[03:55] <jamesh> robert_ancell: thanks for the snapd/polkit patch.  I had found your old pull request, and reused some of the ideas and code in my own pull request
[03:55] <jamesh> robert_ancell: Here's the result if you're interested: https://github.com/snapcore/snapd/pull/3581
[03:55] <robert_ancell> jamesh, ah good, I wasn't sure if the old PR still existed
[03:57] <jamesh> robert_ancell: I used godbus instead of libsystemd, so hopefully that makes things a little more palatable
[03:59] <robert_ancell> jamesh, last time there were issues raised that godbus wasn't reliable enough. I never worked out what the issues were; my conclusion was the team had been burnt by it in the past and were wary.
[03:59] <robert_ancell> I feel like I've been along this path before :) I hope you are more successful!
[04:01] <robert_ancell> I like how you just did /v2/login. Now we have snapd-login-service in the wild this feels like a logical step that should make sense.
[04:07] <jamesh> yep.  And if we can get this through, it is a much smaller battle to say polkit auth should be treated like root access for "snap install"
[04:09] <jamesh> but on its own, it hasn't removed the need for macaroons.
[04:09] <jamesh> so hopefully the discussions won't go down that related rabbit hole
[04:11] <robert_ancell> Small steps.
[04:13] <robert_ancell> jamesh, Something I was concerned about is the Polkit failure case. In this case it will just fall through and use the existing behaviour. I think that's safe because a) polkit is assumed to be an essential service if it is installed and b) the fallback is safe
[04:14] <robert_ancell> On an Ubuntu Core system I guess you risk seeing errors about Polkit not existing, but that shouldn't be too much of a problem.
[04:14] <jamesh> Right.  I'm just logging the error and continuing, which is essentially failing closed.
[04:15] <jamesh> if polkitd isn't available, you'll just need to be root to use that action
[04:15] <jamesh> which is the status quo
[04:15] <robert_ancell> yeah
[04:16] <robert_ancell> The snap design seems to be big on highly predictable behaviour, so I hope they're OK with graceful degradation.
[04:17] <robert_ancell> I figure it can't be compile time due to the way snapd is shipped, though it could be a config option of some sort if that was desirable.
[04:17] <jamesh> I think they'd take the ability to share the same snapd binary between core and classic over stricter error handling
[04:17] <robert_ancell> You'd think so :)
[04:18] <jamesh> another possible improvement for this would be to add pkttyagent support to the snap client
[04:18] <jamesh> that would give sudo-less access outside of a graphical session
[04:18] <robert_ancell> I think that every time I use systemctl
[06:09] <oSoMoN> good morning desktoppers
[06:37] <didrocks> good morning
[06:37] <jamesh> hi oSoMoN, didrocks
[06:38] <didrocks> hey jamesh
[06:38] <didrocks> how are you?
[06:39] <jamesh> good.  I got my snapd/polkit branch working, so I'm hoping we can keep pushing that forward.
[06:40] <jamesh> It implements only enough to let you run "snap login" without sudo, which is a start.
[06:42] <didrocks> yeah, reading that on the forum :)
[06:42] <didrocks> let's see what answers we'll get
[07:19] <seb128> good morning desktopers
[07:19] <seb128> hey oSoMoN jamesh
[07:19] <flexiondotorg> Morning seb128 jamesh oSoMoN didrocks
[07:20] <seb128> hey flexiondotorg, how are you?
[07:20] <didrocks> hey flexiondotorg, how are you?
[07:20] <flexiondotorg> Very well thanks. Had a long weekend away. Sunshine and boats :-)
[07:21] <didrocks> ok, indicator-messages should now be off the iso with those changes
[07:21] <didrocks> remaining is indicator-applications
[07:21] <didrocks> onboard doesn't have a good upgrade story
[07:21] <didrocks> (basically, no systray/indicator on G-S)
[07:22] <didrocks> I can patch it easily so that the "auto-detect" shows up the systray
[07:22] <didrocks> look for a "onboard" extension
[07:22] <didrocks> and only show systray if no extension is installed
[07:22] <didrocks> (standard upgrade)
[07:22] <didrocks> that way, I can dowgrade the recommends on indicator-application to recommends
[07:22] <didrocks> shouldn't be hard
[07:22] <didrocks> BUT
[07:22] <didrocks> there is update-notifier as well which deps on it
[07:23] <didrocks> (I meant "downgrade indicator-application to suggests")
[07:23] <didrocks> and on update-notifier, as it's C, it's harder to make a runtime dep detection…
[07:23] <seb128> it's not because they use libappindicator that they need a renderer for it
[07:24] <seb128> we should enable https://extensions.gnome.org/extension/615/appindicator-support/
[07:24] <didrocks> does it use the same dbusmenu stack?
[07:25] <didrocks> to communicate
[07:25] <seb128> ?
[07:25] <seb128> I think that protocol is common to unity & kde
[07:25] <seb128> I don't think it's dbusmenu specific?
[07:25] <didrocks> libappindicator pull dbusmenu to communicate over dbus (I thought we switched to the gtk/glib menu at some point, though?)
[07:26] <didrocks> and dbusmenu is pulling part of the libunity stack
[07:26] <seb128> could be, but that's orthogonal to the renderer/indicator-application
[07:26] <seb128> or are you talking about getting ride of dbusmenu from the iso now?
[07:26] <seb128> my point was that we could have clients using libappindicator still
[07:26] <didrocks> I'm talking about getting ride of most of the unity stack/dee and such dbusmenu from the iso
[07:26] <seb128> and the rendering in the shell done by that extension ^
[07:26] <seb128> k
[07:27] <seb128> I though you were talking about indicator-application
[07:27] <didrocks> sorry, as a source package
[07:27] <didrocks> wasn't clear enough
[07:28] <seb128> well indicator-application is just the applet rendering appindicator icons right?
