21:02:38 #startmeeting crossproject 21:02:40 Meeting started Tue May 5 21:02:38 2015 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:41 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:02:44 The meeting name has been set to 'crossproject' 21:02:56 Our agenda for today: 21:02:59 #link http://wiki.openstack.org/Meetings/CrossProjectMeeting 21:03:07 #topic Kilo release postmortem (ttx) 21:03:16 I'd like to start with a quick Kilo release postmortem 21:03:27 Overall I think we really needed the 6 weeks between feature freeze and final 21:03:36 ++ 21:03:36 We were slightly early on getting FFEs in, which was good 21:03:44 Then RC1s were generally on time 21:03:45 how many rcs did we end up spinning? 21:03:53 But we lost out safety-valve week with the stable/kilo requirements mess, which delayed the RC2s 21:04:02 And then we were a bit in a hurry on release week with late RC3s 21:04:09 \o 21:04:21 dhellmann: we had 3 RC3s 21:04:27 Nova, Neutron and Barbican 21:04:33 ok, that's what I remembered 21:04:53 we have a session for the summit to figure out the requirements management dance better for next time 21:04:54 (which raises another question onto how far coordination should reach in a bigtent world 21:05:10 We also have a cross-project session about release models 21:05:18 we also did some stuff earlier, like the upgrade testing before the release was out, which exposed at least one critical bug that forced some of the RC 21:05:38 sdague: on the good side, cutting stable/kilo pre-release sure made post-release calm 21:05:48 so, it was both pain, as well as catching a regression that we'd never have caught before 21:05:52 I hope we'll fix the process around requirements and the cutting of release branches before Liberty release and that we'll get it right this time 21:05:59 But in the mean time the Liberty release cycle maintains 6 weeks between FF and final 21:06:03 #link https://wiki.openstack.org/wiki/Liberty_Release_Schedule 21:06:20 yes, I think tightening up the acls on those branches during the pre-release period, and recruiting a few folks to help with the work, will make that go more smoothly next time 21:06:43 Personally I think that the stable branches on the libs/clients will most probably make the life much easier in the future 21:06:48 We need to find ways to distribute release management a bit more, the centralized thing doesn't work so well now 21:06:49 ttx: did we decide the library freeze a week before L3 was "official" enough to add to that schedule? 21:07:17 dhellmann: I'd be fine with that. I kind of wanted to wait until that design summit session about library releases 21:07:25 ah, right, I'll wait 21:07:34 Any other remark on the release process ? 21:08:03 * nikhil_k_ sneaks in 21:08:13 sarob: you around? 21:08:19 hello 21:08:26 oh I see you 21:08:29 #topic Project Work Group - Charter, Deliverables and Work Flow discussion 21:08:35 o/ 21:08:37 sarob: floor is yours 21:08:42 Hi all! Nice to meet most of you. 21:08:45 thanks 21:08:50 Greetings! 21:08:58 hi 21:09:03 anyone else from project wg? 21:09:14 Product WG,I think 21:09:14 okay lets start 21:09:21 oops 21:09:27 product, yes 21:10:01 so we started working on the idea that there are "hidden influencers" 21:10:07 Hey! 21:10:16 about a year plus ago 21:10:31 started as a side board meeting discussion 21:10:45 and it turned into the product team working group 21:11:00 we had a F2F about 5 months ago 21:11:16 and worked out what we were attempting to accomplish 21:11:43 its to create a multi-release roadmap 21:12:07 i spoke to the board some about the concept 21:12:44 so we became a "working group" like win the enterprise, application, and others 21:13:09 barrett: you want to give a bit too here? 21:13:18 sarob: Sure 21:13:47 The work group has drafted a mission and objectives 21:14:10 In Vancouver we're planning a couple of conference sessions and 2 working sessions. 21:14:38 We wanted to come to this group and share what we've developed and get your feedback to improve it 21:14:58 You can find this info at this link: https://docs.google.com/presentation/d/1pSvfn9IOMLw1KTawzlxSlu1R_bTXWmAjh0IRIifN4Jo/edit?usp=sharing 21:16:00 so these slide go into great detail 21:16:02 We'd also like to gain support for folks from this group to join our cross work group/project working session at the Summit, to help us refine our work flow and deliverables, and how we work together 21:16:05 slides that is 21:16:45 Slide 4 might be a good one to look at 21:16:57 we have a summarized proposed workflow https://docs.google.com/document/d/13JPDDiBGGXf5dtP0u8C-1So2Mjb3yEmGhv_ijVqyEf0/edit?usp=sharing 21:17:19 barrett: did you retain the idea of acting like a funnel to coalesce all the downstream stakeholders input and present it upstream ? 21:17:32 ttx: yes 21:17:37 o/ here now 21:17:57 because I know part of the difficulty of the PTL prioritization job is to have clear input to work from, and there are many special interest WG out there now :) 21:18:11 barrett: great 21:18:17 We are looking to connect with community groups at the summit and figure out how we work together. 