| *** mriedem has quit IRC | 02:03 | |
| *** liujiong has joined #openstack-tc | 02:34 | |
| *** harlowja has quit IRC | 03:18 | |
| *** chkumar246 is now known as chandankumar | 04:59 | |
| *** harlowja has joined #openstack-tc | 05:12 | |
| *** ykarel has joined #openstack-tc | 05:44 | |
| *** ykarel has quit IRC | 06:19 | |
| *** harlowja has quit IRC | 06:29 | |
| *** chandankumar has quit IRC | 06:35 | |
| *** chandankumar has joined #openstack-tc | 06:36 | |
| *** liujiong has quit IRC | 06:50 | |
| *** liujiong has joined #openstack-tc | 06:50 | |
| *** liujiong has quit IRC | 06:50 | |
| *** ykarel has joined #openstack-tc | 07:50 | |
| *** cdent has joined #openstack-tc | 08:45 | |
| openstackgerrit | Merged openstack/governance master: Retire horizon-cisco-ui project https://review.openstack.org/541809 | 08:45 |
|---|---|---|
| openstackgerrit | Merged openstack/governance master: replace tracking sections with storyboard links for rocky goals https://review.openstack.org/545130 | 08:47 |
| openstackgerrit | Merged openstack/governance master: Update status of 'API deployed in uwsgi' goal https://review.openstack.org/545484 | 08:47 |
| openstackgerrit | Merged openstack/governance master: [rally] Add a separate repo for OpenStack plugins https://review.openstack.org/545827 | 08:49 |
| openstackgerrit | Merged openstack/governance master: Add openstack-ansible-nspawn.* to OpenStack-Ansible https://review.openstack.org/546091 | 08:49 |
| openstackgerrit | Merged openstack/governance master: Add puppet-monasca to Puppet OpenStack https://review.openstack.org/546190 | 08:49 |
| openstackgerrit | Merged openstack/governance master: Add openstack-ansible-os_panko to OSA deliverable https://review.openstack.org/546572 | 08:49 |
| openstackgerrit | Merged openstack/governance master: Add roles into openstack-ansible deliverable https://review.openstack.org/546575 | 08:49 |
| openstackgerrit | Merged openstack/governance master: Add Glance completion info for policy in code goal https://review.openstack.org/546981 | 08:52 |
| * ttx grabs coffee | 08:58 | |
| cmurphy | morning | 08:59 |
| ttx | tc-members: good morning | 09:05 |
| ttx | I expect most to still be in post-snowpenstack recovery | 09:06 |
| * cmurphy finally got home yesterday afternoon after getting stuck in Amsterdam overnight :( | 09:07 | |
| ttx | Quick thing I wanted to mention: the OpenStack Foundation is likely to join the Open Container Initiative (OCI) to show its support for the spec (which is used by Kata Containers but also by OpenStack LOCI). Let me know if you have questions on that | 09:08 |
| ttx | Also PowerStackers will be approved on Friday unless new objections are posted | 09:09 |
| cdent | I'm sort of here, good morning everyone. | 09:12 |
| cmurphy | o/ | 09:13 |
| cdent | Joining OCI seems reasonable, but I worry a little bit about perceptions that the infra side of the house is being de-emphasized (emphasis on the "perceptions"). | 09:13 |
| ttx | cdent: I think that should be fixed by the messaging around open infrastructure, which paints the overarching story | 09:16 |
| cmurphy | i've also noticed a perception that the foundation is chasing the new container hotness rather than sticking close to its infra roots | 09:16 |
| cmurphy | so yes better messaging would be good | 09:16 |
| ttx | In that case it's just supporting the OCI standard, which we happen to leverage in LOCI (our OCi container packaging for OpenStack) | 09:17 |
| ttx | and that Kata Containers is compatible with. | 09:17 |
| ttx | but yes I can see how that needs to be counterbalanced | 09:17 |
| cdent | I think it is a confluence of factors, none of which in themselves is a big deal, but in a stressed out or paranoid mind feels differently: | 09:18 |
| cdent | the ptg is too expensive to continue | 09:18 |
| cdent | let's do kata | 09:19 |
| cdent | let's do oci | 09:19 |
| cdent | "my presentation at summit is far more likely to get accepted if it is about nfv or containers" | 09:19 |
| cdent | etc | 09:19 |
| cmurphy | where did "the ptg is too expensive to continue" come from? | 09:19 |
| cdent | board meeting | 09:19 |
| cmurphy | that is new to me | 09:19 |
| cmurphy | :( | 09:19 |
| cdent | they decided to punt on that one | 09:19 |
| cmurphy | i guess i'll listen in on the next board meeting then | 09:20 |
| cdent | but yeah, never ever do the board meeting on a ptg day | 09:20 |
| cmurphy | for real | 09:20 |
| ttx | cmurphy: we definitely need to find a more sustainable model, but there is also the question of inclusivity. Lots of complaints from Chinese contributors that they are unable to attend | 09:20 |
