14:09:02 #startmeeting meeting 14:09:02 Meeting started Mon Nov 28 14:09:02 2016 UTC. The chair is mvollmer. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:09:02 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:09:02 The meeting name has been set to 'meeting' 14:09:10 .hello mvo 14:09:11 mvollmer: mvo 'Marius Vollmer' 14:12:36 nobody here? 14:13:56 .hello stefw 14:13:57 stefw: stefw 'Stef Walter' 14:15:00 i'm here, no agenda points though 14:15:05 same for me 14:15:31 * Debian Jessie 14:15:56 * Exporting DBus Objects from javascript 14:22:55 [cockpit] petervo pushed 1 new commit to master: https://git.io/v1LLQ 14:22:55 cockpit/master eefdc0a Stef Walter: ostree: Don't overflow the 'Check for Updates' button... 14:24:08 [cockpit] petervo pushed 1 new commit to master: https://git.io/v1LtJ 14:24:08 cockpit/master c0114e7 Stef Walter: ostree: Display signature data at full panel width... 14:27:52 * stefw wonders if he got disconnected or something 14:28:04 mvollmer, are my agenda points up there? 14:28:06 i think everyone just went quiet 14:28:14 they are 14:28:48 yep, sorry, started doing something else... 14:28:51 F25 COPR? 14:29:15 okay, i repeat the agenda 14:29:20 * debian jessie 14:29:29 * Exporting DBus Objects from javascript 14:29:41 * F25 COPR 14:29:53 * UDisks and Storaged news 14:30:14 okay, let's try to go through this quickly 14:30:24 #topic Debian Jessie 14:31:17 So we've done tons of work to test and fix Cockpit on Debian Jessie 14:31:37 Does it work to actually have a repo that lets people use the work that we did by next release? 14:31:42 it seems that people use it way more than unstable 14:31:55 even if we had to ditch the debian unstable repo, and replace it with a jessie repo we'd come out a head 14:32:02 what do you think larsu? 14:32:20 if it's too hard to do both? 14:32:33 I've been meaning to ask pitti about that 14:32:39 but haven't yet 14:33:16 but yeah, if all else fails, let's ditch unstable 14:34:54 ok, so you'll let us know when all else fails? 14:34:54 or just by next release? 14:34:55 if by next release there's no progress we ditch unstable? 14:35:35 sounds good to me. I'll try to reach pitti in the meantime 14:35:38 or just two repositories with two different directories? 14:36:03 I'd rather see different version numbers in the same repo 14:37:00 if you have a concrete way to do that, i'd be all for it 14:37:05 i mean fedora has that 14:37:13 f25 and f24 in the rpm names 14:37:19 well in the revision 14:37:30 yeah, seems to be common in debian and ubuntu package names as well 14:37:50 we couldn't use the same source package (which is part of your PR) in that case 14:38:03 otherwise that should be easy to do 14:38:44 ok, sounds good ... if both ubuntu, and each debian release need to build their own source package 14:38:46 then fair enough 14:39:05 but at the end of the day ... i think this is pretty important 14:39:10 it prevents the work we're doing from actually being used 14:39:15 right 14:40:16 ok, next topic? 14:41:36 #topic Exporting DBus Objects from javascript 14:42:50 So a lot of the code is done for exporting DBus objects from javascript 14:42:56 this isn't going to be creating real system services in DBus 14:43:07 but for when a system service expects to talk to something else exported 14:43:12 on the bus, by the "client" 14:43:19 for example a polkit agent 14:43:39 in the process the DBus client stuff is also getting cleaned up 14:43:45 are there other things driving this, or only being a polkit agent? 14:43:56 ahh, now I get what's going on... 14:43:59 just that one, and also the finishing of the API 14:44:26 some of my changes are running into NetworkManager race issues ... which mvollmer put in a heroic effort to solve a long time ago 14:44:33 sounded a bit weird that we need this in cockpit 14:44:45 that we need a polkit agent? 14:44:52 no, services in general 14:44:53 or that we need exporting dbus objects? 14:44:57 polkit makes a ton of sense of course 14:46:25 stefw, what races are those? 14:46:57 the ones where consistency of state cannot be guaranteed unless at certain points in the code 14:47:06 but anyway, i'm not complaining 14:47:10 ahh, races in our own JavaScript? 14:47:16 no in NetworkManager 14:47:24 ... but this means there was a regression in my pull requests with regard to ordering or something 14:47:34 so hopefully i can find the actual thing that regressed soon 14:47:40 okay 14:48:01 stefw, ping me if you need help 14:48:08 thanks 14:48:11 okay, that's it on that topic 14:48:23 unfortunately, I can't recall the details right now... 14:48:42 okay 14:48:50 #topic F25 COPR 14:49:00 sgallagh, that was your topic, right? 14:49:08 yes 14:49:38 COPR no longer automatically forks the releases, so if you want an F25 test repo, you need to manually add it to the COPR config 14:50:03 Right now it's building F23, F24 and Rawhide 14:51:03 thanks for the info 14:51:12 i don't know what to do, unfortunately. 14:51:38 mvollmer: It's a checkbox in the COPR admin interface for that project 14:51:52 (Though it also may require changes to your CI infra; that part I don't know0 14:51:58 stefw, do we release via coprs still? 14:52:01 i guess so 14:52:07 let's check that box then 14:52:07 yes, it happens automatically 14:53:39 i guess that's up to me? 14:53:43 let me see 14:54:30 done 14:54:43 * stefw notes that he just went here and edited it: https://copr.fedorainfracloud.org/coprs/g/cockpit/cockpit-preview/ 14:55:16 thanks 14:55:50 we have zero downloads... 14:56:04 anyway 14:56:14 #topic UDisks and Storaged news 14:56:38 Martin Pitt has proposed to merge udisks2 and storaged back togetehr again 14:56:44 so that's great news 14:56:56 looks like it is going to happen 14:57:21 woo 14:57:40 mvollmer: RE zero downloads. I think there's a bug there, because I use the COPR :) 14:57:43 pitti is the same person, right? 14:58:26 Is it going to keep the (better) storaged name? 14:58:36 martin pitt is pitti, yes 14:58:39 https://lists.freedesktop.org/archives/devkit-devel/2016-November/001764.html 14:59:00 sgallagh, i don't know, that's the difficult part 14:59:10 sgallagh, you can join the bikeshed at that post :D ^^ 14:59:26 personally, I wouldn't mind going back to udisks2 and freeing the storaged name for the real next step 15:02:34 alright, done? 15:02:54 sorry for messing up the start 15:03:20 no worries 15:03:47 #endmeeting