15:00:50 <bcotton> #startmeeting Council (2019-01-30)
15:00:50 <zodbot> Meeting started Wed Jan 30 15:00:50 2019 UTC.
15:00:50 <zodbot> This meeting is logged and archived in a public location.
15:00:50 <zodbot> The chair is bcotton. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:50 <zodbot> The meeting name has been set to 'council_(2019-01-30)'
15:00:59 <bcotton> #meetingname council
15:00:59 <zodbot> The meeting name has been set to 'council'
15:01:14 <bcotton> #chair jonatoni bex contyk dgilmore dperpeet langdon mattdm sumantrom tyll bcotton pbrobinson stickster
15:01:14 <zodbot> Current chairs: bcotton bex contyk dgilmore dperpeet jonatoni langdon mattdm pbrobinson stickster sumantrom tyll
15:01:24 <contyk> .hello psabata
15:01:25 <zodbot> contyk: psabata 'Petr Ĺ abata' <psabata@redhat.com>
15:01:29 <bcotton> #topic Introductions, Welcomes
15:01:38 <bcotton> hi, contyk
15:02:13 <bcotton> i know jonatoni and bexelbie are unlikely to join and mattdm may be late
15:02:25 <bcotton> anyone else around or will contyk and I get to have all the fun ourselves?
15:02:32 <mattdm> bcotton: I am here
15:02:36 <stickster> .hello pfrields
15:02:37 <zodbot> stickster: pfrields 'Paul W. Frields' <stickster@gmail.com>
15:02:40 <contyk> damn
15:02:48 <mattdm> haha
15:02:55 <mattdm> langdon is also on the way
15:03:03 <mattdm> we don't have to have a long meeting :)
15:03:06 * stickster is here lightly, just got back last night from Brno and I have a metric poopload to do in the two days this week I'm here before I'm gone again next week
15:03:31 * mattdm does not ask questions about imperial vs metric pooploads
15:03:51 <stickster> "keep calm and continue shoveling"
15:04:24 <mattdm> stickster scatalogy aside, are you feeling better from the fosdem death plague?
15:04:43 <mattdm> i mean devconf
15:04:48 <mattdm> which we are now going to bring to devconf
15:04:49 <stickster> I'm about 75% back to strength
15:05:23 <stickster> just an incessant cough now, yay
15:05:33 * pingou sends honey to stickster
15:05:50 <mattdm> sorry to hear. I think I managed to keep asamalik from infecting me at group dinner :)
15:06:06 <bcotton> speaking of devconf, i suppose we might want to talk about that briefly today. anyone else have topic proposals?
15:06:20 <mattdm> bcotton: nice segue
15:06:43 <mattdm> Yeah, I think brief report from devconf things might be useful
15:06:45 <bcotton> my segues are so nice, i could be a mall cop
15:06:57 <langdon> .hello2
15:06:58 <zodbot> langdon: langdon 'Langdon White' <langdon@redhat.com>
15:07:21 <mattdm> stickster do you want to talk briefly about the infra team prioritization meeting?
15:07:29 <stickster> Indeed
15:08:19 <stickster> There were several days of team meetings by the Community Platform Engineering (CPE) team which includes Fedora + CentOS folks
15:08:25 <bcotton> two topics is plenty of topics. are you ready now, stickster, or should we do devconf first?
