14:00:00 #startmeeting Workstation WG 14:00:00 Meeting started Wed May 25 14:00:00 2016 UTC. The chair is stickster. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:00 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:00:00 The meeting name has been set to 'workstation_wg' 14:00:02 #meetingname workstation 14:00:02 The meeting name has been set to 'workstation' 14:00:03 #topic Roll call 14:00:04 .hello pfrields 14:00:05 stickster: pfrields 'Paul W. Frields' 14:00:14 .hello otaylor 14:00:15 otaylor: otaylor 'Owen Taylor' 14:00:15 .hello catanzaro 14:00:20 mcatanzaro: Hey, you made it! 14:00:20 mcatanzaro: catanzaro 'None' 14:01:17 rdieter: rdieter_work: mclasen___: cschalle_: juhp: kalev-afk: ping 14:01:22 .hello mclasen 14:01:23 mclasen___: mclasen 'Matthias Clasen' 14:01:33 pong 14:01:36 * stickster gives a couple minutes for others to come in 14:01:42 .hello rdieter 14:01:43 rdieter: rdieter 'Rex Dieter' 14:01:57 .hello kalev 14:01:57 kalev-afk: kalev 'Kalev Lember' 14:02:26 .hello skrzepto 14:02:28 skrzepto: skrzepto 'Szymon Mucha' 14:03:17 #chair otaylor mcatanzaro mclasen___ cschalle_ rdieter kalev-afk 14:03:17 Current chairs: cschalle_ kalev-afk mcatanzaro mclasen___ otaylor rdieter stickster 14:03:17 .hello cschalle 14:03:18 cschalle_: Sorry, but you don't exist 14:03:24 nooo 14:03:30 existential crisis commences 14:03:36 try '.hello cschaller' ? 14:03:42 I think it's keyed off your FAS name 14:03:52 .hello uraeus 14:03:53 cschalle_: uraeus 'Christian Schaller' 14:03:56 there we go 14:03:57 oh there you go 14:04:50 #topic GNOME Boxes/libvirt default network bug 14:04:56 #link https://bugzilla.redhat.com/show_bug.cgi?id=1164492 14:05:04 decause: Also ping, since you're up next 14:05:10 .hello decause 14:05:11 decause: decause 'Remy DeCausemaker' 14:05:26 stickster: should I go now? 14:05:29 So on the topic of this Boxes/libvirt bug... this is a longstanding issue AIUI 14:05:35 decause: hold please, next topic is you 14:05:37 kk 14:05:39 thanks 14:05:43 stickster: it's pretty incredible to me that we've been doing this weird dance for multiple releases 14:05:57 it just seems wrong to set network configuration via package dependency 14:07:23 stickster: I admit to not having read the entire backlog in the bug - is there a proposed f24 solution ? 14:07:25 Long term, I think we need to ecncourage libvirt+kvm+networking people to figure out a better way to handle networking for the user libvirtd 14:08:12 I'm not fully convinced that Boxes is something that should be on the live media in the first place, it has pretty heavy deps and is probably only be useful for a small subset of users 14:08:15 it's also a big problem for vagrant-libvirt, that needs to use the system libvirtd, which causes other things to work weirdly (like shared directories have to have to be read/write the qemu user) 14:08:39 mclasen___: There's a fix suggested in the original bug from F21 timeframe, https://bugzilla.redhat.com/show_bug.cgi?id=1146232 14:09:05 adamw has been doing the dance of removing the dep for last few releases, could easily do the same again 14:09:12 kalev-afk: The issue there is that the Live media contains the image that's going to be transferred to the Workstation that most people install 14:09:29 kalev-afk: we missed it for f23, i think. 14:09:33 I don't think we want to start defining a separate live image package set 14:09:59 kalev-afk: So, seeing as how the Live image is the definition of "what you'll get by default installation," and we prioritize that as the installable source people download, we can't really drop it and then expect people with Workstation to end up using Boxes 14:10:12 It sounds like Cole Robinson rejected adding the CondVirtualization to f24, so the only approach would be to do the dependency dance again 14:10:41 otaylor: Correct, but I don't think he ruled it out for F25 14:10:52 otaylor: the idea was always that someone would come up with a better fix, it's just that it never actually happened, so we kept coming back to the same workaround. 14:11:05 well, given that this has been going on since f21, we should figure out a way to ensure that it happens this time around 14:11:16 I can volunteer 14:11:17 cole not ruling it out doesn't seem to be sufficient... 