Monday, 2023-02-20

JayFhttps://review.opendev.org/c/openstack/governance/+/872769 was landed while I was out with COVID16:19
JayFI would've voted -1 to it16:20
JayFand I do not believe it represents the consensus from the PTG16:20
JayFWhat is the action I need to take to get this re-evaluated / voted on before the full TC? 16:20
JayFcc: gmann ^16:20
JayFIt's my believe the "OpenStack" prefix was added post-agreement16:21
JayFe.g. I agree with 2023.1 (Project x.y.z) 16:21
JayFbut not OpenStack 2023.1 (Project x.y.z)16:21
JayFas it will create confusion for projects and utilities that can be use standalone from OpenStack16:21
opendevreviewJay Faulkner proposed openstack/governance master: Revert "Add guidelines about naming versions of the OpenStack projects"  https://review.opendev.org/c/openstack/governance/+/87443916:25
JayFhttps://governance.openstack.org/tc/reference/house-rules.html#rolling-back-fast-tracked-changes I reverted it per ^^ tc-members please land this per policy 16:26
JayFty rosmaita, I'll propose a revert-revert for us to have a discussion on once that lands16:48
rosmaitasounds good16:48
opendevreviewMerged openstack/governance master: Revert "Add guidelines about naming versions of the OpenStack projects"  https://review.opendev.org/c/openstack/governance/+/87443916:55
opendevreviewJay Faulkner proposed openstack/governance master: Revert "Revert "Add guidelines about naming versions of the OpenStack projects""  https://review.opendev.org/c/openstack/governance/+/87444616:56
JayFI started a discussion ^ there with a comment I would've put on the original change had I not been ill. Thanks for circling back o/17:01
gmannJayF: checking17:57
gmannJayF: rosmaita: we cannot merge the revert of any motion approved change without formal-vote. any change merged under mentioned 'house-rule' not motion then they are applicable to merge the revert under  fast-tracked category.18:05
JayFI was unaware something could merge with less than a majority of TC members' voting except in cases of fast track18:06
gmannJayF: original change was merged as per the TC majority as per TVC charter motion.18:08
JayFWhat does this mean? > TVC charter motion 18:08
gmannhouse rules are fast tracked things which can be reverted with another TC approval but not motion changes18:08
JayFI do not care about the process of getting there; but I know the policy that was landed while I was sick is harmful and will alienate Ironic users and contributors18:08
JayFand whatever we need to do to get that resolved is my concern18:09
JayFI'm happy to follow a path you lay out to get there if the one I'm taking is not correct18:09
gmannthis change has formal-vote and required vote4 as per motion and https://review.opendev.org/c/openstack/governance/+/87276918:09
gmannJayF: I am concern on revert was merged by breaking the process and motion rule18:10
gmannwe want to correct the doc/ update is another discussion 18:10
JayFI will not be bothered if we have to revert the revert directly if there was something incorrect; but it's still not clear to me what the difference is18:10
gmannand original change was open since Feb 0618:10
JayFand bluntly it obviously wasn't clear to rosmaita either18:10
JayFso I don't think this is just a personal misunderstanding18:10
gmannthis talk about only changes merged under this house rules not the motion https://governance.openstack.org/tc/reference/house-rules.html#rolling-back-fast-tracked-changes18:11
*** ralonsoh is now known as ralonsoh_ooo18:11
gmannthe charter motion does not have any revert or fast tracked things https://governance.openstack.org/tc/reference/charter.html#motions18:11
JayFack; I see the difference now. I'm happy to have a revert of the revert landed18:12
JayFbut then I'll need to propose further revision as the policy as it was landed is not acceptable18:12
gmannJayF: yeah, let's do that and yes agree to append/correct or even revert the original with formal-vote, whatever you prefer18:13
JayFSo what's the specific path forward you'd like?18:13
JayFland the revert I already posted? do another revert explicitly saying the first revert was out of policy?18:13
gmannI will say modify the doc as separate change under forml-vote18:14
JayFthat will be step 2, what is step 1?18:14
gmannstep1 is to merge the revert-of-revert 18:15
