19:00:27 <tonyb> #startmeeting releaseteam
19:00:28 <openstack> Meeting started Thu May 30 19:00:27 2019 UTC and is due to finish in 60 minutes.  The chair is tonyb. Information about MeetBot at http://wiki.debian.org/MeetBot.
19:00:29 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
19:00:31 <openstack> The meeting name has been set to 'releaseteam'
19:00:33 <diablo_rojo_phon> ...or maybe I've shortened it to 4min.
19:00:54 <diablo_rojo_phon> o/
19:00:57 <smcginnis> o/
19:01:00 <tonyb> :)
19:01:03 <fungi> i should look into having my reminders fire 5 minutes early
19:01:12 <smcginnis> tonyb: You were 27 seconds late today.
19:01:21 <tonyb> smcginnis: true
19:01:22 <fungi> being notified the second a meeting starts is probably suboptimal
19:01:32 <tonyb> smcginnis: I'll do better next time
19:01:37 <smcginnis> :D
19:01:52 <smcginnis> fungi: That's why I like ping lists. ;)
19:03:24 <fungi> oh, i got pinged as soon as the #startmeeting was added
19:03:30 <tonyb> We're flying without an agenda today because I've been distracted
19:03:44 <fungi> but my reminders fired at 19:00z on the dot too
19:04:16 <tonyb> Next week is m-1 week
19:05:15 <tonyb> We didn't have any tasks that needed to be done by m1 (well we pushed them back to m2)
19:05:36 <tonyb> The only open one we have is https://storyboard.openstack.org/#!/story/2005702
19:05:44 <tonyb> which diablo_rojo_phon started
19:06:09 <diablo_rojo_phon> Was waiting for feedback before going further
19:06:14 <tonyb> but between kubecon and my questions I figure we should discuss it here
19:06:31 <smcginnis> We have release requests for cycle-with-intermediary to do for m-1.
19:07:17 <tonyb> #link https://review.opendev.org/#/c/660781/
19:07:23 <smcginnis> Process doc also says send reminder about completing responses to community-wide goals, but I don't think we have those yet?
19:07:51 <tonyb> smcginnis: Yeah I don't think they're decided on
19:09:21 <tonyb> So, lets look at diablo_rojo's 660781 quuckly, then come back to the cycle-with-intermediary realeases
19:09:45 <tonyb> smcginnis: Did you/can you look at the review?
19:09:57 <smcginnis> fungi: Is there a better way to store that to be able to link to the accounts? ^
19:10:06 <smcginnis> I did look. It seems a little too... verbose.
19:10:31 <diablo_rojo> I figured by adding the emails it would be easier to find/add them to reviews in gerrit
19:10:35 <fungi> checking the change out
19:10:45 <diablo_rojo> (I was pulling gerrit email addresses)
19:10:50 <smcginnis> Not sure we need separate first and last name fields. And if we don't need a specific value to link to an account, then I'd rather see it all on one line and be able to have a list of them.
19:11:05 <fungi> smcginnis: you want to be able to link to specific gerrit accounts?
19:11:13 <smcginnis> Then maybe just name and email?
19:11:25 <smcginnis> fungi: There was some talk about being able to automate adding folks to reviews.
19:11:42 <fungi> e-mail should be all we need for that, and feed into a gerrit rest api query if you want
19:11:58 <diablo_rojo> I figured the names would be nice for when we need to reach out to them directly
19:12:09 <diablo_rojo> But I can make it just name and have first + last
19:12:14 <diablo_rojo> instead of breaking it up
19:12:51 <diablo_rojo> I just threw something together since I figured everyone would have an opinion about how it should look ;)
19:13:08 <smcginnis> So I think something like:
19:13:11 <smcginnis> liaisons:
19:13:18 <smcginnis> - name: Elmer Fudd
19:13:18 <tonyb> Like I said on the review I like name, email and IRC as data (and have the structure match the ptl record)
19:13:25 <smcginnis> email: elmer@fudd.com
19:13:27 <fungi> i'd need to fiddle with the gerrit rest api for a moment
19:13:30 <smcginnis> - name: Bug Bunny
19:13:36 <diablo_rojo> I can add irc nics too
19:13:39 <smcginnis> email: carrots@bb.com
19:13:56 <tonyb> that means we can do all the things nicely
19:14:11 <fungi> https://review.opendev.org/accounts/fungi@yuggoth.org
19:14:19 <fungi> that should work even anonymously
19:14:22 <tonyb> yeah that's what I do already
19:14:49 <tonyb> fungi: acctually ... https://review.opendev.org/accounts/?q=fungi@yuggoth.org
19:15:14 <fungi> both work for me
19:15:25 <diablo_rojo> SO the next question is do we want them all in a single liaisons file?
