2014/10/30/#ubuntu-desktop.txt

[01:56] <thumper> what ever happened to webapp integration with the sound menu?
[01:56] <thumper> like spotify?
=== m_conley is now known as m_conley_away
=== duflu_ is now known as duflu
=== oCrazyLemn is now known as CrazyLemon
[07:51] <didrocks> morning
[08:16] <pitti> Noskcaj: if the version you branched off in the PPA is the same as in vivid, and you used distro-compatible version numbers, we can sync the sources; otherwise needs merges/uploads
[08:19] <pitti> Noskcaj: upower synced; it'll stay in -proposed until the transition is done
[08:22] <seb128> hey pitti
[08:22] <seb128> good morning desktopers
[08:24] <pitti> hey seb128
[08:31] <seb128> pitti, did we decide to start the upower transition?
[08:31] <pitti> seb128: well, it was started long ago, and Noskcaj has everything ready in his PPA? I don't see why not?
[08:31] <pitti> and so does Debian
[08:31] <seb128> pitti, not sure, we decided to not go for it before utopic because the xubuntu team was unhappy about it
[08:32] <seb128> but I didn't follow the details/what problem it created for them
[08:32] <seb128> I just know they reached to the release team to have the ffe refused
[08:32] <pitti> seb128: the only long-time blocker was indicator-power, but Charles ported it recently
[08:32] <seb128> oh well, it's a start of cycle, we can figure out the remaining bits
[08:32] <pitti> yes, because it was too close to release indeed
[08:32] <pitti> so let's not slip it again :)
[08:33] <seb128> I'm just a bit worried about piling too many transitions at the same time in proposed
[08:33] <seb128> because they interact and we end up with months of untangling before being able to get anything through then
[08:34] <seb128> reading backlog
[08:34] <seb128> Noskcaj, gtk "transition"? I do hope the new GTK is not ABI incompatible/doesn't require a transition
[08:34] <pitti> seb128: well, as it stands the old upower makes things more complicated wrt. merging from Debian or updating to new upstream versions, so I think it's one of the first that we should do
[08:34] <pitti> it's all ready, after all
[08:35] <seb128> k
[08:35] <seb128> well, some of the transitions are already in proposed by autosyncs
[08:35] <seb128> so we don't get to decide to start or not those
[08:35] <seb128> anyway, let's see how it goes
[08:36] <seb128> reading backlog still
[08:37] <seb128> didrocks, @qtcreator on unity8/mir, bregma had a look previous cycle and know what is not working iirc
[08:37] <seb128> so maybe check with him first
[08:37] <seb128> it might spare some work
[08:37] <didrocks> seb128: oh sure, will do :)
[08:37] <didrocks> thanks
[08:37] <seb128> yw
[08:38] <seb128> didrocks, http://irclogs.ubuntu.com/2014/04/17/%23ubuntu-desktop.html#t14:56
[08:39] <didrocks> thx
[08:39] <seb128> "bregma	hey tedg I'm trying to run Qt Creator in my Unity 8 Desktop Preview session but it apparent fails because its subprocesses are not allowed to connect or something, what's the best approach to analysing that problem"
[08:39]  * didrocks opens
[08:40] <willcooke> morning
[08:40] <seb128> hey willcooke
[08:40] <seb128> willcooke, ^ btw, about qtcreator
[08:40] <willcooke> ah, cool
[08:41] <seb128> so it's rather a process management issue, if the problem is still the same one
[08:41] <seb128> not a rendering one
[08:41] <willcooke> hmmmm
[08:41] <willcooke> that sounds complicated
[08:42] <didrocks> yeah, seems that we depend on the API team :/
[08:42] <didrocks> and they want to know if they implement that using upstart or systemd
[08:42] <pitti> Noskcaj: you don't have distro compatible version numbers there (quite rightfully for a PPA), so that'll need uploads; also, several of the gnome-* are in bzr
[08:42] <pitti> Noskcaj: also some merges with newer versions
[08:46] <seb128> I've an appointment at 10, need to go, back a bit later (going to have some shifted hours in the evening to compensate)
[08:47] <willcooke> seb128, catch you later
[09:03] <Laney> morning
[09:03] <Laney> darkxst: The problem was that new clutter-1.0 had Breaks on the current mutter
[09:13] <Noskcaj> seb128, Just the random visual stuff that always breaks
[09:14] <Noskcaj> pitti, That's the utopic stuff, i wasn't sure if i should re-do it in PPA.
[09:14] <pitti> Noskcaj: nah, let's just upload it to vidid now
[09:14] <pitti> vivid, too
[09:14] <Noskcaj> ok
[09:14] <Noskcaj> I'll make a heap of MPs in the next few days then
[09:17] <Noskcaj> need sleep for now though
[09:20] <xnox> seb128: Laney: i think we need to discuss ssh/gpg agents for the default desktop. I'm getting complaints, after re-enabling gnome-keyring by default via SRU.
[09:44] <Laney> hi xnox, got some time off? ;-)
[09:44] <Laney> what are the complaints?
[09:45] <Laney> I have swapped this problem out of my brain state I'm afraid
[09:46] <xnox> Laney: gnome-keyring's ssh/gpg agents are subpar compared to openssh/gnupg agents.
[09:46] <xnox> Laney: it can't handle eliptic curve keys, keys that are protected by a better encryption algorithms, smartcards, ssh certificates, etc.
[09:46] <Laney> what happens without upstart?
[09:47] <xnox> Laney: but, it's still the default via xdg autostart, and it's now default for ssh/gpg with upstart.
[09:47] <xnox> in xdg autostart it's .desktop file per component (secrects, pkcs11, ssh, gpg), in upstart it's a single job.
[09:47] <Laney> I guess before people could just untick it in the startup applications thing?
[09:47] <Laney> or is it being non split the problem?
[09:48] <xnox> yes. and the untick doesn't migrate to upstart job overrides.
[09:48] <Laney> is it lack of UI?
[09:48] <Laney> don't understand really
[09:48] <xnox> and the upstart job is unsplit so one can't start pkcs11/secrets by gnome-keyring, without it also taking over ssh/gpg agents.
[09:48] <xnox> and yeah, there is no UI for it.
