Tuesday, 2018-03-13

persiaclarkb: For clarity, you mean after announcement of end of support?  As an example, I believe Ubuntu LTS has examples at both 58 months and 63 months.00:29
clarkbpersia: I mean no longer than the upstream support period whatever that may be00:30
clarkbfedora's moves around because it is based on getting the n+2 release out for example00:30
persiaApologies: I think I failed to ask the question in a way that makes sense.  As alternatives I meant "release day + support claim on release day" vs. "release day until EOS day".  My first read of your response made me think you might mean the former, but your detail on fedora now makes me think you mean the latter.00:32
clarkbya I think we would support it until the actual EOS day even if that changes from the original plan upstream. Either because next fedora release is slow to go out (which has happened) or ubuntu decides they are done supporting a release early (I don't think this has ever happened)00:34
*** mriedem has quit IRC00:36
*** liujiong has joined #openstack-tc00:50
persiaclarkb: Based on last updates shown at old-releases.ubuntu.com, I beleive the last update to 6.06 LTS was 2011-04-21 (58 months).  There exists some chance there just weren't any bugs the last couple months, of course :)  I don't expect it will happen again, and some releases as the other way (e.g. apparently 66 months for 10.04).00:56
*** liujiong has quit IRC01:08
*** harlowja has quit IRC01:10
*** liujiong has joined #openstack-tc01:16
*** zaneb has quit IRC01:54
*** zaneb has joined #openstack-tc01:57
*** rosmaita has quit IRC02:49
dhellmannclarkb : that seems like a detail we should write down somewhere so people aren't surprised, but I don't think the resolution is necessarily the place for it. Maybe as a note in the stable team section of the project team guide?03:07
*** kumarmn_ has quit IRC03:18
*** kumarmn has joined #openstack-tc03:30
*** harlowja has joined #openstack-tc03:48
*** kumarmn has quit IRC04:20
*** kumarmn has joined #openstack-tc04:33
*** harlowja has quit IRC05:59
*** david-lyle has quit IRC06:08
*** cdent has joined #openstack-tc08:46
*** jpich has joined #openstack-tc08:59
ttxhttps://wiki.openstack.org/wiki/Technical_Committee_Tracker just updated09:00
cdentgood morning tc-members09:00
ttxcdent: o/09:00
ttxLooks like we are slowly getting closer to resolution on the EM and the interop test location proposals09:00
* cdent nods09:03
ttxDoug proposed to track conversation starters at https://etherpad.openstack.org/p/tc-office-hour-conversation-starters09:03
* johnthetubaguy waves hello09:04
cdentAs I understood the idea wasn't to use the etherpad as yet another tracker, but more as a way to list things that maybe we should chat about.09:06
*** liujiong has quit IRC09:07
ttxAlright then09:07
ttxRe: tracking with Storyboard, I plan to create stories for TC $stuff09:07
ttxWhat is the ideal level of granularity ? I was thinking trying to keep it reasonably high-level09:08
ttxusing https://etherpad.openstack.org/p/PTG-Dublin-TC-topics as a base09:09
ttx(and the vision)09:09
ttxMaybe I should first list potential stories on an etherpad before creating them, see what that level of granularity would look like09:10
johnthetubaguyWhere you thinking a story per project, or a task for each project?09:10
ttxdepends what you mean by project09:10
johnthetubaguyyeah, I wasn't sure when I wrote that09:11
johnthetubaguyI think I mean per governance project09:11
ttxbrainstorming at https://etherpad.openstack.org/p/rocky-tc-stories09:13
johnthetubaguyah, now I understand your question better09:14
cdentShould we talk today about tony's concerns with the powerstackers group? I feel like I'm missing some critical info.09:16
ttxcdent: happy to09:17
johnthetubaguyah, so vendors already have Power KVM products, which are different to PowerVM products?09:18
ttxThis team is working on PowerVM support09:22
ttxWe suggested that its scope should be on all Power, since "PowerStackers" sounds so much better09:22
ttxBUT that failed to take into account that some other people are working on POWER arch support09:23
ttxand those are not really in the Power[VM]Stackers team09:23
ttxAt this point it's probably just simpler to align the mission/name with the current team scope09:23
ttxIt's a cheap rename anyway09:24
cdentThat doesn't seem like it is really addressing tony's concerns which is that we have groups that cluster around brands rather than projects. (At least it seems like that's what he's saying. I'm not certain)09:25