19:15:34 <diablo_rojo> Or to be in the deliverable files like we originally said?
19:15:40 <smcginnis> fungi's has the advantage that all we would actually need is an email address. The results of the query has everything else we need.
19:15:46 <smcginnis> Well, I guess not IRC nick.
19:15:49 <fungi> tonyb: yours just gets you the account id number, while mine gets you name and stuff
19:16:11 <tonyb> True
19:16:17 <fungi> smcginnis: we have an osf rest api for that which can also be queried by e-mail address if that's preferred
19:16:23 <smcginnis> diablo_rojo: Are we going to track per release cycle, or just current state? We've always done current state, which would mean separately.
19:16:30 <tonyb> append '&o=DEATAILS' to mine to get them
19:16:42 <fungi> granted, it requires people to set up an openstack.org profile and include an irc nick in their profile
19:16:51 <diablo_rojo> smcginnis, how much churn is there from cycle to cycle?
19:16:54 <fungi> we use that in the openstack technical election tooling
19:16:59 <diablo_rojo> If its a small to 0 number I would say current state
19:17:13 <tonyb> fungi: I was planning to borrow a small amount of that code ;P
19:17:23 <diablo_rojo> heh
19:17:51 <smcginnis> Small, but I would say especially if it was large we would want to do current state.
19:18:19 <diablo_rojo> ...so single file?
19:18:24 <diablo_rojo> Where do we want that to live?
19:18:26 <tonyb> I'm not sure we need to do the lookup in gerrit if it's user supplied data, we can just add that to the git-review command
19:18:40 <fungi> so anyway, there are places folks could manage their names and irc nicks independently and all the release liaisons list would need is e-mail addresses
19:18:51 * tonyb likes next to series status
19:19:10 <tonyb> fungi: Yeah true
19:19:13 <smcginnis> tonyb: That would seem like a logical place for it.
19:20:35 <tonyb> so chnage the task from 'add liasions to delivererable files' to 'add liasions to releases repo and deprecate wiki'
19:20:53 <tonyb> and add them all to deliverables/release_liasions.yaml ?
19:21:16 <diablo_rojo> works for me
19:21:18 <smcginnis> That sounds reasonable to me.
19:21:33 <diablo_rojo> and we do or dont want nick with name and email?
19:21:46 <smcginnis> Sounded like all we needed was email.
19:22:38 <diablo_rojo> All we need yes, but do we want anything else?
19:22:55 <tonyb> All we need is email, can we see what it looks like with all 3?
19:23:06 <tonyb> in the wiki at the monment we have name and IRC
19:23:29 <diablo_rojo> I can throw together a mock up for review
19:23:43 <tonyb> diablo_rojo: thanks
19:24:03 * tonyb dislikes data duplication but likes the consistency
19:24:29 <diablo_rojo> Okay so I have my plan to move forward then :)
19:24:43 <tonyb> \o/
19:26:05 <tonyb> So I'm on the hook to to do the c-w-i releases next week
19:26:52 <tonyb> process.rst is pretty clear on what's needed so I'll use that to re-test the new automation tools and then upload them again too
19:26:53 * smcginnis is afraid we are going to find a lot of broken repos due to requirements updates
19:29:15 <tonyb> smcginnis: You mean like sphinx or like the constraints URLs?
19:30:27 <smcginnis> Sphinx and ones like that.
19:30:39 <smcginnis> I've seen a lot of updates, but I'm sure there are still some lingering.
19:32:16 <tonyb> :(
19:33:12 <tonyb> I was kinda assuming that $projects would have landed those chnages if they were blocking merging any code but maybe I'm over stating/simplifying the problem
19:33:43 <smcginnis> The active ones should all be good. There's a bunch of low volume repos that might not have noticed it yet.