| cdent | I continue to think representation of individual daily developers in board and foundation activities is way low | 09:20 |
| cmurphy | ttx: due to expense? | 09:22 |
| ttx | Which is why there are two options on the table: continue as separate but with some alternate funding model (like increased ticket price), or co-locate with summit again so that it's less travel | 09:22 |
| ttx | cmurphy: I feel like money is not the only factor. Having to ask for visas every 3 months is not fun | 09:23 |
| cmurphy | oh right :/ | 09:23 |
| ttx | We are preparing a dev survey on the topic | 09:23 |
| * cmurphy reminded of her privilege | 09:23 | |
| ttx | not just PTG attendess, all -dev | 09:23 |
| cdent | I suspect the power of the rumor mill is being underestimated | 09:24 |
| ttx | I think if we did not receive that many complaints against the separated event, the question of continuing PTGs would not even be on the table | 09:24 |
| cdent | Is there a clear picture of how those complaints divide up amongst regular contributors, casual contributors, managers of contributors, companies wth contributors, $other? | 09:25 |
| ttx | BUT that all may be an illusion. If people need to meet regularly to do their work, they will. Unofficial meetups or PTG. Which will always screw up some portion of contributors | 09:25 |
| * johnthetubaguy wonders in late and reads scrollback | 09:27 | |
| ttx | cdent: I received that feedback from regular Chinese devs in Sydney | 09:27 |
| ttx | not their management... but I know others at the Foundation received that feedback from managers as well | 09:28 |
| ttx | I'd say from significant/intermediary-sized Chinese contributing companies | 09:28 |
| ttx | anyway, let's collect data first | 09:29 |
| ttx | cmurphy: at the board meeting the conclusion was that we should definitely organize the next noe, and make the call for 2019 by Vancouver joint meeting. | 09:29 |
| ttx | one* | 09:29 |
| cdent | ttx: I was meaning about the "separated event", was that what you meant? | 09:30 |
| cdent | I think the PTG is the only openstack event I've experience in the last 3.5 years that has felt like community | 09:30 |
| cdent | and the only one that actively drives cross project work | 09:31 |
| ttx | cdent: yes, those contributors were advocating for co-location with summit again | 09:31 |
| johnthetubaguy | ttx: +1 on waiting for more data. I certainly heard complaints about the expensive of it all since it was introduced | 09:31 |
| cdent | the expense concern, to me, is bogus. How much money are companies making off the back of contributor labor? | 09:31 |
| cdent | travel difficulties and amount is certainly a thing to think about, though | 09:31 |
| ttx | cdent: it's bogus because the event makes you more productive, yes | 09:31 |
| ttx | BUT unfortunately not many companies think sanely when it comes to money | 09:32 |
| johnthetubaguy | cdent: I am curious how it feels different to the design summits? | 09:32 |
| ttx | With design summits we traded productivity for not engaging with the rest of the community | 09:32 |
| johnthetubaguy | ttx: +1 on the lack of sanity, I found it was more a lack of clear precedent, everyone else manages without the travel etc... | 09:33 |
| cdent | johnthetubaguy: I'm not sure I can explain that in clear terms. It just does. :) | 09:33 |
| ttx | so yes we could be productive, but by isolting ourselves from the others | 09:33 |
| johnthetubaguy | ttx: actually the whole opps vs dev sessions is coming back to me now | 09:33 |
| ttx | cdent: the calculation should be obvious. Attending the pTG makes you at least 10% more productive for the rest of the cycle. Which has an obvious RoI | 09:34 |
| johnthetubaguy | cdent: thinking back, it is like the midcycle, but with friends and not isolation, I think my memory was playing tricks on me there | 09:35 |
| ttx | Sending a team of 10 people to the PTG is a better move than keeping an 11-people team at home | 09:35 |
| cdent | Yeah, midcycles have never been a positive experience for me | 09:35 |
| ttx | Unfortunately travel budgets and sponsoring budgets don't work that way | 09:35 |
| johnthetubaguy | tried so many ways to ring fence travel budget, every one fails, and its always considered optional, but oddly not for sales people... | 09:38 |
| cdent | johnthetubaguy: yeah, the travel budget for sales thing is fun | 09:39 |
| johnthetubaguy | it feels a lot of this is about lack of precedent, we are strangely diverse in leadership, etc. | 09:43 |
