Tuesday, 2026-03-10

mnasiadkaspotz[m]: There are questions in the Scientific SIG (and in the mail thread on ML) if the RDO community has enough contributors to maintain old releases like Caracal or Epoxy and should they expect there will be updated packages if a CVE is uncovered? I think a simple message on the RDO main page would be extremely helpful. Sorry to add to your plate :-)05:13
opendevreviewStephen Finucane proposed openstack/governance master: Retire requestsexceptions  https://review.opendev.org/c/openstack/governance/+/97981511:55
spotz[m]So some folks might be stepping up to help, I spoke with folks from both Alma and Rocky this weekend and they don't have the resources either. I'll see what I can do about getting an update added to the page. I sent the email out before I left for SCale and just got back and I leave next week for Kubecon. I think there will be an email from someone on the engineering side to answer some questions going out shortly as I honestly can't14:14
spotz[m]answer them.14:14
spotz[m]I'm doing everything I can to help, y'all know that14:16
mnasiadkaspotz[m]: Thank you - I was just relying messages :)15:03
spotz[m]I know:)15:03
gouthamrtc-members: a gentle reminder that the weekly IRC meeting will be held here in ~50 minutes16:11
gouthamrthe agenda is here: https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee#Next_Meeting 16:12
gouthamr#startmeeting tc17:00
opendevmeetMeeting started Tue Mar 10 17:00:32 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
opendevmeetUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.17:00
opendevmeetThe meeting name has been set to 'tc'17:00
gouthamrWelcome 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
gouthamrToday's meeting agenda can be found at https://wiki.openstack.org/wiki/Meetings/TechnicalCommittee17:00
gouthamr#topic Roll Call17:00
dansmitho/17:01
frickler\o17:01
bauzaso/ (fried but here)17:01
gouthamrcourtesy-ping: noonedeadpunk, spotz[m], cardoe, mnasiadka17:02
cardoeo/17:02
gouthamrnoted absence: t o n y b17:02
mnasiadkaO/17:02
noonedeadpunko/17:02
spotz[m]o/17:03
gouthamrgreat, that's everyone.. let's get started17:03
gouthamr#topic Last Week's AIs17:04
gouthamrwe took an AI around process - we'll talk about chair/vice-chair/liaisons today17:04
gouthamrhere's the poll to pick a time for this meeting:17:05
gouthamr#link https://rallly.co/invite/FCz849xYhriU (Poll for TC meeting - 2026.2)17:05
gouthamrthis is somewhat suitable for the western hemisphere.. 17:06
noonedeadpunkum... how to set the timezone to utc in there...17:06
gouthamri enabled an option to see it in your timezone :D is that confusing?17:06
noonedeadpunkit really is17:06
bauzasoh, timings are in local TZ ?17:07
noonedeadpunkespecially given that there's a saving time incoming17:07
gouthamrsigh, sorry - i can fix that17:07
bauzasso I need to calculate it correctly then (europeans have daylight savings only in 2 weeks)17:07
noonedeadpunkbut also all other meetings I have I've converted to UTC, so some metal math is needed to acocunt for that and daylight saiving17:07
gouthamrgood stuff, hold off from taking this, i'll fix it in a little bit17:08
bauzasI'm used to count in UTC, even sheeps at night17:08
fungikeeping my computer's time set to utc helps trick annoying applications like that into using utc as my "local" time17:08
noonedeadpunkhehe17:08
gouthamri wanted to gauge interest in continuing our alternate meeting17:09
noonedeadpunkit's weird then I can't select UTC from the list explicitly17:09
noonedeadpunkneither gmt+017:09
spotz[m]I'll admit I went with the later times as my mornings are already booked. I could do later on Friday but there's always the possibility I'm traveling or PTO on a Friday17:09
bauzasI have a specific no-Friday afternoons meetings hard policy that I usually enforce :)17:10
mnasiadkaOn Friday evenings my brain melts :)17:11
mnasiadkaAnd I support what bauzas said :)17:11
noonedeadpunk++17:11
spotz[m]I save it for emergencies not scheduled regular meetings17:11