21:18:41 product wg is meant to help with communication between different groups 21:19:38 Potentially, we could also play a role in the cross-project communication process... 21:19:41 Quick background info for PTLs listening: the product WG could reach out to various stakeholders (large deployment WG, Win the Enterprise WG, ops, Telco WG etc...) and try to extract key priorities so that we don't have to wade through a pile of disconnected data 21:20:10 and produce a unified report for us to include in our prioritization work 21:20:37 yup 21:20:51 said it better than i could have 21:21:05 is work already ongoing to do that for the ops meetup etherpads? 21:21:58 gathering information for PTLs? 21:22:00 PTLS: would such a unified report be useful input ? 21:22:06 We started data mining the previous Ops Meetup etherpads to identify use cases 21:22:46 sarob: yes - there are a bunch of etherpads with notes that sort of only make sense if you were in the room. It would be useful to have someone turn those into prose, in focused chunks. 21:23:00 In vancouver, we're working on a schedule to cover different sessions in the Ops Summit to gather more use cases and enlist Operators help in documenting them and checking them. 21:23:07 dhellmann: yup, that def part of it 21:23:09 so, dhellmann, yes. 21:23:22 ok, good! 21:23:28 the workflow is very similar to defcore 21:23:46 ttx: I think a report might be useful. I don't know. is there any feedback the other way? how do the PTLs figure out if implementing the stuff the product WG says has made any measurable impact on anything? 21:23:46 to get this to be part of the schedule 21:24:21 I get the feeling (so far) that this will be another voice asking for "stuff". there's no shortage of that now 21:24:23 sarob: I hope the turn-around time is quicker? :-) 21:24:42 notmyname: yeah, I think the idea here is to collect all of those voices so there's a bit more coherence 21:24:45 :x 21:25:00 notmyname: I hope that the product WG could reduce the noise rather than increase it 21:25:23 ttx: ++ 21:25:48 the workflow allows for the community to vote on use cases 21:25:59 with a deadline 21:26:02 who does the work in the end though? 21:26:17 then the PTLs get a crack at them 21:26:48 by distilling the asks to a readable format, it allows for better prioritzation for implementation 21:26:49 the PTLs would need to decide if the use cases make sense 21:26:50 PTLs != the boss of a project they cannot mandate things being done 21:27:02 they are the chief cat herder 21:27:04 jogo: developers in the community. Some members of the Product WG have the ability to influence and align the development priorities in our companies with the priorities of the community 21:27:35 barrett: +1 21:27:51 jogo: CCH 21:27:56 barrett: is there a breakdown on what companes are represented? 21:28:37 gordc: right now we are only reflecting back what 21:28:47 Yes - Though I don't think I can produce a complete one right now. But it includes: Intel, Huawei, EMC, Red Hat, Dell, 21:28:49 gordc: the PTLs have told us 21:28:53 Cisco 21:29:07 and Akanda! 21:29:08 geoffarnold: Thanks 21:29:17 ditto sarob 21:29:37 but the companies involved have not created use cases 21:29:47 yet, and thats the red mat 21:29:48 meat 21:29:51 sarob: ++ 21:30:05 Red Mat. I like it 21:30:06 we have only created the process 21:30:06 sarob: barrett: cool cool. i guess if all the politics is shielded from us, that'd be nice. :) 21:30:18 Our goal is to increase the number of companies participating and get a direct connection with Operators too - This is how the use cases will get developed 21:30:24 * sarob thinks in redmat 21:30:42 sarob: the red mat goes with the Red Hat, right? 21:31:13 i think its WWF sneaking in 21:31:26 sarob, barrett : it might be useful to have your input on this ML thread: http://lists.openstack.org/pipermail/openstack-dev/2015-May/thread.html#63065 21:32:11 jogo: I agree ptl is Cheif cat herder. 21:32:27 on the red mat 21:32:54 dhellmann: we've already been discussing this on our ml 21:32:58 and TC members are global cat herders 21:33:07 dhellmann - Who votes for TC members? 21:33:12 barrett: yes 21:33:14 barrett: the cats 21:33:15 * jogo wonders who cleans out the litter box 21:33:19 Rockyg: why not on the thread itself? 21:33:23 jogo: I assumed you did 21:33:41 ttx: yeah jogo and sdague 21:33:43 yeah, reading thread 21:33:57 barrett: that's what lets us be cat herders. The fact that cats elect us. 21:34:00 strong user committee, is part of this for sure 21:34:17 Well, I'm trying to get them to respond. I don't want to be the only one posting on the dev ml in the group ;-) 21:34:20 otherwise cats would just bite us. 21:34:25 ttx: gotcha 21:34:30 We're with you Rockyg. :-) 21:34:32 plus getting product minded people that are making openstack stuff to sell 21:34:45 we need to get them involved as well 21:35:03 so they are asking for features to be landed at the right time 21:35:09 I think the Product WG can support this by enlisting the hidden influencers in each company that provides development resources to the community 21:35:15 and not when they are freaking out 21:35:19 So the alternative is that every downstream group (win the enterprise, operators, nfv, etc.) winds up lobbying/negotiating with every PTL individually. Does this idea help the process? 