[07:28] <seb128> I mean it's a one binary source and that can go to universe easily
[07:28] <didrocks> correct, but the source package is libindicator, which provides that libs, which has this long trail of deps
[07:29] <seb128> https://launchpad.net/ubuntu/+source/indicator-application
[07:29] <seb128> I was talking about
[07:29] <seb128> let me look at the libs
[07:29] <didrocks> oh that, yeah
[07:29] <didrocks> sorry, read the results the wrong way, not enough coffee :p
[07:29] <seb128> np :-)
[07:29] <didrocks> and this libindicator vs libappindicator
[07:29] <didrocks> grrr, afer all those years :p
[07:29] <seb128> ahah
[07:30] <didrocks> so yeah, talking about libappindicator
[07:30] <seb128> libappindicator depends on libindicator and libdbusmenu
[07:30] <didrocks> yeah
[07:31] <seb128> it's another of those tradeoff discussion
[07:31] <seb128> libappindicator is used by a bunch of packages in main
[07:31] <seb128> so either we drop its use and regress a bit more the unity session
[07:31] <seb128> or we keep them building with it
[07:32] <seb128> which is low space on the iso but more unmaintained code we carry on the iso
[07:32] <didrocks> yeah, how can we come to a conclusion on this? Would be great to have a discussion with involved parts
[07:32] <seb128> though the code is mostly stable and low maintainance
[07:32] <didrocks> (and patches)
[07:32] <seb128> but still technical debt
[07:32] <didrocks> I don't really care about the space vs the patch/diff maintainance on apps
[07:33] <seb128> step 1 is to list it on the wiki
[07:33] <seb128> one we think the wiki list is complete then we can have the discussion
[07:34] <seb128> with interested parties (including Mark?), on what we do from unity7
[07:34] <didrocks> yeah, I guess that, with the list of packages on the iso at least as example of regression vs perf
[07:34] <didrocks> yep
[07:34] <didrocks> seb128: maybe we'll be more productive doing a HO (like tomorrow), together and go to the list?
[07:34] <seb128> tomorrow sounds good, +1
[07:35] <didrocks> let's do that :)
[07:36] <seb128> cool
[07:57] <willcooke> morning all
[07:58] <didrocks> hey willcooke
[07:58] <willcooke> hi didrocks!  Is part 2 of your tutorial out soon?  Want to reshare, but I think it would be good for them to have just a little more
[07:58] <willcooke> in other news... I built a PPA!  (thanks kenvandine for the help)
[07:59] <didrocks> willcooke: I can publish it in a few minutes, let me do a last typo reread :p
[08:00] <willcooke> didrocks, woot
[08:00] <duflu> willcooke, seems one debhelper regressed 7 hours ago. I just grabbed a proposed fix so in theory can test your branch proper now
[08:00] <willcooke> duflu, afternoon!
[08:00] <duflu> willcooke, and happy Wednesday
[08:01] <willcooke> duflu, I dont think we'll merge it yet, spoke to s_eb128 about it yesterday.  I think speaking to upstream PAudio for a "real" fix is the right thing to do atm
[08:01] <Laney> yo
[08:01] <willcooke> duflu, I'll talk about it some more in the meeting
[08:01] <willcooke> morning Laney
[08:02] <duflu> willcooke, OK nevermind. BTW I noticed a new feature highlighted in bold here. Seems to be very timely - http://www.android-x86.org/releases/releasenote-7-1-rc1
[08:02] <willcooke> duflu, but if you're interested:  https://launchpad.net/~willcooke/+archive/ubuntu/gdm-nobluetoothaudio
[08:02] <duflu> willcooke, ta but I was trying to do the proper thing and test the branch directly
[08:03] <willcooke> duflu, ah right.  Thx
[08:03] <seb128> hey willcooke Laney, how is u.k today?
[08:03] <willcooke> rainy
[08:03] <willcooke> which is great
[08:03] <willcooke> :)
[08:03] <seb128> wonder if it's raining as much for you that it is here
[08:03] <seb128> lol
[08:03] <seb128> :p
[08:04] <willcooke> My grass might turn green again
[08:04] <willcooke> ooooooh!  Software update notifications on the lock screen, I like that
[08:05] <duflu> Welcome to the world of upstream shell collaboration :)
[08:06] <willcooke> \o/
[08:07] <Laney> hey seb128
[08:08] <Laney> not raining here, we had enough yesterday for a week ...
[08:08] <Laney> how are you?
[08:09] <didrocks> willcooke: just git push, should be up in few seconds
[08:09] <willcooke> didrocks, merci
[08:09] <willcooke> didrocks, can I suggest a link from pt 1 to pt 2 and vice versa?
[08:10] <didrocks> ah, I have done pt2 to pt1
[08:10] <didrocks> I can add from pt1 to pt2, one sec
[08:12] <duflu> Umm, I no longer have a mouse on the login screen (but Gnome Shell does). Is that a feature?
[08:13] <didrocks> willcooke: link pushed
[08:13] <willcooke> thx didrocks
[08:13] <didrocks> yw
[08:13] <willcooke> duflu, hum.  Bluetooth mouse?
[08:13] <duflu> No, USB. It's visibly powered but no cursor in gdm
[08:14] <duflu> OTOH I am testing a custom gdm
[08:14] <willcooke> duflu, testing here
[08:15] <duflu> But Bluetooth audio is fixed  :)
[08:15] <duflu> Forget it, unless the problem occurs in official builds
[08:21] <willcooke> duflu, yeah, mouse works here.
[08:22] <duflu> Cosmic rays
[08:22] <duflu> Moving on...
[08:22] <willcooke> :)
[08:22]  * willcooke goes to try the auto connect fixes
[08:22] <willcooke> "switch-on-connect"
[08:27] <willcooke> duflu, bah, no good.  Could be this particular speaker though
[08:28] <duflu> BT and USB hotplugging works great for me now. The USB part drove me crazy for years and I guess BT is a big deal for others
[08:28] <duflu> Strangely also fixed a separate USB DAC problem... somehow
[08:30] <willcooke> hehe
[08:32] <willcooke> so, toggling the speaker on and off in the control panel seems to work well,  just powering off the speaker causes a bit of a lock up and eventually switches back to the internal speaker (probably fair enough, since there isn't a handshake, it's just OFFFFFF)
[08:33] <willcooke> powering the speaker back on doesnt re-connect automatically
[08:33] <willcooke> I'll test with another speaker, since this one is super-cheapo
[08:33] <Laney> Setting up gdm3 (3.24.2-1ubuntu5) ...
