| *** ralonsoh_out is now known as ralonsoh\ | 05:46 | |
| *** ralonsoh\ is now known as ralonsoh | 05:47 | |
| *** ralonsoh_ is now known as ralonsoh | 09:21 | |
| cardoe | I'm good with lazy consensus. | 12:46 |
|---|---|---|
| * frickler isn't sure what exactly "the matter" is. release the quantum name on pypi? or rather handover to the specific person that asked for it? or something completely different? | 12:48 | |
| noonedeadpunk | I'm not sure about handing over to a specific person, but would be totally fine with releasing it | 14:08 |
| fungi | "releasing" a project on pypi isn't really a thing. we can delete it, and then anyone who wants to reuse it would have to work with the pypi admins to free it up, but by default projects that have been deleted can't be registered by anyone | 14:17 |
| noonedeadpunk | ah | 14:18 |
| fungi | for reddwarf i deleted the one empty release that had been used to register the project many years ago, then added the requesting individual as a co-owner after which they deleted us as an owner | 14:18 |
| noonedeadpunk | ok, I actally never tried doing that flow, so unaware if that was possible or not | 14:18 |
| noonedeadpunk | yeah, ok, then we can indeed follow the same flow | 14:19 |
| noonedeadpunk | we'd just need to pass it through some vote process first I'd assume | 14:19 |
| fungi | yeah, i have no opinion either way, just happy to instrument whatever the tc decides as long as it's possible | 14:20 |
| gouthamr | frickler: did you have any objections to hand this over to the specific person asking? | 15:09 |
| gouthamr | we can wait until EOD for anyone else here, and i'll respond on the thread | 15:10 |
| frickler | gouthamr: I only object to fast-tracking the process after 10y of idling. IMO a formal vote would be better for documenting consensus and tracking the decision | 15:14 |
| frickler | maybe "object" is too strong even, I'd just prefer the other solution. if ppl do not want to invest more time, then maybe that's fine, too | 15:17 |
| gouthamr | a formal-vote through a resolution? | 15:17 |
| gouthamr | or here, on IRC? | 15:17 |
| gouthamr | or email? | 15:17 |
| frickler | formal vote would be in gerrit, likely in form of a resolution, yes | 15:18 |
| gouthamr | hmm, looks like a bit of an overkill for me | 15:19 |
| JayF | There is value in having a document in git describing what's happening | 15:20 |
| gmaan | I agree to have a formal vote which actually help as record as well as getting more eyes | 15:32 |
| gmaan | I am sure resolution is needed as must but it can be in meeting or so | 15:32 |
| opendevreview | Merged openstack/election master: Create candidates/2026.1 placeholder directories https://review.opendev.org/c/openstack/election/+/949554 | 16:29 |
| gmaan | * I am *not* sure resolution... ^^ | 16:30 |
| * gmaan hate missing *not* in many sentence, very bad typo | 16:30 | |
| opendevreview | Goutham Pacha Ravi proposed openstack/governance master: [resolution] Relinquish "quantum" project on PyPI https://review.opendev.org/c/openstack/governance/+/949783 | 16:55 |
| gouthamr | haha :) okay, i posted one anyway ^ lets discuss there.. we'll eventually wait until next week because of our 7+ day soak time.. | 16:56 |
| fungi | might be nice to reply on the ml thread noting where the decision is taking place | 16:59 |
| gouthamr | yes doing that | 17:00 |
| fungi | thanks! | 17:00 |
| opendevreview | Goutham Pacha Ravi proposed openstack/governance master: [resolution] Relinquish "quantum" project on PyPI https://review.opendev.org/c/openstack/governance/+/949783 | 17:05 |
| gouthamr | what is it about badges that prevents "tox -e docs" to work properly locally :/ | 17:06 |
| gouthamr | on the governance repo | 17:06 |
| clarkb | what failure do you get? | 17:07 |
| clarkb | looks like the extension to sphinx for the bagdes is in the repo itself so unlikely to be a dependency issue with the extension itself | 17:08 |
| gouthamr | https://paste.opendev.org/show/bAlLHHirb2pmkJJZM6ra/ | 17:08 |
| gouthamr | weird, and i end up commenting https://opendev.org/openstack/governance/src/commit/8cdd19ac689ec9a0ffddcd8dfd704df90ad72d2e/doc/source/_exts/badges.py#L200 | 17:09 |
| gouthamr | to get it to build | 17:09 |
| gouthamr | (and have wondered if it is a /me problem :D) | 17:10 |
| clarkb | gouthamr: do you have an doc/build/badges dir? | 17:11 |
| clarkb | I think that is where it is trying to write the content (and it attempts to mkdir that dir too) | 17:11 |
| gmaan | I did not check what all deps are which version in my local env but it work fine for me | 17:11 |
| clarkb | but there may be clues in there | 17:11 |
| gouthamr | nope, that directory isn't created | 17:12 |
| gmaan | works fine with latest governance source | 17:12 |
| gouthamr | oh, ty.. lemme try this on an ubuntu/Linux VM | 17:12 |
| clarkb | oh it may be doc/build/html/badges | 17:13 |
| gouthamr | yes, that exists, and is empty | 17:14 |
| gouthamr | sigh, this works fine on ubuntu.. | 17:14 |
| gouthamr | its a mac thing, sorry for the alarm; the sphinx extension may have some issue with the env | 17:15 |