14:11:30 another case of perfect being enemy of the good it seems :-/ 14:12:21 * stickster leaves https://bugzilla.redhat.com/show_bug.cgi?id=1146232#c55 14:12:51 stickster: +1 14:13:06 I understand and agree that we probably need to do the dependency workaround for F24... not introducing new problems at this point in F24 is smart 14:13:55 adamw: I put a NEEDINFO on that bug and can follow up in a few weeks to see that it went in, if you think Cole is OK with this -- I can of course touch base with him to make sure of that, don't need to ASSume it 14:14:20 I fear Cole is not keen on this solution for F24 or for F25 either, but it seems like a good approach to me... Workstation users really do not need virtualization to work inside of VMs.... 14:14:44 * stickster needs to take some time to more thoroughly understand the problem anyway 14:15:17 We don't need Russian stacking doll level virtualization 14:15:36 * rdieter agrees 14:16:16 it seems a little off the beaten path, yeah 14:17:22 So... is the next action here for zeeshan to drop the dep and rebuild? 14:17:33 seems so 14:17:54 I'll deliver that message 14:18:11 #action mclasen___ work with zeeshan to do dep-dropping dance in gnome-boxes for F24 14:18:11 I just did that, http://pkgs.fedoraproject.org/cgit/rpms/gnome-boxes.git/commit/?h=f24&id=dbf30611915da77a96a7680002a87509fad34787 14:18:20 #undo 14:18:20 Removing item from minutes: ACTION by stickster at 14:18:11 : mclasen___ work with zeeshan to do dep-dropping dance in gnome-boxes for F24 14:18:25 ah, okey dokey, thank you kalev-afk 14:18:31 kalev++ 14:18:31 rdieter: Karma for kalev changed to 6 (for the f23 release cycle): https://badges.fedoraproject.org/tags/cookie/any 14:18:58 kalev you did it on master too, that's not very temporary ;) 14:19:12 #action kalev-afk do dep-dropping dance, rebuild, work with zeeshan to restore in rawhide as needed 14:19:44 #action stickster do add'l research to understand problem, work with adamw, zeeshan, mclasen___, et al. on workable F25 solution 14:19:55 kalev-afk++ 14:20:02 kalev++ 14:20:03 stickster: Karma for kalev changed to 7 (for the f23 release cycle): https://badges.fedoraproject.org/tags/cookie/any 14:20:07 yay, cookies 14:20:10 nom nom, cookies 14:20:19 Anything else on this one then? 14:20:26 stickster: can you add me on a cc if you send out mails - I have opinions here (which I think more are NetworkManager <=> libvirt than gnome-boxes) 14:21:01 otaylor: ack 14:21:10 ok moving on 14:21:16 #topic Workstation demo at Red Hat Summit 14:21:36 * decause steps up 14:21:40 So just by way of context... the Red Hat Summit is a huge conference that a lot of Red Hat customers, partners, and users attend 14:21:40 so 14:22:26 and this year, Fedora has a prominent placement within "community central" where a number of projects will have their booths 14:22:27 There will be a Fedora booth in a "community central" type area, along with other associated community booths. At that booth there are scheduled to be systems running Fedora editions, so we have the opportunity to do something with Fedora Workstation there 14:22:34 ha, I yield to decause :-) 14:22:36 * decause waits for stickster to finish 14:22:38 ok 14:22:40 14:23:20 #link https://fedoraproject.org/wiki/Red_Hat_Summit_2016 14:23:36 This is the wiki page we are using to organize and socialize our plans for RHT Summit 14:23:51 decause: Is it better to have a recorded demo, rather than just running through things live? 14:24:00 there is a variety of activity that has already been identified and carried out 14:24:25 otaylor: That's my opinion, yes 14:24:39 that is safest 14:25:05 I have it on good authority that we'll have power and net, but aiming for the least common denominator is prolly a good call 14:25:09 safest and also means that it doesn't count on somebody being at the booth who knows how to run it continually 14:25:16 otaylor: also agreed 14:25:18 and 14:25:28 I don't think we can be 100% certain of someone manning the demo at all times who's conversant with all parts of Workstation... Not that I fit that criterion myself, but althought I'll be at Summit, I won't be able to be a booth-babe (booth-bozo?) much 14:25:31 we're going to have 4 laptops (we've requested them at least) 14:25:43 what length video are we aiming for here - 5 minutes ? 