Tuesday, 2018-03-20

*** rosmaita has quit IRC00:32
*** kumarmn_ has joined #openstack-tc01:04
*** kumarmn has quit IRC01:06
*** annabelleB has quit IRC01:11
*** liujiong has joined #openstack-tc02:57
*** kumarmn has joined #openstack-tc03:05
*** kumarmn__ has joined #openstack-tc03:07
*** kumarmn_ has quit IRC03:07
*** kumarmn has quit IRC03:09
*** liujiong has quit IRC04:22
*** liujiong has joined #openstack-tc04:23
*** liujiong has quit IRC04:58
*** liujiong has joined #openstack-tc04:59
*** kumarmn__ has quit IRC05:38
*** kumarmn has joined #openstack-tc05:47
*** liujiong has quit IRC07:57
ttxTracker updated08:46
*** openstackgerrit has joined #openstack-tc08:46
openstackgerritMerged openstack/governance master: Add config_template to OpenStack-Ansible  https://review.openstack.org/54667908:46
openstackgerritMerged openstack/governance master: Remove ironic-inspector-tempest-plugin  https://review.openstack.org/55053808:46
openstackgerritMerged openstack/governance master: Remove Zuul from OpenStack governance  https://review.openstack.org/55263708:46
*** cdent has joined #openstack-tc08:50
cdenttc-members it is time to engage in office hours, let's hope we have visitors from near and far09:00
ttxI'll be grabbing coffee09:00
ttxslow start this morning09:00
cdentme too09:01
cdentI have a comment/question as SIG leader, related to executive power in the sigs. there's a subthread on a recent api-sig weekly posting about working on api schema. Gilles eventually comes round to hoping that the SIG can more actively drive things: (he starts things here http://lists.openstack.org/pipermail/openstack-dev/2018-March/128414.html then I respond, and then he responds this morning)09:04
cdentI'm asking about it because it is an example of the general case of needing/wanting to do multi-project change in the absence of any one "authoritative" asserting the change is right09:05
cdentWe see that a lot.09:05
cdentI think it is probably just the nature of our beast, but...09:06
ttxThey obviously have limited power beyond the respect that their members trigger in the general population09:07
*** jpich has joined #openstack-tc09:07
ttxThey could derive authority from some delegation of power from the TC09:07
ttxbut not by just being a SIG09:07
ttxand we've found softer suggestions (like release goals) to work better than top-down mandates09:08
cdentwell in this specific case he's not really got support of the SIG either, but he's hoping that by convincing us we can then convince the world, and I'm trying to say "convince the world and we will be convinced"09:08
ttxyes, I agree with your position there09:09
cdentBut (as usual), I'm intrigued by the underlying pattern here, as it related to getting stuff done.09:10
ttxIt is (as usual) quite subtle.09:10
ttxThe way I framed it in teh past is that you amass influence tokens by helping others and making yourself useful to the project09:11
johnthetubaguyits a community dynamic that the PWG were finding hard to understand I think09:11
ttxThen you can spend some of those influence tokens by weighing on decisions in ML threads and/or reviews09:11
ttxand asking people to do things09:12
johnthetubaguycdent: every time I reflect on this stuff I seem to decide its our biggest advantage and biggest disadvantage at the same time09:12
cdentjohnthetubaguy: yeah, exactly09:12
johnthetubaguynot sure if that cancels each other out :p09:12
ttxThe title/position is really not the key. Like TC members can get others to do things more because they are respected individual members of the community than because they were elected09:13
johnthetubaguyI like that bind man seeing the elephant story, that makes me happier about it09:13
ttxwhich I think is what some people overlook09:13
johnthetubaguyI mean the PTL had trust that got them elected, its not really the other way around09:13
ttxAlso you need to periodically make yourself useful again, or you will run out of influence tokens09:14
johnthetubaguyso that elephant thing, its really easy to make the tree trunk you want to build and ignore everyone, but you get unstuck when you realised you needed the full elephant after all09:14
ttx"Be the tree trunk you want to see the elephant of"09:15
* johnthetubaguy thinks ttx has gone very yoda09:15
ttxor is it the other way around09:16
johnthetubaguyyeah, works both ways09:16