[08:33] <Laney> dpkg-query: package 'ubuntu-session' is not installed and no information is available
[08:33] <duflu> willcooke, Yeah bouncing bluetooth doesn't seem to be robust after a few attempts. But it's seemingly getting better. I think this is as good as it's ever been
[08:33] <Laney> /o\
[08:34] <willcooke> duflu,  it's a clear improvement in my book
[08:34] <willcooke> found another missing icon
[08:36] <willcooke> duflu,  new speaker - a Sony one - works perfectly!!!
[08:36] <duflu> I can also say "It's a Sony"
[08:37] <andyrock> willcooke: by what time you need the demo ready? End of day or before?
[08:37] <willcooke> andyrock, EOD is fine for me
[08:37] <andyrock> and good morning
[08:37] <willcooke> hi!
[08:37] <andyrock> kk thanks
[08:37] <andyrock> :)
[08:38] <willcooke> duflu, lulz..  this cheapo one... "protable":  http://imgur.com/a/9WUE0
[08:44] <willcooke> didrocks, soo.. rebuilding those blender files for the help movies, I left it runinng for 3 days and it never finished.  Can't tell if it's broken, or if it just takes a long time to run.  Do we upload the videos as binaries. or are they built in LP?
[08:47] <didrocks> willcooke: right now, they are all pre-built
[08:47] <didrocks> distributed with the source
[08:47] <willcooke> didrocks, oki, that's probably good.  Do you have a machine in the cloud somewhere that I could set to work on them?
[08:47] <willcooke> I'm not sure I can stand the fan noise any more :D
[08:48] <didrocks> willcooke: I don't, maybe you can try canonistack?
[08:48] <willcooke> didrocks, nod
[08:51] <didrocks> willcooke: jibel: FYI, the gdm3 simple patch is uploaded to favor X11 sessions over wayland when no default is set
[08:51] <didrocks> (still in proposed)
[08:51] <willcooke> didrocks, nice one, thanks!  You got to the bottom of it then?  Just gdm prefers Wayland?
[08:52] <didrocks> willcooke: yep, when there has been no session selected, which still makes me wondering for "your" case
[08:52] <didrocks> as you did select "ubuntu" on purpose and not "ubuntu on wayland"
[08:55] <willcooke> didrocks, ah right.  I'll do another fresh install (maybe tomorrow) and try again
[09:01] <jibel> didrocks, okay, I'll try it. I'll install the package at the end of a fresh installation and before first login
[09:02] <didrocks> jibel: this is gdm3 3.24.2-1ubuntu6
[09:18] <ricotz> didrocks, hi, added thunderbird-gnome-support to Breaks of ubuntu-session?
[09:22] <didrocks> ricotz: urgh, was probably not enough coffee, wanted recommends
[09:22]  * didrocks checks
[09:24] <didrocks> ricotz: correct, more coffee was needed. Thanks for spotting it! uploaded.
[09:25] <ricotz> didrocks, it was hard to miss here due suggested package removals
[09:26] <didrocks> ricotz: yeah, I guess ;)
[09:26] <didrocks> (I've already removed it here to test in a G-S session)
[09:39] <duflu> willcooke, oops. Your fix works permanently as the new file doesn't uninstall
[09:39] <willcooke> duflu, heh, yay
[09:39] <willcooke> duflu, I've made a mess of that branch now by hacking the changelog, so if we decide to use it I'll make a nice clean one.
[09:40] <duflu> willcooke, it's fine. bzr doesn't make history very discoverable after a merge
[09:41] <willcooke> so should I consider putting something in the postrm file as well?
[09:42] <duflu> willcooke, probably? Although I thought debs had better config file management...?
[09:42] <jamesh> "bzr log -n0" will show all your sins
[09:43] <duflu> Yes, but my point was nobody looks there too much
[09:43] <jamesh> (assuming you consider the process of software development to be a sin)
[09:44] <duflu> sinners use spaces :)
[09:45] <Laney> willcooke: The gdm user's home directory is already removed in the postrm
[09:45] <Laney> (when purging)
[09:45] <willcooke> Laney, the file lives in /var/lib/gdm3/.config/pulse
[09:45] <willcooke> (or something like that)
[09:46] <Laney> I know, that's the home directory
[09:46] <didrocks> /var/lib/gdm3 is gdm user's home
[09:46] <willcooke> oh
[09:46] <didrocks> as Laney pointed :)
[09:47] <willcooke> the moral of this story is that I will ask someone who knows what they are doing to help me if we decide to actually go with this workaround
[09:47] <duflu> willcooke, I'm looking at the upstream docs for how to do it. Gimme a sec
[09:48] <willcooke> Laney, btw - we just had our Bluetooth meeting.  Decision is that seb128 will speak to upstream and see if we can get a proper fix in place.  If that doesn't go anywere we can reconsider the proposed work-around and do it properly
[09:49] <Laney> k, that sounds right to me
[09:49] <duflu> OK, that didn't work. But anyway willcooke there is https://www.freedesktop.org/wiki/Software/PulseAudio/Documentation/User/PerfectSetup/
[09:50] <willcooke> blimey, I remember reading that page yeeeears ago
[09:51] <willcooke> is ESD still a thing?
[09:51] <willcooke> oh, no, it says about replacing esd with pulse
[09:52] <duflu> Hmm, yes. If you kill and restart the gdm pulseaudio then the user's own one gets priority and works
[09:52] <duflu> Or maybe that's a race
[09:53] <willcooke> I think thats not going to work perfectly.  a11y would be unavailble at the lock screen then
[09:53] <willcooke> or, if you killed the users pa on lock, then music would stop playing when you locked the machine
[09:56] <duflu> willcooke, did you say you saw the mention of the problem on Arch wiki?