[09:49] <xnox> (and the xdg autostart override is not honored by the upstart job)
[09:49] <Laney> upstart job parsing .desktop file? :)
[09:49] <xnox> can the UI ticks/unticks be queried somehow? cause i think upstart job shouldn't run if it was disabled.
[09:50] <xnox> and we should split gnome-keyring uptart job into 4 components - keyring, pkcs11, gpg, ssh - just like xdg autostart desktop files are split
[09:50] <Laney> it puts a desktop file in .config/ somewhere with Hidden=true, so you could read that
[09:51] <xnox> what about giving up on keyring's ssh & gpg agents all together?
[09:51] <Laney> don't think so
[09:51] <xnox> and only use gnome-keyring for secrets & pkcs11.
[09:51] <xnox> cause it can unlock/save the ssh/gpg passwords, right?
[09:52] <Laney> I guess I'm a simple ssh key user but it works well enough for me
[09:52] <xnox> Laney: any idea what gnome-keyring's pkcs11 component does?
[09:54] <Laney> smart card support isn't it?
[09:58] <seb128> back
[09:58] <seb128> Laney, hey, wie gehts?
[10:00] <darkxst> Laney, ok, I didnt see that
[10:00] <Laney> hey seb128, doing good thanks, went to a ska show last night :-o
[10:00] <Laney> then got SOAKED on the way back
[10:01] <seb128> "ska"?
[10:01] <Laney> no waterproofs
[10:01]  * seb128 googles and learn a new thing
[10:01] <darkxst> Laney, but landing new clutter with current mutter is trivial
[10:01] <seb128> Laney, better to get soaked on the way back than on the way there...
[10:01] <Laney> darkxst: maybe, but it's better for you to do it than me
[10:01] <Laney> I didn't know if you can have new mutter and old gnome-shell
[10:02] <darkxst> Laney, no you can't do that
[10:02] <Laney> so then you have to upgrade gnome-shell which requires gtk
[10:02] <darkxst> but new clutter with old mutter will work
[10:02] <Laney> if you drop the breaks?
[10:02] <Laney> why are they there then?
[10:03] <Laney> seb128: yep, but if it was raining when I left I would have taken the waterproof stuff :p
[10:03] <Laney> some guy standing outside a pub laughed quite a lot at me
[10:03] <darkxst> yes, for api/abi break, which there was a minor one
[10:03] <darkxst> I suppose
[10:04] <Laney> anyway, this can wait until next week to do it properly
[10:05] <Laney> did we get a rebased mir patch for gtk 3.14 yet?
[10:05] <darkxst> Laney, yes next week is fine
[10:06] <darkxst> Laney,  I think we will need upower and maybe gnome-desktop before we can land gnome-shell 3.14 ;(
[10:07] <Laney> upower is happening or going to happen
[10:07] <pitti> darkxst: new upower is in -proposed, we need to upload Noskcaj's PPA changes now
[10:07] <Laney> is g-desktop hard?
[10:07] <pitti> Laney: ^
[10:07] <Laney> cool
[10:07] <pitti> we should do this quickly to reduce delta from debian and unblock other transitions
[10:08] <seb128> Laney, I think larsu's vcs had that "rbeased", read it applies/build, unsure it works
[10:08] <Laney> didn't when I looked
[10:08] <Laney> maybe it does now?
[10:08] <seb128> didn't work?
[10:09] <seb128> where did you try? unity8 desktop?
[10:09] <Laney> was disabled
[10:09] <seb128> hum
[10:09] <darkxst> Laney, I think gnome-desktop should be pretty straight forward this time
[10:09] <Laney> yeah it is
[10:09] <seb128> k, dunno then
[10:09] <Laney> I thought desrt was doing it, might be wrong
[10:09] <seb128> Lars said it applies/builds
[10:10] <seb128> iirc
[10:10] <seb128> but he's not here this week
[10:10] <seb128> that can wait monday imho
[10:10] <darkxst> I'm also working on gnome-session merge, which is a little tangled trying to revert old dialogs]
[10:10] <Laney> darkxst: mitya57 / Alberts work upstream doesn't help?
[10:11] <darkxst> Laney, how far have they got? it only helps if they stop using the gnome-session dialogs
[10:11] <Laney> I thought that was the point
[10:11] <seb128> "they"?
[10:12] <seb128> is that gnome-panel?
[10:12] <Laney> yep
[10:12] <seb128> do we care about gnome-session sessions not using gnome-shell/unity/gnome-panel?
[10:12] <seb128> is there any of those in use?
[10:12] <Laney> does that exist?
[10:12] <seb128> like running gnome with cairodock or something
[10:12] <seb128> instead of panel
[10:12] <seb128> dunno, that's what I'm asking
[10:13] <darkxst> seb128, gnome-panel has its own dialogs, but they are not used for keyboard shortcuts, last time I checked
[10:14] <Laney> I would check with thowe two
[10:14] <seb128> bah, packages.ubuntu.com is buggy/incomplete
[10:14] <Laney> before doing any complex reverts or whatever
[10:16] <seb128> wasn't p.u.c able to look for packages containing a directory, like /usr/share/gnome-session/sessions/ ?
[10:18] <darkxst> Laney, will do, I kind of stopped once I saw how tangled it was anyway
[10:29] <seb128> happyaron, hey, did you see my fcitx error yesterday?
[10:32] <seb128> happyaron, bug #1387382
[10:32] <ubot5> bug 1387382 in fcitx (Ubuntu) "package libfcitx-config4 (not installed) failed to install/upgrade: trying to overwrite '/usr/lib/x86_64-linux-gnu/libfcitx-config.so.4.1', which is also in package fcitx-libs:amd64 1:4.2.8.5-1" [High,Triaged] https://launchpad.net/bugs/1387382
[10:55] <pitti> seb128, Laney, Noskcaj: starting to upload some upower transition bits now, if that's ok?
[10:56] <pitti> (with merging/pushing to bzr, etc.)
[10:56] <pitti> e. g. gnome-applets is now a fakesync
[10:57] <Laney> fine by me
[10:57] <Laney> thanks for helping!
[10:57] <pitti> I'm doing -proposed cleanup today anyway, so I might just as well :)
[10:57] <pitti> and this has been lingering far too long
[10:58] <Laney> we planned to do it at the start of this cycle
[11:25] <Laney> hey didrocks, CI train question ;-)
[11:25] <Laney> if I have a branch with a tagged (uploaded) revision and another commit on top
[11:25] <Laney> can CI train figure this out?
