=== lifeless_ is now known as lifeless === doko_ is now known as doko === duflu_ is now known as duflu [05:24] Bonjour tout le monde, bonne nouvelle année ! === OutOfControl is now known as benonsoftware === tjaalton_ is now known as tjaalton === Tm_K is now known as Tm_T [07:03] Hi everyone :) [07:04] I try to compile bamf-qt (there is a Qt5 branch) with Qt5.2 on non-ubuntu Linux [07:04] Is it known to work with ubuntu and Qt5.2? [07:04] I use the Qt 5.2 RC1 from the Qt website [07:10] Branch owner is Florian Boucault I can't reach him via mail, do you know him? Is he in some IRC Channel? :) [08:37] good morning desktopers! [08:39] hey seb128! good morning :) [08:39] didrocks, lut, en forme ? [08:39] ça va bien! dormi 10h ;) [08:39] et toi? [08:40] ça va bien, dormis 7h :p [08:40] t'étais fatigué on dirait ;-) [08:40] ouai, pas beaucoup dormi de dimanche à lundi ;) (donc dodo à 22h hier :p) [08:45] bonjour! [08:45] larsu, guten tag! [08:45] * larsu really should do duolingo to learn French instead of Spanish [08:45] you should! [08:45] or both ;-) [08:46] two languages in ONE HEAD? How is that supposed to work?! [08:53] larsu, you can even go up to 3, you don't need to erase german from there ;-) [08:55] seb128: German is easy. I can just relearn it if I happen to need it again :P [08:56] lol [08:56] * seb128 slaps larsu with trout [08:56] +a [08:57] "german is easy" [08:57] because what is more easy than 1 or 2 genders ... 3! :p [08:58] haha [09:00] tkamppeter_, xmbc ping, do you still care about this package? === psivaa_ is now known as psivaa === lan3y is now known as Laney [09:03] morning === Laney is now known as Guest1882 [09:04] Laney, good morning, how are you? [09:04] or Guest1882 I should say? [09:04] suits me, I think I'll keep it [09:04] * seb128 slaps Guest1882 with an old trout [09:05] umm I forgot how to let freenode allow me to change it back :P === Guest1882 is now known as Laney [09:06] aha [09:06] seb128: good thanks! I was tired after being back yesterday so napped and played games last night :( [09:06] so ruined [09:06] how are you? [09:06] I'm good thanks [09:07] I stopped early yesterday, went to play some tennis, watched some TV and had a good night of sleep ... and bonus point, it's sunny here today [09:07] with some 8°C, wth weather [09:08] * larsu has the feeling he lives in the same city as seb128. Always the same weather. [09:08] larsu, Germany is a small country? ;-) [09:08] lol [09:08] indeed [09:11] heh [09:15] hey seb128, larsu, and Laney -- happy new year to you! [09:16] did you have some good holidays? [09:17] oh, it's a pitti! [09:17] Oh, it's a pitti! [09:17] woah [09:17] wtf [09:17] pitti, happy new year, good to see you ;-) [09:17] Laney, seb128: nice coordination :) [09:17] pitti, wie gehts? had good holidays? [09:17] haha [09:17] yes, absolutely; a week in Dresden with family and friends, and a week in Stockholm with my wife and a friend [09:17] amazing city [09:18] nice [09:18] very good holidays thank you: lots of food, walking & games including getting my mum to play cards against humanity (hahaha) [09:18] oh those reminiscences of Abba! (I used to be a fan when I was a kid) [09:18] ;-) [09:18] * seb128 spent time with friends and family, playing some tennis and video games in between [09:19] good time for eoy holidays ;-) [09:19] pitti: happy new year! [09:20] I had great holidys as well. The usual: hanging out with family and eating a lot [09:20] Laney, lol, how did she find the cars against humanity? ;-) [09:20] cards [09:21] she didn't want to play properly so just did the card tsar bit [09:21] i think 1 game was enough :P [09:21] hehe === ochosi_ is now known as ochosi === vrruiz_ is now known as rvr [09:46] chrisccoulson, i just got an FF update on precise ... afterwards FF refused to start [09:46] chrisccoulson, i had to remove firefox-locale-de to make it work again ... something seems out of sync === tkamppeter_ is now known as tkamppeter [09:48] doko, no, I had only done some experiments for the Nexus 7 Ubuntu desktop, but nowadays I do not package XBMC any more. [09:51] doko, one thing which also made me stopping with that is that there are live TV modules with non-suitable licenses and so these models need appropriate package splitting. [09:52] doko, if you are packaging XBMC I have a simple patch which makes it working on the Ubuntu desktop with touch screen (with mouse it works perfectly). [09:52] tkamppeter, I don't care, just merged because it was ftbfs. kept your two arm patches === greyback_ is now known as greyback === czajkows1i is now known as czajkowski [10:47] Mirv: mitya57: qtcreator's RPATH looks borked [10:48] no wait, lies, $ORIGIN is a real thing [10:48] my brain was borked! [10:50] By the way, do I get it right that qtcreator is the only remaining qt5.2 blocker? [11:17] mitya57: it's one of many, a lot of packages have problems if you see the recipe builds at https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta2/+packages - I have tried to parse the dependency chain to find current blockers at http://pad.ubuntu.com/qt52-dependencies [11:18] bug #1258057 suggests a newer stable branch snapshot of at least qtdeclarative would be needed [11:18] Launchpad bug 1258057 in Unity API "unity-api fails to build against Qt 5.2" [Critical,In progress] https://launchpad.net/bugs/1258057 [11:19] Mirv: Thanks. Currently most