15:08:30 <bcotton> i guess he's ready :-D
15:08:32 <stickster> oh sorry
15:08:37 <mattdm> yeah let's go with that :)
15:08:42 <stickster> bcotton: didn't mean to jump the gun
15:08:48 * stickster waits for topic
15:08:52 <bcotton> #topic Community Platform Engineering (CPE) meeting recap
15:09:00 <stickster> There were several days of team meetings by the Community Platform Engineering (CPE) team which includes Fedora + CentOS folks
15:09:01 <langdon> bcotton++
15:09:01 <zodbot> langdon: Karma for bcotton changed to 11 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:09:03 <bcotton> all you, stickster. i appreciate your enthusiasm
15:09:06 <langdon> for the segue joke
15:09:14 * sumantro is here
15:09:37 <mattdm> oh hi sumantro :)
15:09:47 <stickster> I wasn't there for the first three days due to travel and all-day internal Red Hat meetings
15:09:49 <sumantro> hey mattdm :)
15:10:21 <stickster> But the fourth day was planning priorities for the CPE team for calendar year 2019 (or Red Hat's fiscal year 2020 if you prefer)
15:10:48 <stickster> lgriffin, the new co-manager of the team along with Evolution, was there as well
15:11:53 <mattdm> #info Leigh Griffin (lgriffin) is co-managing the Community Platform Engineering team at Red Hat along with Jim Perrin (Evolution)
15:12:36 <mattdm> #info we care about this about this because CPE is the team that employs Fedora Infrastructure and app development people to work on Fedora @ Red Hat
15:13:00 <stickster> also for more information see https://fedoraproject.org/wiki/Community_Platform_Engineering
15:13:07 <mattdm> #link https://fedoraproject.org/wiki/Community_Platform_Engineering
15:13:08 <stickster> pretty standard stuff
15:13:15 <mattdm> so what as the outcome of that meeting?
15:13:40 <contyk> everyone should smile more
15:13:52 <langdon> contyk: happy weds!
15:13:59 <stickster> We had guests dperpeet, afedorova, ttomecek, dzickus, and some others as well, where we discussed a broad new goal for Fedora that was also (AIUI) discussed at DevConf.cz afterward in several talks
15:14:04 <contyk> langdon: <3
15:15:15 <stickster> Writ large that looks like "Fedora becomes an overarching umbrella to be the git-forge of choice for OS/platform in the wider open source community"
15:15:59 <stickster> Not coincidentally, this plays into the priorities we've already been developing under lifecycle + CI objectives
15:16:18 <stickster> Chief among them getting Rawhide gating to be a REAL, WORKING THING
15:16:25 <mattdm> stickster: or, rather, "what github is to git-forges, we want to be to operating system development and open soruce packaging"
15:16:26 <langdon> "git-forge"? like we want gnome sources to be in dist-git-pagure?
15:16:36 <stickster> mattdm: +1
15:16:45 <langdon> ahhh ok.. that makes more sense
15:17:01 <mattdm> because the way you wrote it looked like we were deciding to prioritize being a github competitor
15:17:04 <stickster> no
15:17:07 <langdon> english is a terrible language... stickster's sentence says both of those things
15:17:14 <mattdm> yes. :)
15:17:19 <bcotton> #info the goal is to be to operating system development & open source packaging what GitHub is to git forges
15:17:25 <stickster> ^ yes, that.
15:17:26 <mattdm> bcotton++
15:17:43 <mattdm> so what are the main focuses of CPE and Fedora Infra to enable that?
15:17:45 * langdon disappointed there was no mention of rolling transport
15:18:11 <stickster> So the team also faces a number of challenges in that we have a very large surface of work and the prioritization is somewhat loose, decentralized
15:18:31 <langdon> (that was for bcotton)
15:19:01 <stickster> Whereas the imperative here is to get a few VERY CRITICAL things done timely in partnership with the teams of dperpeet, afedorova, ttomecek, et al.
15:19:38 <lgriffin> there is a thread opened on infrastructure@lists.fedoraproject.org called Project Details which will give an insight into the volume of projects under our ownership at present....with a lot more responses to come!
15:20:02 <mattdm> oh hi lgriffin!
15:20:05 <mattdm> lgriffin++
15:20:05 <zodbot> mattdm: Karma for lgriffin changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:20:11 <mattdm> have a cookie :)
15:20:25 <lgriffin> :)
15:20:41 <bcotton> #link https://lists.fedoraproject.org/archives/list/infrastructure@lists.fedoraproject.org/thread/WYLII4UG3RVDJZAGIXUF4M2GQEDDXANS/
15:20:48 <mattdm> bcotton++
15:21:10 <mattdm> (ben doesn't get a cookie because of the fascist cookie limit rules)
15:21:13 <lgriffin> thanks bcotton I didn't have the link at hand!