[11:26] <Laney> i.e. will it notice the extra commit and let me upload it?
[11:40] <pitti> darkxst, Noskcaj: gnome-session would need heavy backport patching for upower (in https://launchpad.net/~noskcaj/+archive/ubuntu/upower/+packages); do you actually want to do this, or rather just update to 3.14?
[11:43] <pitti> seb128, Laney: unity-control-center/settings-daemon are in a "magic" bzr; is this supposed to go through CI train, or normal commit/push?
[11:43] <pitti> .../dput?
[11:44] <Laney> pitti: ci train, but dput if you want, I don't mind
[11:44] <pitti> and one makes up a new sensible version number?
[11:44] <Laney> just commit/tag/push to the branch
[11:44] <Laney> just ubuntu2 is fine
[11:44] <pitti> 14.10.0+14.10.20140922 would be quite wrong, though
[11:44] <pitti> (obvious how to bump, of course)
[11:45] <Laney> I don't know how the upstream upstream part is generated
[11:45] <Laney> but they're 1.0 so you can just edit the source
[11:53] <didrocks> Laney: CI train will always take latest branch content
[11:54] <didrocks> so yeah, it will notice this extra commit and let you upload it
[11:54] <pitti> so I build the orig.tar.gz from everything except debian/, I suppose?
[11:54] <Laney> pitti: just bzr bd -S should do it
[11:55] <didrocks> yeah, it's in splitmode
[11:55] <Laney> or you can drive the train if you want, all core-devs should have permission
[11:55]  * willcooke . . o ( tickets please )
[11:56] <didrocks> willcooke: on my original airplane vision, there are some tickets concepts :)
[11:56]  * pitti hides in the toilet
[11:56] <Laney> is there status and upgrades?
[11:56] <Laney> and lost baggage?
[11:56] <Laney> :P
[11:56] <didrocks> pitti: no smoking please!
[11:57] <willcooke> is there a white zone which is for the loading and unloading of passengers only?
[11:57] <pitti> didrocks: je ne fume jamais !
[11:57] <didrocks> Laney: well, now, I would implement business class at least ;)
[11:57] <willcooke> :D
[11:58] <Laney> that's called being a core-dev
[11:58]  * Laney breezes right through all of the checks
[11:58] <didrocks> heh
[12:02] <pitti> Laney: oh, I see https://code.launchpad.net/~laney/ubuntu-system-settings/upower0.99/+merge/230988 ; that never landed, is that broken, or needs an ack, or it was just forgotten?
[12:03] <Laney> the last two
[12:05] <darkxst> pitti, not entirely sure what what you are talking about gnome-session, but I will have g-s 3.14 ready soon enough
[12:06] <pitti> darkxst: for the upower transition; applying the patches to 3.8 is a bit hackish IMHO
[12:06] <pitti> darkxst: it's in the PPA, but I figure you'd rather want to upgrade to 3.14 anyway?
[12:07] <darkxst> pitti, gnome-session is at 3.9.90ish in that archives no?
[12:07] <pitti> darkxst: right, that's what I meant (not 3.8)
[12:08] <pitti> Laney: oh sorry, unity-system-settings != unity-control-center :)
[12:08] <Laney> ITYM ubuntu-system-settings
[12:09] <darkxst> pitti, and yes we should just go with 3.14 straight up, if possible
[12:09] <pitti> darkxst: OK, so I'll leave those two bits of bug 1330037 to you and Noskcaj?
[12:09] <ubot5> bug 1330037 in xfce4-settings (Ubuntu) "[FFe] upower 0.99.1 transition" [Undecided,In progress] https://launchpad.net/bugs/1330037
[12:12] <darkxst> pitti, I wont have time to look until the weekend, but ok
[12:12] <pitti> darkxst: that sounds fine
[12:14] <darkxst> pitti, ok
[12:15]  * darkxst must sleep now
[12:16] <Laney> night!
[12:23] <didrocks> willcooke: Laney: seb128: so, I have the ubuntu desktop next image booted here, I have the password prompt asking me for a password, empty or "ubuntu" doesn't work
[12:23] <didrocks> seb128: is it the thing we saw together and so, the blank password doesn't really work under Qt?
[12:25] <didrocks> hum, updated to "ubuntu" in a tty and still doesn't work
[12:26] <didrocks> weren't we supposed to autolog in that live image?
[12:26] <seb128> didrocks, just go to a vt and change the passwd
[12:26] <didrocks> 13:25:11  didrocks | hum, updated to "ubuntu" in a tty and still doesn't work
[12:26] <didrocks> seb128: ^
[12:26] <seb128> didrocks, or edit https://code.launchpad.net/~seb128/unity8/drop-workaround-empty-pwd-login/+merge/239401 by hand
[12:27] <seb128> didrocks, it's one line to delete
[12:27] <seb128> didrocks, we do autologin, you get the unity8 lock screen
[12:27] <seb128> and we don't use "no password"
[12:27] <seb128> we use the empty password set as a passwd
[12:27] <seb128> remember, that's what you helped me with in washington ;-)
[12:28] <didrocks> hence my 13:23:49  didrocks | seb128: is it the thing we saw together and so, the blank password doesn't really work under Qt?
[12:28] <didrocks> so yeah, I do remember about it :)
[12:28] <didrocks> it's still weird to have that prompt, even with empty password
[12:28] <seb128> didrocks, try deleting the line from that mr ^
[12:28] <didrocks> and that it doesn't work if you passwd
[12:28] <didrocks> yeah, doing that (restarting, it hanged)
[12:29] <seb128> didrocks, well, prompt is normal, because from a pam viewpoint you have a password
[12:29] <seb128> the text just happens to be the empty string
[12:29] <seb128> still unsure why the livecd guys did that
[12:29] <seb128> rather than using passwd -d
[12:30] <didrocks> seb128: unity8 isn't an upstart job?
[12:30] <seb128> it is
[12:30]  * didrocks can't sudo restart unity8
[12:30] <seb128> no sudo
[12:30] <seb128> it's an user job
[12:30] <didrocks> unknown job as well (from a tty)
[12:31] <seb128> same user?
