[07:19] <ricotz> good morning desktopers
[07:20] <ricotz> https://launchpad.net/~ricotz/+archive/ubuntu/staging/+sourcepub/7567774/+listing-archive-extra
[07:20] <ricotz> https://bugs.launchpad.net/ubuntu/+source/network-manager/+bug/1650972
[07:20] <ubot5> Ubuntu bug 1650972 in network-manager (Ubuntu) "No .vapi in Zesty" [Undecided,Confirmed]
[09:02] <Laney> hey HEY
[09:02] <davmor2> Morning all
[09:03] <willcooke> o/
[09:06] <seb128> hey u.k band
[09:07] <b4n> andyrock, Trevinho: any news on https://code.launchpad.net/~banw/compiz/compiz.a11y-shotcuts/+merge/320091 ? :)
[09:07] <davmor2> seb128: oh we are a band I hope it's a rock band
[09:08] <seb128> it's up to you to decide!
[09:11] <Laney> we're west country folk
[09:11] <Laney> the wurzels
[09:14] <davmor2> Laney: I don't even mind that as long as we aren't a rap band or a boy band
[09:18] <Laney> uk garage it is
[09:21] <davmor2> Laney: /me always hears UK rap like this https://www.youtube.com/watch?v=JAIOzM7SsMo
[09:22] <davmor2> Laney: Nice cover of Zombie
[09:26]  * Laney dies
[09:26] <Laney> made about 10 seconds of that
[09:27] <desrt> hi hackers
[09:27] <davmor2> Laney: wuss that was a number 4 hit
[09:28] <davmor2> Laney: it is pretty much how UK rap sounds in comparison to US rap though :)
[09:29] <Laney> hi desrt
[09:29] <Laney> how's the london?
[09:39] <seb128> hey desrt
[09:40] <davmor2> Laney: also that's still not a bad song in comparison to some I know :D
[09:47] <Laney> :((((
[09:53]  * Sweet5hark prefers french rap to UK rap for the most part. French rap appears like something of its own, while UK rap cant escape to appear wannabe to me.
[09:54] <Sweet5hark> ^ Sentence best consumed with some spice girls rapping in the background.
[10:08] <ogra_> Sweet5hark, is that because one is european and the other is brexit ?
[10:41] <desrt> Laney: wet.
[10:41] <desrt> seb128: hi :)
[10:43] <Laney> desrt: you seem to have come during the second winter
[10:47] <desrt> ya..... happy to be outside of the continent, as not to shatter my illusions about what's going on there
[11:23] <willcooke> Right, that does it.  5ghz wifi access point is going in the bin.
[11:49] <allison[m]> so, matrix is the hot new thing in gnomeworld
[11:49] <allison[m]> time to cancel my irccloud account, apparently
[11:50] <Laney> ya, everyone's around with these [m] nicks
[11:50] <allison[m]> i wonder if it's possible to get rid of them....
[11:50] <Laney> and if you tab complete someone else on matrix it substitutes their full name on IRC
[11:50] <Laney> which looks curious :-)
[11:51] <allison[m]> seems that i'm the only matrix user in this channel...
[11:51] <allison[m]> the web client is substantially nicer than irccloud
[11:54] <allison[m]> (to say nothing of the phone app)
=== JanC_ is now known as JanC
[11:57] <Laney> ricotz: is that nm patch forwarded?
[11:57] <Laney> or is that code part of a distro delta?
[11:58] <ricotz> Laney, see https://bugzilla.gnome.org/show_bug.cgi?id=780032
[11:58] <ubot5> Gnome bug 780032 in gobject "Add missing attributes to two functions" [Normal,Resolved: fixed]
[11:58] <ricotz> Laney, and I didn't forward it
[12:00] <Laney> might be nice to
[12:00] <Laney> and patch headers would be good in future too
[12:02] <ricotz> Laney, hmm, I thought I mentioned the bug report at the top of the patch
[12:38] <seb128> Laney, allison[m], what's the point of that [m]?
=== hikiko is now known as hikiko|ln
[12:43] <Laney> I think matrix puts it on for you
[12:44] <seb128> for a technical reason?
[12:44] <seb128> or just to as free-advertizing for their service?
[12:47] <seb128> jbicha, was enabling the packagekit backend in http://launchpadlibrarian.net/294723961/gnome-software_3.22.2+git20161108.0.a58dfc7-0ubuntu3_3.22.2+git20161108.0.a58dfc7-0ubuntu4.diff.gz voluntary? it's not documented in the changelog
[12:47] <seb128> Laney, ^ you might know?
[12:48] <Laney> no I don't, it wasn't anything to do with me
[12:48] <Laney> is there a problem?