[09:56] <willcooke> duflu, yeah.  This debian wiki:  https://wiki.debian.org/BluetoothUser/a2dp
[09:56] <willcooke> then links to this arch wiki:
[09:57] <willcooke> https://wiki.archlinux.org/index.php/Talk:Bluetooth_headset#GDMs_pulseaudio_instance_captures_bluetooth_headset
[09:59] <duflu> I wonder if there is an a11y problem at all. Before today Pulse would default to the PC speaker or whatever is plugged in to that. I doubt a11y users could have been using BT in GDM if they wanted to
[10:00] <duflu> Although I forgot about static priorities and don't remember testing booting with the device on previously
[10:00] <willcooke> AIUI, the Pulse settings are system wide, so.. if you logged in, paired the headset/speaker, logged out again - then it /might/ try and use it at greeter - so maybe it's possible?
[10:01] <willcooke> quite a lot of guess work there ^ :)
[10:01] <duflu> willcooke, no I don't think we use that mode and it's mildly not recommended https://www.freedesktop.org/wiki/Software/PulseAudio/Documentation/User/SystemWide/
[10:02] <duflu> That should solve it tho
[10:04] <duflu> Anyway, I feel terrible. Time to try this new medication...
[10:13] <ricotz> didrocks, better check gnome-session again, the changelog mentions unity-session and you touched ubuntu-session
[10:15]  * didrocks *sighs*, ok no more upload apart from that fix today :/
[10:15] <jamesh> willcooke: I'm still not sure whether we'll be able to convince Gustavo to go fully macaroon-less for local snapd auth, but I've got a working first stab at polkit integreation for snapd: https://github.com/snapcore/snapd/pull/3581
[10:16] <ricotz> didrocks, having things reviewed helps ;)
[10:16] <ricotz> ... before pushing it
[10:16] <didrocks> ricotz: yeah, especially when doing too many things at the same time :p
[10:16] <jamesh> this essentially just lets snapd do what we use snapd-login-service for
[10:16] <didrocks> well, at least it's a minor typo, not breaking people compared to other issues ;)
[10:17] <didrocks> (gnome-session has too many binary packages, easy to scroll up and be in a different one)
[10:20]  * didrocks did a debdiff + bzr diff this time
[10:31] <willcooke> jamesh, sorry was otp
[10:31] <willcooke> reading
[10:32] <willcooke> jamesh, ah nice one!
[10:45] <willcooke> What's the wayland tag for bugs again?  Is it just "wayland"?  I had a feeling there might have been a different one
[10:45] <seb128> "wayland"
[10:45] <willcooke> thx seb128
[10:45] <seb128> yw
[10:49] <Laney> jamesh: hope you manage to get buy in for that work
[10:49] <Laney> & hi!
[10:51] <jamesh> Laney: I do too.  It sounds like there is a desire to tighten up the API security a bit, so hopefully they'll see this as part of it.
[11:05]  * Laney nods
[11:51] <jibel> willcooke, it's wayland but apport adds the tag wayland-session if wayland is running
[11:53] <jibel> https://bugs.launchpad.net/ubuntu/+bugs?field.tag=wayland-session
[12:04] <jibel> seb128, what do you think about using wayland-session instead of wayland for the official tag
[12:04] <jibel> ?
[12:04] <jibel> and merge both lists
[12:05] <seb128> either way is fine to me, just pick whatever you think is easier
[12:05] <seb128> but maybe check with jbicha as well
[12:08] <jibel> I'd keep wayland-session, easier to change the official tag than updating and releasing apport
[12:08] <jibel> jbicha, what do you think? ^
[12:12] <jbicha> I think it's good to have 2 different tags
[12:13] <jbicha> one tag is because the user happens to be running wayland but doesn't mean the problem is specific to wayland
[12:15] <jibel> so someone has to review the bugs tagged wayland-session and add the wayland tag if the problem is specific to wayland
[12:16] <jbicha> yes
[12:17] <jibel> jbicha, why not use 1 tag for wayland eg wayland-session and remove the ayland-session tag if it is not specific to wayland?
[12:17] <jibel> I don't see any benefit in having 2 lists for wayland
[12:18] <jbicha> it's like the 'amd64' tag; it's useful to know what architecture someone is running but it's rare for the bug to be because they're using amd64
[12:18] <jbicha> very few bugs are because of Wayland but we still want to know whether the user is running Wayland
[12:20] <jbicha> it's only because Wayland is not the default yet that the number of wayland-session bugs is so small, and maybe someone triaged some of those to remove the tag too (but that's needless busywork)
[12:23] <seb128> jibel, one list doesn't let you make the difference between triaged bugs which are confirmed to be wayland specific (=bugs to look at for us)  and the stack of incoming reports from users on the wayland session (need triaging)
[12:24] <jibel> seb128, that's fine as long as someone triage them. Several bugs tagged wayland-session seem wayland specific and are not tagged wayland. These bugs are useful to help with the decision to go or not go with wayland
[12:28] <jibel> didrocks, just to confirm, with latest gdm3, autologin, the session is running ubuntu-x11
[12:30] <didrocks> \o/
[12:35] <seb128> jibel, fair point, we should triage them regularly
[12:49] <willcooke> Can anyone work out where the heck this icon is coming from?  http://imgur.com/a/r5CLa
[12:49] <jbicha> seb128: is it worth proposing your evolution change upstream? and remember to do it for e-d-s too :)
[12:49] <willcooke> I'm pretty sure it's called "audio-headset" and I've looked at all the ones on disk, and none of them match
[12:52] <jbicha> willcooke: it's probably from adwaita-icon-theme
[12:53] <jbicha> I opened up gtk3-icon-browser and looked for it, then I ran  apt-file search audio-headset
[12:54] <jibel> willcooke, looks like it's from adwaita
[12:54] <jibel> willcooke, /usr/share/icons/Adwaita/24x24/devices/audio-headset.png
[12:55] <willcooke> thanks!
[12:56] <willcooke> so, related to that, you can see on that screenshot the missing icon - IMO a direct copy / symlink to that headset one is good enough.  So do I need to try and fix that in Adwaita, or can we bodge it at an Ubuntu level?