[12:31] <didrocks> ubuntu-desktop-next@
[12:31] <Laney> you won't be in the upstart session there
[12:31] <seb128> you need to export the right env
[12:31] <seb128> easier to sig9 unity8
[12:33] <didrocks> seb128: doesn't really restart after a sigkill
[12:33] <seb128> weird
[12:33] <didrocks> ok, rebooting then
=== MacSlow is now known as MacSlow|lunch
[12:34] <didrocks> hoping that the key is persistent :)
[12:38] <didrocks> humok, not persistent and unity8 definitively doesn't restart after a sigkill. Let me try an installation on that machine…
[12:38] <seb128> didrocks, I usually sudo restart lightdm
[12:38] <seb128> that for sure work
[12:38] <didrocks> ah, the whole login…
[12:39] <seb128> yes
[12:39] <didrocks> ok, doing
[12:39] <seb128> well, it works on utopuc
[12:39] <seb128> utopic even
[12:39] <seb128> didn't try vivid yet
[12:39] <didrocks> I will tell you if this works with vivid :)
[12:40] <didrocks> however, booting is quite unreliable…
[12:41] <seb128> :-/
[12:41] <seb128> usb stick issue you think?
[12:42] <didrocks> really unsure, worked well with other live recently, never got a glinch
[12:43] <didrocks> ok, restarting lightdm works and I can log in
[12:46] <seb128> great
[12:46] <seb128> now we just need to nag Saviq to get that one liner unity8 change in vivid
[12:46] <seb128> that shouldn't be blocker by rtm
[12:47] <Saviq> seb128, yeah, will do
[12:47] <seb128> Saviq, thanks
[12:48] <Saviq> seb128, I've one vivid landing that I just ACKed (sync with rtm really)
[12:48] <seb128> Saviq, it doesn't include that change though?
[12:48] <Saviq> seb128, not yet
[12:48] <seb128> :-(
[12:48] <Saviq> seb128, I need to get a clean-ish slate first
[12:48] <seb128> it's a one liner and safe, shame it takes so much effort to get it landed :/
[12:49] <seb128> it doesn't even impact codepath used on the phone
[12:49] <Saviq> seb128, well, it's not that one that takes so much effort, it's everything else that got out of sync between vivid and rtm
[12:49] <Saviq> seb128, yeah, which is why it won't land in rtm at all
[12:49] <seb128> well, you could have batched with whatever else was in there
[12:49] <seb128> it's not going to conflict with rtm work
[12:49] <Saviq> seb128, except I already have like 30 branches in there
[12:49] <seb128> so it should apply fine on whatever you upload to vivid
[12:50] <seb128> k, anyway just being grumpy about desktop-next login being buggy since july
[12:50] <seb128> sorry for the ranting
[12:50] <Saviq> seb128, it'll happen, really, it will, I just couldn't land that silo yesterday 'cause we had a different issue
[12:50] <seb128> I just wish desktop would get some attention
[12:50] <Saviq> seb128, well now, bug was only filed in september
[12:50] <seb128> Saviq, k, thanks
[12:51] <Saviq> seb128, MP is ap since when, a week?
[12:51] <seb128> Saviq, yeah, I discussed it a couple of times with mterry and others
[12:51] <seb128> Saviq, right, I did look at fixing because nobody was looking at it since <ages>
[12:51] <Saviq> seb128, it's not like we're trying to ignore that actively
[12:51] <seb128> but I guess I could have looked earlier, I'm as much to blame as others
[12:51] <Saviq> seb128, we're basically excluded from what happens in desktop-next
[12:52] <seb128> Saviq, well, it's more than we let land things that regress the desktop and then don't bother trying to sort out the regressions
[12:52] <seb128> Saviq, we need the test plans to account for desktop-next
[12:52] <seb128> and stop the line when we bug this one
[12:53] <didrocks> argh, I have no way to disable the touchscreen on Mir
[12:53] <didrocks> and this hardware gives random inputs, so I needed on X to disable it
[12:53] <seb128> didrocks, we don't have any on unity7/xorg either I think
[12:54] <didrocks> seb128: xinput
[12:54] <seb128> oh, right, I though you mean GUI
[12:54] <didrocks> and then toggle the correct property
[12:54] <didrocks> no
[12:54] <didrocks> just want to disable it whatever
[12:54] <seb128> k
[12:54] <didrocks> here, my mouse is moving randomely
[12:54] <didrocks> clicking
[12:54] <didrocks> not really nice to experiment…
[12:54] <seb128> on what device did you install it?
[12:54] <didrocks> XT2
[12:55] <seb128> oh, so you have a test laptop :-) does that have bt?
[12:55] <didrocks> it doesn't
[12:55] <seb128> I see
[12:55] <didrocks> it claims it has one, but doesnt' work
[12:55] <didrocks> tried with fedora/ubuntu
[12:55] <didrocks> but yeah, the unity8 image isn't usable at all if I can't turn off the wrongly touch screen :(
[12:55] <seb128> if you are going to work on bluetooth I guess it would be worst expensing a stick
[12:56] <didrocks> seb128: well, seems priority changes, so not sure yet, but yeah, it would make sense
[12:56] <seb128> :-(
[12:56] <seb128> did you ask on the mir channel about turning off touch?
[12:56] <didrocks> not yet, will try
[12:56] <didrocks> have few hope though
[12:57] <seb128> I first had to create some text file for the touch screen to work on the latitude
[12:57] <seb128> not sure if they moved away from those description
[12:57] <didrocks> seb128: how would you try starting any random .desktop file?
[12:57]  * didrocks goes on #mir
[12:57] <seb128> if they didn't you might be able to delete a file to make it stop work
[12:58] <seb128> didrocks, X-Ubuntu-Touch=true in the .desktop and restart the session
[12:58] <seb128> (or maybe just the dash process is enough, or a lens)
[12:58] <seb128> that makes the .desktop listed in the dash
[12:58] <seb128> I can run gedit like that
[12:58] <didrocks> ok, let's try
[12:59] <didrocks> seb128: perfect, thanks! :)
[13:00] <didrocks> so, confirming Qtcreator is quite, unusable
[13:01] <didrocks> can't even switch back to it clicking to the .desktop file
[13:02] <seb128> didrocks, can you right swipe back to it?