| johnthetubaguy | the academics seem to get the idea of going to talk to others about what we are doing and getting feedback and new ideas | 09:43 |
| * cdent nods | 09:44 | |
| *** dims has quit IRC | 09:45 | |
| *** dims has joined #openstack-tc | 09:49 | |
| cdent | A survey of "attitudes to the ptg" will be useful, especially if lots of people respond. | 09:49 |
| cmurphy | anyone thinking of attending https://events.linuxfoundation.org/events/open-source-summit-europe-2018/ ? they are having "open collaboration conference" and "diversity empowerment summit" tracks in their CfP which sounds relevant to this discussion and the OpenStack TC in general | 09:50 |
| cdent | The other thing I think we ought to consider is more active interaction between the TC and the foundation's planning. I'm not sure what that would look like though. I'd like to see more explicit info sharing rather than tacit. | 09:50 |
| ttx | Agreed, which is why I'm raising the OCI topic here :) | 09:50 |
| cdent | cmurphy: sounds interesting, timing may be problematic for me | 09:51 |
| cdent | ttx, yes, thanks for that | 09:51 |
| ttx | cmurphy: been considering it yes | 09:52 |
| cdent | I think there should be more PTGs and less conferences :) | 09:54 |
| EmilienM | hello | 09:55 |
| ttx | EmilienM: hola! | 09:56 |
| cdent | EmilienM: morning | 09:56 |
| cmurphy | I feel like I recall the selling point of having PTGs was that devs would no longer have to attend the summits, which did not end up being the case at all it feels like | 09:56 |
| cmurphy | if there was no forum at the summit then maybe the goal of having less travel would actually be achievable with PTGs | 09:57 |
| cmurphy | EmilienM: I hope you are not still in Dublin | 09:57 |
| ttx | Nah, the selling point is that devs present at summits would be able to engage and share with others | 09:57 |
| ttx | rather than be holed up in a room with their team for 3 days | 09:57 |
| EmilienM | cmurphy: thanks! I made it to my parent's place | 09:58 |
| cmurphy | EmilienM: oh good :) | 09:58 |
| EmilienM | I'm currently baby sitting a 3 weeks old baby | 09:58 |
| EmilienM | while trying to get some work done. But he's not helping :-P | 09:59 |
| ttx | We also said that it changed dev travel to summit from a MUST to a SHOULD | 09:59 |
| cmurphy | :'D | 09:59 |
| cdent | ttx we say that's the selling point, but I don't know that everyone feels that way (which is always the problem) | 09:59 |
| cdent | ttx should is bogus though: if you want to be relevant (for your employer or yourself) you need to be there, because other people will be, and they will be making decisions (OMG) without you! | 10:00 |
| johnthetubaguy | cdent: I felt disconnected after not being in syd, although that wasn't strictly budget reasons, more family reasons | 10:00 |
| cdent | johnthetubaguy: exactly. that disconnectedness is difficult to manage given the modes of activity in some of the projects | 10:01 |
| johnthetubaguy | I mean lots of people make heroic efforts to document what happened, but there is nothing like being in the discussion | 10:01 |
| * johnthetubaguy nods | 10:01 | |
| cdent | if we want to save money we need to change the modes of behavior | 10:01 |
| cmurphy | it seemed like a "should" before, or it still feels like a "must" now, but somehow it doesn't feel like the value of being there has changed | 10:01 |
| cdent | cmurphy: yeah, agree | 10:02 |
| johnthetubaguy | +1 both of you | 10:02 |
| cdent | the old modes were based on a professionalization of openstack development. that mode is still true for some (me for example) and still true for some projects (nova for example) but is not sustainable as is | 10:03 |
| cdent | Or rather, in order for it to be sustainable, contributing companies need to step up with $$ | 10:04 |
| * cdent checks on chickens | 10:06 | |
| *** rosmaita has joined #openstack-tc | 10:29 | |
| * persia ponders the effect of having PTG in China: reverses the visa requirements, which may offer perspective | 11:52 | |
| cdent | would be interesting. But I'd hope that if were exercising any kind of political attitudes in our PTG destinations we'd avoid China (and the US too) | 11:54 |
| cmurphy | I think there are countries that are not China but also easier for Chinese residents to travel to, Singapore for instance looks like it has an easier visa process for Chinese passport holders | 12:06 |
| persia | I object to your use of "political" in that sentence, but agree with the intent I read into the words. Also, US tax codes make it hard for US entities to pay for conference attendance unless they can demonstrate that the conference is as likely to happen inside the US or outside the US. | 12:07 |
