| opendevreview | Merged openstack/governance master: Update Ansible SIG co-chairs https://review.opendev.org/c/openstack/governance/+/988435 | 04:44 |
|---|---|---|
| opendevreview | Merged openstack/governance master: doc: Fallback to default font if DejaVuSans.ttf not installed https://review.opendev.org/c/openstack/governance/+/988470 | 05:11 |
| opendevreview | Merged openstack/governance master: Add devstack-plugin-lustre to QA https://review.opendev.org/c/openstack/governance/+/987314 | 05:13 |
| opendevreview | Merged openstack/openstack-manuals master: Add inactive project badge https://review.opendev.org/c/openstack/openstack-manuals/+/987163 | 05:43 |
| opendevreview | Merged openstack/openstack-manuals master: Add inactive flag to project schema and listing https://review.opendev.org/c/openstack/openstack-manuals/+/987164 | 05:43 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Migrate packaging configuration to pyproject.toml https://review.opendev.org/c/openstack/governance/+/989151 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: docs: Use ``literals`` https://review.opendev.org/c/openstack/governance/+/989152 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Fix bug in projects schema https://review.opendev.org/c/openstack/governance/+/989153 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Bump JSON Schema draft version https://review.opendev.org/c/openstack/governance/+/989154 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Make leadership_type a required key https://review.opendev.org/c/openstack/governance/+/989155 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Add release liaisons for PTL-led projects https://review.opendev.org/c/openstack/governance/+/989156 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Use quotes for all people https://review.opendev.org/c/openstack/governance/+/989157 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Allow irc values to be null https://review.opendev.org/c/openstack/governance/+/989158 | 15:12 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Make appointed an attribute of ptl https://review.opendev.org/c/openstack/governance/+/989159 | 15:12 |
| noonedeadpunk | hey folks. Won't be able to join the meeting today. From my side update, that doc changes and acl changes for venus retirement were proposed | 15:24 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Make leadership_type a required key https://review.opendev.org/c/openstack/governance/+/989155 | 15:46 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Add release liaisons for PTL-led projects https://review.opendev.org/c/openstack/governance/+/989156 | 15:46 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Use quotes for all people https://review.opendev.org/c/openstack/governance/+/989157 | 15:46 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Allow irc values to be null https://review.opendev.org/c/openstack/governance/+/989158 | 15:46 |
| opendevreview | Stephen Finucane proposed openstack/governance master: reference: Make appointed an attribute of ptl https://review.opendev.org/c/openstack/governance/+/989159 | 15:46 |
| gouthamr | tc-members: a gentle reminder that our weekly IRC meeting will be hosted here in ~55 minutes | 16:05 |
| gouthamr | agenda is here: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda | 16:06 |
| opendevreview | Stephen Finucane proposed openstack/governance master: pti: Update with references to bindep profiles https://review.opendev.org/c/openstack/governance/+/989178 | 16:38 |
| bauzas | gouthamr: I'll probably need to skip it | 16:55 |
| gouthamr | ack bauzas | 16:57 |
| gouthamr | #startmeeting tc | 17:00 |
| opendevmeet | Meeting started Tue May 19 17:00:20 2026 UTC and is due to finish in 60 minutes. The chair is gouthamr. Information about MeetBot at http://wiki.debian.org/MeetBot. | 17:00 |
| opendevmeet | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 17:00 |
| opendevmeet | The meeting name has been set to 'tc' | 17:00 |
| gouthamr | Welcome to the weekly meeting of the OpenStack Technical Committee. A reminder that this meeting is held under the OpenInfra Code of Conduct available at https://openinfra.dev/legal/code-of-conduct. | 17:00 |
| gouthamr | Today's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Agenda | 17:00 |
| gouthamr | #topic Roll Call | 17:01 |
| frickler | \o | 17:01 |