[12:57] <willcooke> Could I add that to, say, Humanity
[12:58] <jbicha> willcooke: please check if installing adwaita-icon-theme-full makes a difference
[13:00] <willcooke> jbicha, will do, and thanks for the point to the icon browser, very useful indeed
[13:01] <seb128> jbicha, I guess I can try to upstream them, technically what upŝtream does is a bit better but in practice it shouldn't make a difference, I'm just unsure "we have tools parsing that file which are not smart enough to understand variables" is a good justification
[13:19] <Laney> didrocks: kenvandine: I made a change to the gdm3.config - want to review it? https://bazaar.launchpad.net/~ubuntu-desktop/gdm/ubuntu/revision/421
[13:20]  * Laney goes to lunch
[13:21] <kenvandine> Laney, we had wanted to do the check based on ubuntu-session being installed, not for all users of gdm3
[13:21] <kenvandine> based on our original discussion
[13:22] <kenvandine> but i guess if ubuntu-session isn't installed, users won't have their session set to ubuntu
[13:24] <willcooke> jbicha, full icon theme doesn't help, and I dont see it in icon browser
[13:24] <willcooke> https://bugzilla.gnome.org/show_bug.cgi?id=696402
[13:24] <willcooke> makes mention of the headset one, and says its in gnome-icon-theme (comment #7)
[13:25] <willcooke> grr,ignore that.  I'm looking for handsfree
[13:25] <willcooke> which is says is to come
[13:28] <didrocks> Laney: +1 on the change. It covers all cases we discussed together and I like it better than previous implementation. My only nitpick is that you revert if people reselected "lightdm" since last week as you compare with your own version, but I guess this is fine :) For backport, maybe compare with 3.24.2-1ubuntu7~ rather? (if people backported it to xenial), minor nitpick :p
[13:33] <seb128> willcooke, what's the issue with that headset icon? doesn't represent the device well?
[13:33] <seb128> willcooke, also you are investigating the missing icon for handsfree?
[13:33] <willcooke> rather the handsfree icon is missing.  I think it could just be a copy of the handsfree one.  I can't find a "real" one anywhere.
[13:33] <willcooke> *headset
[13:33] <willcooke> handsfree could be a copy of headset
[13:34] <popey> jibel: yo, got time for a hangout this afternoon about desktop testing?
[13:34] <willcooke> (seems two words begining with the same letter is more than my brain can handle)
[13:35] <seb128> willcooke, that's sort of what was suggested in https://bugs.launchpad.net/ubuntu/+source/gnome-control-center/+bug/1130137
[13:35] <willcooke> wfm
[13:36] <jibel> popey, sure, when?
[13:36] <jibel> popey, just add something to my calendar
[13:36] <willcooke> seb128,  so would that "fix" be in Ubuntu?
[13:36] <seb128> willcooke, ideally upstream but we could add the symlink to our package meanwhile
[13:38] <willcooke> seb128, ack, thx
[13:38] <popey> jibel: kk
[13:54] <Laney> kenvandine: yeah, but I don't think that can work properly if the check isn't in ubuntu-session itself
[13:56] <Laney> didrocks: yeh, but I was worried about previous versions not working properly
[13:56] <Laney> not sure about the ~, what's the idea there?
[13:56] <didrocks> Laney: like for a lot of upgrade story, if someone backport the same version to a ppa, like ~ppa1, we don't want to remigrate after this
[13:57] <kenvandine> Laney, ok, +1 from me then
[13:57] <didrocks> but again, that's in the case people backport the stack (to xenial), not sure if you care, I don't that much
[14:01] <Laney> nod, I think that makes sense
[14:30] <jibel> how can I disable the snap plugin of gnome-software?
[14:30] <Laney> delete it
[14:31] <jibel> :) ok
[14:33] <jbicha> jibel: if you're using Artful, you can look up GNOME Software in GNOME Software and turn off Snappy support
[14:33] <jbicha> or just uninstall gnome-software-plugin-snap
[14:34] <Laney> ...
[14:34] <didrocks> Laney: btw, the gnome software icon isn't the ubuntu one (I didn't check why yet), did you notice?
[14:35] <didrocks> IIRC, there was an onlyshowin=ubuntu
[14:36] <jbicha> it shows up as Ubuntu Software in 'Ubuntu' but not in 'Ubuntu on Wayland'
[14:37] <didrocks> the icon isn't Ubuntu Software though
[14:37] <didrocks> which was the point made in the trello card
[14:37] <Laney> actually I see both
[14:37] <didrocks> oh, I don't
[14:38] <didrocks> only one entry "ubuntu-software"
[14:38] <jibel> jbicha, ah thanks. Sounds better.
[14:38] <didrocks> but with upstream icon
[14:38] <Laney> k, well it'll only work in a session called ubuntu
[14:38] <didrocks> as told:
[14:38] <didrocks> $ echo $DESKTOP_SESSION
[14:38] <didrocks> ubuntu
[14:38] <Laney> k
[14:38] <jibel> the snap plugin seems to have a huge negative impact on performance of gnome-software
[14:39] <Laney> if you want to debug it that would be welcome
[14:39] <didrocks> yep, will do :)
[14:51] <seb128> jibel, on what actions? startup? search?
[14:52] <didrocks> willcooke: how do you import bug reports to trello automatically?
[14:52] <didrocks> I don't see a tag
[14:52] <willcooke> didrocks, add "desktop-trello-import" tag and then tell me the project name
[14:52] <willcooke> package name
[14:53] <willcooke> whatever it's called
[14:53] <willcooke> the thing
[14:53] <willcooke> :)
[14:54] <didrocks> willcooke: the things are: gdm3 gsettings-desktop-schemas gnome-software
[14:54] <didrocks> :)
[14:54] <willcooke> didrocks, same bug for all of them?
[14:54] <willcooke> looks like gnome-software was the right one
[14:55] <willcooke> https://trello.com/c/nUBUEkyG/188-bug1703849-gnome-software-doesnt-show-ubuntu-icon-in-the-ubuntu-session
[14:55] <didrocks> willcooke: no, it's 3 different bugs I wanted to attach
[14:55] <willcooke> didrocks, that will appear as 3 different cards
[14:55] <willcooke> is that ok?