gouthamrwe hosted an alternate meeting once every month at 0800 UTC.. 17:11
gouthamrdo we continue that? has it helped?17:12
dansmithany chance  we could wait until after the DST chaos ends to pick a meeting time17:12
spotz[m]Oh true that, I already missed one meeting today17:12
dansmithbecause I've got a few others that conflict (like the one I'm on now) that might change during/after the DST flip17:12
* noonedeadpunk constantly forget when the morning meeting actually happens17:13
gouthamrdansmith: ^ we could; so i could move that poll to occur in the first week of April17:13
bauzasnoonedeadpunk: if we don't change anything, next one will be next week17:13
dansmithgouthamr: just seems like the worst time to ask people what time is good for them :) (not your fault of course)17:14
mnasiadkagouthamr: I don’t believe it helps, and the person that we’ve done that for is mostly awol?17:14
spotz[m]He was in the board meeting this morning17:15
gouthamri am unsure, i haven't attempted chatting with tonyb lately either.. i honestly hoped to pick up on reviews, chatter here and on the ML etc.. (hoping this ping will elicit a response during a decent time for him :D) 17:15
gouthamrokay, any objections to keeping the Tuesday 1700 UTC meeting until April? this means that we meet next week at this time (NOT at 0800 UTC), and do the same on 24th and 31st of March17:17
mnasiadkaAnd usually I have to remind everybody that this is the week we’re having a EU morning meeting :)17:17
spotz[m]+117:17
mnasiadkaNo objections from me and I’m happier because I don’t have to run the meeting :)17:17
gouthamrhaha, easy questions17:18
noonedeadpunk+117:18
dansmith+117:18
bauzaslet's keep the existing quorum for now17:18
gouthamrokay, hearing no objections, lets move to the next AIs17:18
gouthamrthese were regarding leaderless teams17:18
bauzas(even if that means drinking beers during TC meetings after savings)17:19
gouthamrwe can dive into them at detail in a bit, i've up-wip'ed most of the changes on gerrit17:19
gouthamri'll guide them through merge conflicts when the election results are posted17:19
gouthamrthe results are due in ~8 days 6 hours 17:20
gouthamr#link https://governance.openstack.org/election/ 17:20
gouthamrin the meantime, a reminder to review these open changes:17:21
gouthamr#link https://review.opendev.org/q/hashtag:%222026.2-leaderless%22+(status:open%20OR%20status:merged)17:21
gouthamrbarring any exceptions, these can merge after the results change merges - so please post any concerns you have on these patches 17:22
gouthamrwe have an AI from last week that needs a volunteer: initiating the retirement for venus, vitrage17:24
gouthamrwould anyone like to take a shot at this? 17:24
noonedeadpunkI can go with this process17:25
gouthamrty noonedeadpunk 17:25
gouthamrit's a careful song-and-dance, well documented17:26
gouthamr#link https://docs.opendev.org/opendev/infra-manual/latest/drivers.html#retiring-a-project17:26
gouthamr#undo17:26
opendevmeetRemoving item from minutes: #link https://docs.opendev.org/opendev/infra-manual/latest/drivers.html#retiring-a-project17:26
gouthamr#link https://docs.openstack.org/project-team-guide/repository.html (Retiring an OpenStack Repo)17:26
noonedeadpunkyup, thanks. I did that bunch of times already, but don't remember by heart17:27
gouthamrnice17:27
gouthamrsome scope creep for you: fungi brought up additional pypi cleanup recently17:27
noonedeadpunkI think it's not automated?17:27
gouthamrthey now have a way to tag packages as retired.. so that might be a good thing to follow for these 17:27
noonedeadpunkand needs some manual archival process from ones who has an interactive access to pypi?17:28
gouthamrand take the learning to do that with all retired openstack packages from the past17:28
gouthamryes17:28
gouthamrmanual17:28
fungithere is (very recently) now an api for doing some package management on pypi, but i'm not sure if archival is supported by the api yet17:28