| gouthamr | courtesy-ping: noonedeadpunk, spotz[m], dansmith, cardoe, mnasiadka | 17:02 |
| mnasiadka | o/ | 17:02 |
| mnasiadka | noonedeadpunk mentioned he won’t be here today | 17:03 |
| gouthamr | ack mnasiadka; missed it if he mentioned it here | 17:03 |
| gouthamr | PSA: please do edit https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee and add to the "Absence" section if you'd like | 17:04 |
| dansmith | o/ | 17:04 |
| gouthamr | not you, mnasiadka - you do that diligently :D | 17:04 |
| gouthamr | alright, small crowd today.. but lots to cover.. let's get started | 17:05 |
| gouthamr | #topic Last Week's Action Items | 17:05 |
| gouthamr | mnasiadka took an action item regarding Ansible SIG's leadership | 17:05 |
| gouthamr | he proposed: | 17:06 |
| gouthamr | #link https://review.opendev.org/c/openstack/governance/+/988435 (Ansible SIG leadership update) | 17:06 |
| gouthamr | it was merged today, ty for reviews... and quickly because we had aligned already in past meetings and the openstack-discuss ML | 17:06 |
| gouthamr | so now you folks can control "ansible-collections-openstack-release" and any other gerrit groups; please do work with #openstack-infra in case you need to be added by them | 17:07 |
| mnasiadka | I can add myself with my infra powers | 17:07 |
| fungi | yep, happy to help there | 17:07 |
| mnasiadka | Unless that’s politically wrong :) | 17:07 |
| fungi | nope, go for it | 17:07 |
| gouthamr | great power, great responsibility and that sort of thing | 17:08 |
| gouthamr | no one dead punk had several action items.. we can pile on more now that he's taken a day off | 17:08 |
| spotz[m] | o/ | 17:08 |
| gouthamr | he meant to push a galaxy publishing role to "openstack-zuul-jobs", and follow up on ACL changes in project-config to allow us to make progress on retirement of venus and deprecation of vitrage | 17:09 |
| gouthamr | we'll check on these next week | 17:09 |
| gouthamr | next, frickler: we mentioned the AC resolution for core reviewers in passing last week | 17:10 |
| cardoe | o/ | 17:10 |
| cardoe | whoops | 17:11 |
| * gouthamr marks cardoe and spotz[m] tardy | 17:11 | |
| spotz[m] | I get distracted easily?:) | 17:11 |
| frickler | gouthamr: yeah, on my todo list | 17:11 |
| gouthamr | :D okay, some more action items: | 17:12 |
| gouthamr | ty frickler | 17:12 |
| gouthamr | project tracker management guide (mostly launchpad): | 17:12 |
| gouthamr | #link https://review.opendev.org/987433 (Add project tracker management guide) | 17:12 |
| gouthamr | this was merged, thanks for the reviews | 17:12 |
| gouthamr | "inactive" projects badge changes: | 17:13 |
| gouthamr | #link https://review.opendev.org/q/hashtag:%22inactive-projects-badge%22+(status:open%20OR%20status:merged) | 17:13 |
| gouthamr | one last change here that needs reviews from openstackdocstheme reviewers | 17:13 |
| gouthamr | this one looks related too? https://review.opendev.org/c/openstack/openstackdocstheme/+/987166 | 17:14 |
| mnasiadka | That one is probably a nice improvement - I’ve had people complaining the link gets them to docs.openstack.org/2026.1 instead back to the project-docs but for proper release | 17:15 |
| gouthamr | ah | 17:15 |
| gouthamr | they're in merge conflict with each other.. so needs a rebase | 17:15 |
| gouthamr | thanks for working on this mnasiadka.. love the visibility when we need it | 17:16 |
| mnasiadka | Ok, will do, once one of them gets merged :) | 17:16 |
| cardoe | They've got a +1 from me. | 17:16 |
| gouthamr | alright, that's all the action items i was tracking.. was anyone working on anything else to note here | 17:17 |
| gouthamr | i'll note that there's no PQC related channel yet (mharley[m], fyi) and bauzas is trying to kick everyone out of #openstack-agentic-work flows (intention, don't join :P) | 17:18 |
| gouthamr | so he can register the channel | 17:18 |
| mnasiadka | fun | 17:20 |
| gouthamr | if you downloaded a calendar invite for this meeting | 17:20 |
| gouthamr | please download it again | 17:20 |
| gouthamr | #link https://meetings.opendev.org/#Technical_Committee_Meeting (TC meeting ical) | 17:20 |
| gouthamr | the event was updated again | 17:20 |
| gouthamr | alright, let's move to the next topic: | 17:21 |
| gouthamr | #topic Updates to PTI (stephenfin) | 17:21 |
| gouthamr | stephenfin's been proposing a number of governance changes relating to the PTI (mostly fixing outdated info in the python PTI) | 17:22 |