[14:55] <didrocks> willcooke: as I expect, they are 3 different things :)
[14:55] <didrocks> willcooke: yes
[14:55] <willcooke> kk
[14:56] <didrocks> thx!
[14:56] <jibel> seb128, search for sure, maybe startup but I don't have any data and startup is already really slow just with debs because it refreshes the package cache
[14:56] <willcooke> didrocks, https://trello.com/c/fX8lP0VF/189-bug1703326-lock-screen-wallpaper-is-a-plain-blue-screen
[14:57] <didrocks> willcooke: doesn't do assignement matching, I'm soooooooooooooooooooo disappointed :)
[14:57] <willcooke> didrocks, hm, the gdm3 one isnt being found, got a bug number?>
[14:57] <willcooke> didrocks, its on my todo list
[14:57] <willcooke> :)
[14:57] <didrocks> willcooke: it's fix released, maybe you don't import them? (it was just for tracking)
[14:57] <willcooke> ah
[14:57] <willcooke> oki,
[14:57] <willcooke> --force to the rescue
[14:57] <didrocks> ahah :)
[14:58] <willcooke> yay!
[14:58] <willcooke> https://trello.com/c/1wEWXt9p/190-bug1703601-use-ubuntu-x11-as-default-session
[14:58] <didrocks> merci beaucoup !
[14:58] <willcooke> bien sur
[17:15]  * willcooke found the kids easters eggs in the cupboard.  Now they are mine.
[17:15] <willcooke> nom
[17:20] <sarnold> is it your fault they didn't find them just because they were hidden on the top shelf way in the back? they had their chance
[17:22] <willcooke> I'm sure I don't know what you're talking about
[17:22] <willcooke> :D
[17:23] <willcooke> but you're right, out of sight out of mind.
[17:23] <jbicha> fossfreedom: Budgie doesn't actually need the 'mutter' binary package, does it? (it doesn't have that dependency on Debian)
[17:29] <fossfreedom> jbicha - hmm - no it shouldn't need it
[17:29] <fossfreedom> done a test on the live ISO - removed mutter logged out and logged back in successfully
[17:30] <jbicha> fossfreedom: does today's live iso boot to Budgie or to GNOME Shell?
[17:30] <fossfreedom> gnome-shell is still installed.
[17:31] <fossfreedom> I note - there is a mistake in our seeds
[17:32] <fossfreedom> we have mutter listed in the desktop seed
[17:32] <jbicha> see my comment on LP: #1703685
[17:34] <fossfreedom> interesting.
[17:35] <fossfreedom> ok - I still need to remove mutter from our desktop seed though
[17:35] <jbicha> we could try having metacity just not recommend gnome-session then
[17:38] <jbicha> Budgie looks like the only Flavor that uses metacity for ubiquity, MATE & Kylin use marco and Xubuntu uses xfwm4
[17:39] <sarnold> polo
[17:39] <jbicha> I wonder what would be needed for Budgie to handle the wm for ubiquity?
[17:43] <fossfreedom> the window manager is basically mutter as you know - budgie-wm is a "plugin" to mutter
[17:47] <jbicha> maybe you could look at what ubiquity does for gnome-shell and try doing the same for budgie?
[17:51] <Laney> ah man, my gdm3 ist falsch
[17:53] <Laney> or is it
[17:53] <Laney> root     22177  0.0  0.0   4496   764 pts/2    S+   18:49   0:00                      |                               \_ /bin/sh /tmp/lightdm.config.cQpVrt configure 1.22.0-0ubuntu2.1
[17:55] <Laney> guess lightdm's .config has to be neutered somehow
[17:56] <Laney> kenvandine: ^- if you have any ideas, otherwise I'll look tomorrow
[17:56] <fossfreedom> jbicha - in the control file for ubiquity  gnome-shell is the alternate - so not really sure what more I should be looking at
[17:56] <Laney> that's: start lxd zesty container, install ubuntu-desktop, make sure an AS user file exists, do-release-upgrade -d, see prompt apparently coming from lightdm
[17:57]  * Laney waves
[17:57] <kenvandine> Laney, no ideas
[17:57] <willcooke> night Laney
[17:57] <Laney> kenvandine: me neither right now, would have to think about it a bit ;-)
[17:57] <Laney> tara
[17:57] <jdstrand> kenvandine: hey, fyi, I just added snap decls for your uploads
[17:58] <jdstrand> kenvandine: curious, do this run with wayland if install with --devmode?
[17:58] <kenvandine> jdstrand, haven't tried with wayland
[17:58] <kenvandine> just strict
[17:58] <kenvandine> with X
[17:59] <jdstrand> kenvandine: I'm dangerously close to picking up the wayland interface work and found that the existing snaps in the store would not work with wayland if installed in devmode
[17:59] <jdstrand> kenvandine: is this something that the desktop team is looking to do? (getting these apps to work with wayland)
[17:59] <jdstrand> iirc, we want wayland by default for 18.04
[18:00] <fossfreedom> jbicha - I presume gnome-shell as a dependency for gnome-session is a recent change for artful - I don't see a recommendation or dependency in previous versions of gnome-session
[18:00] <jdstrand> not sure if I misheard of if that has changed
[18:00] <kenvandine> jdstrand, i haven't been in any discussions around the snaps with wayland
[18:00] <kenvandine> jdstrand, but clearly we want them to work
[18:00] <jbicha> fossfreedom: in bin/ubiquity-dm, it already apparently tries to run budgie-wm, so just try adding budgie to the alternate depends list
[18:04] <jbicha> jdstrand: I assume having snaps work in Wayland would be a high priority even if we don't use Wayland by default
[18:04] <jbicha> wayland by default is currently undecided
[18:06] <jdstrand> jbicha: I see, well, I'm just asking cause I may pick up the snappy interface work this week and may need some assistance with a working devmode snap
[18:11] <kenvandine> jdstrand, that's exciting
[18:12] <kenvandine> jdstrand, ping me if you need anything
[18:12] <kenvandine> if i can't help i can redirect as needed
[18:23] <willcooke> how can I find the build logs for: https://launchpad.net/ubuntu/+source/sound-theme-freedesktop
[18:23] <willcooke> I want to know if it's using dh_link.
