Thursday, 2017-10-19

*** hyakuhei has joined #openstack-tc00:12
*** flwang1 has quit IRC00:41
*** emagana has quit IRC00:58
*** flwang1 has joined #openstack-tc01:07
amrithdtroyer ... this is the way it has been forever, prospective TC members who pay attention (ahem) know about this :)01:37
*** openstackgerrit has joined #openstack-tc01:45
openstackgerritEric Kao proposed openstack/governance master: Add congress planning and completion artifacts  https://review.openstack.org/51321401:45
openstackgerritEric Kao proposed openstack/governance master: Add congress planning artifact  https://review.openstack.org/51321601:48
*** kumarmn has joined #openstack-tc02:44
*** rosmaita has quit IRC02:55
*** kumarmn has quit IRC03:03
*** kumarmn has joined #openstack-tc04:02
*** emagana has joined #openstack-tc04:15
*** kumarmn has quit IRC04:22
*** alex_xu has quit IRC05:54
*** alex_xu has joined #openstack-tc06:00
*** emagana has quit IRC06:17
*** kumarmn has joined #openstack-tc06:22
*** kumarmn has quit IRC06:27
*** jpich has joined #openstack-tc06:59
*** gcb has quit IRC07:55
*** gcb has joined #openstack-tc08:28
*** dtantsur|afk is now known as dtantsur08:32
*** cdent has joined #openstack-tc08:56
*** sdague has joined #openstack-tc09:38
openstackgerritMerged openstack/governance master: Add senlin-tempest-plugin to project list  https://review.openstack.org/50655009:42
*** cdent has quit IRC09:53
*** cdent has joined #openstack-tc10:31
*** cdent has quit IRC12:10
*** cdent has joined #openstack-tc12:10
*** cdent has quit IRC12:17
*** dtantsur is now known as dtantsur|brb12:26
smcginnisamrith: ;)12:27
*** rosmaita has joined #openstack-tc12:30
*** cdent has joined #openstack-tc13:02
*** kumarmn has joined #openstack-tc13:10
*** rosmaita has quit IRC13:39
*** rosmaita has joined #openstack-tc13:42
*** hongbin has joined #openstack-tc14:03
*** lbragstad has joined #openstack-tc14:04
cdenttc-members, looks like it is office hours time15:01
fungiyay! so it is15:01
dhellmanno/15:01
* fungi is... sort of around15:01
EmilienMo/15:01
cdentsuch enthusiasm, I’ll have what you’re having.15:01
fungihere, have some zuul15:01
* dhellmann hands cdent the broken tag-releases job15:02
EmilienMoops15:02
* cdent falls over15:02
smcginnis:)15:03
dhellmannwhat are folks thinking about pre-summit/forum?15:04
dhellmannwhat issue is top of your mind for discussions that week?15:04
cdentheh, I wish I had more headspace to think about it15:05
cdentso I would guess it is the same issue as ever: how to ensure people have headspace15:05
fungii'll likely focus on a lot of the community track sessions at the summit again, hopefully get to most of them this time15:05
EmilienMwhile responding to some [tc] emails this week, I thought about proposing a change in governance to force some sort of rotations at the TC (not sure about the frequency)15:05
EmilienMbut before doing that, I want to talk with humans about this one :)15:06
dhellmannI went through the schedule yesterday and ended up with a scattered set of topics.15:06
dhellmannEmilienM : I am opposed to "term limits" on principle.15:07
johnthetubaguypart of me is wondering how do we ensure we don't exclude all the folks that can't make it (including me), we have got better at PTG summary emails15:07
cmurphyEmilienM: I think it would be a shame if the wisest/most experienced members were forced out of their seats15:08
dhellmannit seems like in boston we did a pretty good job of setting expectations that sessions with actionable outcomes needed to be summarized on the ML by the moderator (or a designated person).15:08
smcginnisEmilienM: Were you thinking term limits? Or some other idea?15:08
*** dtantsur|brb is now known as dtantsur15:08
fungidebian revisited term limits for their tc a few years ago. worth reading some of the discussions and drawing parallels: https://lwn.net/Articles/623851/15:09
fungiworth noting, this term we have at most 50% incumbents reelected to open seats15:09
dhellmannfungi : nice, thanks for that link15:09
fungipotentially fewer15:09
fungiso it doesn't seem like term limits solve a problem we have15:09
mugsiefungi: this is a very new trend though15:10
cdentmugsie: very true15:10
cdentand importantly, it is very rare for an incumbent not to retain their seat. whether this is good or bad is unclear.15:11
mugsieterm limits have a use. Personally I am against life time term limits, but can be pursuaded to support sequential term limits15:11
cdentso it is often the case that new blood only comes in when someone chooses to step aside15:11
fungii wouldn't say "very" rare, there have been at least a few15:11