15:21:43 <mattdm> stickster: so, those "VERY CRITICAL" things are mostly around Gated Rawhide?
15:22:00 <stickster> The team did a really useful exercise of "budgeting" in which we enumerated as many pending projects as possible, giving them a relative complexity rating, and then "laying our money down" on which were the most critical. Rawhide gating is chief among them (and 1-2 of the others are related).
15:22:31 <mattdm> I know bexelbie is felling a bit frustrated but he's not here so you get my take :) which is that one thing 100% done is *infinitely* better than ten things 10% done.
15:22:42 <stickster> I have photos of the outcome board: https://photos.google.com/photo/AF1QipNNTPBBiLOny2f8Q4sjm_pv0hJW78gFHP4XcQJ1
15:22:48 <mattdm> #info Rawhide gating is the teams top priority
15:22:53 <bcotton> #link https://photos.google.com/photo/AF1QipNNTPBBiLOny2f8Q4sjm_pv0hJW78gFHP4XcQJ1
15:23:03 <lgriffin> mattdm the team did talk about and mention things like the Definition of Done, so getting things to 100% and not 'almost there' is going to be a priority going forward
15:23:04 <mattdm> AIUI from lgriffin this is still an initial prioritization and wanting more feedback
15:23:18 <mattdm> lgriffin: awesome
15:23:23 <langdon> url seems to be busted for image
15:23:25 <pingou> https://fedoraproject.org/wiki/Changes/GatingRawhide *unfinished*
15:23:49 <bcotton> #undo
15:23:49 <zodbot> Removing item from minutes: <MeetBot.items.Link object at 0x7fe544f4ac90>
15:24:00 <contyk> langdon: re:english -- toki pona ftw
15:24:08 <lgriffin> the prioritisation within Brno was only a limited view, so more input from stakeholders is certainly needed to get the bigger picture
15:24:08 <stickster> The information was transcribed down by smooge and the team is working on breaking down each of the key projects, but here they are for easier reference in the notes:
15:24:21 <stickster> * Rawhide gating
15:24:36 <bcotton> stickster: can you please #info these as you go?
15:24:43 <stickster> oh yeah, sorry bco
15:24:49 <bcotton> np
15:25:08 <langdon> contyk: you would use all emoji if you could
15:25:09 <lgriffin> There is also an open question of a landing page for storing all of these Requirements, so somewhere on the wiki for the community to track things and have input on them as well
15:25:12 <stickster> #info --- Team priorities (in *rough* priority order) ---
15:25:18 <stickster> #info Rawhide gating
15:25:24 <stickster> #info Team self improvement
15:25:35 <stickster> #info F30/F31 release steadiness
15:25:38 <stickster> #info EPEL 8
15:25:50 <stickster> #info Fedora messaging fixes
15:26:17 <stickster> #info Automation of release tooling
15:26:21 <pingou> stickster: it's porting our apps from fedmsg to fedora-messaging
15:26:21 <stickster> #info repoSpanner completion
15:26:33 <stickster> #info --- END ---
15:26:37 <stickster> pingou++
15:27:01 <mattdm> those all sound good to me
15:27:06 <bcotton> stickster: what does "team self improvement" mean?