fungithey did add the ability to query collaborator types though17:29
fungiwhich i know would have helped a lot in earlier cleanup17:29
gouthamrty.. alright, no other AIs that i can spot17:30
gouthamrany others that you were working on?17:30
gouthamr#topic Chair/Vice Chair and Liaisons17:31
gouthamrin the past week, we took an AI to add TC chair nominations.. 17:32
gouthamrthere was just one, and so no election is necessary.. 17:32
gouthamrunless there are any objections, dissents? :) 17:33
spotz[m]I'm good:)17:33
gouthamrty for your confidence, please note that i, like past Chairs, have no desire to do this in perpetuity... except that i've energy to pick up some more for now17:35
gouthamrwould anyone like to be vice-chair?17:35
mnasiadkaI can continue if nobody wants to relieve me from that duty17:36
gouthamrty very much mnasiadka 17:36
gouthamralright, you're it! i'll push up a patch after the meeting17:38
gouthamrnext order of business, need to refresh VMT, Election and DPL liaisons17:38
bauzasI can continue to run as a VMT liaison17:39
fungithanks bauzas! gouthamr will you be continuing on that too?17:39
gouthamris anyone interested in these? 17:39
gouthamrbauzas and I were VMT liaisons, tonyb was the last election liaison, and DPL liaisons are here: 17:39
gouthamr#link https://opendev.org/openstack/governance/src/commit/057c04eb426d8e1c57fc7b195f81d5aa2da8ad59/reference/projects.yaml17:39
gouthamrsure can do fungi, i'd like to actually join the VMT though.. so i was going to ping JayF and you in case you need assistance17:40
spotz[m]I can do election17:40
gouthamrty spotz[m]!17:40
gouthamrty for volunteering again for the VMT, bauzas 17:40
fungigouthamr: even better. hit me up any time17:40
JayFWe'll have to see if we have decoder rings and smoking jackets in your size ;) 17:40
gouthamrhaha17:40
gouthamrperfect, if any more liaison interest exists, please do chime in.. 17:41
gouthamr#topic 2026.2 Election17:42
gouthamrlast little bit of governance to-dos for today17:42
gouthamri'll update 17:43
gouthamr#link https://etherpad.opendev.org/p/2026.2-leaderless 17:43
gouthamrwe're converging on a solution for adjutant, mistral, skyline, requirements and oslo17:44
gouthamrnoonedeadpunk will initiate the retirement process for venus and vitrage17:44
gouthamrfrickler will propose a resolution/charter change regarding reviewers being credited as contributors in our election process17:44
gouthamrnoonedeadpunk: has there been any update to share on masakari? 17:45
noonedeadpunkno, not really, I think there was not a meeting this week as it's bi-weekly17:45
spotz[m]Reviewers being included as contributer is a great addition!17:45
gouthamrack noonedeadpunk17:46
fungiyeah, i'm hoping to work on the implementation for that in the election tooling very soon17:46
gouthamr++ nice, ty fungi 17:46
bauzasprovided reviews are not gamified :)17:46
fungibauzas: it's maintainer review activity specifically17:47
fungiso code-review +/-2 and workflow +117:47
fungifor the rare cases we hit in some projects where some core reviewer can't run as ptl because they were only reviewing/approving changes and didn't have any of their own merge17:48
spotz[m]I know that's where I spend a lot of my time so it's appreciated17:48
bauzasfungi: I missed the point then, if that only includes already recognized maintainers17:48
bauzasthen I'm supportive without concerns17:48
fungithat was my takeaway from the discussion in last week's meeting, at least17:48
bauzasmy attention dropped obviously then ;)17:49
gouthamryou're correct, but a resolution and discussion will reflect what we'd land on.. 17:49
fungiyeah, obviously i wouldn't finalize any implementation in tooling until there's a resolution approved17:49
gouthamrthink we're okay with the leaderless teams then, a good thing to wrap this up alongside elections! 17:50
gouthamranything else to note wrt $topic?17:50
gouthamr#topic A check on gate health17:51
gouthamrwe're nearly at the RC1 deadline.. the theme at the tail end has been a bunch of fire fighting with breakages/regressions upstream of us.. 17:53