smcginnisThe only thing I think might be ok is saying no more than 50% of an election can be incumbents, but that's got all kinds of issues too.15:11
cdent“very” is a very subjective term, verily15:11
mugsiewell, one of the leadership goals was to grow new leaders, wasn't it?15:11
fungiyes, growing new leaders requires some mentorship in those roles, so swapping out experienced tc members too frequently can leave newcomers unprepared15:12
mugsiewhich is hard, when the people at the "top" (for lack of a better term) don't change very much15:12
mugsiefungi: sure, but they have to see there is a path at all15:13
fungii suppose it would help to have statistics on how many incumbents are elected each term15:13
cdentEmilienM: you’re not suggesting that term limit be one term are you? If not, then I’m not sure the experience thing is a huge (verily) issue?15:14
cdentAlot of it is perception too, not just the real numbers. Sean and Monty having the grace to step away opens up a huge sense of opportunity.15:14
fungii mean, all the incumbents were new tc members at some point (for example, i'm an incumbent this term, but was not last term)15:14
* cdent is undecided15:15
* EmilienM back in 10 min15:15
cdentI think term limits is an action which may not be directly tied to a problem, or the problem may be insufficiently defined.15:15
EmilienMok I'm back sorry for that15:17
* EmilienM reads15:17
EmilienMcmurphy: this isn't the goal. I guess the goal is to ensure we have new people (with new ideas) joining the TC at every cycle15:19
dhellmannI was surprised at how many people ran this term. I really didn't know (or have any way to know) who would be allowed by their companies to serve. Some of the folks running contribute a *lot* but not necessarily in ways that obviously translate (for me) to the sort of things the TC gets into, so it wasn't clear there were people *interested* in running either.15:19
EmilienMfungi: I agree, a lot of new people ran this term, which is very cool15:19
dhellmannif that set of people remains engaged, I predict more turnover15:20
cdentI wonder if we need a GC (governance committee) and a TC (technical committee). The GC woud be as now, with no term limits and is the judiciary that the current TC mostly behaves as. And the new TC would be the thing that many people how are new to the TC wish it could be: a _technical_ committee that drives change/improvement/innovation/etc15:20
EmilienMfungi: and https://lwn.net/Articles/623851/ sounds like a nice reading15:20
cdents/how are/who are/15:21
smcginniscdent: That's not a horrible idea.15:21
cdent(not saying that idea is realistic, simply throwing around ideas)15:21
cdentsmcginnis: it’s entirely contrary to the way projects are currently constituted as autonomous15:21
smcginniscdent: Yes, but it also addresses what some expect the TC to be, which it is not.15:22
fungii do agree that the people who are on the tc aren't there just to be there, and many would probably be happy to move on to other things _if_ they were confident there were enough good new candidates that they didn't need to worry about things falling apart15:22
fungibut hard to know until the start of the election that will be the case, unless there's an established history of more people volunteering to do the work15:23
EmilienMI guess the whole intent of my question is, how can we create new leaders in our community, who promote our values that are reflected in the work we do at TC. So how we ensure experimented TC members train candidates (in different ways) so we ensure regular rotations15:25
EmilienMI'm not interesting in kicking out people because they are here forever. In fact it's the opposite: how can we make more experimented people, and transfer knowledge & experiences15:26
EmilienM(aka stewardship)15:26
fungithat doesn't seem like something enforced term limits would cause to happen15:27
dhellmannfungi : yes, that's a much better way of saying what I was trying to say15:28
cdentIf the champions top 5 entry gets shaped appropriately, I think that can be a big factor: people are able to find the time to volunteer “to do the work” and are shepherded by existing tc folk…15:28
dhellmanncdent : yes, that will be a good source15:29
EmilienMindeed15:29
cdentBut people have to have the license. We’ve only got mugsie on the candidates list because he’s got a new employer, right?15:29
dhellmannI would like for us to find other ways to delegate things like that. Have a TC member be responsible, but have someone not on the TC lead the work or committee or whatever.15:29
