16:03:35 <jzb> #startmeeting Atomic SIG
16:03:35 <centbot> Meeting started Thu Apr  2 16:03:35 2015 UTC.  The chair is jzb. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:03:35 <centbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
16:03:50 <jzb> #chair walters Evolution imcleod kbsingh
16:03:50 <centbot> Current chairs: Evolution imcleod jzb kbsingh walters
16:04:10 <jzb> let's go over action items from last meeting (two weeks ago)
16:04:15 <jzb> http://centos.org/minutes/2015/march/centos-devel.2015-03-19-16.00.html
16:04:23 <jzb> #topic signing
16:04:36 <jzb> imcleod: you had the AI on getting signing ready
16:04:44 <jzb> imcleod: I'm guessing that's probably still WIP?
16:04:51 <jzb> given we were all holed up in meetings all last week...
16:05:05 <imcleod> jzb: I'm afraid so, though I probably should have taken the opportunity to work the issue in person.
16:05:37 <jzb> imcleod: OK
16:05:39 <imcleod> jzb: Can we please pull that forward for next week?
16:05:47 <jzb> imcleod: we ca
16:05:48 <jzb> n
16:06:00 <jzb> #action imcleod  take lead on getting signing ready for imminent "stable" release
16:06:03 <imcleod> jzb: It dovetails nicely with other pending work about the mechanics of our 4 weekly and 6-weekly builds.
16:06:36 <jzb> #topic document release checklist
16:06:39 * kbsingh is here 50%, be 100% in ~ 10 min
16:06:50 <jzb> jbrooks had the AI for this, and he has put one up here:
16:07:49 <jzb> http://wiki.centos.org/Atomic/ReleaseManagement
16:07:56 <jzb> #link http://wiki.centos.org/Atomic/ReleaseManagement
16:08:04 <jzb> so we can check that off
16:08:07 <kbsingh> can we move this to be under the SIG stuff where everything else is
16:08:18 <jzb> kbsingh: pardon?
16:08:51 <kbsingh> http://wiki.centos.org/SpecialInterestGroup/Atomic has a bunch of stuff under there, maybe this can also live there ?
16:09:00 <kbsingh> http://wiki.centos.org/SpecialInterestGroup/Atomic/Download as an example
16:09:23 <kbsingh> http://wiki.centos.org/Atomic/WeeklyMeeting
16:09:37 <kbsingh> looks to be the only other content under /Atomic and is quite a bit stale
16:10:06 <walters> is there a way in this wiki to see the hierarchy?
16:10:15 <jzb> kbsingh: I suppose. I just tried moving it but it yelled at me that the page exists
16:10:29 <jzb> #action jzb move Release Management page after meeting
16:10:46 <kbsingh> walters: i believe so, but will need to check
16:11:01 <kbsingh> jzb: let me know if it doesnt move - we can try a bigger hammer
16:11:01 <jzb> #topic release scheme for image kickstarts and Dockerfiles
16:11:09 <jzb> kbsingh: ack
16:11:21 <jzb> imcleod: that was on your plate too. Any update?
16:11:50 <imcleod> Yes.  I know where these things live internally, and they are, at the moment, entirely inappropriate for external consumption.
16:11:56 <imcleod> Baked-in hostnames.
16:12:02 <imcleod> And a few other things.
16:13:22 <jzb> imcleod: OK, should we carry the AI forward a week or is it now "clean up RHEL kickstarts and Dockerfiles"?
16:13:32 <jzb> or clean up & publish
16:13:50 <imcleod> Hrm.
16:14:19 <imcleod> Ideally we want a publishing process, not individual point publishing.
16:15:17 <imcleod> Which is really what I wanted to sort out.  So, sigh, can you carry that forward as well (as "Establish ongoing release process for non-RPM artifacts, ideally into g.c.o")?
16:15:26 <jzb> imcleod: yep
16:15:47 <jzb> #action imcleod Establish ongoing release process for non-RPM artifacts, ideally into g.c.o
16:15:56 <jzb> imcleod: thanks!
16:16:17 <walters> this question also intersects with whether centos atomic related work is a RHEL rebuild or is an upstream itself,  as if we're doing the latter then maybe it would make more sense to derive from Fedora
16:16:38 <jzb> walters: that touches on the next topic :-)
16:16:43 <imcleod> walters: Agreed.  And it may well end up being both.
16:17:11 <jzb> #topic get final dates/info on two and four-week update plans.
16:17:17 <jzb> that was on my plate
16:17:27 <jzb> the good news is we have some additional clarity, but it's not entirely done yet.
16:17:42 <jzb> it looks like we *are* going to do a rebuild of RHEL Atomic
16:18:10 <jzb> we do not yet have all dates and such sorted out for aligning the 4-week cycle
16:18:36 <jzb> kbsingh and imcleod - I talked to mmcgrath earlier today and I think you three, plus walters and myself probably, need to sync on that sooner than later.
16:18:51 <jzb> it's probably going to take a little massaging to get all the pieces aligned.