ttxcdent: I think renaming the team and rewording te scope would address Tony's concerns. That is what he proposed after all09:27
cdentRight, it is his proposal but he admits in his response to me that it is a compromise09:28
cdentand it may be a fine compromise09:28
ttxIt moves from clustering around a brand (Power*) to clustering around a technology (PowerVM), which is probably better09:29
ttxAt one point I was wondering if those (WinStackers or Power[VM]STackers would not be better set up as SIGs09:30
ttxsince they could also band more directly with their users09:30
cdentBut there are two types of groupings: topical (nova, ceilometer, etc), branding  (power, etc)09:30
ttxagree that the verticals conflict a bit with the Guild set up09:30
cdentI personally think (I've just read the great virt drive split thread from 2014) that grouping by brand could have some advantages09:31
ttxPowerVM and Winstackers (or VMWareStackers if that is ever proposed) are really transversal guilds09:31
ttxWould actually make a lot of sense as SIGs09:31
cmurphythat does seem to fit a little better09:32
ttxthey still own repos, but I think it's ok as long as those are... peripheral/optional repos09:33
ttxLike the Security SIG maintaining bandit09:34
ttxor winstackers maintaining Hyper-V support plugins09:34
ttxWinstackers was actually next in my long list of SIG candidates09:34
ttxBut I was OK with PowerStackers being set up as a team first, and then convert both to SIGS :)09:35
ttxsince conversion takes some amount of convincing09:36
* ttx looks for other guild-like topic-driven or brand-driven or tech-driven project teams09:36
ttxSecurity was already converted...09:37
ttxI could see Rally turn into a Performance SIG, but it makes sense as a project team as well09:38
ttx(as long as it keeps its Rally-centric scope)09:38
ttxso yes, WinStackers / PowerStackers are the outliers in the current project team list09:39
cmurphyrally as a SIG doesn't seem right to me, that seems like leading down a road where all horizontal projects become SIGs09:39
ttxIf InteropWG turns into a SIG I would just fold RefStack under it09:40
ttxsince it's really a tool for the WG, not part of the openstack deliverables09:40
ttxcmurphy: yeah...09:40
ttxcmurphy: I think it has SIG potential when it's both an upstream and ddownstream concern09:41
ttxso QA or release management are very upstream, would stay as project teams09:42
ttxSecurity on the other hand is much better as a SIG, since expertise is all over09:42
cdentttx: Is "degree of upstream" the right heuristic? I always thought it was breadth of potential participants09:43
ttxPerformance, or Documentation... it's a bit more fuzzy I guess09:43
cmurphythat's a really difficult line to draw09:43
ttxcdent: breadth of potential participants is a much better way to describe it09:43
cdentSIGness was supposed to be a way to get different people in the same room09:43
ttxthat is what I actually meant :)09:43
* cdent contacts the ministry of information09:44
ttxcmurphy: yeah, opens up the Pandora box a bit, which is why I haven't been pushing too hard to convert WinStackers to a SIG yet09:44
ttxBut I feel like that group would be a lot more relevant if it as the OpenStack-on-Windows group rather than the maintainers of the hyperv plugin09:45
ttxIdeally the ask should come from them09:45
cmurphywinstackers fits as a SIG in my mind because it really is "special interest", it's a bit niche09:46
ttxcmurphy: yes, you either care or you don't09:46
cmurphya performance SIG doesn't make sense to me becaue everyone should be at least somewhat concerns about performance09:46
ttxthat's a good way of looking at it yes09:46
cdentgeneral interest instead of special interest?09:47
ttxhmm, not really, but the idea that perforamce should be everyon's concern rather than just a specific team's09:47
cdentI was kidding09:47
cdentI do, however, want to be sure that we don't wield SIGs as a multi-tool/cureall kind of thing09:51
ttxIt's not really a question of whether the topic is important for everyone or not... It's more a question of how specific is the group working on it. Does it take a special kind of person. As much as I think that everyone should care about security, I realize it takes a specific mindset09:51
ttxcdent: agreed, we just added that new construct and need to check that everything that was supposed to benefit from transitioning to it already moved09:52