15:27:22 <lgriffin> Just a note on Team Self Improvement, this is something that will be realised via other priorities. So the team are going to look at ways to improve (e.g. pair programming, tracking work, definition of done etc.) so this is a priority but not in an ordered sense like the others. It was also raised that in any transformation that things tend to get worse before they get better, so a dip in output /
15:27:22 <lgriffin> performance is expected which ultimately gives rise to a better team down the line
15:27:26 <mattdm> "repospanner" is a component for sharing fedora and centos distgit
15:27:40 <stickster> ha, asked and answered bcotton ^ :-D  lgriffin++
15:27:41 <zodbot> stickster: Karma for lgriffin changed to 2 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:27:48 <langdon> bcotton: self improvement is the entire team learns toki pona ;)
15:27:58 <mattdm> langdon++
15:28:13 <pingou> and Team Self Improvement will not have a Definition of Done :)
15:28:15 <bcotton> lgriffin++
15:28:15 <zodbot> bcotton: Karma for lgriffin changed to 3 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:28:26 <contyk> langdon++
15:28:26 <zodbot> contyk: Karma for langdon changed to 5 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:28:56 <mattdm> So, I'll represent Bex a little bit. or my spin on Bex. :)
15:29:06 <mattdm> This really seems good and well aligned
15:29:08 <langdon> pingou++
15:29:08 <zodbot> langdon: Karma for pingou changed to 12 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:29:18 * langdon notes the volume of cookies is high today
15:29:18 <mattdm> It is, however, very engineering/platform focused
15:29:31 <mattdm> We've got a lot of things that need doing there and that's the main focus of the team, so that makes sense
15:29:37 <lgriffin> pingou it may not :) The thing is the team self improvement is really being driven by the team, so things they want to focus on / work on improving will fall out of discussions, a one size fits all approach is not going to work for any team
15:29:52 <mattdm> but there are a lot of other *community infrastrcture* things that we need in order to make the bigger vision of Fedora successful
15:29:58 <mattdm> and we're not really getting those here
15:30:06 <langdon> lgriffin: will you be adding a measured percent to your "sprints" (or equiv) of "team reqs" going forward so we can maintain a high level of performance and not need future "dips"?
15:30:20 <mattdm> (Specifically: badges, elections, docs tooling, websites)
15:30:31 <lgriffin> mattdm certainly other stakeholder views will be solicited and most welcome
15:30:33 <mattdm> We need to, as a project, figure out how to get those things done.
15:30:39 <mattdm> </endofbex>
15:30:53 <pingou> elections is being ported to openshift, which should reduce the time between commit and prod
15:31:13 <mattdm> One thing that isn't there that is on the platform engineering side more is self-service spin/lab creation
15:31:23 <lgriffin> langdon tbd on how we will protect the team from that, first thing is forming teams, second thing is monitoring / inspecting so we have data to go off of to gauge whether we are over / under performing and ultimately to start tracking our output and successes
15:31:40 <mattdm> (from devconf converations, maybe using composer?)
15:31:41 <lgriffin> how the team realises that, tracks that and plans for that is going to be decided by the team
15:32:20 <mattdm> maybe that's part of "automation of release tooling" but it's not obvious to me
15:32:41 <mattdm> And the other thing is "getting compose time down to < 4 hours"
15:32:46 <langdon> lgriffin: +1 .. i would point out though that it is a well understood concept to do so.. so i would definitely encourage using a defined percentage
15:32:46 <pingou> automation of release tooling is more automate all the tasks releng does today
15:33:45 <lgriffin> langdon I'd welcome that view / thoughts on the various threads that will emerge in time, I'm a fan of that kind of approach but the team will self-guide this in the long run
15:33:54 <contyk> pingou: if you package this, don't forget to include Obsoletes: releng
15:34:07 <mattdm> pingou: which is awesome, but I'd like it to at least be done with an eye towards self-service for a lot of things
15:34:08 <pingou> contyk: noted
15:34:25 <pingou> mattdm: I can understand the wish
15:35:12 <stickster> The compose time problem still underlies things like Rawhide gating, if you think about it, because we need to have a more efficient compose for gates to be on a usable time sacle
15:35:14 <mattdm> pingou: good :)
15:35:16 <stickster> *scale even
15:35:53 <mattdm> stickster: that makes sense -- let's make sure it's spelled out in the epic or whatever
15:36:57 <stickster> Anyway, I think that's it for my summary. But each of these items has one person assigned from the team to do a breakdown of tasks that gives us more concrete information about the work to do
15:37:05 <mattdm> stickster++
15:37:10 <stickster> By the end of this week that stuff should be built out into a wiki tree
15:37:28 <mattdm> stickster, lgriffin Do you have any updates on when the fedora taiga instance will be in place?