| persia | cmurphy: Meeting in a neutral country has value, but meeting in a country with a complex visa process for many of the folk who often attend PTG may help build shared understanding of the issues. I know of one person whose inconvenience by snow in Dublin was related to a requirement to leave the country no later than Saturday, adding more stress than for many who might just need to rebook flights or hotels. | 12:08 |
| cdent | persia: how did you interpret my use of the term? | 12:10 |
| cmurphy | I can agree in principle on the benefits of reversing perspectives (Sydney was great for that) but practically I think that's a hard sell | 12:11 |
| persia | Sydney was terrible for nearly everyone: nearly as much hassle for most folk in asia as for folk in the americas. Excellent for australians. | 12:12 |
| persia | cdent: I read your statement as encourgaging reduction in economic incentives as a means to express opinion about the practices of specific sovereign states. In my glossary, "political" is reserved for things related to policies that apply to polities. As many of us are not properly members of a polity involved with either of the sovereign states you mention, I don't think our economic expression of opinion should use that term. | 12:14 |
| cdent | ah, hmm. I wasn't really thinking of it in terms of economic incentives or expressions more as a (mostly ineffectual but still symbolic) vote of no confidence, thus an expression of power (albeit limited), thus "politics". | 12:16 |
| persia | For those of us that are properly within the polity for those sovereigns, it is very likely that our personal economic influence is directly related to inflows, thereby causing us to be personally wounded by such economic action. | 12:16 |
| cdent | I tend to use politics as an overriding term for anybody's expression of power, not just polities as polity, as you say, are non-inclusive | 12:17 |
| persia | For me, politics is the art of compromise. Power is an expression of might. In practice, one can have (limited) power by controlling the policies by which might is expressed, which can be done through discussion and compromise with those to whom power has been delegated. | 12:17 |
| persia | I do not believe we can effectively practice politics if we do not appreciate this distinction. | 12:18 |
| cdent | I expect my interpretation is overly informed by critical theory | 12:18 |
| persia | And, as a collaborative distributed community who compose both code and policy regarding how the world consumes our output, I think it is unfortunate when we don't practice politics. | 12:18 |
| persia | (but most of the time we seem to do reasonably well, and most of our policies support active political engagement by anyone in the community, which is promising) | 12:19 |
| cdent | (and in that context, I'm a radical in the "get to the root of" definition) | 12:19 |
| persia | My definitions are etymologically defensible, but yes, most theory is written from the perspective of those attempting to cause change, which implies they seek power, so it is convenient to conflate the terms. Broadly speaking, in a policy-controlled environment (follows the rule of law), politicians are more effective at providing minor change (serving constituents), which makes the narrative "politicians are the problem" useful for those who | 12:23 |
| persia | are more practiced in managing segmented communications to multiple audiences than in politics, as it provides them a base from which to develop sufficient constituency to overthrow the prior power brokers. | 12:23 |
| dims | EmilienM : awwwww | 12:40 |
| *** cdent has quit IRC | 12:44 | |
| *** cdent has joined #openstack-tc | 13:12 | |
| *** cdent has quit IRC | 13:33 | |
| EmilienM | dims: :) (not mine though) | 13:37 |
| dims | :) good to practice! | 13:50 |
| *** cdent has joined #openstack-tc | 13:52 | |
| fungi | seems i missed another excellent office hour, but i do not regret sleeping | 14:04 |
| cdent | good life choices | 14:06 |
| persia | +4 | 14:09 |
| * dtroyer finally enjoyed a sleep cycle that matches the sun cycle | 14:14 | |
| dtroyer | I think one of the things we need to keep in mind is the balance of the complaints for any given decision (such as PTG vs Summit or release cycle length) is that good compromise will have nearly everyone unhappy about something. And those that are happy aren't as vocal | 14:15 |
