15:35:52 <dgilmore> #startmeeting RELENG (2016-04-18)
15:35:52 <zodbot> Meeting started Mon Apr 25 15:35:52 2016 UTC.  The chair is dgilmore. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:35:52 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:35:52 <zodbot> The meeting name has been set to 'releng_(2016-04-18)'
15:35:52 <dgilmore> #meetingname releng
15:35:52 <zodbot> The meeting name has been set to 'releng'
15:35:52 <dgilmore> #chair dgilmore nirik tyll sharkcz bochecha masta pbrobinson pingou maxamillion
15:35:52 <zodbot> Current chairs: bochecha dgilmore masta maxamillion nirik pbrobinson pingou sharkcz tyll
15:35:55 <dgilmore> #topic init process
15:36:25 <nirik> morning
15:36:42 * sharkcz is here
15:37:48 <maxamillion> .hello maxamillion
15:37:49 <zodbot> maxamillion: maxamillion 'Adam Miller' <maxamillion@gmail.com>
15:37:55 <maxamillion> sorry I'm late, was dealing with something
15:38:48 * pbrobinson is here
15:39:41 * threebean 
15:39:46 <dgilmore> so we have no meeting tickets
15:40:02 <dgilmore> we really need to go through the tickets at some point
15:40:59 <dgilmore> #topic Fedora 24 Beta
15:41:24 <dgilmore> we had a hickup last week in that masta pushed things to stable that broke dependencies in KDE
15:41:25 <maxamillion> dgilmore: +1
15:41:42 <nirik> and a kernel update that was in a weird state.
15:41:47 <nirik> but it should be all back on track now.
15:42:38 <dgilmore> there was a bunch of bodhi issues
15:43:45 <nirik> well, just the one new one that I know of.
15:44:17 <dgilmore> there was a bunch of updates locked bceause bodhi lost communication to the db
15:44:24 <dgilmore> and did not roll back a transaction
15:45:35 <dgilmore> anyway
15:45:57 <dgilmore> Fedora Media Writer for windows is being pushed to f25
15:45:59 <nirik> oh yeah, forgot about that one
15:46:11 <dgilmore> too many unanswered questions
15:46:48 <dgilmore> I think there is a lesson for us all in it
15:47:07 <dgilmore> that is to not assume how something will be built and shipped
15:47:42 <nirik> things take time to figure out.
15:47:47 <dgilmore> yeah
15:48:07 <dgilmore> the change owner there assumed they would just build it on their desk and we would ship it
15:48:31 <pbrobinson> and it's been an overall very busy cycle with pungi4 and other such features for rel-eng
15:48:48 <dgilmore> we have not had spare cycles
15:48:56 * nirik nods
15:49:11 <dgilmore> additionally legal has an opinion on how it is shipped
15:49:24 <maxamillion> +1
15:50:20 <maxamillion> there's definitely a mismatch between the policies for shipping content, both legal and releng influenced, and what people perceive as how to ship content
15:50:51 <dgilmore> they wanted all sources for all binaries in the app to be available
15:51:06 <pbrobinson> primarily the first two are paid to give a shit, the later just don't....
15:51:07 <nirik> well, and this is new... so there may well be newer rules or different ones.
15:52:04 <dgilmore> I am sure things will change as we move forward
15:52:32 <dgilmore> Other than all that I think we are ready for Beta
15:53:11 <dgilmore> #topic openh264
15:53:43 <dgilmore> On Thursday pfrields and I had a meeting with some people at cisco
15:54:12 <dgilmore> we sorted out how to have cisco ship the binaries and let everything work
15:54:44 <dgilmore> On friday I sent cicso a tarball with the latest f24 builds
15:55:05 <dgilmore> and nirik setup codecs.stg.fedoraproject.org
15:55:25 <dgilmore> we verified that the redirects work
15:55:49 <dgilmore> #info fedora to host and serve the repo metadata, Cicso hosting and serving the rpms
15:56:20 <maxamillion> +1
15:56:21 <masta> who signs the rpms? Cisco?
15:56:22 <maxamillion> that's awesome
15:56:58 <masta> and yeah, awesome!
15:57:00 <dgilmore> masta: we build and sign the rpms
15:57:11 <dgilmore> we sign the repo metadata
15:58:07 <dgilmore> the end goal is to provide fedora built openh264 codec support, there is a gstreamer plugin and firefox integration
15:58:29 <dgilmore> cisco has to distribute the binaries to comply with the patent grant
15:58:51 <masta> ah, okay... was wondering about that.
15:59:15 <dgilmore> we have glue in place so that people can not get the rpms from koji
15:59:53 <dgilmore> so its pretty cool
16:00:08 <dgilmore> there is one thing I am not sure about
16:00:21 <dgilmore> https://fedoraproject.org/wiki/Third_Party_Repository_Policy
16:00:26 <masta> that is a big understatement... I think this will make a lot of people very happy.
16:01:17 <dgilmore> the plan was to have the repo ship with enabled=0 and enabled_metadata=1
16:01:23 <dgilmore> so you can get the metadata
16:01:40 <pbrobinson> masta: it's only h264 video, I don't believe it covers things like mp3 audio. It's not a complete solution AFAICT. It's a single video codec
16:03:48 <masta> pbrobinson, ah, okay.... well it's progress :-)
16:06:24 <dgilmore> it is just the one codec
16:08:00 <dgilmore> nirik: maxamillion: do you think we should ask fesco about shipping it enabled?
16:08:20 <maxamillion> probably
16:08:25 <nirik> hum. I guess we could...