ttxor decided to keep using the construct they prefered09:53
ttxThe main issue with SIGs at this point is communications, due to the silly split of MLs09:53
ttxbut that's not really a SIG-specific issue, just hinders their progress09:54
ttxas it hinders all dev/ops artifical barrier reduction efforts09:55
johnthetubaguyI think I had always made the distinction more on deliverables, do they deliver python code, but that seems incorrect...09:56
ttxjohnthetubaguy: SIGs can own repos and produce code, but in vase of code that would be more support/optional code09:57
ttxsince the production of "OpenStack" is under the TC's responsbility, and uses the Project Team structure09:57
cdentI think we've got at least three interpretations, which is both not suprising also propbably fine09:57
johnthetubaguyttx: Right, that seems to put winstackers and power(vm)stackers in governance, I think09:58
ttxIt's one thing to have a Security SIG maintaining bandit. It's another to have a WinStackers SIG producing nova-hyperv09:58
*** dtantsur|afk is now known as dtantsur09:59
johnthetubaguyor os.win or whatever it is09:59
johnthetubaguyso I think this makes me +2 on tonyb's patch... is that where other folks are thinking?10:00
ttxI agree that's pushing it a bit too far, which is why despite WinStackers's guild-like / brand-centric approach I refrained from suggesting that they transition to SIG yet10:00
ttxjohnthetubaguy: I'm fine with it as well10:00
cdentI'm neutral, but I'm not really sure why I feel that way.10:03
cdentproabably because matthew is neutral10:04
cmurphywhether or not this group moves to a SIG, the complaint about its name and declared scope is valid10:04
cmurphyit's not inclusive of existing work that it claims to include10:05
johnthetubaguyI am leaning on the side of correctly representing the reality of the group10:06
cdentso we got in this state because we wanted to _permit_ the group to be inclusive of more stuff10:06
cdentand now we're saying "it's not already in there, so the name is no good"10:07
johnthetubaguythe Power KVM folks have little overlap with the Power VM folks, in code and group make up. Asking them to join forces doesn't seem to help anyone at this point, or am I missing someone?10:07
cdentIf that's true, then yeah, a rename is right, but we should try to learn our lesson (whatever it is) from this quick turnaround.10:08
johnthetubaguyI totally forgot about that distinction, I always forget how many options there are with Power10:22
ttxsmcginnis: do you have anyone who stepped up to fill the "Designate contributors" help-most-needed item that the Foundation should send a thank-you letter to ? (asking you since you are the TC sponsor for that item)10:48
ttxflaper87: do you have anyone who stepped up to fill the "Glance contributors" help-most-needed item that the Foundation should send a thank-you letter to ? (asking you since you are the TC sponsor for that item)10:48
smcginnisttx: No, I never had anyone that contacted me about designate.10:51
flaper87ttx: yeah, smcginnis did :D10:53
flaper87and, TBH, the entire glance community became more active10:54
flaper87the few there are10:54
smcginnisThat's true. It's a small team, but it really seems like they've been able to pick things up again.10:54
ttxok, so nobody in particular. Thanks!11:07
*** kumarmn has quit IRC12:47
*** mriedem has joined #openstack-tc12:51
*** rosmaita has joined #openstack-tc12:57
fungiclarkb: what we had suggested about stable support timeframes is that it's highly unlikely infra will commit to building/running images for distro releases which fall out of (distro upstream) maintenance12:58
fungiif there are people interested in those ancient branches, that means they in theory have some supported platform they're interested in continuing to run them on12:59
fungiso it's up to them to work with infra to get images building for those platforms if they want them to continue to be tested in our ci system12:59
fungithey might also opt for running third-party ci systems reporting on changes for the respective extended maintenance branches if that makes sense13:00
*** kumarmn has joined #openstack-tc13:04
cmurphythat made a little more sense when the proposal involved an explicit handoff13:05
mugsiettx: I have a few people that came to me and stepped up for Designate, I can send on a list and what they did13:07