| gouthamr | stephenfin: here? and would you like to take us through this? | 17:23 |
| stephenfin | Sure. There are a few closely related changes here. | 17:23 |
| stephenfin | 1. requirements.txt files are slowly going away in Python packaging. One of the replacements is dependency-groups. I've proposed updating the PTI to capture how to use them and to try ensure projects use similarly named dependency groups, just as we use similarly name tox testenvs etc. | 17:25 |
| stephenfin | tbc, I haven't covered the reason for dependency-groups themselves in the changes since the relevant PEPs already do that. The focus here is purely cross-project consistency | 17:26 |
| gouthamr | ack, this looks good to me.. what's the timeline around the removal? and are dependency groups supported in all python versions in our current PTI? | 17:27 |
| clarkb | PBR directly supports requirements files so the main removal to worry about is probably support in tools like pip to install requirements file? | 17:28 |
| stephenfin | No idea. I suspect not soon, since it's still a widely used pattern, but we have been burned before when it comes to the PyPA | 17:28 |
| stephenfin | clarkb: yes exactly | 17:29 |
| stephenfin | dependency-groups do admittedly come with some advantages also, particularly when it comes to granularity of dependencies for different types of thing. For example, you can have different targets for functional tests vs. unit tests. This is handy when building images. | 17:30 |
| stephenfin | but I haven't suggested mandating their use for now. For now, we just want to have consistency as some projects start adopting them | 17:31 |
| gouthamr | ack, that's good direction.. ty for proposing it | 17:32 |
| stephenfin | Moving on. 2. cardoe has noted that we would benefit from codifying the same expectation wrt bindep and its profiles. This already exists in people's heads, so it's just a case of writing it down. I've another patch up to do that. | 17:32 |
| cardoe | Overall this is a good way of codifying / programmatic-ifying our dependencies in one place/way vs convention. | 17:33 |
| fungi | yeah, we already have an established pattern of matching up bindep profile names to tox env names, for example | 17:34 |
| fungi | just because it makes things like writing ci jobs easier | 17:34 |
| gouthamr | +1 | 17:35 |
| stephenfin | cardoe: Yes, exactly. By writing this down, we get this knowledge out of people's heads. It's also likely useful for agents. (iirc the outcome of the TC PTG session on this was that we should avoid agent-focused docs in favour of human docs that we can point the agents to) | 17:35 |
| stephenfin | 3. I've noted that the golang PTI is massively out-of-date, and references tools that no longer exist plus patterns that don't make sense in modern Go. I've proposed adding a warning to this effect, but an alternative would be to just delete the doc and recreate it if/when someone wants to use Go again. I'm okay with either approach | 17:36 |
| gouthamr | yes | 17:36 |
| gouthamr | +1'ed the warning for now | 17:36 |
| gouthamr | deletion is fine as well.. because it's useless to have outdated/unused information | 17:37 |
| fungi | right, it was written 9 years ago and then essentially never exercised since the need for it quickly evaporated | 17:37 |
| cardoe | I'm good with nuking it as well. | 17:37 |
| gouthamr | okay any objections to dropping the golang doc from our PTI? | 17:39 |
| mnasiadka | If nobody is using it, let’s just drop it? | 17:39 |
| stephenfin | the only argument I can see for keeping it is that someone refreshing it is likely to be an easier/less contentious lift than what will then look like a wholly new proposal, but that's a social/communication issue, not a technical one | 17:39 |
| fungi | the main benefit it has provided is serving as an example in case anyone wants to add another language in the future | 17:39 |
| fungi | it's been useful to point at, but we can also just point at old copies from the git repository history too | 17:40 |
| gouthamr | i meant to draft a resolution and updates to the "adding new languages" guide.. we have context and can share it with people that want that | 17:40 |
| gouthamr | +1 | 17:40 |
| stephenfin | as in, people will discuss the merits of supporting go and rehash arguments from 9 years ago rather than assuming that to be a settled debate and focusing on the actual change | 17:40 |