| persia | Being unhappy with the status quo is not necessarily the same as having a constructive opinion towards effective change. At least for me, cancelling PTG would avoid making people feel excluded, but at the cost of avoiding making people feel included. | 14:20 |
| mugsie | Also, on the PTG, if we cancel it, I would bet that a large amount of projects will start midcycles again, which will be just as exclusionary as the PTG. | 14:26 |
| mugsie | and adding a week to the forum is probably a non starter for most people | 14:26 |
| dtroyer | mugsie: generally I agree, but for those with visa complications that is one event rather than two and may be a better situation. | 14:28 |
| persia | I actually like that idea. I'm not sure I have the stamina for two weeks, but it simplifies scheduling. | 14:28 |
| mugsie | dtroyer: yeap, for visas it is probably a win win | 14:28 |
| cdent | I barely have the stamina for 1 week | 14:28 |
| cdent | but I'm a delicate flower | 14:28 |
| dtroyer | I've wondered if we could do 3 days of non-overalpping dev-sessions and a 3 day conference | 14:29 |
| mugsie | but 2 weeks away from home / family / $responsibilities (and companies paying for staff to have a weekend off in $city) is a lot to ask | 14:29 |
| dtroyer | basically make 2 4-5 day trips and 2 6-7 day trips into 2 8-10 day trips? | 14:29 |
| dtroyer | it is a lot | 14:30 |
| dtroyer | or maybe we do the combined once a year? | 14:30 |
| mugsie | cdent: you are not alone | 14:30 |
| dtroyer | and then 2 separate events | 14:30 |
| mugsie | I think that would cause us to look at the length of a cycle again, wouldn't it? | 14:30 |
| persia | mugsie: Companies paying for weekend in $city can often be negotiated into companies not providing time-in-lieu for weekend flights. | 14:31 |
| mugsie | persia: I know my company tries to do that anyway, regardless of the situation | 14:31 |
| * mugsie had to negotiate that out of his contract | 14:32 | |
| *** mriedem has joined #openstack-tc | 14:43 | |
| *** hongbin has joined #openstack-tc | 14:44 | |
| TheJulia | good morning everyone | 14:46 |
| zaneb | much scrollback. so jetlag. | 14:47 |
| TheJulia | ++ | 14:47 |
| mugsie | zaneb: yeah, the jet lag is killing me | 14:52 |
| mugsie | :D | 14:52 |
| zaneb | mugsie: :P | 14:52 |
| pabelanger | o/ | 15:16 |
| pabelanger | Our polling for S release names closed yesterday. I'm going to be looking this morning on the next steps | 15:17 |
| pabelanger | ttx: fungi: I think the foundation looks over the suggestions first, but not sure who or how to do that step | 15:17 |
| zaneb | mugsie: thanks for updating the interop testing resolution. I have comments :) | 15:20 |
| cdent | pabelanger: there are some excellent choices in there | 15:20 |
| * cdent is shocked, SHOCKED, by zaneb | 15:21 | |
| * zaneb fears he is becoming typecast | 15:21 | |
| pabelanger | cdent: yah, it seems to have went well | 15:22 |
| cdent | zaneb: I jest. I, at least, very much appreciate your comments | 15:23 |
| zaneb | pabelanger: how does "Words that describe the feature should not be included" apply when all of the words describe a feature? :D (e.g. Seestraße == Lake Street) | 15:24 |
| cdent | Shelhaus makes me giggle | 15:27 |
| pabelanger | zaneb: indeed | 15:28 |
| ttx | zaneb: separate words that describe the feature should not be included ? | 15:28 |
| ttx | Basically that rule was so that you would not include "Street" is "Icehouse Street" | 15:29 |
| ttx | I think we said ASCII set though | 15:30 |
| cmurphy | in german the words all run together | 15:30 |
| cmurphy | so we'd have to chop off straße even though it's part of the word | 15:30 |
| cmurphy | ß can be converted to ss though | 15:31 |
| ttx | bah | 15:31 |
| ttx | Ich spreche sehr gut Deutsch | 15:32 |
| ttx | (Ich glaube) | 15:32 |
| zaneb | ttx: schön | 15:33 |
| * cmurphy will lol if S-Bahn is chosen | 15:34 | |
| ttx | that '-' will likely disqulify it | 15:35 |
| ttx | Name must be composed only of the 26 characters of the ISO basic Latin alphabet | 15:35 |
| SamYaple | i missed the PTG chats above... but everyone ive talked to prefers the ptg to the design summit, and a large porition of them only go to the PTG not the summit (unless they have a talk) | 15:37 |
| SamYaple | i also love the PTG because we get so much accomplished. both talking about new ideas and coding | 15:38 |
| SamYaple | i would be sad to see it go back to the old way | 15:38 |
| cdent | SamYaple++ | 15:39 |
| cmurphy | ++ | 15:40 |
| cmurphy | the keystone team gets a lot out of all four events actually | 15:40 |
| cmurphy | but the PTG more so | 15:40 |