ttxmugsie: that would do it, yes. thanks!13:09
smcginnisWould be good to recongnize responses to that "most wanted" list to try to encourage more.13:10
dmsimardfungi: It probably becomes a bit sensitive to have first party "support" for upstream unmaintained distribution images. Even if someone maintains it... If we host it, we're probably the ones on the hook and liable for issues/security.13:12
fungidmsimard: yep, which is why we would only want to do that for distro versions which are actually being maintained, just acknowledging they may not be the same distros we started testing those releases on in the beginning13:14
dhellmannttx, cdent : the idea with the etherpad was that not every "conversation starter" is necessarily a "task" to be completed or tracked13:28
*** Ryushin7BX8NR has joined #openstack-tc13:32
zanebI've been thinking a bit about the project vs. SIG question in the context of the Kingbird thread...13:33
zanebit seems to me that successful projects are defined by their solution-space13:34
zanebif you have a group that's defined by the problem-space, then you should be asking why it's not a SIG instead of a project13:34
*** Ryushin7BX8NR has quit IRC13:35
cdentthat's potentially useful13:37
cmurphywhat's the difference between a problem space and a solution space? those seem like two sides of the same coin to me13:39
cmurphyi'm trying to form an articulate question about the difference and i can't even form it without the phrase "solving a problem"13:40
*** jroll has quit IRC13:41
*** jroll has joined #openstack-tc13:42
*** dtantsur is now known as dtantsur|brb13:45
dtroyercmurphy: I see it as solutions can fix more than one problem, problems may have more than one solution.  Scope depends on which one you pick13:45
dimshey y'all13:49
*** nzfrenetically has joined #openstack-tc13:50
smcginnisMorning dims. Snowed in yet?13:50
dimssmcginnis : 4 inches or so as of now. possibly 12-16 by the time it wraps up this evening13:52
smcginnisdims: Sounds like a snow day for the kids. ;)13:53
dimsy the teenager is not even awake yet :)13:54
*** nzfrenetically has quit IRC13:55
zanebcmurphy: e.g. in the context of Kingbird, their problem-space is "multi-region". but there isn't a coherent solution to 'fix' multi-region, it's a bit of work on a lot of different solutions (images in multiple regions, keypairs in multiple regions, &c.). IMHO it's a mistake to try to form a _project_ around that14:02
cmurphyzaneb: what about winstackers? is their problem-space OpenStack on hyperv or is their solution-space the hyperv-related deliverables they produce?14:05
zanebI don't know because I really don't care about hyperv ;)14:07
zanebbut I'd say that if they have one deliverable, or related set of deliverables, (e.g. a hyper-v driver for Nova) then it's a project14:08
zanebif it's about making every user interaction in OpenStack hyperv-friendly (I don't think it's about that) then it's a SIG, because you can't fix every project in OpenStack by starting your own project to compete with bits of all of them14:10
fungigranted, that's what we told the tricircle team, which is why they split off kingbird (unless i'm misremembering the history there)14:17
*** dtantsur|brb is now known as dtantsur14:24
*** hongbin has joined #openstack-tc14:31
openstackgerritJeffrey Zhang proposed openstack/governance master: Move kolla-kubernete project under governance of openstack tc  https://review.openstack.org/55253114:33
*** annabelleB has joined #openstack-tc14:40
clarkbfungi: ya its fine if they want to migrate onto some supported platform they are interested in keeping going. Just want to make it clear we likely won't run unsupported releases even if that is the desire (and I expect that there will be a non zero desire for that just based on knowing what random people run in production)14:49
fungiclarkb: i fully agree we should make that clear. maybe https://docs.openstack.org/infra/manual/testing.html is begging for a new section about images?14:58
clarkbya that likely is the best location for this info14:59
fungidmsimard: ^ maybe you're interested in trying to draft that?15:08
*** david-lyle has joined #openstack-tc15:25
*** harlowja has joined #openstack-tc15:46
*** david-lyle has quit IRC15:57
*** dklyle has joined #openstack-tc15:57
*** harlowja has quit IRC16:06
*** dtantsur is now known as dtantsur|afk17:18
openstackgerritJames E. Blair proposed openstack/governance master: Remove Zuul from OpenStack governance  https://review.openstack.org/55263717:22