14:25:56 with text overlay / audio / both ? 14:26:09 mclasen___: We could do several of them and have them run in a loop 14:26:13 at least one should absolutely have the Workstation Demo as a dedicated activity 14:26:29 mclasen___: in my experience text/subtitle overlay is a very good thing, because of noise factor 14:26:42 stickster: +1 14:26:58 my experience is that doing videos in quality we had for rh workstation in past years is a lot of effort 14:27:08 I suspect shorter (roughly 2 minutes) would work better than longer 14:27:09 as in weeks of jimmac time 14:27:40 We probably can just skip audio entirely unless we want to reuse the demo 14:28:39 mclasen___: When I did these for anaconda for a previous Summit, I spent on the order of 1.5-3 hours per video, depending on the complexity of the procedure 14:28:47 that included audio + subtitling 14:28:49 mclasen___: I don't think that it needs to look entirely polished and professional - probably could just be a live recording of someone with a few overlayed titles - getting jimmac to do something low-fi may be hard though :-) 14:28:50 + video editing 14:29:26 * mclasen___ had videos rendered in blender with nice overlays and transitions 14:29:26 A screencast of feature demonstration would probably be sufficient 14:29:31 it was really nice 14:29:36 mclasen___: yes, those were beautiful, I remember them well 14:30:01 however, if we don't intend for these to have long shelf life, something with a lower bar would still be fine 14:30:15 nostalgia aside, can collect a few topics / apps / use cases to cover ? 14:30:38 mclasen___: Sure... and if you're looking for a participant, I can set some time aside to help, since I've done it before 14:31:00 * stickster would also like to share his method/madness with others 14:31:01 Basic question is are we trying for these to be *F24 Workstation* videos or *Fedora Workstation* videos 14:31:58 If there is some super defining characteristic of F24 that would be awesome to highlight, great! 14:32:06 cschalle_: this is a great place for you to chime in, but IMHO we may want one or two that are general (the latter), setting out the reason Workstation exists, and highlighting some general characteristics 14:32:33 then maybe a few that are specifically the latest and greatest stuff e.g. in GNOME 3.20 as part of the Workstation 14:33:02 wfm 14:34:26 I would maybe make it future oriented so make videos advertising Wayland, Flatpak and so on. it is always easier to make people excited about the future than the present 14:34:46 cschalle_: good point 14:35:12 So it sounds like we're agreed that a few video demos shouldn't be too hard to do -- we have about 4 weeks 14:35:44 The Fedora booth is one place we can do that at the summit without making anybody grumpy! 14:35:51 otaylor: exactly :-) 14:35:55 this being for the summit maybe highlighting rishis work on subscription manager integration in fedora would be a plus, definetly the target audience 14:36:18 well, that work hasn't happened yet 14:36:31 mclasen___, yes, but we are selling the future here :) 14:36:33 so its hard to highlight it now 14:36:47 otaylor: cschalle_: I've also asked decause elsewhere to touch base with some Fedora-oriented Summit speakers, so they can point out the demos at the booth (not just Workstation) e.g. in the last slide of their respective talks 14:36:48 subscription manager? (Kerberos?) 14:37:18 stickster: nod nod 14:37:20 cschalle_: I think something like that would work better as mentioned *in* an associated talk, as opposed to trying to demo that which doesn't exist 14:37:31 mcatanzaro, rishi is looking into integrating the RHEL subscription manager with GOA, so that you can get automatic entitlements in docker for RHEL development 14:38:15 stickster, well we would definetly not demo it :) but in the world of animations and slides everything already exists :) 14:39:07 Right now I'm betting that Flatpak, Wayland (maybe? after all, shouldn't *look* different), Software distro-upgrade would work OK 14:39:26 Anyway, we can dump those ideas somewhere in writing 14:39:48 #action mclasen___ start a public etherpad or wiki page we can pull together the concepts in the next day or two 14:40:04 stickster: https://fedoraproject.org/wiki/Red_Hat_Summit_2016 14:40:13 there is a section at the bottom for ideas/brainstorming? 