of Qt auto-synced stuff and pyqt5 is blocked by that transition, so I *am* interested in moving it forward [11:21] OK, that build overview will be a good start for me. [11:24] buh [11:24] I can't give the password box focus to unlock my desktop === alan_g is now known as alan_g|tea [11:25] try keyboard navigation? (e.g tab) [11:26] nope but it came back after I let it blank again [11:26] attach a kinect and dance your password :) [11:26] or smile to the webcam to unlock! [11:26] :) [11:26] some of the symbols would be quite tricky to dance [11:27] * Laney still has that picture of attente to hax his phone [11:27] https://bugzilla.redhat.com/show_bug.cgi?id=905935 [11:27] bugzilla.redhat.com bug 905935 in gnome-screensaver "virt-viewer - unable to enter gnome-screensaver password" [Urgent,Closed: errata] [11:33] * Laney takes that patch [11:35] guess g-s is really really dead upstream, that's not committed there [11:37] Laney: has there been a decision yet on whether light-locker will be used over gnome-screensaver? [11:37] (in 14.04) [11:37] I think l-l was the intention, but I don't know about doing the integration [11:38] don't think there's really much to do for you [11:38] (update the session-indicator and session itself i suppose) [11:38] ochosi, light-locker sends you to the greeter no? [11:38] seb128: yes, it uses the greeter as un/lock screen [11:39] ochosi, right, that's not going to happen for this cycle then [11:39] ok [11:39] because of VT switching? [11:39] ochosi, the greeter is on another VT, which means your seat becomes unactive, which means audio playback, mic, etc would stop on lock screen [11:39] can't that be fixed somehow? [11:40] which is especially annoying if you let your computer play music or are in a call [11:40] yeah, otoh in a multi-user env it's what's supposed to happen... [11:40] yes, when you user switch it's intended [11:40] but probably not for screen locking though [11:41] well, it's not intended if you idle and the screensaver kicks in [11:41] yeah, true. for some strange reason this always worked for me here [11:41] my music keeps playing when the screen gets locked [11:41] on the greeter? [11:41] i know that it doesn't work for others, but it seems it's theoretically possible [11:41] yes [11:41] weird [11:41] and since lightdm does distinguish between switch-to-greeter and lock_hint, it could probably be tackled [11:42] righ [11:42] right [11:42] it could [11:42] it's just not safe enough for a LTS cycle [11:42] i recently patched lightdm-gtk-greeter to blank the screen upon locking (to reduce vt-switching flickering) [11:42] that's the sort of changes you want to do some cycles in advance to have time to get feedback on the potential side effects [11:42] you know that light-locker is a fork of g-s 3.6 i suppose? [11:42] yes [11:43] basically g-s without the GTK UI [11:43] yup [11:43] well, in any case we would like to use the greeter as lock screen [11:43] it's just too much change for a LTS cycle [11:43] ok, understood [11:43] it's easy enough to opt in to do that for those who want it though... [11:44] who would know a bit about keeping stuff like music playing on locking? robert_ancell? or someone else as well? [11:44] s/locking/VT-switching/ [11:45] well, that's a feature I think [11:45] only the active seat has access to the devices [11:45] right, makes sense [11:45] but it could be made conditional [11:45] right [11:46] well, my guess is that it's going to be fixed in unity8/ubuntu "touch" [11:46] Mir [11:46] e.g in the futur when playing is done through a service and not by the app itself [11:46] the service is probably not going to be linked to an active session [11:47] hm, but it'll be important for desktops as well, and i guess ppl won't be happy if only the default player supports this [11:47] ogra_, I think it's more the music-service that's going to resolve that particular issue [11:47] well, in the futur all apps/players are going to use the service [11:47] yeah, but it is Mir keeping the seat and tellig the service to go on i think [11:47] ogra_, ok, wfm [11:47] I just know that's something resolved on the ubuntu touch stack [11:48] yeah [11:49] seb128: well thanks for the heads up. so i guess if i do find a solution for this with the current stuff, that wouldn't change anything for you right? [11:49] ("this" being audio playback etc) === alan_g|tea is now known as alan_g [12:00] ochosi, if you find a solution that doesn't make the security team unhappy I guess we could rediscuss it [12:01] ochosi, we first wanted to use the greeter as lock screen for the LTS, but we decided that stuff like calls/playback cutting on idle->lock screen wouldn't be something users would be happy with, and that maybe it was too much untested arch/changes for a LTS === jjardon_ is now known as jjardon [12:21] i don't see how you can guarantee that "in the futur all apps/players are going to use the service" unless you specifically disallow direct access to the audio device [12:22] which would basically make ubuntu unusable for playing games or audio/video production [12:22] well, apps that want to go on playing while the screen is locked will need to have special permissions [12:23] which are provided via the service [12:23] other apps will simply stop playing when you lock the screen [12:24] also how did this even get implemented? i thought it was impossible