[13:02] <desrt> good morning
[13:02] <didrocks> yeah, this is working
[13:02] <didrocks> morning desrt
[13:03] <desrt> Laney: what am i supposed to have been doing?
[13:03] <seb128> hey desrt, how are you?
[13:03] <desrt> sleepy :)
[13:03] <pitti> hey desrt
[13:04] <desrt> pitti: good morning
[13:04] <seb128> desrt, rebasing the gtk-mir patch on current stable gtk I think
[13:04] <desrt> oh.  i did that and pastebinned it to larsu
[13:04] <desrt> it was pretty simple -- only had one conflict in configure.ac
[13:04] <seb128> who is off this week
[13:04] <desrt> i can do it again
[13:04] <seb128> so maybe he did it but didn't push his work
[13:04] <seb128> don't bother
[13:05] <seb128> that can wait monday for Lars to be back
[13:05] <desrt> sure
[13:06]  * desrt acquires coffee
[13:06] <Laney> I was just saying that I thought it did happen
[13:07]  * didrocks waits for an answer on #ubuntu-mir…
[13:07] <seb128> didrocks, east is sleeping, u.k is eating, u.s is sleeping, not the best time of the day ;-)
[13:08] <didrocks> seb128: let's see if they will reply when being back
[13:08] <seb128> yeah
[13:08] <didrocks> I'm afraid from past experience though :)
[13:09] <didrocks> ok, great, I have a name to bother now :)
[13:09] <seb128> didrocks, see :-)
[13:09] <seb128> I was going to suggest racarr as well
[13:09] <seb128> nice to see you got a reply though
[13:09] <didrocks> yeah, things are changing with years :)
=== alan_g is now known as alan_g|lunch
[13:13] <desrt> seb128: us may be sleeping, but .ca is drinking coffee
[13:25] <didrocks> interesting, the tty was in fr, and unity8 used US layout
[13:26] <didrocks> I guess we'll have quite a lot of such bugs…
[13:28] <seb128> "input" under Mir is not something that got lot of work yet, that's on the list for this cycle though
[13:30]  * didrocks waits on tedg to get up now! :)
[13:32] <ChrisTownsend> seb128: Hey, I was reading the backlog here and noticed some discussion about https://code.launchpad.net/~seb128/unity8/drop-workaround-empty-pwd-login/+merge/239401 which is linked to bug #1360307.  I think that bug is slightly different than what your fix is addressing.
[13:32] <ubot5> bug 1360307 in unity8 (Ubuntu) "Logging in to the desktop session brings up the lock screen" [High,In progress] https://launchpad.net/bugs/1360307
[13:33] <seb128> ChrisTownsend, hey, shrug
[13:33] <ChrisTownsend> seb128: Correct me if I'm wrong, but your fix doesn't get rid of the unity-greeter prompt at that point, right?  It only accepts an empty password, right?
[13:33] <seb128> ChrisTownsend, did you open another one for "can't log in from unity8 log screen"?
[13:33] <seb128> ChrisTownsend, no, it doesn't but I though your bug was about the "can't get to the session"
[13:33] <ChrisTownsend> seb128: Hmm, I don't remember.  Let me go look around.
[13:33] <seb128> feel free to change for another one
[13:34] <ChrisTownsend> seb128: Ok, if I didn't then I'll create one.  But thanks a bunch for fixing that!
[13:34] <seb128> ChrisTownsend, thanks
[13:34] <seb128> yw!
[13:35] <desrt> so i did some research last night
[13:36] <desrt> looks like halifax has a few good hotels with lots of conference space
[13:37] <desrt> it also has (relatively short) direct flights from london and frankfurt
[13:40] <willcooke> hummm. Do we need a GStreamer sink for Mir?
[13:40] <desrt> "doesn't _really_ work like that"
[13:41] <willcooke> k
[13:41] <willcooke> :)
[13:41] <ogra_> desrt, well ...
[13:41] <desrt> you need a gstreamer sink for drmable video memory
[13:41] <ogra_> there are x11 sinks ... so you can take screenshots
[13:42] <desrt> ogra_: *sources
[13:42] <desrt> and this wouldn't work in mir from what i understand (for security reasons)
[13:42] <ogra_> we kind of lack the same thing for mir
[13:42] <ogra_> oh, right .. wrong term
[13:43]  * ogra_ was looking into doing screen captures locally on the phone before ... sadly you need to convert the raw data somehow 
[13:43] <desrt> willcooke: in short, you're probably never going to want gstreamer talking directly to the display server, but rather displaying something that's wrapped at least a little bit by some toolkit-provided chrome (controls and such)
[13:43] <desrt> unless you want a really trivial demo output for debug/demo purposes (and indeed this exists for x11)
[13:44] <desrt> and the way that gstreamer talks to the toolkit for showing video in wayland is going to be more or less exactly the same as for mir
[13:45] <desrt> and it has more to do with the acceleration capabilities of the graphics card than it has to do with the display server
[13:48] <tedg> didrocks, ping
[13:50] <didrocks> tedg: pong
[13:50] <tedg> didrocks, You were waiting for me to wake up, but I can't figure out context.
[13:50] <didrocks> tedg: ah no worry :) some questions on what next steps we need to take to make qtcreator running on the desktop next image a reality
[13:51] <didrocks> tedg: I know you talked a little bit about it with bregma a while back
[13:51] <didrocks> I see that there are few X11 calls (mainly to raise the main window)
[13:51] <tedg> didrocks, I think that the part I was blocking on was landing the cgroups support in UAL, which we've done.
[13:51] <didrocks> but apart from multi-window managements, it seems you thought about other issues to the multi-process needs?
[13:52] <tedg> didrocks, That way Mir can detect all the process connections for an app.
[13:52] <didrocks> tedg: ok, this part is now over. and all processes are tagged for Mir to know it's part of this app
[13:52] <tedg> didrocks, I don't think that bregma is focused on Qt Desktop apps right now, they're trying to get something generic for legacy X apps.
[13:52] <bregma> the problem still happens but evidently cgroups was not the answer
[13:53] <didrocks> bregma: what was the problem you are talking about?
[13:53] <bregma> I haven't looked in to it lately, other than to check that it's still a problem
[13:53] <didrocks> the fact that one surface is displayed but doesn't receive any inputs?