opendevreviewJay Faulkner proposed openstack/governance master: Revert "Revert "Add guidelines about naming versions of the OpenStack projects""  https://review.opendev.org/c/openstack/governance/+/87444618:15
gmannso this is unmodified of revert of revert right?18:16
JayFYes; and I updated commit message to reflect why it is immediately being un-reverted18:17
JayFand commented in the PR with a link to this IRC conversation about why18:17
gmannthanks, I will approve it18:17
gmannJayF: and for merging during any TC not available. I do not merge those unless any urgent change. for example this guidelines one was open for 14 days and I think I have asked for review in two meeting. but still you giving a heads up on waiting till you review and I could have waited.18:18
gmannfor example this one where it is ready to merge as per motion but noonedeadpunk want to review and I am holding until that https://review.opendev.org/c/openstack/governance/+/87223318:19
JayFI never saw the change in the first place; which since it was up since 2/6 is my fault; just bad timing with that and other things in my life (including getting sick)18:19
gmannbut as you know we hardly get all TC votes on all changes it is not possible to wait for all vote once it reach motion required votes 18:20
noonedeadpunk(and we don't require all votes)18:20
JayFI'm surprised that 3 or 4 is enough for a motion; that's where the root of my misunderstanding was18:20
gmannJayF: i understand, sorry about that. but let me check PTG discussion which seems confusing here18:20
noonedeadpunkBut I don't think it's a big deal to revert if there're concerns being raised afterwards18:20
gmannohk, 3 is required which is 1/3 of total TC18:21
noonedeadpunkAs concerns might be raised by some end-user as well once they see new docs and have valid reason why this should have not be done at the first place18:21
JayFYeah TBH my goal is to get it fixed before people made work based on what was written that I'd want to change18:21
JayFI thought the process I found was right; it was wrong; I'll pay closer attention to the tags in the future18:22
gmannnoonedeadpunk: yes we can do but again we need formal-vote motion to follow process unless majority of TC agree to merge ASAP otherwise it is again process violation and more issue if any othwer TC object to revert fast merge 18:22
JayFthe feeling of trying to quickly "catch up" a week of missed work certainly fed into that mistake18:22
noonedeadpunk++18:22
JayFas soon as the revert-revert lands (which gmann said he's landing), I'm going to push up the edit you suggested noonedeadpunk 18:22
JayFand start the clock on a fix :D 18:22
noonedeadpunkJayF: then I won't need to fix a thing in OSA :p18:23
JayFlol18:23
JayFI just don't want metal3 users going into three layers of WTF clicking through a release note for "OpenStack 2023.1 (Ironic x.y.z)" 18:23
noonedeadpunkWell, I'm not saying it's best thing we can do, but it sounded like that would perfectly cover your usecase18:23
JayFyour suggestion was better than anything I had thought of for sure18:24
JayFso I'm on board :D18:24
gmannJayF: noonedeadpunk: approved the revert-of-revert, please propose the modification in doc. I still need to recall the PTG discussion on that after my coffee 18:27
JayFyeah; I re-read that whole PTG discussion, I believe there was some contention even in the ptg discussion18:27
JayFI certainly do not recall, nor believe I would have, agreed to the item listed in the etherpad as "AGREED"18:27
JayFbut that was also a while ago and shortly after I started on the tc18:28
gmanni see different color in AGREE for 'openstack' but let me check via timeline 18:28
JayFgmann: exactly what I noticed18:28
JayFgmann: timeline? can you see edits in etherpad?18:28
JayFthat would be awesome to be able to track changes in etherpad18:28
gmannJayF: you can see via timeslider https://etherpad.opendev.org/p/tc-2023-1-ptg/timeslider#1459618:29
gmannbut in my summary email also it is 'openstack *' https://lists.openstack.org/pipermail/openstack-discuss/2022-October/030954.html18:30
JayFThis is stunning18:30
gmannok, I will check in details and update but this does not stop even propose the modification in what we agreed in PTG18:30
gmannso feel free to propose the change and in parallel we can try to collect the PTG data18:31
JayFGiven the PTG consisted of TC members trying to reach consensus; I'm not sure there's value in looking back versus trying to come to an agreement on something we can all be happy with18:31