cdentthat causes me to have a sad and a happy at the same time17:25
smcginnisI was thinking the same thing.17:28
smcginnisKind of like a kid growing up and moving off to college?17:29
fungiexcept it got a scholarship so we don't need to worry about coming up with tuition17:59
fungijust beer money17:59
*** jpich has quit IRC18:07
dmsimardAre "top-level projects under the OpenStack Foundation" documented somewhere ? Mostly curious.18:10
cdentdmsimard: I believe they are supposed to be responsible for that themselves. I'm not aware of a unified view. ttx probably has more accuate info.18:12
dmsimardcdent: I guess my question is about the "split" between openstack projects, openstack foundation projects and then projects that are not in either but still hosted by the foundation (don't want to spark a stackforge debate)18:14
dmsimardSo I guess this would be the foundation's third project ? Between OpenStack, Kata and Zuul.18:14
fungiyeah, not counting the edge computing strategic focus area which last i heard still has no code projects, only working groups18:14
dmsimardOpenStack has it's projects, Kata has it's own and then Zuul has others.18:15
cdentyes, but at the foundation layer they are something like strategic areas: edge, infra, ci/cd, containers18:15
clarkbcdent: ya that18:15
dmsimardah, interesting18:15
fungii expect zuul will likely copy the formatting of openstack's governance model for tracking repositories, mainly so that they can identify our constituent electorate and be able to reuse some of the election tooling we've developed in openstack18:17
dmsimardSo then, theoretically speaking, a new project could be "hosted" (or "governed") by any of these top level projects18:17
fungiit's unknown at present whether there will be a unified project governance across all projects in a given strategic focus area, or multiples in parallel18:18
cdentdmsimard: yeah, that's one of the topics we didnt quite reach at the PTG: how, when someone applies to be an official openstack project do we guide them to a foundation friend if that makes better sense18:18
fungiright now the only mature example we have for precedent is openstack, which is the only project within the open infrastructure strategic focus area18:18
fungiall of this is at such an early stage i think we're all afraid to smother it in too much policy and process but rather allow it to evolve a bit more organically until we see the problems/roadblocks which demand more structure18:19
clarkbalso worth noting I don't think there is a stackforge debate. As of today if you aren't an openstack project but are hosted by infra you are part of stackforge iirc18:20
fungidefining too much structure up front could result in steering these projects in the direction we predict they would go rather than in the direction which is actually best for them18:21
dmsimardclarkb: Yeah, that's where I was getting at by talking about "openstack, kata, zuul ... and then the 'others'". I'm trying to stay up to date on these kind of changes so I'm not surprised by something I didn't see coming for ara ;)18:21
fungimy only debate wrt "stackforge" is that i'd like to not reuse that name, since it already has baggage/connotation for a lot of people18:22
smcginnisI think it's a very good idea not to impose unnecessary structure or process, at least at this point.18:22
clarkbdmsimard: ya and kata largely isn't hosted by infra (we host their mailing lists, that is it)18:22
fungii think we've sufficiently burned the name "stackforge" such that people will make (likely somewhat incorrect) assumptions about what we're talking about if we call something "stackforge"18:22
smcginnisIt would be interesting to have some sort of high level Foundation view though that shows all the things under its governance.18:22
smcginnisIf for no other reason that curiosity.18:23
clarkbdmsimard: so that is decoupled from infra even further18:23
dmsimardImposing anything is never fun for anyone anyway. We can suggest things like we did to Kata for their CI and infrastructure, for example, that's totally fine.18:23
smcginnisWe stored the "stackforge" bits somewhere in the big tent, right? :)18:23
dmsimardsmcginnis: The linux foundation has such a list: https://www.linuxfoundation.org/projects/18:23