mugsiecdent: yeah - my previous job would not have been happy if I ran15:30
dhellmannthat has been a problem in the past, yes, and I think quite a few of our candidates this term are in that situation15:30
dhellmannmugsie : where are you now?15:30
mugsieVerizon15:30
dhellmannah, good15:30
* dtroyer sneaks in late15:30
dtroyerEmilienM: I remember seeing your initial thought on th ML, was there something specific that triggered it for you?15:31
EmilienMdtroyer: not at all, my only goal was to initiate the idea and see how bad it is15:32
mugsiefungi: if you knew you had at most 3 years per stint on the TC, would succession planning not be on your mind?15:32
EmilienMdtroyer: it exists in some other communities fwiw15:32
cdentthe flip side, is that if we are encouraging people to say “the tc thinks that stewardship work is an important part of participation in an open source community, please give me some time for that” to their employers, then there also needs to be fairly clear outcomes and value from that participation15:32
mugsie(3 years is a random number)15:32
dtroyerEmilienM: ok, good.  I was hoping it wasn't because of a specific event or comment15:33
fungione of the concerns raised in the debian debate was... consider if we already had term limits now. three former tc members stepped down and some number were _also_ forced to resign due to reaching the term limit. how much tc turnover can we safely absorb in one go?15:34
persiacdent: Generally, the benefit is soft, in that it increases influence of the person doing the stewardship.  This is not always of interest to employers, as the benefits often do not reflect on the employer, but rather increase the market worth of the participant.15:34
dhellmannyeah, continuity is definitely a concern15:34
cdentpersia: yeah, I know, that’s a lot of why I’m raising the point15:34
dhellmanncdent : do you think the top 5 list helps there?15:35
dtroyerfungi: good point.  our rate of change is slow (we are a re-election heavy community), but it does change15:35
cdentdhellmann: I think it helps gets things started, but it opens up an expectation15:35
dhellmannwe have the split term thing in place to help with continuity, but we have a *lot* of history for people to absorb, too15:35
persiaSome continuity is guaranteed by one-year terms with six-monthly elections.  Other continuity can be supported by having folk remain active in stewardship and mentoring when leaving the TC (if they have time).  [I am against term limits]15:35
dtroyercdent: I do know in my case I was hired _before_ i decided to run for the TC but they were _really_ happy that I did.  Some companies still see that as a status thing, although that is less than it used to be15:36
* dtroyer is slightly behind15:36
fungibasically, i worry that term limits would have the opposite effect intended... by forcing out some number who might have been willing to serve again, it puts pressure on others who may have been wanting to resign but feel compelled to stay on because they have not reached the forced limit15:36
fungiso as to avoid too much discontinuity in leadership15:37
dhellmannred hat doesn't see it as a "status" thing and I would be expected to do a lot of the same work I'm already doing if I was not on the TC. My position may be a bit unusual, though.15:37
EmilienMdtroyer: no it wasn't on a specific event or comment, at all15:38
dtroyerI see a lot less of the status talk, especially with PTLs as there are now so many of them  When there were only 8-10 and most of them worked in 2 or 3 companies it was more pronounced15:39
persiadhellmann: I think your situation will become more normal as contributing organisations mature in their interaction with open source communities.  Presumably, by being on the TC, some of your other activities are easier (or you have more deposited in the favour bank).15:39
dhellmanndtroyer : that's an interesting observation about the number of ptl roles15:40
smcginnisIt should be noted, in certain newer/growing regions, PTL-ship and TC-ship are still seen as "achievements". But that should change over time.15:40
*** kumarmn_ has joined #openstack-tc15:40
fungii, similarly, will generally work on the same things whether or not i am elected to the tc15:41
dtroyersmcginnis: I've seen some of that too, that seems to be a function of 'time within the community' or 'age of community'15:42
cdentIt’s sometimes weird to me the extent to which we talk about “being on the tc” and “the work we do” as separate15:42
dhellmannpersia : that may be true. I hope the current long candidate list is indicative of a new crop of cross-project contributors ready to step up a level15:42