[18:23] <willcooke> Do all packages just do that by default, or do I need to call it in the rules?
[18:23] <sarnold> click on the 0.8-1 then i386 then buildlog
[18:27] <willcooke> ahah! Hiding in plain sight, thanks sarnold
[18:28] <sarnold> you're welcome :)
[18:30] <immu> why do you end up in such a situation when its few weeks from release date and you find out that stuff its not working willcooke
[18:31] <willcooke> immu, care to give an example?  could be a millon reasons
[18:32] <fossfreedom> jbicha - budgie-wm isn't a binary so I can't add as an alternate
[18:33] <fossfreedom> jbicha can metacity downgrade gnome-session to a suggests rather than a recommendation ?
[18:35] <jbicha> fossfreedom: add whatever package provides budgie-wm as an alternate depends
[18:40] <fossfreedom> jbicha - ok - budgie-wm is provided by "budgie-core" - how is looking after ubiquity pull requests at the moment ?
[18:40] <fossfreedom> who
[18:46] <jbicha> probably cypherm_ox
[19:05] <immu> willcooke, wayland to include it or not, seeing the huge number of ubuntu users so that effect would be quite risky and big
[19:06] <willcooke> indeed, and that's why we're giving it a lot of careful consideration now
[19:10] <fossfreedom> thanks jbicha for the info
[19:14] <fossfreedom> cyphermox, when you have an opportunity - please can you review my merge request for ubiquity on behalf of Ubuntu Budgie?  TIA https://code.launchpad.net/~ubuntubudgie-dev/ubiquity/ubuntu-budgie-ubiquity/+merge/327323
[19:20] <jbicha> fossfreedom: as for metacity, I'm asking Debian whether it should recommend gnome-session-flashback or just drop that recommends completely
[19:41] <willcooke> jbicha, any clues as to what this missing icon might be called, or suggestions for how I can find out?  http://imgur.com/a/L6zpg
[19:43] <willcooke> I think it might be a in control center itself
[20:21] <willcooke> yeah, found it
[20:21]  * willcooke patches gnome-control-center
[20:30] <willcooke> I'm tryng to build gnome-control-center now
[20:30] <willcooke> but it's complaing about:
[20:30] <willcooke> No package 'libsystemd-login' found
[20:30] <willcooke> but it's installed, and the dev packages
[20:32] <seb128> willcooke, dpkg -l | grep libsystemd-login-dev
[20:32] <willcooke> nothing
[20:32] <willcooke> also, hi seb128!
[20:32] <seb128> so it's not installed it seems
[20:32] <seb128> hey :p
[20:34] <willcooke> oh, apt-file says my cache is empty - perhaps apt borked
[20:34] <willcooke> updating
[20:35] <seb128> that's on artful I guess?
[20:36] <willcooke> seb128, yeah
[20:36] <willcooke> gah, updating apt is running at 120kB a sec.  slowwwwww
[20:40] <willcooke> seb128, package name is libsystemd-login-dev  right?
[20:43] <seb128> it was in xenial, let me look at artful
[20:44] <seb128> can you pastebin the exact error?
[20:44] <seb128> that lib seems deprecated
[20:45] <seb128> it should try to look for it, weird
[20:46] <seb128> or do you build an old g-c-c?
[20:46] <jbicha> willcooke: how are you building it? with jhbuild? sbuild? regular ./configure & make ?
[20:47] <willcooke> jbicha, ./autogen.sh and then make I guess, not got that far yet.
[20:47] <willcooke> seb128, http://paste.ubuntu.com/25077567/
[20:48] <jbicha> sudo apt build-dep gnome-control-center
[20:48] <jbicha> try that then ^
[20:49] <willcooke> jbicha, yeah did that first.
[20:49] <willcooke> ahh
[20:50] <willcooke> maybe I branched the wrong thing
[20:50] <willcooke> I used this: https://code.launchpad.net/gnome-control-center
[20:53] <jbicha> I think LP has problems with git repos that use submodules (g-c-c uses submodules)
[20:59] <seb128> willcooke, https://code.launchpad.net/~vcs-imports/gnome-control-center/trunk
[20:59] <seb128> willcooke, most recent commit is from 2012-11-26
[20:59] <seb128> so yeah, not a good idea :p
[20:59] <seb128> willcooke, better take the source package from artful or git from upstream
[21:03] <willcooke> seb128, as in apt-get source?
[21:03] <seb128> willcooke, if you are on artful yes
[21:04] <seb128> or debcheckout gnome-control-center
[21:04] <seb128> the package is maintained in a bzr with the debian dir only
[21:04] <seb128> which is recombined with the tarball using bzr-builddeb
[21:04] <seb128> but that might be too much packaging for one day :p
[21:04] <seb128> +info
[21:04] <willcooke> and then how do I propose a branch?  Do I just make a diff and stick it on a bug?
[21:05] <seb128> basically
[21:05] <seb128> bzr get lp:~ubuntu-desktop/gnome-control-center/ubuntu
[21:05] <seb128> cd ubuntu
[21:05] <seb128> bzr bd-do
[21:05] <seb128> do your changes but that needs to be in the debian dir
[21:05] <seb128> e.g if you change the upstream source then you need to create a patch using quilt
[21:06] <jbicha> willcooke: are you asking how to submit it to Ubuntu or to GNOME?
[21:06] <seb128> or just cp the diff in ubuntu/debian/patches from the checkout and edit the serie by hand
[21:06] <willcooke> jbicha, it'll be a hack, so prolly just ubuntu
[21:06] <willcooke> seb128, oki, I'll test my patch works first, and then call it a day, and try again tomorrow :) thanks!