fungii think the latest disruption was the new tox release?17:53
frickleryes, we did some pinning https://review.opendev.org/c/openstack/openstack-zuul-jobs/+/97911817:53
fungiand then fixed the role that was using the old option to branch its behavior based on which tox version is in use (since old branches need the old syntax on tox v3)17:54
frickleralso noonedeadpunk did https://review.opendev.org/c/zuul/zuul-jobs/+/979136 , still need to verify the pin can be reverted now17:55
fricklerbut in the longer run we might want to discuss moving away from tox17:56
fungiopendev's tools and upstream zuul happily switched to nox years ago, fwiw, instead of upgrading past tox v317:57
fungibut there was definitely a lot of work involved in making that switch, not wanting to downplay oit17:58
fungiit17:58
clarkbI think the primary benefit to that switch was that we continue to use standard packaging tools with nox. Whereas tox has reimplemented its own package build and install tooling17:58
clarkb(and some differences in interpretation of the spec have been found)17:58
fungiand nox vendors in a lot of the tools that it does reuse17:59
fungier, tox i mean17:59
fungiwhich makes it harder to pin them17:59
gouthamrworth doing if we can coordinate it as a community goal perhaps18:00
noonedeadpunkfrom my humble usage of tox, I hardly faced real issues post 4 upgrade tbh18:00
fungiit would not be a painless transition18:00
noonedeadpunkbut learning a new tool doing the same thing - not sure where it stands in list of priorities...18:01
funginoonedeadpunk: right, in opendev and zuul we opted to put the effort into nox instead of tox v4, so there was some reduction in overall cost to the transition at that time18:01
noonedeadpunkwe're running toxv4 in openstack everywhere now I guess18:01
fungiswitching from current tox to nox is harder to justify18:01
noonedeadpunkto it's just cost18:01
* gouthamr time check18:01
gouthamrwe're past the hour18:02
gouthamrdo you want to note anything else in the minutes today?18:02
bauzas-18:03
gouthamralright, lets wrap it up here.. ty all for participating18:03
gouthamr#endmeeting18:03
opendevmeetMeeting ended Tue Mar 10 18:03:39 2026 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)18:03
opendevmeetMinutes:        https://meetings.opendev.org/meetings/tc/2026/tc.2026-03-10-17.00.html18:03
opendevmeetMinutes (text): https://meetings.opendev.org/meetings/tc/2026/tc.2026-03-10-17.00.txt18:03
opendevmeetLog:            https://meetings.opendev.org/meetings/tc/2026/tc.2026-03-10-17.00.log.html18:03
fungicorrecting myself wrt tox vendoring its dependencies, it doesn't seem to do that now but it does require and regularly bump the minimum version of a lot of things, particularly virtualenv which is where lots of problems tend to arise, but also libs like packaging, platformdirs, filelock, and the recently cursed chardet18:06
fungiand also fairly aggressively drops support for old python versions18:07
mnasiadkaI’m not sure we can orchestrate a transition easily, it would need to be a community goal - or we can ,,just’’ sort of support both, if that’s an option.18:10
noonedeadpunkit kinda is, but that would bring quite some mess, imo18:13
noonedeadpunkand a headache for occasional contributors18:14
clarkbyes mixing things is probably worse because then you never know why something has broken18:18
fungieven just not fully cleaning up stale references to the old thing in random places is enough to confuse new contributors into trying to use the wrong now-broken thing18:21
fungiwhich wastes a lot of their time and can lead to an unpleasant turnoff experience18:22
clarkbthat said I think it remains important to call out deficiencies in tools and that there are alternatives. It may be the case that tox (or nox or chardet or etc) make changes that are more painful to address than switching to the alternative18:22