[12:49] <seb128> yeah, I know it doesn't have to do with you, but you follow a bit what's going on so I asked in case, no worry
[12:49] <seb128> that's what leads to sessioninstaller not working
[12:49] <seb128> g-s owns the packagekit session bus name when pkgkit is enabled
[12:50] <jbicha> seb128: the pk backend is not actually enabled in Ubuntu's gnome-software
[12:50] <seb128> jbicha, build log disagrees with you
[12:51] <seb128> https://launchpadlibrarian.net/294733318/buildlog_ubuntu-zesty-amd64.gnome-software_3.22.2+git20161108.0.a58dfc7-0ubuntu4_BUILDING.txt.gz
[12:51] <Laney> it certainly is enabled
[12:51] <seb128> https://launchpadlibrarian.net/294733318/buildlog_ubuntu-zesty-amd64.gnome-software_3.22.2+git20161108.0.a58dfc7-0ubuntu4_BUILDING.txt.gz
[12:51] <Laney> it just doesn't normally cause a problem because the apt backend conflicts it out
[12:51] <seb128>         PackageKit support:        yes
[12:51] <Laney> seems it does actually do something though
[12:51] <seb128> well since that version the deb ships /usr/share/dbus-1/services/org.freedesktop.PackageKit.service
[12:51] <seb128> so at least that's one issue
[12:52] <jbicha> seb128: ok, let me explicitly disable it, thanks!
[12:52] <seb128> thanks
[12:53] <seb128> so that was an overlook?
[12:53] <seb128> in any case we should make it work
[12:53] <seb128> but it's a bit late for this cycle to start on that
[12:53] <Laney> make packagekit work?
[12:53] <jbicha> we can either use the pk backend or the apt backend, we can't have both
[12:53] <seb128> removing that .service makes code install work
[12:53] <seb128> codec
[12:54] <seb128> I just did a rm locally
[12:54] <seb128> Laney, but we don't use packagekit so it's rather "implement that feature in the apt backend"?
[12:54] <Laney> asking what you mean by 'make it work'
[12:54] <seb128> or do we use the pkgkit backend for some bits like codecs?
[12:54] <seb128> I don't know yet
[12:55] <seb128> I just figured out that it was the issue
[12:55] <seb128> I'm poking at g-s code next
[12:55] <seb128> I keep you updated
[12:55] <Laney> jbicha is going to turn off packagekit, so then it will stop producing the .service file
[12:55] <seb128> that fixes it
[12:55] <seb128> but I don't know enough about current g-s to know if packagekit is useful/used in some other way atm
[12:55] <Laney> but
[12:55] <Laney> it would be good to make it work so we can drop sessioninstaller next cycle
[12:56] <Laney> so feel free to continue on that path
[12:56] <Laney> I think you need gstraemer1.0-packagekit at least
[12:58] <Laney> I should guess it's because /usr/share/dbus-1/services/sessioninstaller.service has the same Name= as /usr/share/dbus-1/services/org.freedesktop.PackageKit.service
[13:00] <Laney> well done for finding it :-)
[13:00] <jbicha> +1
[13:00] <seb128> yeah
[13:01] <jbicha> if we switch gnome-software to use pk instead of apt, g-software can handle codecs, you do not need gstreamer1.0-packagekit
[13:01] <seb128> unsure what is blocking us to use the packagekit backend
[13:01] <jbicha> you can try it out with https://launchpad.net/~ubuntu-desktop/+archive/ubuntu/gnome-software/
[13:01] <davmor2> jbicha: it can't now
[13:01] <jbicha> well, right now, it's Feature Freeze :)
[13:01] <seb128> not only
[13:01] <seb128> one of the blockers was performances
[13:01] <seb128> and I don't think that got resolved
[13:01] <Laney> somebody testing it and committing to doing any work required
[13:01] <Laney> it's fine and trivial to do the packaging side
[13:02] <seb128> whoever wants to switch should at least properly benchmark
[13:02] <Laney> but it's not acceptable to drop a shitty experience on us and then stand back
[13:02] <davmor2> currently if you open a video without a codec it fails to open software center at all so I think worrying about if it can install them is a mote point
[13:03] <Laney> yes we know
[13:03] <seb128> davmor2, can you read the backlog?
[13:04] <seb128> your comment suggest you didn't
[13:04] <davmor2> seb128: no racing between machines too much
[13:04] <seb128> we are being discussing solutions for that issue
[13:04] <davmor2> seb128: ah cool
[13:05] <seb128> davmor2, k, so please just step out of the discussion, you are commenting about things which were already raised/resolved
[13:05] <davmor2> seb128: \o/
=== hikiko|ln is now known as hikiko
[13:28] <jbicha> the g-s- fix is in the zesty unapproved queue
=== allison[m] is now known as allison
[13:33] <jbicha> g-s-with-pk actually did a clean start faster for me than gs-with-apt but then it takes a while to fill out the Installed page for instance
[13:35] <jbicha> please add concerns or comments to LP: #1643134
[13:35] <ubot5> Launchpad bug 1643134 in gnome-software (Ubuntu) "Switch gnome-software to use PackageKit backend" [Wishlist,Triaged] https://launchpad.net/bugs/1643134
[13:39] <Laney> jbicha: I don't think it's helpful to think of it as a packaging driven change. Someone needs to own the work, figure out the problems and then drive a process of fixing them.
[13:39] <Laney> The mechanical stuff is relatively trivial.