[13:53] <bregma> QtCreator hangs after presenting the startup screen, witing for two subprocesses to complete
[13:54] <didrocks> how did you identify it was waiting on some subprocesses to complete?
[13:54] <didrocks> I see the same number of processes between my Xorg and Mir sessions
[13:55] <bregma> maybe there's a new problem then
[13:55] <bregma> it's the usual whack-a-mole problem getting stuff to run
[13:55] <didrocks> seems this likely won't be trivial
[13:55] <didrocks> also, I wonder why only the welcome screen is displayed
[13:55] <didrocks> not the chrome
[13:56] <didrocks> it's like it was sending the wrong surface to Mir?
[13:56] <bregma> I'll take a quick look again today, now that I've unbroken my test system again
[13:56] <didrocks> bregma: unbroken doesn't mean fixed? :)
[13:57] <didrocks> bregma: keep me posted if I can help you in any mean ;)
[13:57] <bregma> evidently you can DOS upstart with a well-crafter config file
[13:57] <didrocks> ahah
[13:57] <didrocks>  stop calling upstart within upstart! :)
[13:57] <bregma> and an SSD can full up really fast with log messages
[13:58] <didrocks> yeah ;)
[14:02] <bregma> 13158 ?        Ssl    0:01  |       \_ qtcreator
[14:02] <bregma> 13221 ?        Z      0:00  |       |   \_ [cmake] <defunct>
[14:02] <bregma> that's the problem there
[14:03] <bregma> dunno whu cmake gets run, that would require looking at code again
[14:04] <kenvandine> seb128, who takes care of ubuntu-themes these days?
[14:04] <seb128> kenvandine, you!
[14:05] <seb128> thanks for stepping up
[14:05] <seb128> ;-)
[14:05] <willcooke> attente_, pingaling
[14:05]  * kenvandine disconnects
[14:05] <seb128> kenvandine, you are looking at the indicator updated icons?
[14:05] <kenvandine> bzr says larsu has been active
[14:05] <seb128> kenvandine, I replied to Pat yesterday that I was ok taking that one
[14:05] <kenvandine> https://code.launchpad.net/~tiheum/ubuntu-themes/new-indicators/+merge/235598
[14:05] <kenvandine> seb128, great
[14:05] <seb128> but if you want to do it feel free
[14:05] <kenvandine> it doesn't merge cleanly though
[14:05] <seb128> oh? :-/
[14:05] <kenvandine> i replied on the MP
[14:06] <kenvandine> i'd rather not :)
[14:06] <seb128> tiheum, ^ known issue? you might need to rebase
[14:06] <seb128> kenvandine, thanks
[14:06] <seb128> kenvandine, well, larsu has been maintaining the desktop theme/updating for newer GTK versions
[14:06] <seb128> kenvandine, I'm fine handling those landings
[14:07] <attente_> willcooke: hi
[14:08] <kenvandine> seb128, thanks :)
[14:08] <willcooke> attente_ in the house!!
[14:09] <didrocks> bregma: what are you using? I don't see that with pstree -p -s
[14:09] <kenvandine> seb128, i'm creating a vivid landing for everything not blocked by the unity8 landing
[14:09] <kenvandine> seb128, but no free silos right now
[14:10] <didrocks> bregma: with ps axf, I see the same (2 subprocess defunct, named qtcreator), as when run on X11
[14:11] <didrocks> nice to see you back attente_!
[14:11] <didrocks> attente_: how are things in Canada? :)
[14:11] <bregma> didrocks, after a while the zombies seem to get reaped and regular Mir buffer swaps start happening, so it may be a non-problem there
[14:11] <bregma> as in, the problem is something else
[14:11] <attente_> didrocks: pretty good, how was japan?
[14:12] <didrocks> attente_: excellent! Really worthed the travel and double jetlag :)
[14:12] <didrocks> attente_: I'll show you some photos if you want at next sprint (but not all of them, 2100+ ;))
[14:13] <didrocks> at least, was nice to compare SNCF and a real train organization in a civilized country
[14:13] <attente_> wow, you photographed everything there, ha
[14:13] <didrocks> yeah, it was really really nice, and had some splendid weather
[14:13] <didrocks> even discussed with an ubuntu fan at a shrine :)
[14:14] <didrocks> bregma: it seems that a subsurface is rendered anyway, I wonder if the inputs are just not matching the surface
[14:15] <didrocks> bregma: do you think we should ask the Mir team to look at this?
[14:15] <seb128> kenvandine, excellent
[14:15] <didrocks> attente_: is it cold in Canada already?
[14:16] <bregma> didrocks, it wouldn't hurt: it may be pure Qt needs integration that's already in qt-mir or something
[14:16] <didrocks> yeah
[14:16] <attente_> didrocks: not particularly, in toronto at least
=== MacSlow|lunch is now known as MacSlow
[14:17] <attente_> it's about 5 degrees i think
[14:17] <willcooke> do we have an official IRC client?
[14:17] <willcooke> s/official/default
[14:17] <bregma> didrocks, forecast is for snow on Hallowe'en  (Friday) here
[14:18]  * didrocks reads "not particularly cold" and "5" in the same sentence -> doesn't compute! :)
[14:18] <didrocks> bregma: waow, lovely ;)
[14:18] <didrocks> willcooke: we don't, but most sane people don't use xchat-gnome and rather a CLI one
[14:18] <didrocks> willcooke: the kind of people using xchat-gnome are emacs users generally btw
[14:18] <willcooke> didrocks, BUUUUUULLLLLLLLLLSHHIIIII........
[14:18] <didrocks> :p
[14:19] <willcooke> :D
[14:19] <willcooke> so I was chatting with attente_ about this
[14:19]  * seb128 doesn't feel concerned
[14:19] <willcooke> and seb128
[14:19] <willcooke> and didrocks
[14:19] <seb128> who would use emacs and xchat-gnome?
[14:19] <willcooke> well lots of people
[14:19] <seb128> crazyness ;-)
[14:19] <didrocks> seb128: isn't it? ;)
[14:19]  * mdeslaur slaps didrocks
[14:19] <willcooke> right - so we (I) think we should get XChat Gnome working on Mir
[14:20] <willcooke> the reason for that is because I assumed thats what "most people" use
[14:20] <willcooke> if that's not the case, then perhaps we shouldnt do that
[14:20] <willcooke> my gut feeling is that we should though
[14:20] <seb128> willcooke, didrocks was mostly trolling there ...