clarkbI think it also helps provide feedback to the upstream. THough in the case of tox they have been not super receptive to the feedback on issues like the package building discrepency18:23
clarkbin that particular case they felt that they were spec compliant and that was the end of the story and didn't care that a different spec compliant tool allowed for some different behavior that was activeily used in the wild18:24
sean-k-mooneyclarkb: fungi  for what its worth proting from tox to nox is soemthign ai woudl likely be very good at.19:20
sean-k-mooneyit woudl still need a lot of effort to do that and review it but less then it once might have19:20
sean-k-mooneyalso i suspect in many cases it woudl be pretty forulaic19:20
fungiwell, ai or sed19:20
fungithey're both very similar though19:20
fungii'm personally more familiar with sed, fwiw19:21
clarkbyes I think it can be largely mechanical. The main issue is likely all the special targets19:21
clarkbwhere doing a straight up replacement might want a bit more intelligence19:21
sean-k-mooneyya we do use tox wehre just or make would prorbly be more approtiate19:22
JayFif we did such a thing we could make minor improvements19:22
JayFe.g. the fact that it feels like 50% of lint targets are "lint" and the other 50% are "pep8"19:22
JayFnot sure it's worth it generally, but if we added in more consistency cross project19:22
* JayF loves a good scope creep19:22
sean-k-mooneyJayF: i think linter is actully the althernitive target name but yes that partly inconsitent19:23
sean-k-mooneyfor what its worht i think we are usign tox v4 allready in most case19:23
sean-k-mooneywe just have nto raised our min version19:23
fungiexcept we also do other static analysis in linter (and pep8/flake8) targets too ;)19:23
sean-k-mooneyin the project i mean since we do not cap tox in any way19:23
* fungi suggests new colors for the shed19:24
JayFwe are just tox, even 4.x, is breaky and generally doesn't always present stable apis or behavior... 19:24
JayFclaude suggested the best colors to use for the shed are periwinkle, razzmatazz, mauvelous, or goldenrod /s  19:26
fungithe up side is we have the benefit of an amazing ci system that implements pre-merge testing for changes to test frameworks and even job definitions, with inheritence where needed, so such a transition could be end-to-end tested before it ever merges19:26
sean-k-mooneyboth can also be supproted for a tiem but the mucel memory will kill me if we ever to cahnge althogh i could alias tox=nox perhaps19:31
fungireplace tox.ini with one that runs an echo with a reminder to use nox ;)19:32
clarkbsean-k-mooney: the command syntax is different. nox is a lot more verbose which I don't like. But I do like that it is far more explicit about things like python versions which tox will silently run the wrong python version and report success in certain scenarios19:33
clarkbvs nox which can be forced to use a python version and if that isn't present will fail19:33
JayFI don't like how, e.g. unless we force basepython that docs jobs run with whatever py version tox is installed with.19:33
JayFI assume given the shape of nox's CLI that issue wouldn't exist there19:33
sean-k-mooneyclarkb: i think that changed in tox 4 for what its worth19:33
sean-k-mooneythe basepython stuff is gone19:33
sean-k-mooneyor rather not required i think19:34
sean-k-mooneystephen told me it impoved but i dont recal the details19:34
spotz[m]REDACTED21:26
JayFspotz[m]: change your password :( 21:59
* JayF has done the exact same thing in another #openstack- channel before22:00
spotz[m]yeah already done22:00
fungii'll redact the public logs once they're rotated. editing them before utc midnight (in ~81 minutes) will cause the log to be truncated for the rest of the day22:39
spotz[m]Thanks23:04
opendevreviewGoutham Pacha Ravi proposed openstack/governance master: Add TC chair and vice-chair for 2026.2  https://review.opendev.org/c/openstack/governance/+/97992223:09

Generated by irclog2html.py 4.1.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!