[13:40] <jbicha> well I've done the testing in the bug description
[13:41] <jbicha> gs-with-apt has bugs too that haven't been fixed either; switching to pk solves some of those (but at the expense of making it even less likely someone will get around to fixing those bugs in xenial)
[13:45] <didrocks> I know some fedora contributors/users not really happy with the pkgkit usage on g-s
[13:45] <didrocks> really bad experience, like having to refresh multiple times the package list, no feedback of upgrade happening behind the scene
[13:46] <didrocks> (sometimes, you click refresh and it looks like you are done, but reclicking then retrigger a request, blocking for a good 5 minutes, and you have your package list)
[13:46] <didrocks> you can have to hit the refresh 5 times
[13:46] <didrocks> the issues were identified in the packagekit using from gnome-software
[13:46] <didrocks> usage*
[13:46] <didrocks> I did test it myself yesterday in a vm, and can confirm
[13:47] <didrocks> the benefit for free is the offline update though
[13:47] <jbicha> I believe we can disable gnome-software's Updates by default in 3.24 which I assume we'll do since we have update-manager
[13:47] <didrocks> yes, but we would still have those refresh package list bugs
[13:47] <didrocks> that are unresolved as of fedora 26 at least
[13:48] <didrocks> so, proper and long time benchmarking/testing is really needed before taking any decision IMHO
[13:48] <didrocks> coming from fedora fanboys and contributors themselves, I wouldn't take their concerns lightely
[13:49] <jbicha> didrocks: did you test gs-with-packagekit on Ubuntu? https://launchpad.net/~ubuntu-desktop/+archive/ubuntu/gnome-software/
[13:49] <Laney> Essentially we should take responsibility for it after the switch
[13:49] <didrocks> jbicha: I didn't, I don't work on desktop anymore, I'm just giving here some hints of the words on the street from people close to those upstreams
[13:51] <seb128> jbicha, start I don't know about, it's just computation of upgrades/install and actions that we slower, part of the issues were due to aptcc from what the debian maintainers said iirc
[13:51] <Laney> It's all well and good to test it, but then the next steps need to happen too
[13:51] <seb128> but what Laney says
[13:51] <seb128> whoever does the switch should own the transition/to get things in shape
[13:51] <Laney> Fair enough you might think the apt backend is crap, but our team *has* contributed there, and we have fixed bugs even if there are others remaining
[13:52] <Laney> The same will have to happen with the packagekit backend
[13:52] <Laney> This piece of software requires more integratin than most
[13:59] <jbicha> let's take a look again after zesty
[14:00] <jbicha> installing apps doesn't feel slow to me; and if we don't even use the Updates part then slowness there won't matter except for people who intentionally enable that
=== dandrader is now known as dandrader|afk
[14:01] <seb128> jbicha, by "update" I mean installing pending updates, which we can do from our UI
[14:01] <jbicha> seb128: I don't think we want both GNOME Software and update-manager handling pending updates
[14:03] <seb128> no we don't
[14:03] <seb128> but the update tab is there in g-s
[14:03] <seb128> so it should be usable for users who manually decide to go there and use it
[14:03] <seb128> or we should remove it
[14:04] <jbicha> https://git.gnome.org/browse/gnome-software/commit/?id=81dbe62
[14:07] <seb128> ah, ok
[14:07] <seb128> I'm undecided on that
[14:09] <seb128> I still think a package management UI should allow you to update installed packages, synaptic of software-center allows that as well
[14:13] <jbicha> yes but it's not well-integrated yet (for instance because of phased-updates, gnome-software shows you updates that you don't see in update-manager)
[14:15] <seb128> well I guess we could disable the tab
[14:15] <seb128> as long as you can do "update" from a detailed view of a specific software
[14:17] <Laney> what about firmware updates, snap updates (do they always happen automatically?), flatpak updates?
[14:19] <seb128> good point
[14:21] <jbicha> oh, you're right, it's complicated :(
[14:21] <Laney> and then is it weird that there's one type of update you don't see in there (deb packages)?
[14:21] <Laney> tough stuff!
[14:23] <seb128> I don't think it's a big deal to have g-s listing updates and still having update-manager defaulting to be the tools to apply those
[14:23] <seb128> users who just apply those don't see g-s
[14:23] <seb128> those who go to g-s expect a package management tool
[14:25] <jbicha> gs-with-pk only does offline upgrades now (at least for system updates) which is "interesting"
[14:28] <Laney> it's all fixable
=== dandrader|afk is now known as dandrader
=== carlolo is now known as clobrano
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
[18:01] <willcooke> quttin' time.  I'm off tomorrow (going to the NEC) so back Friday.  Toodles
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
[23:20] <jbicha> FJKong: good morning, for LP: #1644781 you'll want to look at the extra dependencies added in 3.21.91-2 but it looks like Debian missed the evince one
[23:20] <ubot5> Launchpad bug 1644781 in gnome-sushi (Ubuntu) "Sushi-start fails with 16.10 upgrade, both Unity and Mate" [High,Incomplete] https://launchpad.net/bugs/1644781