| mugsie | zaneb: thanks for the comments - they all make sense | 15:42 |
| mugsie | But overall it looks like what we talked about right? | 15:42 |
| *** cdent has quit IRC | 15:46 | |
| fungi | pabelanger: in nthe past we've run the election first and then foundation legal/marketing have crossed out the first n choices as unworkable, leaving is with the n+1 most consensual choice from the ranked poll | 15:53 |
| pabelanger | fungi: okay, that makes sense. | 15:53 |
| fungi | basically, we have one poll either way, but the research needed to rule out names is expensive so polling first cuts down that cost | 15:54 |
| fungi | as they only have to research the top picks rather than every single option | 15:55 |
| pabelanger | yah | 15:55 |
| pabelanger | okay, I'll now look into public polling software over CIVS | 15:57 |
| dhellmann | pabelanger : I think we can still use civs as long as we make the poll public instead of feeding it a list of foundation member emails | 15:58 |
| dhellmann | I thought that's what we said we'd do, for simplicity | 15:58 |
| ttx | yes, that was the suggestion, use CIVS public polls | 15:58 |
| pabelanger | okay, that makes sense | 15:59 |
| ttx | "A Condorcet election is held to rank the names. The poll will be public, with the voting URL communicated through OpenStack mailing-lists..." | 16:00 |
| ttx | See https://governance.openstack.org/tc/reference/release-naming.html | 16:00 |
| pabelanger | Ack, see that now | 16:00 |
| zaneb | mugsie: yes, +1 for the general direction | 16:03 |
| dims | off-topic : fyi, https://landscape.cncf.io/ | 16:30 |
| mugsie | https://landscape.cncf.io/selected=zuul :'( | 16:33 |
| pabelanger | seen that | 16:34 |
| mugsie | https://github.com/Netflix/zuul/issues/315 made me even sadder | 16:35 |
| *** ykarel has quit IRC | 17:14 | |
| *** cdent has joined #openstack-tc | 17:22 | |
| *** ykarel has joined #openstack-tc | 17:43 | |
| *** dtantsur is now known as dtantsur|afk | 18:00 | |
| *** rosmaita has quit IRC | 18:16 | |
| *** ykarel has quit IRC | 18:18 | |
| *** harlowja has joined #openstack-tc | 18:39 | |
| *** rosmaita has joined #openstack-tc | 19:20 | |
| *** rosmaita has quit IRC | 19:23 | |
| *** portdirect has quit IRC | 19:41 | |
| *** portdirect has joined #openstack-tc | 19:41 | |
| *** david-lyle has joined #openstack-tc | 20:08 | |
| *** david-lyle has quit IRC | 20:15 | |
| dmsimard | yeah it's not like we didn't try | 20:17 |
| dmsimard | mugsie: oh wow I didn't see Netflix's Zuul on there though | 20:18 |
| *** david-lyle has joined #openstack-tc | 20:39 | |
| *** david-lyle has quit IRC | 20:52 | |
| -openstackstatus- NOTICE: The infrastructure team is aware of replication issues between review.openstack.org and github.com repositories. We're planning a maintenance to try and address the issue. We recommend using our official supported mirrors instead located at https://git.openstack.org. | 21:19 | |
| *** david-lyle has joined #openstack-tc | 21:24 | |
| *** bmace_ has joined #openstack-tc | 21:33 | |
| mordred | mugsie: https://github.com/cncf/landscape/pull/315 https://github.com/cncf/landscape/pull/316 https://github.com/cncf/landscape/pull/317 | 21:40 |
| *** david-lyle has quit IRC | 21:41 | |
| *** david-lyle has joined #openstack-tc | 21:51 | |
| *** bmace_ has quit IRC | 21:57 | |
| dmsimard | Re: PTG/Summit split: something I haven't seen mentioned here is that prior to the split we had a design summit *and* an ops summit. Having devs and ops meet has a lot of value and I feel we kind of lost that with the PTG. | 22:02 |
| dmsimard | Ops are not really incentivized to go to the PTG and devs aren't incentivized to go to the (now split) ops meetup. | 22:03 |
| persia | For me, ops summit is largely replaced with the expanded and formalised "ops midcycle". | 22:04 |
| persia | And the ops<->dev stuff seems to be scheduled at forum, in parallel to the product tracks. | 22:04 |
| persia | Whereas my memory of the old model was that ops summit was just before design summit, and ops would provide feedback into devs at that time. | 22:05 |
| persia | I also tend to see a good mix of ops, devs, product folk, and others at the larger Days, when I manage to attend. | 22:05 |
| cmurphy | there were ops at the PTG, I'd like to see more of that | 22:06 |
| cdent | One option could be to make the PTG the TG (as technical gathering) and have it be "people who hack on openstack" | 22:08 |
| cdent | that addresses one of the main reasons for creating the PTG (at least as I recall it): stop making me split my attention with marketing | 22:09 |