[21:07] <seb128> yw
[21:10] <willcooke> gah.  After all that, the patch I was considering is exactly what they've already done in g-c-c, so actually it's just a case of more sym links :DDD
[21:10] <willcooke> *done in the new g-c-c
[21:10] <willcooke> so this can be fixed with links
[21:10] <seb128> nice
[21:11] <seb128> jbicha, btw not sure if you saw my comment about e-d-s/evo earlier, but I'm not feeling confortable upstreaming that change, it's an hack to workaround our tools not being smart enough, they shouldn't have to care about that
[21:14] <willcooke> seb128, http://imgur.com/a/L6zpg
[21:14] <willcooke> and also, instead of playing white noise you get a proper sample
[21:15] <willcooke> https://bugs.launchpad.net/ubuntu/+source/sound-theme-freedesktop/+bug/1703946
[21:15] <jbicha> seb128: yes but upstream might not mind the change so maybe I'll propose it for you
[21:17] <seb128> jbicha, if you want feel free
[21:22] <willcooke> morning robert_ancell
[21:22] <willcooke> :)
[21:23] <robert_ancell> willcooke, hi!
[21:23] <robert_ancell> willcooke, are you in a US timezone?
[21:23] <willcooke> that's a sure sign that it's time to log off
[21:23] <willcooke> robert_ancell, be fixing bugs
[21:23] <willcooke> got a bit carried away
[21:23] <seb128> hey robert_ancell
[21:23]  * robert_ancell checks if all my bugs are now fixed by willcooke
[21:23] <willcooke> robert_ancell, they are not :)  Probably doing more harm than good tbh
[21:24] <willcooke> fix -> hack
[21:24] <robert_ancell> seb128, hey
[21:24] <robert_ancell> seb128, I was going to ask you if there's any more state of the art way to make desktop snaps. Trying to snap simple-scan for the nth time and get this interface editor into a snap
[21:25] <seb128> robert_ancell, check with kenvandine, he's pushing some example to the store this week
[21:25] <robert_ancell> seb128, ok, thanks
[21:26] <seb128> robert_ancell, http://bazaar.launchpad.net/~ken-vandine/quadrapassel/snap/view/head:/snapcraft.yaml is one he pushed today
[21:27] <robert_ancell> cool, that's just what I need
[21:27] <seb128> those are getting as complex (if not more) than debian/ packaging
[21:27] <seb128> next we get snaphelper to simply the yaml!
[21:27] <robert_ancell> haha
[21:28] <robert_ancell> I guess this relies on build-packages exactly matching the gnome-platform snap to work?
[21:28] <seb128> yes
[21:28] <robert_ancell> which distro do you build in then?
[21:28] <seb128> which means building on xenial+backport ppa
[21:28] <robert_ancell> aha
[21:29] <seb128> which launchpad let you set up
[21:29] <seb128> so if you auto build in launchpad it's easy
[21:29] <robert_ancell> "easy"
[21:29] <seb128> it's a combo to pick the ppa
[21:29] <seb128> not hard
[21:30] <seb128> but yeah, we need to do better there
[21:30] <seb128> seems the recommended way from the snapcraft team is to tar the prime of the platform snap after build
[21:30] <seb128> and publish that as file somewhere
[21:30] <seb128> and use that as a part to build snaps then
[21:33] <robert_ancell> So no futher thoughts on -dev snaps then
[21:34] <seb128> who feels like arguing about that with the snappy team? ;-)
[21:35] <seb128> that's probably a discussion we should have again at some point
[21:35] <seb128> unsure that's one to have now though
[21:35] <seb128> we might have other items we want to push forward before that
[21:36] <robert_ancell> yeah, it's lower priority for sure
[21:36] <robert_ancell> seb128, I'm having trouble getting SRUs pushed forward, any advise who to annoy for that?
[21:37] <seb128> robert_ancell, in your tz try RAOF?
[21:37] <seb128> or bdmurray
[21:38] <seb128> or maybe infinity would work
[21:38] <robert_ancell> seb128, who manages the unapproved queue?
[21:38] <seb128> robert_ancell, https://launchpad.net/~ubuntu-sru/+members#active
[21:39]  * robert_ancell thinks the correct terminology is probably to push SRUs _backwards_
[21:39] <robert_ancell> Because I think that's stopping them showing up on the SRU dashboard
[21:40] <seb128> btw that polkit locale issue is weird
[21:40] <seb128> why does it pick a random lang?
[21:40] <seb128> just curious how it ends up doing that
[21:41] <robert_ancell> seb128, the .policy file doesn't have a message field without a xml:lang set on it
[21:41] <robert_ancell> So it just seems to use the last / first one
[21:41] <seb128> oh ok
[21:41] <seb128> you should perhaps SRU that fix alone
[21:41] <robert_ancell> It is surely the same amount of work?
[21:41] <seb128> if the snapd-glib updates are too complex to find a SRU team member to review htem
[21:41] <seb128> well ^
[21:41] <robert_ancell> And there's issues that flexiondotorg found with snapd-glib that are fixed in newer versions
[21:42] <seb128> I'm not saying that there is no value at trying to SRU the update
[21:42] <seb128> just that it might be more difficult to find a reviewer and take time
[21:42] <seb128> but yeah, if you can nag somebody to let that in even better
[21:43] <jbicha> I'm annoyed that gnome-software/zesty has yet to clear phased-updates so LP: #1573408 still affects people :(
[21:44] <seb128> what is blocking it?
[21:45] <jbicha> I think right now, it's another unapproved SRU to fix LP: #1702122
[21:45] <robert_ancell> Is there a phased updates page somewhere?
[21:45] <jbicha> https://people.canonical.com/~ubuntu-archive/phased-updates.html
[21:45] <robert_ancell> Yeah, I fixed that bug, also stuck in unapproved queue
[21:45] <robert_ancell> jbicha, ta
[21:48] <sarnold> who gets alerted when the phaser is set to 0% ?
[21:49] <robert_ancell> sarnold, I get an email as the uploader
[21:49] <jbicha> the uploader gets a "Possible Regression" email
[21:50] <sarnold> well, that's something. some of those have been stuck at 0% for a loooong time ;/
[21:59] <willcooke> right, that really is it for tonight
[21:59] <willcooke> ta ta