| stephenfin | though fungi makes a good point about providing a baseline for != python languages too | 17:40 |
| stephenfin | I can propose a follow-up to just remove it and discuss there | 17:41 |
| gouthamr | ty stephenfin | 17:41 |
| gouthamr | anything else for $topic? | 17:42 |
| stephenfin | no, that's all | 17:42 |
| gouthamr | okay, i'll go a bit out of order and pull this one from you: | 17:42 |
| gouthamr | #topic Mandating use of pyproject.toml files (stephenfin) | 17:42 |
| cardoe | Let's just +2/+w this one and move on. :D | 17:43 |
| cardoe | is my support too obvious there? | 17:43 |
| spotz[m] | hehe | 17:43 |
| gouthamr | stephenfin: a lot of repos have been adopting this.. thanks to you/tkajinam/others doing this work | 17:43 |
| stephenfin | as I've written on the agenda, 95% [*] of projects now use these files. Having all projects use them helps contributors the same way the rest of the PTI helps | 17:43 |
| gouthamr | i guess its non-controversial for project teams.. the lift though is in gradually moving more things into pyproject.toml | 17:44 |
| stephenfin | [*] not an actual figure, but it is very high at this point | 17:44 |
| fungi | the sooner we do it, the sooner we'll (eventually) have it in all stable branches and be able to finally pin setuptools when it breaks us again | 17:44 |
| stephenfin | I would like to do this via the requirements check job. That's enabled for all openstack deliverables | 17:44 |
| tkajinam | you are talking about the "minimum" pyproject.toml, right ? | 17:44 |
| tkajinam | not the one with full contents replacing most of the setup.cfg things | 17:44 |
| stephenfin | tkajinam: Yes. I don't see a reason not to do the whole thing at this point, but the minimal one gets us build isolation which is the most crucial aspect | 17:45 |
| cardoe | build isolation as a first step would be huge for cleaning up builds | 17:45 |
| gouthamr | yes ^ although, some of this, our hands are being forced to make a migration because of the packaging/tooling.. it'd be lovely to be proactive | 17:45 |
| tkajinam | yeah I totally agree with the full migration. stephenfin, I agree. I just wanted to make sure that I understand correctly the baseline you are referring to in that "95%" | 17:46 |
| fungi | from a release tooling standpoint, we should be fine since we've (afaik) replaced all prior use of `setup.py ...` with calls to pyproject-build | 17:46 |
| cardoe | worth noting that PyPA has setup.cfg planned for deprecation. They've just not formally said its deprecated NOW so start your timer. | 17:47 |
| fungi | so now we essentially rely on the setuptools implicit fallback in cases where there is no pyproject.toml file | 17:47 |
| stephenfin | cardoe: that shouldn't affect us since we've got pbr | 17:47 |
| tkajinam | makes sense | 17:47 |
| cardoe | stephenfin: we'll have to pull in more code from setuptools into pbr | 17:48 |
| stephenfin | nah, iirc we don't use any of setuptools' ini-parsing logic. pbr's logic all massively predates that | 17:48 |
| fungi | not that much code, just some additional metadata options that aren't plumbed in pbr yet, right? | 17:48 |
| stephenfin | clarkb and fungi will give you a good history lesson on it sometime if you ask nicely | 17:49 |
| stephenfin | fungi: I think *everything* in setup.cfg is handled explicitly by us, but maybe don't quote me on that | 17:49 |
| fungi | remember that pbr's setup.cfg existed long before setuptools grew one of its own (both inherited from distutils, originally) | 17:49 |
| stephenfin | In any case, just because we can continue to support it doesn't mean we should. Many (most?) projects have migrated to pyproject.toml now so we just need to catch the laggards up | 17:50 |
| clarkb | the one corner case is if you use a setuptools setup.cfg directive that pbr won't handle but setuptools will | 17:50 |
| clarkb | but otherwise yes I think it is papered over | 17:50 |
| stephenfin | gouthamr: I'm not sure how to have a "vote" on this one, given my proposal here is to enforce via the requirements-check job (which lives somewhere like openstack/openstack-zuul-jobs or openstack/releases, iirc) | 17:52 |
| gouthamr | stephenfin: yes, so if it's possible to call this out explicitly in the PTI, we should | 17:52 |
| stephenfin | do I just update the PTI again to say all projects must have a pyproject.toml file? | 17:52 |