opendevreviewMerged openstack/governance master: Revert "Revert "Add guidelines about naming versions of the OpenStack projects""  https://review.opendev.org/c/openstack/governance/+/87444618:32
dansmithI think that's what gmann just said18:33
dansmiththe ptg discussion isn't binding, we can modify what we said there (or what we said recently about what we thought we said then)18:33
gmannJayF: agree. just want to check if we overlooked etherpad while merging the original change but PTG agreement is just a direction to start the proposal. Actual agreement is when we merge thngs as per motion/house rule whicih can be same as PTG or modified or completly different 18:34
JayFyeah, I have an edit open right not to make a proposal to modify what was landed while I was sick18:34
JayFit should be up within the hour18:34
gmannyeah, we have many case where we changed the things agreed in PTG which is all good18:34
gmannsure, thanss18:35
gmannthanks 18:35
*** gmann is now known as gmann_afk18:37
opendevreviewJay Faulkner proposed openstack/governance master: Modify release notes, project docs name guidelines  https://review.opendev.org/c/openstack/governance/+/87448418:38
JayFI think that should be right? Patch is up, added topic "formal-vote"18:38
*** gmann_afk is now known as gmann18:51
gmannyeah18:52
gmannWhile checking etherpad timesliders it seems we agreed on the "OpenStack <release version> (<project> <project vesion>)" even 'openstack' was added ~2 min later but i remember that was something obvious to add as we are adding openstack <release version> ('2023.1') there otherwise openstack <release version> without 'openstack'('2023.1') is more confusing  19:22
gmannwe merged the original change as per PTG discussion19:23
gmannbut as we discussed, let's discuss the further modification in this  https://review.opendev.org/c/openstack/governance/+/87448419:24
fungiJayF: trying to understand the objection... currently https://docs.openstack.org/releasenotes/ironic/zed.html just says "Zed Series (20.2.0 - 21.1.x) Release Notes" and you don't want that title prefixed by "OpenStack" i guess? it does say openstack all over that page anyway because of the theme, presumably you include generated release notes somewhere else too? maybe the tc is only19:47
fungiconcerned that the docs.openstack.org site says "OpenStack 2023.1" instead of just "2023.1" but ironic's release notes when published elsewhere could omit the "OpenStack" and maybe even the "2023.1" too since that's the coordinated openstack version, not the ironic project version? seems like it's more a matter of context19:47
JayFfungi: you understand my concern pretty closely; and I am not of the belief that with the policy as writetn Ironic could format their differently.19:48
JayFWe don't currently publish 2x copies of the Ironic release notes, and I don't want to do that. I don't think Ironic has the appetite for picking up that kind of additional work right now19:48
JayFif we forced Ironic to publish in a second place to get a release notes doc published without openstack branding, I suspect we'd see loss of contributors and users19:49
gmannthat is why I am confused that how 'zed' is less confusing and 'openstack zed' is more ?19:49
gmannlater one is even more clear on what is 'zed' means in ironic doc/release notes19:49
JayF'Zed' is not confusing to standalone users because it's essentially an opaque slug; a preface of openstack is confusing to users who may think the docs only apply to openstack-integrated usage of ironic or similar projects19:51
JayFrather than all uses of ironic or similar projects19:51
dansmithPersonally, I think it's a stretch to think that someone would make that assumption, especially if they're not already making it from the release name being there19:52
fungimaybe it would make more sense to just say that when referring to an openstack coordinated release name or number, you should prepend "openstack" (so "openstack zed" or "openstack 2023.1") but when referring to specific project releases outside the scope of the coordinated release itself it's fine to omit both the word "openstack" and the release name or number19:52
gmannwell that can be clarified at the top pf document that "this is for both case, openstack integrated deployment as well as for standalone deployment"19:52
JayFfungi: that is what my change is attempting to do19:53
dansmithJayF: I don't think it does though19:53