16:08:38 <dgilmore> otherwise we are going to have to communicate heavily to have people enable it for it to be there
16:08:39 <maxamillion> I would think it something fesco would want input on
16:08:42 <nirik> Ideally dnf at least would add support for enabled_metadata
16:08:50 <dgilmore> nirik: right
16:08:58 <nirik> I looked, but couldn't find an open bug on it.
16:09:26 <dgilmore> nirik: i found almost nothing on the enabled_metadata option that basically does not exist
16:10:05 <nirik> well, gnome-software/packagekit can deal with it supposedly.
16:10:10 <nirik> (although I didn't test)
16:10:49 <dgilmore> I am not 100% sure that we need to educate the users on the patent grant cisco has
16:10:59 <dgilmore> but it would I guess be nice to
16:11:16 <dgilmore> anyway
16:11:18 <dgilmore> #topic Secondary Architectures updates
16:11:19 <dgilmore> #topic Secondary Architectures update - ppc
16:11:20 <nirik> in the mean time the enabled=0 one should allow people using gnome-software to get it
16:11:31 <dgilmore> nirik: yeah
16:12:05 <dgilmore> pbrobinson: how is ppc?
16:12:18 <pbrobinson> looking good
16:12:42 <pbrobinson> no known issues as we head to beta
16:12:56 <pbrobinson> well nothing major, still awaiting a firefox fix
16:13:08 <pbrobinson> but from my PoV we're in pretty good shape
16:14:14 <dgilmore> #info still waiting on firefox to be fixed but in good shape
16:14:51 <dgilmore> #topic Secondary Architectures update - s390
16:14:59 <dgilmore> sharkcz: how is s390?
16:15:13 <pbrobinson> s390 (non x) disabled on F-25 now
16:15:14 <dgilmore> sharkcz: have we disabled s390 builds for f24 and rawhide yet?
16:15:17 <sharkcz> we are looking good, s390 is dead in f25 since today
16:15:27 <dgilmore> okay
16:15:37 <dgilmore> sharkcz: and f24?
16:15:44 <sharkcz> for f24 I'll wait until close to GA, we will need it internally
16:15:54 <dgilmore> why?
16:15:55 <sharkcz> but GA will be without it
16:16:34 <sharkcz> will try to keep the internal rebuilds with s390, just to be sure for some limited period of time
16:16:59 <dgilmore> okay
16:17:08 <dgilmore> seems like we should not do that
16:17:31 <pbrobinson> from a compose/pungi PoV I plan on closing it out this week, will be s390x only even on f-24
16:17:32 <sharkcz> if there is a need for it, we want something newer than f23
16:17:54 <dgilmore> sharkcz: why would we have a need for it
16:18:01 <dgilmore> it is decided to not have it right?
16:18:23 <sharkcz> only from the engineering point of view
16:18:24 <dgilmore> that decision needs to be final and complete
16:18:41 <sharkcz> no business decision yet
16:18:53 <dgilmore> then we should not disable it
16:19:04 <pbrobinson> LOL... <sigh>
16:19:08 <dgilmore> sharkcz: its eitehr done or it is not
16:19:39 <dgilmore> if the decision to drop it is not complete we should leave it in place until the decision is final
16:19:54 <dgilmore> pbrobinson: I do not disagree
16:20:25 <dgilmore> but if the business can turn around and say we need to keep it then we should at least keep the builds going
16:20:39 <dgilmore> turn off multilib and only ship s390x
16:20:40 <sharkcz> the decision is to drop in f25 (done), and make f24 containing both, at the f24 GA (a stable point) we can drop it there too
16:20:53 <dgilmore> sharkcz: no
16:21:13 <dgilmore> sharkcz: if the decisoion was to ship both in f24 then we ship both
16:21:23 <pbrobinson> if we ship it for F-24, we ship it until EOL
16:21:48 <pbrobinson> we can disable it at any point up until GA, but if it goes GA we support it until EOL
16:21:52 <dgilmore> pbrobinson: right
16:22:16 <sharkcz> yes, that's what I mean, drop s390 in f24 before the GA, don't ship in s390 in GA
16:22:25 <sharkcz> but as close to GA as possible
16:22:27 <dgilmore> sharkcz: then drop it now
16:22:44 <dgilmore> sharkcz: I do not understand why you are saying that
16:22:52 <dgilmore> it is either done or it is not done
16:23:04 <dgilmore> we either just drop it or keep it
16:23:46 <pbrobinson> dgilmore: if some product wants to fork off it at a later point it does kind of make sense to keep it around for longer, even if not sense from a Fedora PoV
16:24:34 <sharkcz> pbrobinson: yes
16:24:46 <dgilmore> pbrobinson: sure, but the decision to drop it was supposed to be done for all
16:25:10 <dgilmore> anyway
16:25:18 <dgilmore> #topic Secondary Architectures update - arm
16:25:25 <pbrobinson> dgilmore: in theory sure.... in practice product managers are always crazy
16:25:25 <dgilmore> pbrobinson: how is aarch64?
16:25:48 <pbrobinson> about the same as PPC, looking good, only real issue is firefox.
16:25:58 <pbrobinson> all ready to go for beta
16:26:04 <dgilmore> awesome
16:26:12 <dgilmore> #topic open floor
16:26:18 <dgilmore> does anyone have anything?
16:26:30 <jkurik> Who is going to represent releng on Go/No-Go & Readiness meeting this Thursday ?
16:26:45 * nirik will be there
16:26:46 <dgilmore> jkurik: I will be
16:26:51 <jkurik> dgilmore: thanks
16:26:52 <dgilmore> and nirik will be there
16:27:04 <jkurik> great, thank you
16:27:44 <dgilmore> anything else?
16:28:02 * nirik has nothing off hand
16:28:02 <dgilmore> #endmeeting