19:35:25 <tonyb> Perhaps is they're low volume they wont need a release?
19:35:30 * tonyb hopes
19:35:46 <smcginnis> True
19:36:11 <tonyb> so *if* we have a bunch that can't release ... and need one what do we do?
19:36:36 <tonyb> I don't want to sign us up for another pike-em style event that took too long
19:38:00 <smcginnis> I guess not much we can do other than notify the teams that they have an issue. Worse case scenario, I think we skip them for m-1 and pick them up at m-2.
19:39:27 <tonyb> So get the reviews out there *early* so we can see how bad it is and inform the teams
19:39:47 <smcginnis> Yeah, that might be a good idea.
19:40:15 <fungi> if software is perpetually unreleasable, that's worth reporting to the tc
19:40:36 <fungi> i don't feel like it ought to be the release team's job to fix projects so they're able to have releases
19:41:22 <fungi> so i agree, let teams know their software has a problem and to submit a new release request when they resolve it
19:41:38 <fungi> if they don't, pretty sure the tc members will appreciate a heads up
19:41:47 <smcginnis> Good idea to share that info with the tc too.
19:42:06 <tonyb> fungi, smcginnis: we can do that
19:42:52 <fungi> the release team already has plenty of responsibilities without also being babysitters
19:43:03 <smcginnis> ++
19:43:06 <diablo_rojo> ++
19:43:23 <smcginnis> I would probably try to pick up some of those, but still shouldn't be expected from the release team.
19:43:49 <tonyb> Yeah
19:44:39 <tonyb> I probably will too but perhaps we shoudl add a footer to the commit message "We're doign this because we're nice ... y'all should've done this already"
19:44:43 <tonyb> #joking
19:46:08 <smcginnis> Haha, yep. ;)
19:46:19 <fungi> "because you couldn't be bothered to fix your stuff..." ;)
19:46:31 <tonyb> :)
19:47:05 <tonyb> okay so that's the stuff I think we have to talk about ... anything more open?
19:47:09 <diablo_rojo> Do enough emoticons make it okay?
19:50:42 <smcginnis> :)
19:50:44 <tonyb> Maybe?  perhaps fake xml tags <message mode='joking'> .... </message> ;P
19:50:44 <smcginnis> Nothing else from me.
19:51:12 <smcginnis> mode='jokingnotjoking'
19:51:24 <fungi> don't forget, gerrit maintenance for project renames commencing at 17:00z tomorrow
19:51:31 <fungi> 15:00z
19:51:35 <fungi> fingers
19:51:37 <smcginnis> Ah, right. Thanks
19:51:43 <smcginnis> fungi: Is that just doing the renames?
19:52:12 <fungi> yeah, and the gerrit restart will also (in theory) fix the "gitweb" links to go to the opendev gitea
19:52:47 <tonyb> fungi: It's thursday for you right?
19:53:00 * tonyb tries to translate "tomorrow"
19:53:07 <fungi> tonyb: yes, but any time i say "tomorrow" i always mean "tomorrow utc"
19:53:32 <fungi> so 2019-05-31 15:00 utc
19:53:53 <tonyb> okay I wont be near a computer then
19:54:02 <diablo_rojo> tonyb, fungi operates in UTC.
19:54:10 <fungi> you will hopefully be having a nice sleep in on a lazy saturday
19:54:13 <tonyb> That's Saturday and I have brekfast plans ;P
19:54:47 <fungi> oh, or breakkie, sure
19:54:59 <tonyb> fungi: "sleep in" my daughter is an early riser ... so I'll be up at 6 ... maybe 7 if I'm lucky
19:55:16 <tonyb> okay I think we're done
19:55:17 <fungi> oof, my sympathies
19:55:54 <fungi> thanks tonyb! it always helps to know that it's friday somewhere even if that somewhere isn't here yet
19:56:02 <diablo_rojo> tonyb, it will fade when she gets to be a teenager ;)
19:56:08 <tonyb> :)
19:56:10 <smcginnis> tonyb: There were many years where we were lucky if our oldest daughter would sleep in until 6. I know how that goes.
19:56:12 <tonyb> #endmeeting