JayFI think there's a perspective difference: as an Ironic user and contributor, with my Ironic PTL hat on, Ironic is an OSS project that is in openstack governance and follows openstack policies, but is used in many ecosystems. Many other projects do not have a strong identity outside of OpenStack.19:53
dansmithJayF: you're referring to "nova 2023.1" in your doc, which doesn't make any sense tome19:53
JayFdansmith: Do you have advice on how to achieve that goal, then? I don't think my change is the end-all-be-all, but I do think there's a problem, as laid out above, that may need solving19:53
fungiit definitely seems like common sense that when talking about a project release (even if it was part of a coordinated openstack release) it's fine not to mention the coordinated release (and so not say "openstack" either)19:54
JayFfungi: ++ I agree, I do not think the docs as they sit in governance repo reflect that currently19:54
dansmithfungi: agree19:54
JayFif I'm reading too much into it, please let me know how you read it19:54
gmannI added that comment in review too, if we want to use openstack release vesion it should be added with 'openstack' otherwise it is more confusing https://review.opendev.org/c/openstack/governance/+/874484/1/reference/release-naming.rst#4819:54
dansmithJayF: so you'd be okay with just adding that sentence then?19:54
fungiso "ironic 21.1.0" or "openstack zed (ironic 21.1.0)" should both be fine, but things like "zed (ironic 21.1.0)" or "ironic zed" would not19:55
dansmith"...and if you're referring to just the nova part of the release that happens to be part of OpenStack 2023.1, then feel free to just use the 29.x.x numbers"19:55
dansmithfungi: exactly19:55
gmannyeah we can add those two option there19:55
JayFI see how that's different from my proposal, but that is also different than what is written ni governance currently19:55
JayFthe idea I put up in that change was suggested by noonedeadpunk in the original revert-revert which is why I started there19:56
JayFI'm not tied to any specific verbiage; just to making sure that Ironic can use version monikers that respect it's independent branding19:56
fungiyes, i agree with you though that what's currently in governance from that motion is a little vague on when it's okay to talk about project releases without talking about them in the context of a coordinated openstack release19:56
gmannlet's add the second option also there, which is what i mentioned in https://review.opendev.org/c/openstack/governance/+/874484/1/reference/release-naming.rst#4819:57
JayFdansmith: would you be +1 to such a change? ^19:57
gmannwell, that was the origin of PTG discussion that what we should use, project version only or with openstack19:57
dansmithJayF: isn't that what I just asked you if you'd be okay with?19:57
gmannand we agreed on using only openstack + project one19:57
gmannbut if w are adding option to use project version alone also that is totally fine19:58
JayFI obviously am not sure or I wouldn't have asked; that's a yes though so I'll take action19:58
fungiso i can understand the concern when interpreting it as saying all projects can only mention their releases in the context of a coordinated openstack release (which wouldn't make sense anyway since not all projects' releases are part of a coordinated openstack release, and some deliverables don't participate in the coordinated release at all)19:58
dansmiththe thing I don't understand is why the language that is there now is being interpreted so strictly.. as it, it *already* refers to nova and neutron by their version numbers and to openstack 2023.1 as a composite, including those20:00
JayFwith the OpenStack coordinated release and brand being placed prominently, and the project name and version being placed in parenthesis20:01
JayFI'd be OK with the option of flipping those around as well20:02
opendevreviewJay Faulkner proposed openstack/governance master: Modify release notes, project docs name guidelines  https://review.opendev.org/c/openstack/governance/+/87448420:07
JayFIt's my desire to have implemented what we discussed here; if I missed the mark please let me know20:07
opendevreviewJay Faulkner proposed openstack/governance master: Modify release notes, project docs name guidelines  https://review.opendev.org/c/openstack/governance/+/87448420:11

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!