ttxRemember your failure at the cave09:16
* ttx creates an etherpad for brainstorming forum sessions09:17
ttx(TC edition)09:17
johnthetubaguycdent: I think in there is something about the key to developer happiness, tricky though.09:18
cdentjohnthetubaguy: can you expand on that a bit?09:19
johnthetubaguywell, the collaboration can be painful and frustrating at times, because people are people09:20
johnthetubaguybut at the other side of it, you do generally get something better and more widely useful, which is rewarding09:20
johnthetubaguywhen you don't, well that is depressing09:20
ttxIt's a two-edged sword. When you want something done elsewhere you kind of like the idea of being able to get other to do things. But when you receive such directives you don't like it much09:21
johnthetubaguyadd in that its a set up in the number of people involved that many folks are used to, with no hierarchy to "enforce" things, there is no real shortcut09:21
johnthetubaguyttx: yeah09:21
ttxWhich is why I prefer a more personal model: make yourself useful to me and I'll likely be happy to be useful to you09:21
ttxIn the ProductWG case they were not very useful to most.09:22
ttxwhich created an imbalance, like why should I care09:22
cdentand here I thought it was because they were suits ;)09:23
johnthetubaguyI tried to say about how early on in the design summit we used to have product managers chip in and say, well that sounds fun but the users are asking for X not Y. ohh... good point, lets flip the conversation, etc.09:24
johnthetubaguynot saying that is easy, but it would have been useful09:25
ttxtrying to find the etherpads we used for session brainstorming in the past, but failing in search foo09:26
cdentOne of the standout sentences in Gilles message is where he says (paraphrase) "I represent a lot of people, most of them are too busy to come upstream"09:26
cdent(in that context he means dev-oriented people)09:27
cdentand I want to write back "they need to come upstream"09:27
* johnthetubaguy nods09:28
johnthetubaguynow there is a truth there, we need it to be easier for non-full time upstream people to engage upstream09:28
* ttx just accidentally fell on https://openstack.nimeyo.com/122095/openstack-dev-forum-sydney-forum-etherpad-list09:28
ttxlooks like a stackoverflow thing built on top of email archives09:29
johnthetubaguyoh my, thats nuts09:29
ttxI'm intrigued09:30
ttxIt's broken, it has dhellmann as #1. Ogh, wait09:31
ttxso it fails a bit in lists where you post statements rather than questions09:34
ttxLike the above link, where it assumes I'm posting a question. while I was posting a link09:34
ttxIf you click on a tag it derives experts for that area09:36
ttxAnyway, enough yak shaving09:38
ttxFeel free to add your suggestions there ^09:38
ttxso we coordinate duplicates and missing sessions before the session submission site opens09:39
ttxI added the two I said I would add. Will post link to ML now09:44
cmurphyyvh? not yvr?09:48
ttxI blame cph09:50
ttxlet me fix it09:50
cdentAre there dates set yet for the next round of TC elections? We talked in Dublin about making sure we contact people and encourage them. It's probably about time for that?09:54
ttxthere is a change proposing dates up for review, let me find that09:55
* cdent blinks and checks his gerrit settings09:55
ttxIt's an election repo change09:55
cdentYeah, I thought I was wathing that one, but apparently I stopped09:55
ttxit gets quite of noisy around ptl election time09:55
cdentHmm. So if the election is over on the 30th, that's only 20 days after the result before summit. There may be candidates (I was one) who will not otherwise be going to summit (because of weird company behavior[1]) who, if elected, will be incurring a cost for late tickets at that point.09:58
cdent[1] no presentation, can't go; not a ptl or tc, can't go. etc.09:59
cdentI was in this boat last year, but I ended up getting myself funded on the travel support plan (for api-sig stuff).09:59
ttxElection needs to happen around summit-3 week per the charter09:59
cdentMaybe we should consider changing that prior to Berlin?10:00
ttxcdent: you should have gotten a pass for free ticket due to PTG attendance, no?10:00
cdentthat's the tiniest weeniest aspect of the cost10:01
ttxoh, I see what you mean10:01
cdentlate plane tickets10:01