[14:20] <didrocks> yeah, I agree it's what most of people are using
[14:20] <didrocks> I was just trolling a half-french-german on the channel :)
[14:21]  * seb128 still not feeling concerned
[14:21] <seb128> :p
[14:21] <didrocks> seb128: try harder!
[14:21] <seb128> or just a bit...
[14:21] <seb128> anyway
[14:21] <seb128> yeah, xchat-gnome running on Mir would be nice
[14:21] <willcooke> the we are agreed
[14:21] <willcooke> \o/
[14:21] <didrocks> so yeah, agreed that xchat gnome is needed as part of what people using IRC are going to need
[14:22] <willcooke> ok attente_ so all that ummming and errring was for nought - looks like you've got a big job on your hands
[14:22] <willcooke> good luck
[14:22] <willcooke> wake me up when you've finished it
[14:22] <willcooke> o/
[14:22] <didrocks> "kthxbye"
[14:22] <willcooke> :D
[14:32] <tiheum> seb128: concerning the mp, how can I know what the conflicts are?
[14:33] <seb128> tiheum, just branch trunk and try to bzr merge your vcs on
[14:33] <seb128> or take you version and bzr merge lp:ubuntu-themes
[14:33] <seb128> your*
[14:38] <tiheum> seb128: sorry but what do you call vcs?
[14:38] <seb128> tiheum, your checkout of lp:ubuntu-themes with your changes
[14:39] <seb128> tiheum, if you are not fluent with bzr, the easier might be to do a fresh checkout from trunk and copy your files over again
[14:39] <seb128> or do whatever you did the first itme
[14:39] <seb128> time
[14:39] <tiheum> seb128: ok, will try
[14:39] <seb128> tiheum, don't hesitate to ask here if you have questions/issues
[14:40] <tiheum> seb128: sure (actually, I am not very fluent with bzr)
[14:40] <tiheum> thanks
[14:46] <tiheum> seb128: can I have two mp for two different branches at the same time? At the moment, there's already this one: https://code.launchpad.net/~tiheum/ubuntu-themes/suru-icons/+merge/231533
[14:47] <tiheum> ready to land since more than one month :/
[15:10] <didrocks> interesting, seems there was not (at least in my installation) sshd host key
[15:11] <willcooke> didrocks, ahh
[15:11] <willcooke> didrocks, on a desktop next install?
[15:11] <didrocks> willcooke: yeah
[15:11] <willcooke> didrocks, yeah, I noticed that too, I assume I'd done something daft
[15:11] <didrocks> I'll need to reinstall to confirm it's not a one-time bug
[15:11] <didrocks> ah, it's not one-time
[15:11] <willcooke> in the end I removed and reinstalled and it sorted it
[15:11] <didrocks> I think we have a hook on ubiquity
[15:12] <didrocks> willcooke: well, this happen on package reconfigure
[15:12] <didrocks> so dpkg-reconfigure openssh-server is enough
[15:12] <willcooke> I did a remove and an install
[15:16] <didrocks> will have a look once achieved the current debugging on qtcreator
=== alan_g|lunch is now known as alan_g|tea
=== alan_g|tea is now known as alan_g
[15:47] <happyaron> seb128: I've uploaded the fix to exp this morning, synced just now
[15:47] <happyaron> Laney: ibus was on my list after fcitx actually, but you are quicker, :)
[15:47] <Laney> I didn't upload that
[15:48] <didrocks> Mirv: hey, I'm trying to find libqt5widgets5-dbgsym, but it doesn't seem to exist?
[15:49] <happyaron> Laney: okay, that's dholbach
[15:50] <Laney> yeah I just told someone off for stealing merges ;-)
[15:51] <Mirv> didrocks: qt packages build their own debug packages, so try installing qtbase5-dbg
[15:51] <didrocks> Mirv: ok, doing that, thanks!
[15:53] <Mirv> np
[16:08] <didrocks> bregma: so that you don't waste your time: https://bugs.launchpad.net/qtmir/+bug/1387762
[16:08] <ubot5> Ubuntu bug 1387762 in qtmir (Ubuntu) "Qtcreator is hanging at startup on the desktop next image" [Undecided,New]
[16:22] <seb128> happyaron, thanks
[16:30]  * seb128 watches pitti happily skipping the CI process and commiting directly to unity-control-center ;-)
[16:30] <pitti> seb128: see scrollback from this morning, apparently that was ok?
[16:31] <seb128> pitti, I'm a bit lost to what is ok or not nowadays to be honest, I though things under CI shouldn't be manually uploaded
[16:31] <seb128> but I'm not going to complain at the same time
[16:31] <seb128> the CI way is more tedious and not always that useful
[16:31]  * desrt hits seb128 with a train
[16:31] <seb128> desrt, :p
[16:32] <pitti> anyway, I'm mostly through the transition now
[16:32] <seb128> pitti, thanks for working on that btw ;-)
[16:32]  * seb128 hugs pitti
[16:32] <pitti> required quite a number of merges, porting, etc., but darkxst really did most of the work already
[16:33] <pitti> seb128: yeah, I'm not sure why we even bother with the train for that; these packages have no autopkgtests, aren't on touch images, etc.
[16:33] <pitti> one needs to manually test them anyway
[16:33] <pitti> so I take some pride in my core-dev powers and skip all the machinery :)
[16:33] <pitti> err, I mean "no autopilot tests", but same difference
[16:34] <didrocks> if the build-deps/deps are well defined and not half of the transition is moving to the release pocket, I don't think we need to have the silo phase
[16:34] <didrocks> (especially if it's not on touch)
[16:35] <didrocks> things will happily stays in proposed while testing
[16:35] <pitti> yeah, that too
[16:35] <pitti> this is going to stay around in -proposed for a few days
[16:36] <pitti> Laney needs to land his system-settings MP, darkxst/Noskcaj their gnome-session bits, etc.