| clarkb | the pillow image lib may not find the deps it needs on osx? | 17:16 |
| clarkb | as a "run linux on all the machines at home" person I'm the wrong one to debug that | 17:16 |
| gouthamr | ack, could be.. i'll debug this :) | 17:17 |
| clarkb | I did recently have to fix a zuul testsuite issue beacuse I run the "wrong" flavor of linux locally: https://review.opendev.org/c/zuul/zuul/+/949527 | 17:18 |
| gouthamr | i sense the angst with the commit message | 17:20 |
| clarkb | ya I'm not super thrilled with how ansible works here. Its basically distro roulette | 17:20 |
| clarkb | but the hard part was getting to the point of undersatnding ^ once I did that it isn't so bad and hopefully now I retain that info for as long as I use ansible and its easy to debug later if necessary | 17:20 |
| gouthamr | ++ | 17:22 |
| clarkb | I do think there is a bug in ansible that could be fixed by sanitizing PATH before trying to find the python installs | 17:23 |
| clarkb | that would align with the table lookup distros more closely | 17:23 |
| clarkb | but I don't have the energy to pursue that and this works | 17:24 |
| opendevreview | Goutham Pacha Ravi proposed openstack/governance master: [resolution] Relinquish "quantum" project on PyPI https://review.opendev.org/c/openstack/governance/+/949783 | 17:26 |
| * gouthamr is sure there are some SUSE contributors to ansible.. | 17:30 | |
| gouthamr | don't know how much of a corner case auto python interpreter discovery thingy is | 17:31 |
| clarkb | I think the main issue there is any changes could break people in the opposite direction to me | 17:32 |
| clarkb | One of those situations where if you change it (even if you're more consistent between distros) someone is likely to break | 17:33 |
| fungi | https://github.com/openstack/skyline-console/pull/11 is a massive code dump from rackspace... cardoe: any chance you know these folks and can remind them that openstack (and so skyline) changes go to review.opendev.org? | 18:44 |
| fungi | that pr is going to get auto-closed in a few hours with that information as well, but some human interface might be preferable in this case | 18:47 |
| gouthamr | pretty cool that there's a lot of code coming :P | 18:47 |
| gouthamr | i do hope they break up the submission into separate commits to make things easier to review, and get some engagement from skyline folks | 18:48 |
| fungi | yeah, which is why i'd love to see some compassionate steering/mentoring of the folks responsible rather than just an impersonal form letter that might alienate them | 18:48 |
| gouthamr | true | 18:49 |
| JayF | fungi: honestly that almost looks like a code leak, like someone pushed to github instead of github.rackspace.com | 18:53 |
| fungi | well, it's a merge request from what looks like an intentionally-created repo in the rackerlabs org | 18:53 |
| cardoe | Hmm | 18:54 |
| cardoe | So I believe they forked it to customize it? | 18:55 |
| cardoe | I really have been trying to get them to contribute upstream and make this stuff configurable. | 18:55 |
| fungi | even if the pr was an accident, rackspace having a huge pile of patches on top of upstream skyline sounds classically untenable, and getting them actually upstreamed would probably be better for everyone, yeah | 18:55 |
| fungi | a lot of the commit titles look like bug fixes | 18:56 |
| JayF | you couldn't `ply` me to work with that many patches :) | 18:56 |
| cardoe | fungi: yep. that's what I say. | 18:56 |
| cardoe | Hence why my group is carrying the water on specs and other things. And if we make a local change, it's gotta have a retirement plan attached. | 18:57 |
| fungi | i can personally reach out to folks there who i know are more directly involvd with flex, but figured you might have better contacts/leverage than e.g. me just pestering cloudnull et al | 18:58 |
| fungi | especially since the community has elected you to represent their interests as part of the tc ;) | 18:59 |
| gmaan | ++ and we have seen skyline sometime (manytimes?) struggling with PTL/leaders role also. Next cycle election coming soon can be good chance for other company to help team | 19:01 |
| fungi | looks like mnaser already manually closed it with a similar message | 19:01 |
| mnaser | i've been doing that anytime i get notifications on github for openstack projects :) | 19:03 |
| mnaser | i wasnt sure if the bot was on some webhook or periodic job | 19:03 |
| fungi | appreciated! it's slightly more friendly than the robot doing it | 19:03 |
| mnaser | so i was just doing them adhoc when i see them | 19:04 |
| mnaser | :P | 19:04 |
| fungi | yeah, it's periodic, about once a day i think | 19:04 |
| mnaser | ah that makes sense, i thought it was maybe a broken webhook/job/thing so i have been doing that here and there, good to know | 19:04 |
| mnaser | speaking of which just for transparency there is a couple of us that are members of the openstack org | 19:05 |
| mnaser | https://github.com/orgs/openstack/people | 19:05 |