smcginniscdent: ++15:42
sdaguedhellmann: was the list really that much longer than past years?15:42
dhellmannsdague : there were a bunch of new names15:42
smcginnisdtroyer: I do expect "time within the community" to correct some of those perspectives.15:42
dtroyerI don't think so.  16 this time, I recall 18 and 20+ IIRC15:43
dhellmannsdague : and quite a few that I didn't expect to see, which I think is a good thing for the community but probably shows that I'm a bit out of touch with some sides of it15:43
cdentthe difference in the list this time is that it is not the usual suspects15:43
persiacdent: Most of current TC spends most of their time doing useful coordination, facilitation, organisation, dispute resolution, etc.  Those activities aren't limited to those who sit on the TC.  The actual TC work is mostly related to reviews and board reporting, which is a tiny min ority of the time for most members (I hope).15:43
dtroyerdhellmann: right, the new names certainly encouraged me15:43
cdentpersia: sure, but being on the TC is also work15:43
persiadhellmann: I don't see it as "levels", and also think many of the candidates have been actively doing the things I listed in my last meesage to cdent, but otherwise "yes".15:44
*** kumarmn has quit IRC15:44
dimso/15:44
persiacdent: Yes: just not as much as the work involved in general openstack leadership (which has large overlap with TC, but isn't TC)15:44
* dtroyer gladly gives up the last-one-in chair to dims15:44
dhellmanncdent : sure. I would probably do less of the TC-specific stuff like project evaluations and worrying about tags or resolutions. I would still do a lot of the other things I do, though.15:44
dimslol dtroyer15:45
fungiif i were not on the tc, i would likely still read and form an opinion on most governance proposals (as i did for years before getting elected), and would still attend public tc functions (meetings, et cetera)15:45
sdaguedhellmann: right, without naming names, this time there are a bunch of medium to long term community members stepping up to run in a role like this for the first time. Which makes me super happy.15:45
dhellmannsdague : ++15:45
fungiyep, whereas we had a lot of long-shot candidates in years past who ran for seats on the tc with little (if any) history with the community at large15:46
dhellmannpersia : I do think it's a level of *commitment* though. Going beyond a couple of projects and thinking about the whole community does take something of a mental shift.15:46
dhellmannfungi : good point15:46
fungii'd like to believe that our recent efforts toward trying to convince people already active in cross-project/community efforts to run for the tc and similar elected positions is actually paying dividends, though i have no evidence to back that up15:47
persiadhellmann: I agree with your second sentence.  I don't see what this has to do with running for the TC, other than being a necessary precondition.15:47
dhellmannTo be blunt, there are a bunch of people running this time who I've never even noticed participating in governance discussions in the past. I think it's good they're interested, but I wish that interest had manifested itself earlier and I hope it continues if they don't win. Because the community is at an age where stepping into the middle of things now is a *lot* harder than it was a few years ago where we had less15:50
dhellmann history, baggage, etc. to consider.15:50
dhellmannActually being around and following conversations is important to the process.15:50
fungitrue, and most of the new tc members elected in the past have tended to follow that pattern15:51
cdentI mostly agree with dhellmann, but there are plenty of sectors of this community where history and baggage are more like anchors than keels.15:51
dhellmannAnd I know a bunch of people weren't necessarily encouraged/allowed before, so I'm not blaming them. Just expressing that I hope this current election is an indication of that changning.15:51
cdents/with/with you/15:51
* fungi can't help but assume cdent is talking about nova when he says vague things like that ;)15:51
smcginniscdent: Was just thinking that - it might be good to have a few people without history and baggage.15:51
dhellmannThat's quite true. But knowing where the shoals are is important if you're going to change course.15:51
* dhellmann may have his nautical metaphor/vocabulary a bit off15:52
dimslol fungi15:52
cdentfungi: not just nova, but of course nova is an obvious place15:52