fungii like being able to come to new projects and say "here's a menu of really amazing resources you can choose from if you want, no hurry, you can always decide later that you want to start using something too"18:24
smcginnisdmsimard: Yeah, something like that would be nice.18:24
smcginnisAlways good to have a nice pretty page of logos.18:24
fungii find that lf logo wall pretty useless18:25
dmsimardsmcginnis: the openstack foundation has a pretty page of logos, it's here :p https://www.openstack.org/foundation/companies/18:25
fungibut i suppose some people relate to it18:25
cdentOn a page like that (of project), we'd have just the OpenStack O, none of the internal project logos, right? (I hope)18:25
fungiyeah, our companies page is more about providing a place for member companies and donors to be able to point and say "see, we're helping!"18:25
smcginniscdent: That's what I would think. Just top level.18:26
smcginnisThe "O" could link to our little menagerie on another page.18:26
dmsimardfungi: the project list is kind of the same thing, "look, I'm a X foundation project!"18:26
cdentyes please18:26
fungithe lf page is definitely "let's see how impressive a logo count we can make this" (no cncf logo, but there's logs for each of the subprojects under the cncf)18:27
fungino, wait, it's worse than that18:28
fungia logo for the cncf _and_ logos for all the projects under the cncf18:28
fungiso i really don't feel that page usefully documents anything18:29
persiaIt is a measure of success problem.  If an organisation has a business model where "success" includes having many well-funded projects, having a project-based logo wall that shows that a given community looking for a legal entity would be in good company to join is useful.18:36
persiaIf an organisation has a business model where "success" includes ensuring that members have a neutral collaboration environment, or members have access to world class infrastructure, or something similar, such a page is less valuable.18:37
persia(and re: cfcf+cfcf "projects": technically, "cncf" itself is defined as an LF project (with appropriate management, controls, etc.), but that project happens to include a facility to initiate projects (unlike most LF projects).)18:38
persiaOh, and if an organisation has a business model dependent on donated funding, hardware, staff, etc., then "success" is usually driven by ensuring donors are well-credited, making it very important to have segmented logo pages for different classes and types of donation.18:40
pabelangerI'm preparing the Condorcet voting for the S release poll, and curious if there is a list of flags I should enable for the poll?19:35
pabelangeror are the defaults (for public) enough19:35
pabelangerfungi: mordred: ^19:35
fungipabelanger: you can always try a draft one and try it out from a few machines or something19:37
pabelangerfungi: yah, that is the current plan now19:38
pabelangerokay, so it seems voting results are live, after votes have been casted now with public19:43
pabelangernot after poll has been closed19:43
smcginnisHmm, is that maybe another option? I kind of like not having influenced votes.19:46
pabelangersmcginnis: yah, we can only expose results to specific email addresses19:47
pabelangerif we do that, there is no public record to users and we rely on officials to say what the results are19:47
*** annabelleB has quit IRC19:51
openstackgerritKiran Thyagaraja proposed openstack/governance master: Add ansible-role-k8s-rabbitmq  https://review.openstack.org/55267319:53
*** annabelleB has joined #openstack-tc20:01
pabelangerokay, seems public results while poll is open, likely fits with the idea of public voting20:11
pabelangerso, I'll keep working on that assumption for now20:11
*** annabelleB has quit IRC20:16
*** annabelleB has joined #openstack-tc20:16
*** annabelleB has quit IRC20:26
*** harlowja has joined #openstack-tc20:37
fungithere's no option to only make results public after the poll closes? that's what we do with the non-open polling for elections20:39
fungihrm, seems there is no explicit option for it, so must be the default behavior (maybe only for private polls?)20:42
pabelangerfungi: yah, no option. And think is the default behavior of private20:43
fungiunfortunately having the results updated in real time while the poll is still open raises the risk of stuffing, as smcginnis suggests20:44
*** cdent has quit IRC20:44
smcginnisI think I would rather keep it private if we have to make that trade off.20:44