| gouthamr | and then reference that in the requirements/zuul changes | 17:52 |
| stephenfin | jinx. okay, I'll do that too | 17:52 |
| gouthamr | ++ | 17:53 |
| gouthamr | anything else for $topic? | 17:53 |
| stephenfin | no | 17:53 |
| gouthamr | okay, we're short on time.. we'll try to keep rolling.. | 17:53 |
| gouthamr | the next couple of topics are intertwined | 17:53 |
| gouthamr | #topic Keeping release liaisons up to date (Duplicate info for DPL teams, inactive liaisons) | 17:53 |
| gouthamr | i was raising an issue where the release team maintains a list of release liaisons | 17:54 |
| gouthamr | and it frequently goes out of date | 17:54 |
| gouthamr | and the governance repo tracks release liaisons only for DPL teams | 17:54 |
| gouthamr | this is a problem, because we don't really know who to ping/poke when we run into release issues.. the release team's pings go into the ether when the person tagged in the release repo is no longer involved with the project | 17:55 |
| gouthamr | #link https://opendev.org/openstack/governance/src/branch/master/reference/projects.yaml | 17:55 |
| gouthamr | #link https://opendev.org/openstack/releases/src/branch/master/data/release_liaisons.yaml | 17:55 |
| gouthamr | while i was thinking of this problem, stephenfin had a solution for it | 17:55 |
| gouthamr | #link https://review.opendev.org/c/openstack/governance/+/989156/1 | 17:55 |
| gouthamr | i'm happy with the solution myself, if this is acceptable to the release team | 17:56 |
| gouthamr | we can orient project maintainers towards keeping project.yaml updated | 17:56 |
| gouthamr | and forget about this inconsistency | 17:57 |
| cardoe | +1 from me | 17:57 |
| stephenfin | the governance change shouldn't be blocked by the release team, given we benefit from having a single source of truth regardless of what they want. However, I will work with them on this and would be (very) surprised if they had objections to this | 17:57 |
| gouthamr | frickler: fungi: you're on the release team and here :) any early feedback? | 17:58 |
| stephenfin | (I would also like to move the branch status data currently stored in openstack/releases to openstack/governance, but that's a job for another day) | 17:58 |
| fungi | there are a couple of openstack/releases changes proposed which depend on it | 17:58 |
| fungi | and yeah, we've got longstanding consensus that deduplicating it to avoid divergence would be good | 17:59 |
| gouthamr | great! | 17:59 |
| fungi | mainly as long as we don't move things from releases to governance that blocks the release team's ability to get work done, it should be fine | 18:00 |
| gouthamr | +1.. | 18:01 |
| gouthamr | this should help, not hurt.. | 18:01 |
| gouthamr | will resolve issues like iurygregory telling us he needs to be release liaison, but ironic's DPL doesn't mention him in the governance repo.. i'd just add him in governance | 18:01 |
| fungi | we just don't want to be in a situation where typical release processes are held up waiting for the tc to approve governance changes | 18:01 |
| fungi | where liaisons are concerned, the release team has a relief valve it can open anyway | 18:02 |
| gouthamr | no, i'd be in favor of this list being covered under "project_updates" - will need 2 tc votes and no waiting period if there's some ack from the release team and/or the project team itself | 18:02 |
| gouthamr | oops | 18:02 |
| gouthamr | time check | 18:02 |
| opendevreview | Stephen Finucane proposed openstack/governance master: pti: Remove Go doc https://review.opendev.org/c/openstack/governance/+/989189 | 18:02 |
| gouthamr | looks like we're trending towards agreement here | 18:02 |
| gouthamr | thank you for bringing these topics stephenfin, and for working on all of this! | 18:03 |
| gouthamr | does anyone want to note anything else for the minutes today? | 18:03 |
| stephenfin | nw | 18:03 |
| * gouthamr googles what that means | 18:03 | |
| gouthamr | https://www.reddit.com/r/fo76/comments/bzh56j/what_does_nw_mean/ lol | 18:04 |
| gouthamr | thank you all for attending.. we'll be back here next week.. | 18:04 |
| gouthamr | #endmeeting | 18:04 |
| opendevmeet | Meeting ended Tue May 19 18:04:29 2026 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 18:04 |
| opendevmeet | Minutes: https://meetings.opendev.org/meetings/tc/2026/tc.2026-05-19-17.00.html | 18:04 |
| opendevmeet | Minutes (text): https://meetings.opendev.org/meetings/tc/2026/tc.2026-05-19-17.00.txt | 18:04 |