fungiyeah, i know. nova makes a convenient punching-bag too sometimes15:52
dims++ cdent15:52
cdentI’m going to have to keep the anchor keel metaphor for future reuse15:52
dhellmannsmcginnis : yep, and that's why I appreciate cdent stepping in to the driver-team discussions15:52
persiaDespite my complaints about the wording, more focus on bringing new folk into governance discussions / stewardship / championing (perhaps leveraging the SWG as the home for those folk) should help make it easier to feel comfortable participating in these sorts of discussions.15:53
persiaThat might allow those not elected (or those considering future election) to have a more obvious place to contribute.15:53
dhellmannthat's a good point15:54
dhellmannI was glad to see a few folks commenting on governance patches (even though I disagreed at least once)15:54
persiaThe disadvantage is that it does make the TC less special, where the TC maps to project "cores", rather than being a magic body.15:54
fungitechnical committee as the core reviewers of the stewardship working group... not a terrible analogy/model15:55
dhellmannI'm not sure how special we need or want to be. We've been trying to act as a court-of-last-resort rather than doing everything ourselves, because 13 people don't scale to managing 2,000.15:55
cdentdhellmann: sure, but that model leaves a leadership void. People complain about that frequently.15:57
persiacdent: Only if one conflates governance with leadership.  If the TC is purely a deliberative governance body (court-of-last-resort), and leadership is granted to the set of leaders (where leaders are defined as such by leading), then as long as a name can be given to the set of leaders, that body can be referenced.15:58
dhellmannI go back and forth on that issue. Sometimes I see the point. Other times I recognize that the TC is not the right body to be making the decisions it's asked to make, and that delegation is the right approach.15:58
persiaIn cases of disagreements between leaders, the TC can resolve.15:58
cdentI’m not declaring that the tc should be that leadership (although maybe it should) but rather that there is a void, people talk about it, from many corners and levels of the community.15:59
dhellmannIt happens that there is a lot of overlap between our "community leaders" and "TC members" right now, as you'd expect. But "community leader" and "distributed system architect" is not the same job description.15:59
persiacdent: I assert that can be resolved entirely with nomenclatural adjustments.15:59
dhellmannwe sure do love to rename things15:59
persiaOne of the issues is that many folk are used to hierarchical systems, wherein one asks permission to do things.  I suspect it healthier to encourage an enabled community, wherein people do things without asking, seeking counsel and support, and raising to governance bodies if/when there is a potential for dispute.16:00
cdentpersia: using the nova battering ram as an example again, that dispute has been there for several years now, but never changes16:01
cdentin large part because nobody seems to know _who_ can change things16:01
persiacdent: I have yet to assert the Nova model is healthy :)16:01
cdentexactly, people have been complaining, forever16:02
persiaI have a biased view of the world, but I think change requires a politician to establish new compromises in Nova.16:02
dimscdent : they did write down what they expect from others and what can be expected from them - https://docs.openstack.org/nova/latest/contributor/how-to-get-involved.html - which other teams have not done yet16:03
persiaI'm not sure anyone has any incentive to cause someone to do that though: those able to effect change today don't want to rock the boat, and the excluded may not perceive this as a political issue.16:03
fungiyeah, i personally feel entirely unqualified to decide how projects should pick a message queuing technology, or what their distributed locking mechanism should be, or what features they ought to implement16:03
persiaOne way in which many governors fail is in believing that such a lack of qualifications can be resolved by technical understanding.  As long as *someone* in the project has the technical background, a good leader is likely to be able to cause a conversation to happen to reach a good answer to those questions.16:05
fungitake the mysql vs postgresql situation (which i know is a mischaracterization unto itself)... i'm cool with helping decide how to document what the community itself has chosen to work on, but i'd be hard-pressed to contribute to any discussion about the pros and cons of the two database backends16:05
