09:01:01 #startmeeting tc 09:01:02 Meeting started Tue Jul 3 09:01:01 2018 UTC and is due to finish in 60 minutes. The chair is ttx. Information about MeetBot at http://wiki.debian.org/MeetBot. 09:01:04 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 09:01:06 The meeting name has been set to 'tc' 09:01:07 #chair cdent 09:01:08 Current chairs: cdent ttx 09:04:20 * ttx copies fragility analysis top results to the health tracker for further analysis 09:05:37 Oddly enough I am awake at the moment... 09:05:38 Looks like my previous message didn't make it through: "My main topic of concern recently is how to make sure that we extract something useful from, and actually do something with, the various threads of discussion on adapting openstack to "now"" 09:06:45 This is not still jetlag is it TheJulia ? 09:06:53 Unsure 09:07:52 good luck? 09:08:10 my sleep sadness story is Hayfever 09:08:45 It might be the same actually, no rain to knock allergens from the air 09:11:10 I guess one thing i wonder is given "now", how do we also foster an environme t where those discussions take place outside of our circle with effect and outcome. 09:11:41 yes, that's an excellent point 09:12:36 It's something I've really been struggling with a lot lately. Feeling very en-bubbled, but not sure how to get out given the constraints of existing obligations. 09:13:48 Trust with ahared purpoae is one way, but if you trust one person and their corporate overlords change their focus... There is not much that can be done. 09:14:11 Shared purpose 09:15:20 So setting the stage or environmenr where rhose discussions of directions taken are the resulting outcomes of our tuning, not our direct actions. If that makes sense. 09:17:29 In my mind that translates to getting people to believe in something (the shared purpose) 09:18:08 To me, getting the existing openstack community to believe in a shared purpose would harden the walls of the bubble. 09:18:48 Rather, I suspect new, fresh input is needed from a wider group that the current smaller group needs to also consider part of "us": those folk probably have new/different purposes 09:18:56 perhaps people != existing openstack community 09:19:06 (but me calling them "those folk" is an example of the problem, really) 09:19:12 cdent: what about shared context instead of purpose 09:20:01 I guess it depends on what you mean by those terms but in my head the first is a requirement for the second, but only the second is something that results in action 09:21:20 that is one people have a shared context they may be able to recognize a shared purpose and choose to collaborate 09:21:25 s/one/once/ 09:22:33 In another context, I'm currently dealing with a small project (tens of developers) that has started to suffer from velocity issues, both in terms of review bandwidth and in terms of progress on shared features (as interest groups start to choose to have "feature branches" as operational forks). This isn't the same problem as OpenStack, but it feels similar in some ways. 09:23:25 One of the things we're doing there is removing the authority of those who have previously been responsible for guiding the main "upstream" direction, and instead just merging everything that passes CI, with the obligation on the interest groups to write tests for anything they care about. 09:24:00 This has caused a number of interesting flame wars, but one of the most interesting aspects to me is that those whose authority was removed seem to be more active in reviews than before their authority was removed. 09:24:32 TheJulia: my feelings these days is that more us need to take a more active role in nudging what the shared purpose might be, by sharing (random) ideas. Which half the time sounds like so much noise, but if any of them land, that's good. 09:24:33 persia: that is interesting 09:24:46 Also, the teams that had been slowly moving towards forks seem to be more active about getting tests for their use cases into "upstream", and getting their features landed. 09:24:57 persia: was it that they found new purpoae in trying to assert authority in code review? 09:24:59 It's early days with the transition, but it seems to be working so far. 09:25:27 TheJulia: No, that those responsible for their time allocation suddenly saw a lot more value in them preemptively reviewing things :) 09:26:17 I see 09:27:22 ttx: I think we need to start thinking in terms of optionally breaking existing users. Things like "people can choose to use nova-compute or if they want they can choose to upgrade to nodelet the all-in-one compute node agent" 09:27:29 So i do agree with you cdent in thay we all, community collectively need to better share our xontext, not purpose. With that in the open, it hopefully becomes easier to negoriate and plan 09:27:31 While the details differ, I think the audience is similar to the "now" issue: the important bit is making sure that things happen visibly to an audience that is able to allocate the resources to accomplish the desired goals. Just getting all the developers excited about things, with shared context and purpose, may not allow progress. 09:27:40 ok, I dropped nuggets of fragility facts onto the Health tracker for TC liaisons to further investigate 09:28:47 ttx: if we don't provide options for "different from how it always was" there's little to perceive as change 09:28:53 * cdent looks at tracker 09:29:28 using data from https://etherpad.openstack.org/p/rocky-2-bus-factor 09:29:48 (which was generated by the script at https://review.openstack.org/#/c/579142/ ) 09:30:00 I'm not sure excited is the desirable target, in the grand state machine of the universe. Excited implies that we convinced them on shared vision instead of ahared context where we understand each other. 09:34:48 * cdent smells brake dust 09:35:13 and there's ringing in my ears from the way that conversation seemed to come to a screeching halt 09:35:17 * cdent sighs 09:36:47 sorry, was still dealing with wiki edits 09:37:53 Also about to propose diversity tag removals -- how well do you think that will fly ? 09:38:07 cdent: sorry, i didnt mean to hit the breaks 09:38:07 cdent: random changes is probably more confusing than useful. An unfiltered firehouse of proposals from all and sundry may provoke better debate. 09:39:23 TheJulia: the implication was unintentional. I think it would be good if participants were excited. I think it would be better if that was organic, rather than directed. 09:41:00 persia: i do agree tbay excitement mighr lead there, but the key is to be excited about the same thing. 09:41:37 cdent: On the "optionally breaking existing users" it should definitely be something that is on the table... Although in that specific case i see no reason why the nodelet would not be a natural evolution of nova-compute :) 09:42:08 ttx: I'm +many on diversity tag removals and I think many people feel the same way: they are at best not meaningful, and at worst damaging 09:42:24 (taken over by a nodeSIG-like entity) 09:42:38 right "natural evolution of nova-compute" is too slow. that's why I think there needs to be concurrent advancement 09:43:23 persia: not random changes. random sharing of ideas from which to discover useful change 09:50:36 Ah, yes, that makes more sense 09:51:14 In other, potentially less controversial news, I updated a story on storyboard related to communication notes in the project team guide to add a few more detailed tasks of things to clarify. There were fewer issues than I feared 09:51:18 #link https://storyboard.openstack.org/#!/story/2001710 09:53:30 lgtm -- I did a pass on currentness on that guide but missed the weekly cross-project meeting mention 10:01:25 we've reached the end of the hour and seem fizzled out so... 10:01:27 #endmeeting