15:37:36 <stickster> transparently for the community to look at it and know what the team is working on, similar to the FY work tables I used to provide on the wiki
15:37:56 <bcotton> stickster++
15:37:56 <zodbot> bcotton: Karma for pfrields changed to 8 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:38:23 * stickster defers questions on taiga to lgriffin/Evolution/mattdm, I'm not really involved with the discussions with the Taiga folks
15:38:45 <lgriffin> I don't have visibility on a timeline mattdm
15:39:02 <mattdm> lgriffin: I see that the site is up and says "login with fas" but that button doesn't actually work
15:39:17 <mattdm> I ask because I agree with robin about "where information kills itself"
15:39:34 <mattdm> wiki is better than not documented, but just barely :)
15:39:43 <mattdm> since this is living, current project work, it'd be nice to use taiga
15:39:57 <lgriffin> yup taiga will be better for it for sure
15:40:01 <stickster> yeah, the intention is that these breakouts make it into true tracking via taiga
15:40:27 <bcotton> anything else from the meeting, stickster?
15:40:31 <lgriffin> 'one source of truth'
15:40:36 <mattdm> lgriffin: perfect
15:40:55 <stickster> bcotton: <eof/>
15:41:15 <bcotton> any final questions for stickster/lgriffin before I turn it over to mattdm for a devconf recap?
15:42:08 <bcotton> #topic DevConf.cz
15:42:15 <bcotton> all you, mattdm
15:42:29 <mattdm> Okay, so, for a change, i did not have any talks. I was just on two panels
15:42:39 <mattdm> Let me tell you, stress and required prep-wise *this is the way to go* :)
15:43:03 <mattdm> The first panel was "the penrose panel". This is a thing from Brendan and no one knows what means
15:43:20 <mattdm> It's a reference to an impossible object that can't be created in the real world
15:43:35 <mattdm> ... and how that is RHEL, Fedora, and CentOS
15:43:48 <contyk> he'll be talking about it at CentOS dojo on Friday again
15:44:08 <misc> (there was a penrose project in rh bugzilla in the past)
15:44:14 <mattdm> The concrete thing here is CentOS sources going into Fedora dist-git. This is awesome and exciting
15:44:26 <mattdm> misc: please tell me it was something completely unrelated
15:44:58 <misc> mattdm: seems to have been related to identity somehow
15:45:01 <misc> back in 2009
15:45:05 <mattdm> We also talked about *other* possible interested parties having branches in dist-git
15:45:11 <misc> https://bugzilla.redhat.com/show_bug.cgi?id=491558
15:45:26 <mattdm> For example, there is a Large Social Media Company that uses a mix of CentOS and Fedora and their own kernel in very large scale production
15:46:01 <mattdm> We spoke with some of their engineers -- they often have changes which are not necessarily appropriate for upstream, but would be very helpful to have branches at a level shared wtih fedora and centos
15:46:09 <langdon> and Fedora as their "company standard build"!!!!
15:46:26 <mattdm> There's really nothing technical preventing them from doing this _today_ (thanks contyk for confirming)
15:46:32 <langdon> they are very excited to use module streams to accomplish this
15:46:36 <mattdm> we just need to figure out the social/political things to make it happen
15:46:41 <misc> langdon: wow, and do they also use our logo ?