persiaeven as a moderator?16:05
fungias a moderator i'd need to be impartial and make sure everyone got a say, not contribute to the discussion16:06
persiaIn my opinion, that's a skill, the exercise of which is often termed "leadership".16:06
persiaand should count as contribution to the discussion, despite a lack of specific opinion, as the discussion will be better with such contribution.16:07
dimspersia fungi : problem i see is that one side does not show up for the discussion or work ...16:07
cdent(I’ve started another meeting elsewhere, so my attention here is drifting, will try to stay engagefd)16:07
persiadims: In such cases, I usually hunt folk down where they lurk, and try to convince them to join me in a smoky room.  This style doesn't work for everyone :)16:08
mugsiepersia: that works for getting people to the discussion, but not always for the actual work :/16:09
cdentdo we acknowledge that there is a difference between a moderator and facillitator and what we really need are the latter?16:09
dimspersia : people who took the initial decision are no longer around (at my employer for example) ... so teams responsible now just plod through it. it wasn't even a technical decision probably at that time16:10
persiacdent: That is perhaps a good distinction.16:10
* cdent has a background on so-called high performance collaboration, the distinction is very important there16:10
persiadims: Those are hard to solve: needs raising to an external party who can act as a counterfoil to the "if I do this, I keep my job, but it's stupid" situation.  I have no unbiased suggestions there.16:11
mugsiecdent: or a facilitator, who may have to be an arbitrator at some point in the discussion16:11
persiamugsie: ++16:11
dimsyep mugsie16:14
cdentthe main role of a facillitator is to get people to realize that what they think they are saying is not actually the same thing as they are saying16:17
cdentand to indicate when the squirm test is being failed: http://eekim.com/2007/02/the-blue-oxen-way/16:17
cdentin other words they reflect what they hear in a concreat fashion so that people can be stunned at amazed at how poorly they have communicated :)16:18
smcginnishah16:20
mugsieyeah - I have been guilty of that ... a lot :)16:20
mugsiebut sometimes, even with shared language / understanding, groups will disagree16:21
mugsies/sometimes/often/ potentinally16:21
persiaWhen there is disagreement, there is often scope for compromise.  Group X can have foo, group Y bar, and group Z baz, and this means that quux doesn't have to happen.16:22
cdentpersia, indeed, but you want to be sur that the disagreement itself is agreed16:22
persiaRequires constituency identification, issue alignment, etc., but these often emerge from such discussions.16:22
cdentotherwise the resulting compromise is one nobody likes16:22
persiacdent: Absolutely.  No potential for compromise if one group thinks everything is fine.16:22
*** jpich has quit IRC16:23
dimspersia : how do we do this say at the PTG, say in a big room for one of the large projects?16:25
mugsiebehind plexiglass?16:27
persiadims: Assuming you are referring to identifying constituencies and building compromise: identify a few folk who seem to be local leaders, and discuss with them in a corner during a break.  Have a few discussions with the quiet folk to confirm they are the followers you think they are.  Once you have some understanding of the various opinions, put it on the agenda.  Start by discussing some of the possible options (rephrasing to identify16:32
persiacommunication gaps).  Name names whilst doing so, so people become attached to the choices.   Continue the discussion until you have some obvious "camps".16:32
persiaIdeally the camps represent alternate use cases: challenge folk from one camp to discuss how their model solves an alternate use case, etc.  Try to find a minimum acceptable compromise "for now".  Do that.  Repeat.16:33
fungimugsie: i think we need to start a new tradition, "tc dunking booth"16:34
persiaI haven't tried this for Nova (the only time I was in a Nova room, the discussion had been prepared enough that it went how I was tasked with having it go without intervention), but I've used that to help reach decisions in Ironic and Neutron before.16:35
persiadims: Also, it helps to be the person who takes the minutes, if you can arrange that.16:36
mugsiefungi: :)16:37
dimsgood tips persia16:40
*** dtantsur is now known as dtantsur|afk17:04
cdentI think we managed to totally derail EmilienM’s line of inquiry17:25
mugsiejust a little17:29