ttxyes i guess we could place it a bit earlier (still between PTG and Summit, but closer to PTG?)10:02
cdenteven -5 or -6 would probably be an improvement10:02
ttxArguably a bit late to change for this one, but we can have something ready before next round10:03
cdentYeah, that's what I'm thinking. I can propose something, if that's a reasonable way to go?10:04
cdentBut back to the original thrust of my question: "We talked in Dublin about making sure we contact people and encourage them. It's probably about time for that?" Is that in people's heads?10:05
ttxI thought we'd do that closer to the nomination period. We also said we would encourage everyone to reach out to people they would consider making good candidates10:08
* ttx checks dates10:09
ttxso nomination is likely to start April 10, I would wait one more week before reaching out10:09
johnthetubaguytoo soon and folks forget I guess10:16
cdentI think that underestimates the amount of negotiation one has to do with oneself and one's employer10:17
cmurphyagreed ^ takes a while to talk yourself into it10:21
ttxAlright, let's do it then :)10:32
openstackgerritJean-Philippe Evrard proposed openstack/governance master: Add systemd roles to OpenStack-Ansible  https://review.openstack.org/55286611:03
*** openstackgerrit has quit IRC12:33
*** kumarmn has quit IRC13:09
*** rosmaita has joined #openstack-tc13:09
*** mriedem has joined #openstack-tc13:18
*** kumarmn has joined #openstack-tc13:29
*** kumarmn has quit IRC13:30
*** hongbin has joined #openstack-tc13:30
*** kumarmn has joined #openstack-tc13:30
*** annabelleB has joined #openstack-tc13:46
*** SamYaple has quit IRC13:52
ttxTC stories added to https://storyboard.openstack.org/#!/project/92313:53
smcginnisHmm. So in lp you can click on a "Bug affects me" link and get notifications of updates. Do we have any kind of equivalent in storyboard?13:55
cdentsmcginnis: thanks for asking that question, I keep having it, and then forgetting that I have it13:56
ttxin Profile there is a "enable notification emails" checkbox13:56
smcginniscdent: That same pattern happens to me almost daily.13:56
cdentWe. Are. Professionals.13:56
smcginnisOh right, once you log in it calls forget_all_history_so_you_can't_hit_back()13:57
smcginnisOK, so once that's enabled, you can "star" a story to subscribe to it. I like it.13:58
*** SamYaple has joined #openstack-tc14:11
* fungi catches up on office hours scrollback, remembers to count out the influence token loose change in his pocket and see if he still has enough for coffee14:17
fungii've previously had to ask nimeyo.com to disable their "feature" which allows their users to "post" messages to those forums (a.k.a. our mailing lists)14:21
fungiand to remove our logo from it14:21
fungiattractive nuisance14:21
fungii think i also had to temporarily force unsubscribe them from some of our lists and block messages from their domain in the mta before they'd notice and respond to the messages i sent them politely requesting they cease14:25
*** annabelleB has quit IRC14:42
*** annabelleB has joined #openstack-tc14:50
*** dtantsur|afk is now known as dtantsur15:00
*** amrith_ has quit IRC17:12
*** amrith has joined #openstack-tc17:13
*** amrith is now known as Guest4626417:13
*** annabelleB has quit IRC17:18
*** annabelleB has joined #openstack-tc17:19
*** kumarmn has quit IRC17:28
*** kumarmn has joined #openstack-tc17:34
*** kumarmn has quit IRC17:38
*** jpich has quit IRC17:50
*** Guest46264 has quit IRC18:10
*** amrith has joined #openstack-tc18:11
*** amrith is now known as Guest5626218:11
cdentwell said fungi18:12
* fungi suspects cdent is crossing the streams18:13
cdentfungi: your email response on making ptg for all contributors18:14
fungithanks! i figured that's what you meant18:14
*** dtantsur is now known as dtantsur|afk18:16
*** diablo_rojo has quit IRC18:22
*** harlowja has joined #openstack-tc18:23
*** Guest56262 is now known as amrith18:30
*** diablo_rojo has joined #openstack-tc18:37
*** kumarmn has joined #openstack-tc18:46
cmurphy"separate branding" would be :(18:51
*** kumarmn has quit IRC18:51
cdenttotally :(18:54
smcginnisMy only concern I've raised is I want to make sure there are distinct tracks or areas so the good things from the gathering and the good things from the ops meetup don't just get mashed together and lose some value.18:54
cdentsmcginnis: also a valid concern18:54