15:46:54 <mattdm> misc: there is some potential confusion :)
15:46:57 <langdon> misc++
15:46:59 <zodbot> langdon: Karma for misc changed to 4 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:47:20 <mattdm> This is important because it's a very big concrete step towards realizing that "we're the place for this kind of thing" vision stickster talked about at the beginning
15:47:21 * langdon owed misc karma anyway.. but a bad joke confirms it
15:47:37 * langdon notes he is trying to remember not to use that term anymore
15:48:00 <mattdm> Oh yeah. There was also a thing where I am Legolas and Evolution is Gimli
15:48:22 <mattdm> This may explain why I'm having trouble recalling specifics of the panel discussion
15:48:45 <mattdm> So anyway we also then had the Fedora Council panel
15:48:52 <mattdm> *I* thought it went well.
15:48:59 <mattdm> Most of the rest of y'all were there too
15:49:07 * bcotton agrees the council panel went well
15:49:24 <langdon> it was packed.. kinda awesome
15:49:31 * misc wonder if mattdm being legolas was a play on the Fedora rings proposal a few year ago
15:49:37 <mattdm> yeah, full small room is better than half-full big room
15:49:42 <langdon> no tomatoes were throne
15:49:46 <mattdm> misc: shouldn't I be frodo????
15:50:24 <mattdm> Are there highlights from the panel that we should highlight here?
15:50:51 <mattdm> Fedora 29 mirror stat numbers are _literally off the charts_
15:51:13 <mattdm> 45% growth over the *peak* of Fedora 28. And we're not near the F29 peak yet
15:51:36 <mattdm> It's so good that I'm suspicious that something is going on influncing the counting rather than real growth
15:51:39 <langdon> i would like to highlight fedora needs to hire an intern to fix our counting .. o rsomething.. it is VERY important and currently lossy and going to fail even worse soon (w/ infra changes)
15:51:55 <bcotton> langdon: gsoc proposals are open through 6 Feb :-)
15:52:05 <mattdm> And langdon, I asked Large Social Media Company, and they confirmed that the way they're doing their thing, their engineers are not counted at all
15:52:11 <langdon> bcotton & mattdm should get on that
15:52:22 <langdon> mattdm: engineers?
15:52:30 <langdon> do you mean servers?
15:52:40 <mattdm> "Engineers using Fedora on their laptops"
15:52:41 <langdon> "CSB"?
15:52:45 <langdon> ahh gotcha
15:53:01 <mattdm> the problem with GSOC is that I'm not qualified to actually *mentor* someone on statsitics
15:53:10 <mattdm> I can help show them how to make horrific pandas scripts
15:53:23 <langdon> ahhh so a) yes.. the uuid counting thing is one part b) i really meant making sure what we *can* count we keep counting even with  infra chnages
15:54:30 <mattdm> yeah. I guess i need to get that onto the CPE priority list somewhere
15:54:36 <langdon> we could do a project proposal w/ Boston Uni Spark! which take non profit projects.. and have their stats people mentor.. but that wont be as soon as gsoc
15:55:06 <mattdm> That seems more likely to succeed even if it's not as soon
15:55:25 <mattdm> And it seems like there's probbaly some interesting stuff here. We do have *a lot* of data
15:55:49 <bcotton> 5 minute warning. anything else from DevConf we want mentioned?
15:56:47 <misc> mattdm: we do have the folks from radanalytics.io who might be interested
15:57:33 <mattdm> misc: I'd like to hear more about that at some point. are you at FOSDEM?
15:57:44 <misc> mattdm: yup
15:57:57 <mattdm> misc: perfect see you there
15:58:00 <misc> (also still in brno for now, but will be at centos dojo)
15:58:31 <bcotton> time to wrap up
15:58:35 <bcotton> #topic Next meeting
15:58:46 <mattdm> thanks bcotton
15:58:48 <bcotton> #info Next week's meeting is a report from D&I, FESCo, and Mindshare
15:59:15 <bcotton> #info 1400 UTC in #fedora-meeting-1
15:59:23 <bcotton> and with that...
15:59:29 <bcotton> thanks for coming everyone
15:59:33 * bcotton bangs gavel
15:59:37 <bcotton> #endmeeting