dtroyermaybe directly, but we explored some of the effects of limiting terms, highlighted the balance of continuity that could be upset (do we have it now? IMHO yes-ish) and wandered through the what is leadership anyway garden, again.  Those are all areas that were there a problem to be solved in this area, I propose we would have seen signs.  I didn't see them.  Anyone else?17:29
mugsieI think the effects remain to be seen - as we grow new leaders, we may naturally rotate, especially as it is a priotity now17:31
mugsiewe have had a very stable leadership in the TC and some projects over the last while, but the corporate shake ups have changed that a bit now17:31
dtroyerthis is not the first time that has happened.  The departure of most of our member startups caused similar rise in turnover in project and community leaders.  It's a bit cyclical, running 4-6 dev cycles maybe?17:33
*** lbragstad has quit IRC17:35
cdentdtroyer: I think a lot of comes down to what we percieve our operational parameters to be.17:37
cdentIf the goal is stability, continuity and sameness, yup, nothing to see here.17:37
cdentI continue to be unclear on whether that’s okay; i.e., are we using the right parameters.17:38
cdentI simply don’t know.17:38
*** lbragstad has joined #openstack-tc17:39
dhellmanncdent : I don't want sameness when I say "continuity". I want a smooth transition and I want new leaders to revisit old decisions with all of the history, so we don't cycle back and forth through the same set of mistakes18:11
cdentack, dhellmann, not disputing that (either that it is wanted or that it is good)18:12
cdentbut that thing that flwang1 said that I put at the top of the latest tc report lodged in my brain pretty hard18:12
flwang1flwang1 is reading the log ...18:15
dhellmannwell, I hope I don't just hold the same beliefs on all issues (and that I'm not perceived as doing that), but neither can we ignore history18:15
harlowjaoh seems like i missed alot, ha18:18
* harlowja reads18:18
harlowjaso ohhh i have a question18:19
harlowjawhat can the TC do to make openstack technical neat to the wider world18:19
harlowjaa roadshow that explains 'this part of nova is doing neat things'18:19
harlowjaor we use 'etcd' like this in a blog18:20
harlowjaor hire that jepsen guy to do some analysis of <openstack> something or other18:20
harlowja^ though not sure i want to see the results, ha18:20
cdentdhellmann: Like so many things in openstack, I don’t think a lot of the behaviors that we as a group sometimes demonstrate can be applied to any individual. Our actions in concert create a perception.18:22
harlowjaor like get interviewed on lwn.net about we are doing this or that awesome (technically) thing18:22
harlowja^ happened once for zookeeper /etcd stuffs18:22
cdentThe etcd question is interesting. Is any project relying on it (as in requiring it)?18:22
harlowjaor perhaps, blah blah, we have microversions stuff, we are working with the openapi (and swagger?) folks to get that support in there, isn't that neat...18:23
harlowja^ idk if its real, ha18:23
harlowja(but could be)18:23
flwang1dhellmann: i agree we shouldn't ignore history. my original point is we need some reviews for the work we have done and especially something we're still doing18:24
cdentharlowja: sdague is gonna present on microversions at apistrat, I’m hoping to hear back on how that goes18:24
flwang1I'm happy to give another sample here: the PTG18:24
harlowjacdent neat18:24
harlowjai think neat, idk, let's see :-P18:24
harlowjabut like, the nova scheduler or scheduelr work for example18:25
harlowjaif u've seen k8s scheduler its simlar to nova's18:25
flwang12 years we think put a lot of developers in the summit is crowd, but now we have lose a lot contributors, is the reason to have PTG still true?18:25
harlowjaso why aren't we like showcasing 'neat' (for some definition) things like that18:25
cdentharlowja: who is “we”?18:25
flwang1and we have more and more contributors from APAC, and all the PTG only happened in US, is that still a low cost travel for most of the openstack contributors?18:26
flwang1s/2 years/2 years ago18:26
harlowjacdent good question18:27
harlowjai am not totally sure i guess as 'we' but looking at http://blog.kubernetes.io/ they seem to do articles from various peoples18:27
flwang1things are changing faster than we react sometimes18:27
harlowjai'd be up for writing an article if there was a place like that18:28
harlowjaEditor's note: today's post is by Aparna Sinha, Group Product Manager, Kubernetes, Google ...18:28
harlowjaand blah blah18:28
harlowja(so its not just one person)18:28