| opendevmeet | Log: https://meetings.opendev.org/meetings/tc/2026/tc.2026-05-19-17.00.log.html | 18:04 |
| gouthamr | oh, "no worries" | 18:04 |
| spotz[m] | Thanks all | 18:05 |
| * gouthamr slaps himself with a large trout, takes a coffee break | 18:05 | |
| cardoe | So stephenfin since you're still around... I'll do this here for visibly. | 18:08 |
| cardoe | What do we think of changing some of the behavior of python-openstackclient in the future and standardizing the arguments around project scoping? | 18:08 |
| stephenfin | That's what gtema has done with their rust CLI. I've no objections to it, but it would be a large, potentially breaking change | 18:10 |
| cardoe | Well we can take a couple of approaches here. | 18:11 |
| cardoe | So for example "openstack port list" vs "openstack server list" defaults. If you're an admin role you get all ports but you don't get all servers. | 18:11 |
| cardoe | We can standardize adding "--all-projects" along with "--project" to commands and not changing the default behavior. | 18:12 |
| cardoe | Some commands also inconsistently have "--project-domain" when they take "--project" | 18:13 |
| opendevreview | Stephen Finucane proposed openstack/governance master: pti: Insist on minimal pyproject.toml https://review.opendev.org/c/openstack/governance/+/989193 | 18:13 |
| spotz[m] | Adding a flag needed kind of does change the default if it's required, if not required ignre | 18:13 |
| stephenfin | The --project-domain thing shouldn't ever happen. If those options are missing, that's a bug | 18:14 |
| gtema | there are 30 options now already, let's add 31th to rule them all. | 18:14 |
| stephenfin | / missing feature, depending on how you want to look at it. An agent would do a good job of finding/fixing all of those | 18:14 |
| cardoe | I think they were mostly in plugins to OSC and not in OSC itself. | 18:15 |
| stephenfin | Ah, that would make sense. This is what you lose by them not all being in a single mono-repo | 18:15 |
| stephenfin | (not technically a mono-repo but you know what I mean) | 18:15 |
| fungi | i would consider those inconsistencies bugs worth fixing in the plugins where they occur | 18:16 |
| stephenfin | Agreed. The `add_*_domain_option_to_parser` helpers from `openstackclient.identity.common` should probably go into osc-lib for this | 18:19 |
| cardoe | Well I've stuck my Claude on re-implementing python-ironicclient's OSC commands using openstacksdk and not its own client. Once that's good enough I'm gonna propose them to python-openstackclient and openstacksdk for the missing bits. | 18:26 |
| cardoe | Then I'll rewrite the standalone ironicclient that folks want to use openstacksdk as well. And we'll just deprecate the standalone API library. | 18:27 |
| cardoe | If gtema adds websso support to the Rust CLI I'll use it. | 18:27 |
| gtema | it is there since a year | 18:28 |
| gtema | but since Keystone sso design is so crap it is not the greatest UX | 18:29 |
| cardoe | https://github.com/gtema/openstack/issues/1673 afaik was the issue around it which is opened still | 18:30 |
| cardoe | ah I see https://gtema.github.io/openstack/auth.html#v3websso | 18:30 |
| cardoe | alright I'll give it a shot. | 18:30 |
| gtema | looking forward to your feedback | 18:31 |
| cardoe | So that leads me to my next question | 18:49 |
| cardoe | Who controls the schema of clouds.yaml? | 18:49 |
| gtema | sort of sdk team, but please also consider https://github.com/gtema/openstack/discussions/1384 | 18:50 |
| cardoe | Yep. I agree with that. | 18:51 |
| cardoe | But we should have the schema for that somewhere on openstack.org | 18:51 |
| cardoe | There's a number of issues with that file | 18:51 |
| gtema | there are number of issues with authn/authz in OpenStack as such, but people are always afraid of touching anything when I talk about that | 18:52 |
| gtema | the schema is constructed from https://opendev.org/openstack/openstacksdk/src/branch/master/openstack/config/schema.json + https://opendev.org/openstack/openstacksdk/src/branch/master/openstack/config/vendor-schema.json + regular untyped stuff | 18:53 |
| opendevreview | Stephen Finucane proposed openstack/governance master: Update tooling for required deliverable_type https://review.opendev.org/c/openstack/governance/+/989217 | 20:11 |
Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!