16:20:04 <jzb> unless there's an objection, I'll take an action to set up a discussion on that early next week
16:20:19 <jzb> #action jzb set up meeting to align 4-week cycle.
16:20:22 <kbsingh> yup
16:20:44 <jzb> #topic atomic info in /etc/os-release
16:20:47 <kbsingh> imcleod and i had a 40 min session earlier today, I've put across the CentOS way of releasing stuff - he agreed with the process.
16:21:12 <kbsingh> there are a few bits we need to workout around who's going to actually do the work - what the SIG interface is going to be - to what CentOS resources and what the end goals we want to focus on are
16:21:38 <kbsingh> eg. do we care about the rpm content side that turns out in the 4 weekly cycle ? we likely care 'care', but we need to encourage folks like walters to own the content
16:22:23 <kbsingh> happy to jump on call, we can then circle back with summary and a proposed plan
16:22:43 <jzb> kbsingh: we do care, and I'm sort of envisioning a week between the development drop and actually releasing the result after testing.
16:22:56 <kbsingh> maybe i should quantify the 'care' word.
16:23:09 <jzb> the problem I see happening, though is that we will have very, very limited resources to actually tend to any bugs
16:23:20 <kbsingh> we certainly care whats in there - but the scope of the sIG might be to just focus on the deliverables rather than curate the rpms
16:24:02 <jzb> kbsingh: <Ricky> 'splain?
16:24:02 <kbsingh> in which case, we'd need to make sure dwalsh and walters and the entire set of folks who 'own' the rpm content set get an easy onramp - but not need the SIG interface to do anything
16:24:21 <gwd> When koji says this:
16:24:24 <gwd> koji: error: This command takes at least two arguments: a tag name/ID and one or more package n-v-r's
16:24:30 <gwd> What does it want for "n-v-r"?
16:24:43 <kbsingh> mostly in line with what you said jzb around  resources - we might not have enough to do the entire loop
16:24:52 <jzb> kbsingh: ack
16:24:52 <kbsingh> gwd: name-version-release
16:25:00 <imcleod> gwd: glibc-2.20-7.fc21
16:25:06 <jzb> kbsingh: we may need to do it a few times and then identify what resources we lack
16:25:23 <kbsingh> jzb: yup
16:25:30 <jzb> kbsingh: hoping we'll have some community folks step up in some places, but also may need to seek additional resources.
16:25:38 <kbsingh> agree
16:25:41 <gwd> Great, thanks!
16:25:42 <jzb> groovy
16:25:54 <kbsingh> i kinda drew a map., will post that along later
16:25:59 <jzb> kbsingh: OK, thanks
16:26:18 <jzb> lalatenduM: are you about?
16:26:23 <jzb> the last topic was on your plate
16:27:00 <kbsingh> is that around os-release ?
16:27:06 <jzb> kbsingh: yup
16:27:08 <lalatenduM> yeah
16:27:25 <lalatenduM> I have not done much progress
16:27:25 <jzb> lalatenduM: do you have an update?
16:27:38 <kbsingh> thats not a simple one - os-release comes from centos-release, so we'd either need to fork that, and build something for atomic - or as langdon had said, find someother way to find out its an atomic host
16:28:21 <kbsingh> I believe, correct me if i am wrong, the issue backing this is - how do tools in the system find out and make decisions based on if they are on a atomic host or not
16:28:23 <lalatenduM> kbsingh, I think I need some of you rtime to come up to speed on this
16:28:43 <jzb> lalatenduM: can I put you down to set up a time that works for you both?
16:28:58 <kbsingh> this needs to really include langdon as well
16:28:59 <lalatenduM> jzb, sure
16:29:30 <jzb> lalatenduM: ^^
16:29:31 <jzb> kbsingh: ack
16:29:34 <kbsingh> walters: is there some other way for the tools to workout if its atomic host or not ?
16:29:48 <walters> https://bugzilla.redhat.com/show_bug.cgi?id=1200122
16:30:08 <lalatenduM> jzb, will do
16:31:29 <kbsingh> walters: ack
16:32:27 * lalatenduM going to the meeting in to calendar , so he dont forget it next time
16:32:39 <jzb> #action lalatenduM set up discussion with kbsingh and langdon about os-release, report back to atomic-devel
16:32:44 <jzb> lalatenduM: thanks
16:32:46 <lalatenduM> s/going/going to put/
16:32:49 <jzb> #topic open floor
16:33:02 <jzb> OK - that's all the ai's from last meeting, current new topics?
16:34:10 <lalatenduM> Any update on stable builds of Atomic sig?
16:34:23 <lalatenduM> I am new to SIG, so might be repeating it
16:34:48 <jzb> lalatenduM: just that we're planning on it, but we don't have a date yet.
16:34:57 <lalatenduM> jzb, ok
16:35:35 <jzb> any other topics?
16:35:47 <jzb> I can tell it's a holiday tomorrow, everybody's super-quiet.
16:36:08 <lalatenduM> yeah
16:36:10 <lalatenduM> :)
16:36:44 <jzb> OK, will close it up. Thanks all!
16:36:47 <jzb> #endmeeting