cdentharlowja: http://superuser.openstack.org/18:28
harlowjawhere are the tech details/meat in there?18:29
harlowjai didn't think superuser (the user part) had much tech details in it18:29
cdentdunno18:29
harlowjaso ya, just an idea, idk if its a solution, ha18:30
harlowjabut create equivalent of http://blog.kubernetes.io/ (that isn't user focused)18:30
* cdent gives harlowja the keys18:31
harlowjarollercoaster here we come18:31
harlowjalol18:31
cdenthttps://www.openstack.org/blog/18:31
harlowjayes, that's also i think slightly different right?18:31
cdentmight be extensible?18:32
harlowjalook at content of http://blog.kubernetes.io/  (which is more like lwn.net) vs superuser and https://www.openstack.org/blog/18:32
* cdent nods18:32
sdagueI agree with what I think harlowja is saying, that we should have folks show up in more external venues and talk about neat parts of openstack18:33
cdentI’m thinking of it more in terms of what’s the minimal effort to get more content under the openstack.org domain. In this case that would seem to be: write something and thingee to post it on /blog18:33
harlowjai'd write something18:33
harlowjaoslo, taskflow, tooz, u pick18:33
cdenttooz + etcd18:34
harlowjai think others could figure out something to18:34
cdentgo18:34
harlowjai'll do it (for real) if there exists a outlet for that18:34
cdentIf you’re okay with trying to add techiness to o.o/blog I suspect there’s room for that and it would be easy to make go. Would need to find some foundation person (thingee does seem like the right person) to help18:36
cdent(they operate that content as far I know)18:36
sdagueharlowja: quite often in the past, I'd just write something on my blog which landed on the planet, and foundation staff asked for republish on superuser18:36
sdagueI would say write the content, get it out there, then use the o.o/blog to amplify18:37
harlowjawhy do we need that barrier18:38
harlowjai mean, i don't have a personal blog (though i've tried a few times, cause just inertia)18:38
harlowjaso why have the bar be that18:38
mugsieyeah, I would say start with content, then see if we need a separate blog - there is nothing worse than a sporadic /empty blog18:38
harlowjai would expect blog.kubernetes.io isn't just copying/pasting from others blogs18:38
harlowjaidk, i'm not sure why this is such a big revelation :-/18:39
harlowja^ honestly, ha18:39
* harlowja keep the bar low?18:40
cdentharlowja: when the foundation folk aren’t busy travelling they’d probably be pretty keen on a “tech blog” either as a category on /blog or something new. but they have the keys to the castle, so the immediate route, if your pen is dying to write, is medium or something. But I suspect it is the idea of having a thing, a place, which matters, and I agree, and for that we need some folk, I guess18:43
harlowjaYa, I mean, I'll write stuff if there is that18:43
harlowjaI'd rather see the foundation folks put some skin into the game, before I put mine in18:44
harlowja(sorry, just how i work, ha)18:44
*** mriedem has joined #openstack-tc18:45
mriedemdims: HI!18:45
dimshey mriedem welcome!18:45
cdentharlowja: I will carry your plea to the castle and see what I can do18:46
harlowjawoot18:46
* harlowja back to grinding flour (by hand)18:46
*** flwang1 has quit IRC18:57
sdagueI'm not sure I understand why it's such a big deal to do draft one without having to engage anyone else, then promote to "production" from there. It's just devops for docs. :)19:26
harlowjai like to dance with others19:54
harlowjalol19:54
harlowjaso if openstack foundation will dance, i'll play, and i hope others will to19:54
harlowjai can draft an rst though, sure19:55
harlowjai'm just not gonna draft much if i won't have anyone to dance with19:55
harlowjacall be a pita, blah blah, its ok :)19:57
*** kumarmn_ has quit IRC20:31
*** flwang1 has joined #openstack-tc20:36
*** flwang has quit IRC21:15
openstackgerritLance Bragstad proposed openstack/governance master: Update policy artifacts for tacker  https://review.openstack.org/51351521:17
openstackgerritLance Bragstad proposed openstack/governance master: Update policy artifacts for tacker  https://review.openstack.org/51351521:18
*** flwang has joined #openstack-tc21:28
openstackgerritLance Bragstad proposed openstack/governance master: Update policy artifacts for tricircle  https://review.openstack.org/51352321:30
*** hongbin has quit IRC21:38
*** cdent has quit IRC21:42
*** sdague has quit IRC22:34
*** kumarmn has joined #openstack-tc22:54
*** kumarmn has quit IRC22:58

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