| persia | I like representatives of operator orgs at PTG, but the vast majority of such representatives I encountered were also devs, which I think is appropriate. | 22:09 |
| persia | Whereas when I go to ops events, there are also representatives of orgs that have no dev component (so might be lost at PTG), who talk about other things. Often these conversations enable those who don't attend PTG to successfully argue for funding to be devs. | 22:10 |
| *** ykarel has joined #openstack-tc | 22:10 | |
| persia | I would be opposed to having folk at PTG who represent promotional or business interests of operator organisations: I think that would change the conversation from "two people are working on something and sit down and do it" to "my company and your company want something: how do we partner". | 22:11 |
| persia | Note that I think all the above applies to vendor organisations, although I think the model changes haven't caused them to be as far from the rest of us as may be described for operators. | 22:12 |
| mriedem | we need operators at the ptg | 22:13 |
| *** david-lyle has quit IRC | 22:13 | |
| mriedem | frequently, "sure, we could do x, but how are operators doing/using foo today? if we did x, what is the pain to ops or the upgrade strategy?" | 22:13 |
| mriedem | ^ comes up | 22:13 |
| mriedem | my guess is, ops need devs at the ops meetup | 22:14 |
| mriedem | i sat in the public cloud wg meeting on monday(?) at the ptg and i was the only dev i think, the rest were mostly operators, and i think it really really helped to chug through their big spreadsheet of doom that they've pushed for the last 18 months | 22:14 |
| dhellmann | cdent : the attention split wasn't just between hacking and marketing for me. | 22:15 |
| mriedem | fewer devs are going to the summit | 22:15 |
| dhellmann | I thought ttx said it well when he said the PTG is for teams to look inward and the Forum is for teams to look outward | 22:16 |
| mriedem | first 2 days at the ptg are totally outward looking things for teams | 22:16 |
| mriedem | being cross-project | 22:16 |
| dhellmann | mriedem : only if you consider cross-project teams not teams | 22:16 |
| mriedem | thursdays have become the defacto cross-project vertical days at the ptg too | 22:16 |
| mriedem | dhellmann: not following | 22:17 |
| dhellmann | mriedem : if their focus is on things they are doing, they are looking inward | 22:17 |
| dhellmann | "what are we doing this cycle?" vs. "what are all of the other demands and inputs coming our direction?" | 22:18 |
| mriedem | as far as i can tell, monday and tuesday at the ptg is nearly the exact same as the "ops summit" first 2 days of the design summit | 22:18 |
| mriedem | except less structured since we don't do 40 minute blocks | 22:18 |
| persia | I think 90 minute blocks are better than 40 minute blocks for most of the discussions that we want to have anyway. | 22:19 |
| * mriedem realizes he has stepped into the age old ptg quagmire and tries to get out | 22:19 | |
| persia | Especially as projects can elect to schedule two (or more) short items in a block if they fit. | 22:19 |
| persia | mriedem: It's sticky for a reason :) | 22:19 |
| dhellmann | I think we could have both types of focus in a single event, but we would have to be very careful with scheduling | 22:20 |
| dhellmann | we could probably even have it concurrent with the marketing stuff, but that would likely result in spending less time on any of the things we do at the PTG or Ops meetups now | 22:21 |
| persia | I do not believe having it concurrent with Expo would work for folk funded by vendors. | 22:21 |
| persia | Works *great* for operators, second-tier support folk, consultants, etc. | 22:22 |
| mugsie | mordred: cool. I was wondering about splitting up OpenStack components on that diagram, but figured it might be too much | 22:22 |
| mriedem | dhellmann: what you just said sounds like you described the summit before the ptg happened | 22:23 |
| persia | Whereas Forum+Expo mostly works because when vendor-funded folk have to go talk up product, it doesn't mean the entire project in which they are involved goes in a direction they didn't expect (at least not without another chance to be involved in the discussion a couple months later). | 22:24 |
| cdent | Let's have the old summit but without presentations, the marketing floor, customer meetings or managers | 22:24 |
| cdent | and call it "PTG" | 22:24 |
| dhellmann | mriedem : yeah, I'm thinking back to San Franciso (I think?) where we had a couple of days of marketing, 1-2 days of overlap, and a couple of days of just design summit | 22:24 |