[16:36] <pitti> and they don't have CI train access
[16:36] <pitti> or I upload the backported patches if it takes too long
[16:37] <seb128> Laney has landig/CI access
[16:38] <Laney> he meant the last two
[16:38] <seb128> gnome-session isn't under CI though
[16:39] <Laney> iiiiiiiiiiindeede
[16:39] <seb128> I maybe mis-parsed the bit about CI and backporting then
[16:39]  * Laney shrugs
[16:39] <seb128> anyway, shouldn't be too difficult to land those
[16:40] <seb128> settings-rtm has its own vcs now and vivid landing are not restricted
[16:40] <pitti> so that leaves powerd
[16:41] <pitti> I'll do an MP and land that via train then, I guess
[16:41] <pitti> (and hope review will be quicker than https://code.launchpad.net/~sforshee/powerd/fix-warnings/+merge/188613 :-) )
[16:43] <didrocks> Laney: do you know of a good documentation explaining how logind works? like concepts and so on
[16:44] <Laney> like http://www.freedesktop.org/wiki/Software/systemd/multiseat/ ?
[16:44] <didrocks> Laney: perfect! thx
[17:08] <didrocks> eod, see you tomorrow or on Monday for those swapping :)
[17:10] <pitti> seb128: I just discovered a stronger reason why the train fails here -- the builds don't seem to use -proposed
[17:10] <pitti> so we can't land that transition in a silo, but use -proposed only
[17:10] <pitti> -ECOMPLICATED!
[17:10] <seb128> pitti, I think that's by design
[17:11] <pitti> seb128: the "complicated"? :-)
[17:11] <seb128> but yeah, the way it should be done is to have the full transition in a silo
[17:11] <seb128> including upower
[17:11] <pitti> (TGIF .. almost)
[17:11] <seb128> and yeah
[17:11] <seb128> overcomplicated
[17:14] <xnox> seb128: Laney: can lightdm use a gtk2 greeter?
[17:14] <seb128> xnox, I guess so
[17:14] <seb128> why not?
[17:14] <seb128> it has qt4 and qt5 greeters
[17:14] <seb128> the interface between lightdm and the greeters is not toolkit dependent
[17:22] <desrt> pro tip: when you hear 'train' or 'silo' just stop paying attention
[17:22] <desrt> saves brain cells that way
[17:23] <Laney> THERE'S A TRAIN COMING
[17:23] <desrt> la la la
=== mpt__ is now known as mpt
[17:30] <mitya57> seb128, darkxst: Re gnome-flashback dialogs: first, this is not in gnome-panel, but rather in a separate application (which can be used in Unity as well).
[17:31] <mitya57> Next, that app basically provides an org.gnome.Shell.something D-Bus interface that gnome-session talks to.
[17:31] <mitya57> So it does not know whether it was called from keyboard shortcut, via UI or just using "gnome-session-quit".
[17:32] <seb128> cool
[17:32] <mitya57> That application also does other useful things like drawing the wallpaper (which we currently do via Nautilus patch).
[17:33] <mitya57> Ubuntu currently has gnome-flashback 3.10, but I am going to upload 3.14 shortly (can't sync as it needs some Ubuntu-specific patches).
[17:34] <Laney> do we use any of those dialogs in unity?
[17:34] <seb128> no
[17:36] <mitya57> But anything that is not gnome-shell or Unity can use that app. So it should be safe to drop these dialogs from gnome-session.
[17:36] <seb128> desrt, in case you don't see launchpad bug pings, https://bugreports.qt-project.org/browse/QTBUG-42189 got some debugging action and a suggested patch and people would welcome you giving it a look ;-)
[17:37] <desrt> i saw that
[17:37] <desrt> not sure i have anything to say about it though
[17:39] <seb128> k, just comment saying that if you don't, so people don't think you didn't see it
[17:46] <darthbunny> anyone knows how you can enable the virtual keyboard on unity8 ?
[17:50] <seb128> bregma, you might know ^ ?
[17:50] <bregma> darthbunny, there is no automated way (yet)
[17:51] <darthbunny> hou about manually?
[17:51] <bregma> mmm, not sure, I think the keyboard designed to work on the phone is click-packaged only
[17:52] <bregma> other keyboards won't work, since they're X11 keyboards
[17:52] <seb128> click work on unity8 desktop though
[17:52] <seb128> you can install things from the click store
[17:52] <darthbunny> yup
[17:53] <seb128> so maybe it's possible to install the osk click?
[17:53] <bregma> yes, but it means I haven't tested it
[17:53] <seb128> something to add to the todolist I guess
[17:53] <darthbunny> but when I click text fields it doesn't auto-show nothing :)
[17:53] <bregma> yep
[17:54] <seb128> desrt, thanks
[17:55] <darthbunny> osk click is a package or store app?
[17:56] <bregma> darthbunny, try 'apt-get install ubuntu-keyboard' and restart Unity 8
[17:56] <bregma> it might *just* work
[17:56] <darthbunny> atm I'm on the live iso
[17:56] <darthbunny> and seems installed
[17:56] <darthbunny> any chance it would work if I undock my tab?
[17:57] <bregma> ah, I see it's installed for me, too, but Unity 8 is not finding it, there must be some other magic required
[17:57] <darthbunny> I'll do a full install
[17:58] <bregma> darthbunny, I don't think undocking will help, I see errors in the log like 'UbuntuKeyboardInfo - socket error: "QLocalSocket::connectToServer: Invalid name"' so I think it needs some development work
[17:59] <bregma> some kinds of configuration that gets statically installed on the phone image perhaps
[18:00] <darthbunny> I also see there is an ubuntu-touch package
[18:00] <darthbunny> I don't suppose that would help...
[18:02] <bregma> darthbunny, ubuntu-touch is a metapackage for installing the phone image, it pulls in things that break the desktop in general
[18:03] <bregma> it's likely that there are some config files that are "seeded" in the phone image, in other words not included in and packages that get pulled in as a dependency, and keyboard support may be one of them
=== m_conley_away is now known as m_conley
[18:04] <darthbunny> I see
[18:04] <darthbunny> should I report this as a bug somewhere?
[18:09] <Laney> right, I'm outta here, see you on Monday
[18:10] <seb128> Laney, enjoy your swap day and w.e!
[18:11] <Laney> cheers, will do ;-)
[18:11] <Laney> there's a halloween climbing competition or something
[18:11] <Laney> ...
[18:11] <willcooke> off too, day off tomorrow
[18:11] <willcooke> ttfn
=== dpm is now known as dpm-afk