pabelangerI don't think I have all the bits in place to work with a private poll in the next ~3hours20:46
pabelangerI'd need a list all our email addresses20:46
pabelangerand would likely need to confirm with mordred on what sort of split was used in the past20:47
pabelangerwe could say results are limited to a single email address, then after the poll closes make the unique URL to view results public20:48
pabelangerI did confirm I got a separate email with URL to voting results20:48
diablo_rojopabelanger, when we do the ptl and tc elections we only enable detailed ballot reporting20:50
pabelangerdiablo_rojo: k, let me try that setting. I didn't toggle that one yet20:51
diablo_rojopabelanger, but dont toggle the next radio box that appears below that unless you want to show in the results who voted for what20:53
pabelangerno luck, results still public before poll closes20:53
pabelangeryah, I only selected top level option20:54
diablo_rojopabelanger, we have the electorate generated for the ptl elections if you want that to use for the private poll20:56
diablo_rojothe actual generation of the rolls takes a long time, but I still have the files from the most recent election20:56
pabelangerdiablo_rojo: yah, I believe part of the reason for trying public polling this time around, was the list of email address was so large, it took a week to just to email everybody their private URL.  But, I am unsure if mordred was aware that results of a public poll could be viewed before it was closed20:59
pabelangerso, I'm a bit conflicted how to proceed ATM20:59
diablo_rojopabelanger, ah got it. I know CIVS has a max number of people you can add at a time, but I don't think its ever taken a week to add the electorate for the TC elections (which I suspect is the same as that for release naming? )21:00
pabelangerdiablo_rojo: I'd have to defer to mordred. But would think TC and release naming used the same source for emails21:01
smcginnisI guess if it's between hiding the real time results and getting the voting out there and a name chosen, for now I would choose the public poll and we can evaluate afterwards how it went and if we think it caused issues and needs to be changed for T.21:02
clarkbrelease naming was open to all foundation members iirc21:02
clarkbnot just TC electorate21:02
pabelangersmcginnis: yah, that is my feeling too21:03
zanebpabelanger: there's no way to vote 'strategically' in a condorcet poll, so I don't think it's a big problem that you can see results. I guess in the worst case it might encourage more people to try voting multiple times?21:04
smcginnisI was thinking of cases where I prefere name A more than name B, but I go to vote and see B is leading so I get discouraged and think there's no way A will win so I might as well vote B too.21:05
*** annabelleB has joined #openstack-tc21:16
pabelangerzaneb: sure, but if feel in a public poll, if results were private or public, that wouldn't deter somebody from voting multiple times.21:16
fungipabelanger: consider testing the private results option. i expect it's just going to e-mail you a url to the results, which you can then publish to the ml21:28
fungiwouldn't be much different than what we do for elections21:29
fungithe only difference is that with elections and a private poll, casting a ballor tells you where the results url will appear once the poll is closed21:29
pabelangerfungi: yah, that is right. I get a dedicated URL to view results. Would can then be shared with people via ML. So, that is also an option. I just means a 2nd URL people need to use to view the results over just a single once the poll closes.21:50
*** mriedem has quit IRC21:55
*** kumarmn has quit IRC22:02
*** kumarmn has joined #openstack-tc22:20
*** kumarmn has quit IRC22:25
fungiwell, even with our normal elections there are separate ballot and results urls (by necessity)22:35
*** kumarmn has joined #openstack-tc22:37
*** Guest78602 is now known as amrith22:52
*** kumarmn has quit IRC22:58
*** hongbin has quit IRC22:59
*** kumarmn has joined #openstack-tc23:07
*** kumarmn has quit IRC23:11
*** kumarmn has joined #openstack-tc23:15
*** kumarmn has quit IRC23:20
pabelangerfungi: okay, I'll proceed with results only viewable with my email address, then post the URL to public ML with results23:32
fungithat seems sane23:32
*** mriedem has joined #openstack-tc23:56
*** dklyle has quit IRC23:57
pabelangerand voting URL sent to MLs23:59

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