| fungi | right, the tc had at one point been divvying up responsibility for managing pinned repos and similar cosmetic settings in the openstack github org | 19:06 |
| fungi | if the current tc cares about openstack's github presence i can help get folks added to do those tasks | 19:07 |
| mnaser | likewise, also i was thinking the other day it would be good to enforce 2fa for the openstack org was well | 19:07 |
| mnaser | i think github is about to start enforcing it site wide soon anyways | 19:07 |
| mnaser | supply chain attacks and all, even though technically, github is _not_ the canonical source, people do often go there | 19:07 |
| fungi | i thought github started enforcing it a year or more ago | 19:10 |
| mnaser | https://github.blog/news-insights/product-news/raising-the-bar-for-software-security-github-2fa-begins-march-13/ you're right | 19:11 |
| fungi | we've been talking about ways to do that. you can opt into 2fa in launchpad but i don't think gerrit's classic openid support has a way to check whether the session was 2fa-authenticated since that's more of an oidc claim. making progress on opendev's sso spec for migrating accounts to keycloak.opendev.org gives us that capability though | 19:11 |
| mnaser | yeah i dont think you can know about that | 19:12 |
| fungi | tonyb was helping wrangle some development work on adding an openid bridge for keycloak so it could support treating launchpad as an identity provider, which would be the next step in migrating from lp to an opendev sso | 19:13 |
| fungi | i think maybe that work has stalled, but if anyone else is interested in pitching in, check with tonyb on that i guess | 19:14 |
| JayF | fungi: how much time would be left in that, you think? What language and is it all coding or some devops? | 19:18 |
| fungi | i have no idea, tonyb could probably answer those questions but there was some coding work, i'm just not sure how much progress was made on what | 19:19 |
| cardoe | fungi: oh I have reached out. | 19:20 |
| cardoe | Which reminds me... I'll be on PTO the next two weeks gmaan. | 19:21 |
| fungi | JayF: for reference we do have a working keycloak server but for now its only production use is to authenticate admins for the zuul webui. the spec is here: https://docs.opendev.org/opendev/infra-specs/latest/specs/central-auth.html | 19:22 |
| cardoe | Interrupt my lunch for something I've lectured that group on multiple times.... | 19:22 |
| fungi | i did *not* mean to interrupt a meal. please accept my humblest apologies | 19:22 |
| cardoe | oh you're good | 19:23 |
| cardoe | everyone in here +1 for pinging me. | 19:23 |
| fungi | no sweat, i didn't want it to get missed | 19:24 |
| * gouthamr isn’t gmaan but ack’s cardoe being away for the next couple of weeks | 19:24 | |
| cardoe | heh gmaan said my name and wires crossed. | 19:25 |
| gmaan | :) | 19:26 |
| cardoe | When I get back, that stuff will be in the skyline pipeline upstream. | 19:27 |
| * fungi offers cardoe a clue-by-four or similar lart | 19:29 | |
| cardoe | oh I've got a great lart. | 19:30 |
| fungi | never leave home without it | 19:31 |
| cardoe | So since I don't know a better place to ask. Looking at some skyline changes... I see that pydantic is in there.. I see it in upper-contraints.txt but not in global-requirements.txt. Does that mean it cannot be used by projects normally but skyline is getting around that? | 19:54 |
| clarkb | cardoe: https://opendev.org/openstack/nova/src/branch/master/.zuul.yaml#L760 this template is what is supposed to check requirements but neither skyline-apiserver nor skyline-console include it (I'm not sure if it makes sense for -console) | 19:57 |
| clarkb | so yes they basically aren't checking and that allows them to add whateverwithout the addtional checks | 19:57 |
| cardoe | Why would we have it in upper-contraints.txt in the requirements repo? | 19:59 |
| fungi | usually if it's relied on by multiple openstack deliverable repos and/or if there are jobs run against changes to the requirements repo which relis on a project depending on it | 20:00 |
| JayF | cardoe: if it's in u-c but not g-r it means it's a transitive dep | 20:01 |
| JayF | cardoe: e.g. we have projectX in g-r, projectX depends on pydantic | 20:01 |
| JayF | cardoe: when something is already in u-c, there's usually a lower bar to getting it into g-r | 20:01 |
| fungi | yes, the constraints file is a complete closure of all dependencies (direct and transitive), while requirements lists are just the direct deps | 20:01 |
| cardoe | makes sense | 20:01 |
| -opendevstatus- NOTICE: Setuptools 80.7.0 broke python package installs for many affecting CI jobs. That release has been yanked and it should be safe to recheck failed changes. | 22:01 | |
| tonyb | It's java, I'm probably about 45% done with the coding and then there will be a non-zero amount of DevOps stuff to deploy it | 22:04 |
| JayF | ack; java coding is not something I personally know or have resources to point to it | 22:29 |
Generated by irclog2html.py 4.0.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!