21:35:26 dhellmann, re: why not on the thread itself : because some times groups need to discuss separately and form opinions as a group 21:35:41 barrett: +1 21:36:00 reed: I guess? It's not like my participation there is on behalf of the TC. 21:36:32 when is this WG planning on producing the first of these documents? 21:36:33 sarob: Any specific action you need out of this discussion ? We shall move on to next topics in ~5min 21:36:44 oh, yeah 21:37:05 so aside from general feedback on our direction 21:37:14 notmyname: we're planning to get a lot of work done at the summit, so shortly after? 21:37:19 the workflow is very actionable 21:37:27 reed: thanks ;-) 21:37:35 ttx: We'd like feedback on the model and agreement from some of the PTLs/TC to join our working session on 5/18 at 3:40 in room 212 21:37:45 sarob: I think we are all at the very least curious to see what you come up with :) 21:37:53 dhellmann: If I don't see a post by this afternoon, I'll do one on that thread. 21:38:03 ttx: i'd like to get feedback on the workflow time and actions 21:38:06 barrett: Woo, PTLs and TC members for 40min in summit week. Do you feel lucky ? 21:38:29 We don't need them *all*. Just the friendly ones. 21:38:33 ttx: Always! 21:38:36 barrett: yeah, that too! 21:38:36 ttx: I'm in Vegas so I do feel lucky. 21:38:40 Also I'm giving my talk at that time, so without me 21:38:52 and without sdague since he is giving the talk with me 21:39:12 thanks for hearing us out 21:39:14 If a couple of folks can join, we'll get a lot more done in that 40 mins then we'll get done in the next 4 weeks! 21:39:24 looking forward to what y'all come up with 21:39:54 Appreciate the time and discussion 21:39:58 Thanks for your time. 21:40:05 thanks 21:40:07 Thanks 21:40:18 sarob, barrett, Shamail: thx for taking the time to present here 21:40:24 * sarob steps off the redmat 21:40:27 #topic Design Summit scheduling 21:40:38 The Design Summit schedule is starting to be more detailed at: 21:40:43 #link https://libertydesignsummit.sched.org/ 21:40:53 About half of the tracks have pushed detailed descriptions at this point 21:41:03 Ideally we would have near-final content before the end of this week 21:41:16 So we can discuss last-minute conflicts at this same meeting next week 21:41:31 That said if you already see obvious scheduling conflicts, we can try to use this meeting to resolve them 21:42:06 Any you can see ? 21:42:29 It looks ok to me ATM. 21:42:37 I'm still seeing a discrepency between that schedule and the one on the main site 21:43:00 In other news, only a couple tracks made active use of the "Also appears in..." feature 21:43:14 if you see any session that could also appear in another track, let me know 21:43:16 ttx: what is the process for pushing the schedule up? 21:43:18 very cool feature. i love it 21:43:20 notmyname: where ? 21:43:28 * notmyname has used ops and zaqar 21:43:34 ttx: getting a link from the other one 21:43:36 stevebaker: you should have received an email from me on that 21:43:45 couple weeks ago 21:43:50 * ttx checks 21:44:12 ttx: found it 21:44:22 Subject: "Design Summit Scheduling -- Introducing Cheddar" 21:44:45 ttx: found it. aparantly it had changed titles. all is ok 21:45:14 notmyname: cool! 21:46:16 OK, so if you haven't pushed your schedule yet, please do so ASAP. We'll use the meeting next week to solve last-minute conflicts and to do that we need all posted 21:46:51 Questions on design summit ? 21:47:31 annegentle: I have a Design Summit 101 session at the start again. Would you be available ? Anyone else interested in moderating that ? 21:47:42 Not sure we'll have Loic around this time 21:47:54 ttx, I'm avaible, Loic told me he won't be in YVR 21:49:11 ok, if we have no questions or conflicts to solve, then... 21:49:12 #topic Open discussion & announcements 21:49:40 I'd like to ask people to sort and tag bugs as low-hanging-fruit 21:50:05 for use in mentoring ? upstream university ? 21:50:07 I'll be asking participants to Upstream Training to select bugs for them, simple ones 21:50:15 ack 21:50:20 so they can be trained on real life scenarios 21:50:36 there are very few lhf bugs at the moment 21:51:10 tag them and don't get tempted to solve them too :) they're quick fix but they're great to teach people about our processes 21:51:17 #info Please sort and tag bugs as low-hanging-fruit so that reed can use them in Upstream Training 21:51:48 I think some projects don't have low-hanging-fruit tag either 21:52:15 reed: anyone can set it, doesn't have to be an "official" tag 21:52:31 ok 21:53:10 Anything else, anyone ? 21:53:33 If anyone wants to mentor in UPstream Training, I would use more help :) 21:53:57 free drinks in YVR for those that sign up on wiki/Mentors :) 21:54:48 all week? 21:55:18 ttx: all the water you can drink 21:55:41 reed: i will be available on sunday 21:55:59 sarob, cool, get to play with legos 21:56:07 reed: first summit sunday in forever 21:56:14 reed: actually ever 21:56:51 indeed, Whoever comes on Sunday plays with Legos! 21:57:02 redmat? 21:57:24 :) 21:57:39 ok, time is up 21:57:57 Thanks for coming 21:58:00 #endmeeting