| dhellmann | that wasn't ideal, but it was better than 100% overlap | 22:24 |
| persia | cdent: +1 | 22:24 |
| mriedem | i believe i was in diapers when that happened | 22:25 |
| mriedem | ATL was my first summit | 22:25 |
| dhellmann | folsom was a lot smaller in terms of # of people, too | 22:25 |
| dhellmann | I forget the argument for changing the schedule. Maybe teams felt like the number of days they had weren't enough? | 22:26 |
| persia | dhellmann: I think we tried to follow that model until Paris. After that, there was too much demand for time to do the workshops, give the presentations, and keep the expo open longer for more meetings. | 22:26 |
| dhellmann | yeah | 22:26 |
| dhellmann | that seems likely | 22:26 |
| persia | Expo+Forum needs 4 days, which makes the entire event much more than a week to do user discussions, then marketing, then developer discusions | 22:27 |
| persia | I think it doesn't quite need two weeks, but it is close. | 22:27 |
| dhellmann | having to meet with customers and partners privately (vs. in the design summit or forum sessions) is the biggest drawback I see to combining the events | 22:27 |
| persia | Unavoidable if we seek funding that way, and we're unlikely to upset the economic model in the next year enough that customers just hire devs to do stuff directly. | 22:28 |
| mriedem | if the ptg is killed, i'll be giving talks, sitting with customers and managers at the summit (as i already have to do today) and then whatever forum sessions i can attend (with only videos later for the talks i actually *want* to attend), and then no deep dive on things to actually get done for a cycle | 22:28 |
| mriedem | and midcycles are likely dead as a result of the ptg killing them | 22:28 |
| dhellmann | yeah, I think a lot of our high-level contributors will have that same issue | 22:28 |
| persia | mriedem: I strongly suspect that would be true for enough of your team that someone would suggest you all fly to $city for a few days (meeting rooms provided by $org), and sort it out. | 22:29 |
| dhellmann | fwiw, we need to make sure our managers hear these arguments from us, not just the foundation | 22:29 |
| cdent | dhellmann++ | 22:29 |
| persia | For those of you funded by vendors, if you have trusted customers who are *also* contributors, having those folk pass the message through professional services can also be helpful. | 22:30 |
| *** david-lyle has joined #openstack-tc | 22:30 | |
| persia | Along the lines of "We were delighted to work with $person at PTG: they really helped us get some of our core concerns understood by $team. We're looking forward to the next one, and hope you can send more folk." | 22:30 |
| * persia may have used such testimonials to fund travel in the past, and hopes the same model may be useful to others | 22:31 | |
| dhellmann | persia : good idea | 22:36 |
| smcginnis | FWIW, I was just encouraging a couple of actively involved operators last night to attend at least the first two days of the PTG. Assuming we still have them. | 22:37 |
| smcginnis | Not that I want it, but if we did have to kill the PTG, having a summit with 3 days for presentations and marketing type activities and 3 days for forum/design summit with a day of overlap would be better than nothing. | 22:39 |
| smcginnis | The key part being trying to avoid as much of the overlap as possible. | 22:39 |
| smcginnis | Having a 3 day summit already seems strange. Especially when your travel to get there and back takes 3 days. | 22:39 |
| *** david-lyle has quit IRC | 22:40 | |
| smcginnis | So even if we don't kill the PTG, I wouldn't mind an extended Forum. | 22:40 |
| *** mriedem is now known as mriedem_afk | 22:41 | |
| dhellmann | ++ | 22:42 |
| dims | right smcginnis | 22:42 |
| cdent | smcginnis for exalted leader | 22:45 |
| smcginnis | Hah :P | 22:48 |
| cdent | oh my, look at the time | 22:50 |
| smcginnis | cdent: It is awfully late for you. And you can't even claim jetlag. | 22:52 |
| cdent | I was writing actual code and got distracted | 22:53 |
| cdent | but yeah, nighty night | 22:56 |
| *** cdent has quit IRC | 22:56 | |
| openstackgerrit | Graham Hayes proposed openstack/governance master: Clarify testing for interop programs https://review.openstack.org/521602 | 22:56 |
| mugsie | zaneb: ^ should have incorporated all your comments | 22:57 |
| *** ykarel has quit IRC | 23:07 | |
| *** david-lyle has joined #openstack-tc | 23:26 | |
| *** david-lyle has quit IRC | 23:42 | |
| *** hongbin has quit IRC | 23:45 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!