as Hangouts on Air (like the bootstrapping hour) so that the video gets automatically posted 20:53:47 sometimes over the next two weeks 20:53:57 thanks 20:53:58 ttx: great idea. 20:54:03 russellb: should that be 9-5 central time, not pacific? (wiki says pacific) 20:54:05 o/ 20:54:08 My video editing skills are about as bad as my web dev skillz, so be perfect 20:54:14 jroll: probably. 20:54:20 heh. k. 20:54:35 i'm terrible at moderating sessions, so mostly interested to see what others come up with in this area ;) 20:54:40 heh old-timers 20:54:46 I'll kick off a thread on openstack-tc to recruit, and if we don't have enough candidates I'll extend to long-time PTLs 20:54:58 ttx: I can participate if you want me to 20:55:07 annegentle: people that have seen a lot of design summits, I guess 20:55:16 annegentle: I know one names Anne 20:55:19 named* 20:55:19 ttx: :) 20:55:24 ttx: I would be happy to help with this 20:55:31 And another named Ed 20:56:02 alright, I'll move that to a thread on -tc and ccing external people who expressed interest like Ed 20:56:16 I can help but have to have a think (I don't often have controversy to tame) 20:56:53 honestly please please introduce yourself and don't assume everyone knows everyone would be my first tip :) 20:57:08 put anteaya and rockyg on the cc list if you would, please 20:57:42 #action ttx to start a ML thread on -tc about the video, adding edleafe, rockyg and anteaya to the cc. 20:58:01 Note that we'll likely keep the group relatively small so that the video doesn't end up being 30 min of mandatory watching 20:58:46 Or maybe we could all come up with a short 30sec advice and then record them all 20:58:47 the microphone is there for reminding people briefly at the start of the session that they should move to te front of the room... then you can turn it off ;) 20:59:08 making sure there are no duplicates 20:59:08 I did add the open discussion agenda of how we want to deal with stale specs http://lists.openstack.org/pipermail/openstack-dev/2016-March/089115.html 20:59:12 fungi: heh 20:59:13 * rockyg hands fungi a microphone 20:59:24 we can talk about it next week tho 20:59:30 since there is a minute left 20:59:33 oh, missed that. Recent addition ? 20:59:41 ttx: yeah sorry 20:59:52 it's not rude to tell people who aren't there to participate in the discussion/work that they should go somewhere else 20:59:54 yeah, next week I guess 21:00:16 Last words, anyone ? 21:00:23 food 21:00:38 Vote for the PTL elections if you still haven't! 21:00:45 #endmeeting