smcginnisWhile still encouraging breaking down the artificial wall between the two.18:54
smcginnisI think even if there's just the chance at the end of the day to have casual conversations over drinks, that will be a big win.18:55
cdentdoes anyone recall where the patches propoesed per time period per project numbers that ttx tends to report every now and again live?18:56
mugsiecdent: I think ttx has local tooling that he runs to get them18:58
cdentthanks mugsie, for the time being I'm use old unreliable stackalytics18:58
mugsiegerrit *might* give some of the same detail, but I am not sure18:59
smcginnisYeah, I had the impression it was some local form of https://github.com/openstack/third-party-ci-tools/tree/master/monitoring/lastcomment-scoreboard19:00
dhellmanncdent , mugsie : there are some tools in the reviewstats repo, but I don't know how up to date they are19:00
cdentdhellmann: thanks19:00
cdentstackalytics is working well enough for now, but I ... don't trust it19:00
dhellmannthat's openstack-infra/reviewstats19:00
cdentor perhaps more accurately: don't want to trust it19:01
smcginnisYeah, you could probably adapt https://github.com/openstack-infra/reviewstats/blob/master/reviewstats/cmd/reviewers.py19:01
dhellmannyeah, it doesn't seem to think anyone has sent email for the rocky cycle and I'm pretty sure that's wrong ;-)19:01
cmurphywe're well known for our quiet mailing list19:02
mugsieI don't know, I have always found -dev to be a source of calm and inner peace :)19:12
* dhellmann wants to join whatever ML mugsie is subscribed to19:13
*** diablo_rojo has quit IRC19:44
*** kumarmn has joined #openstack-tc19:44
*** annabelleB has quit IRC19:58
*** kumarmn has quit IRC20:03
*** kumarmn has joined #openstack-tc20:03
*** kumarmn has quit IRC20:08
*** kumarmn has joined #openstack-tc20:24
*** kumarmn has quit IRC20:28
*** annabelleB has joined #openstack-tc21:15
*** diablo_rojo has joined #openstack-tc21:21
cdentsome weeks are easier than others for writing the tc report21:22
smcginniscdent: Too much material? Not enough? Too hard to decide how opinionated you want it to be?21:24
cdentsmcginnis: I'm not really sure. I think it has more to do with fatigue and thus the effort involved with re-experiencing things (so as to report them) is extra meh.21:32
smcginnisAh, I can totally understand that. Sometimes discussions are best had and moved on from.21:33
cdentsmcginnis: I suspect that force is especially strong this week because a lof the last week was comunicating about how communication often doesn't work and that when we communicate about that not working it often doesn't work and now I need to write about that, yet again21:38
cdentand repeat21:38
cdents/lof/lot of/21:39
smcginnisThat's a lot of communicating.21:39
cdentwhich ought to be good, but if its not bringing about any change...21:39
smcginniscdent: Maybe not a bad thing to include some of that feeling in your update.21:39
cdentindeed, I'll shall communicate about this communication about communication just now in my comunication :)21:40
fungithe official stats gathering tools for affiliation diversity live in the governance repo. as for general contribution stats gathering there's a slippery slope where if you publish the algorithms you're using (or even focus on the sme metrics from one time to the next) there are people who will pollute your statistics by reverse-engineering and gaming them22:13
fungiso they become increasingly meaningless over time22:13
*** diablo_rojo has quit IRC22:13
*** diablo_rojo has joined #openstack-tc22:14
smcginnisVery good point.22:15
persiaAlso, having algorithms in a known place gives insider advantage to folk who investigate such things22:16
persiaIn that someone who has once been active in OpenStack governance (or at least looked at the repo) is much more likely to be able to cause another to be perceived as a contributor than someone who just encountered www.openstack.org for the first time.22:17
*** ianychoi__ is now known as ianychoi22:32
*** hongbin has quit IRC22:44
*** harlowja has quit IRC22:57
cdentwell smcginnis, I may have gone off the rails a bit. doing light proofing and then will publish23:06
*** mriedem has quit IRC23:34
*** harlowja has joined #openstack-tc23:49
*** cdent has quit IRC23:50
*** mriedem has joined #openstack-tc23:52

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!