14:40:26 * decause likes etherpad too though 14:40:36 * stickster goes with whatever Matthias likes 14:40:45 nod nod nod 14:40:56 but we'll pull together the final plan there for sure, decause 14:40:58 thanks 14:41:12 so far it looks like... 14:41:14 #idea Wayland 14:41:17 #idea Flatpak 14:41:26 stickster: Wayland is really hard to demo convincingly 14:41:33 "see, you see nothing!" 14:41:39 mclasen___: yeah, true :-) 14:41:49 mclasen___: That might make for a funny video, actually 14:42:07 " Do you see any difference? EXACTLY" 14:42:14 it would be more convincing maybe to run the demo workstation under wayland and casually mention it 14:42:17 :P 14:42:21 mclasen___: disco 14:42:46 #idea demo definitely should run on F24 GNOME Wayland session! 14:42:59 #idea subscription manager integration with GOA 14:43:18 #idea Software distro upgrade 14:43:34 Anything else we can seed here? 14:44:01 #link http://etherpad.osuosl.org/fedora-rht-summit 14:44:13 this has some of our 'working content' in it already 14:44:26 there's plenty of gnome-software features that are demoable 14:45:09 mclasen___: Agreed, like software ratings, labels, etc 14:45:21 #idea other Software features 14:45:56 Some video that focuses on developer features? Not really sure what that is at the moment other than Boxes 14:47:09 otaylor: kind of makes one think we need to put together a better overarching developer plan... ;-\ 14:47:27 #idea Boxes or other developer features 14:47:45 I have a talk on that accepted at Flock, I better work on it :-) 14:47:51 otaylor: nice :-) 14:47:57 OK, let's take further stuff to wiki -- Matthias will let us know if he prefers elsewhere 14:48:04 shall we move on? 14:48:14 Thank you decause for being here and setting up the conversation 14:48:37 yeah, lets move on 14:48:39 #topic Open floor (all other business) 14:49:15 anyone got a topic? 14:49:25 I have a quick packaging question -- should we ship F24 with wayland 1.10 or wayland 1.11 RC ? 14:50:14 I think 1.10 is fine 14:50:19 ok 14:50:19 stickster: thanks for having me. and more importantly, thank you all in the WG for the work you do to make Fedora amazing for our users 14:50:55 I have a quick update from the workstation ostree effort 14:51:02 Go for it! 14:51:23 #info F24 to ship with Wayland 1.10, save 1.11+ for later 14:51:26 owen and I met again with the rel-eng team yesterday to continue the discussion of koji integration 14:51:54 no concrete progress yet, we've thrown around some of the pros and cons of various approaches 14:52:13 we'll bring this discussion to the koji maintainers to get guidance on the way forward 14:52:34 otaylor: stickster: hope thats a fair summary ? 14:53:04 mclasen___: Sounds right 14:53:20 mclasen___: +1 14:53:46 the great part of this is no surprises for F25 ("wait, you want to do *what*?!?") :-) 14:54:57 #info discussion in progress with koji maintainers on how to properly enable shipping ostree/flatpak deliverables in F25+ 14:55:09 anyone else? holding the door for 1 minute 14:55:42 * mclasen___ added the demo ideas to https://fedoraproject.org/wiki/Red_Hat_Summit_2016#Fedora_Workstation_Demo 14:55:51 Hey, just for my curiosity... has anyone else been experiencing BZ spam e.g. bug 1339549? 14:56:10 Thanks mclasen___ 14:56:24 First BZ spam I've seen 14:56:47 there was bug spamming going on yesterday on bugs.freedesktop.org 14:56:48 Not sure whether to escalate further. I think BZ maintainers are aware of it, but I have no idea what they're doing for countermeasures 14:56:54 mclasen___: I HATE THESE GUYS 14:57:11 Seriously, out of all the targets you could pick... why a FOSS bug tracker?!? 14:57:50 OK, enough on that... thanks everyone for coming :-) 14:57:54 #endmeeting