to revoke sound access once the device is opened, due to linux not supporting revoke syscall? [12:25] device permissions for desktops are handled via policykit [12:25] (which in turn talked to consolekit .... which got replaced by logind) [12:26] CK and logind unset the seat activity when you are locked .... media playback is by default only allowed for active local seats [12:27] that doesn't really answer the question [12:28] ?? [12:28] why not [12:29] because changing the permissions on a device has no effect if i have already opened it, or so i have been lead to believe [12:29] CK/logind revoke the device permissions of the user if you lock the screen [12:29] you dont change the device, you change the users permissions [12:31] so where does this take effect? presumably at a lower level than pulseaudio? [12:32] I'm pretty sure it is pulseaudio monitoring the sessions [12:32] but doesn't PA run per user, as the user? [12:32] its at a lower level ... [12:32] or wait, alsa is routed through pulse in any case [12:33] it can monitor which sessions are active [12:33] well it's alsa (library) -> PA -> alsa (kernel) [12:33] (ignore me) [12:33] right [12:33] so if PA is doing it, what stops me from recompiling a PA that just ignores the session? [12:34] not sure thats possible [12:34] but you could indeed reconfigure alsa for direct device access [12:35] and add your user to the audio group [12:35] that will switch you back to last century permission mgmt and give you constant access [12:37] umm yeah, not sure what assigns the acls actually [12:38] ali1234, nothing blocks you to recompile PA or install another OS on your devices [12:38] seb128: i mean as a user [12:38] ie without root access [12:38] not sure what you try to get at [12:38] seb128, can pulse actually be compiled without CK/PK/logind mgmt ? [12:39] (for user sessions) [12:39] * ogra_ doubts that [12:39] ogra_, I've no idea, and I don't think we care [12:39] that too :) [12:39] seb128: if i open the audio subsystem and start playing and then switch session, precisely which piece of code revokes my access to the audio device? considering that i already have opened the device and linux does not have revoke() it can't be done with normal unix permissions [12:39] ali1234, the bottom line is that an user process is not going to be able to prevent the system to suspend [12:40] ali1234, policykit [12:40] nope [12:40] and which system calls does policykit use to do this? (if any; if not, then how does it do it?) [12:40] ali1234, I assume it's pulseaudio [12:41] ali1234, pulseaudio is what routes the sound, it can decide to stop routing to the device for your user [12:41] seb128: well if you assume it's pulseaudio - pulseaudio runs in the user's session, so the user can just rebuild their own version of it and run it without needing root. their own version can just ignore the session [12:42] ali1234, right, and users can also put whatever OS they want on their devices, how is that revelent? [12:42] You can't modify the udev rules as a user [12:42] which is where the acls are set [12:42] seb128: you need root access to install a new operating system [12:42] no you don't [12:42] you need physical access [12:42] you don't? in ubuntu you can write to /dev/sda without root? [12:42] I can boot on an usb stick or a CD and install whatever I want [12:43] you are missing the point, and i suspect doing so intentionally [12:43] no, I'm not understanding what you are trying to argue there [12:43] sure it's always possible to build/install stuff that are non default [12:43] you claim that the user cannot be allowed access to the audio device because it would be a security issue [12:43] but app writers usually target the normal users [12:44] in which context? [12:44] touch or normal desktop user switching? [12:44] yet i have demonstrated that any malicious user can simply bypass this by installing their own copy of pulseaudio in ~/bin [12:44] normal desktop [12:44] oh, I never claimed it was a security thing, I said to check with the security team in case they have security concerns [12:45] I've no clue why things are done the way they are currently done [12:45] well here's the thing: it *is* a security issue [12:46] the only reason to have the audio cut off when you switch session is in order to mitigate this security issue [12:46] but if it is implemented entirely in software which runs as the user the mitigation is ineffective anyway [12:46] well, it's also to allow the other user to access the device when you switch user I guess [12:46] it's not true [12:47] logind ships udev rules to manage the acls which trigger when the session status changes [12:48] seb128: that's a good point [12:49] i guess it's just to make sound work when you use guest login etc [12:51] it does also protect against maliciousness [12:51] the microphone is in the same subsystem [12:51] right but if i open the microphone and start recording, and then the ACLs on the device change, i am unaffected because linux does not implement revoke() - i can continue to record all i like [12:52] show me [12:52] unless this issue has somehow been fixed [12:52] if so, i would like to know how [12:53] I've no idea about that, mdeslaur or security guys might know [13:05] I may be wrong, but I believe udev removes the device and recreates it [13:05] which invalidates the file handles [13:05] or, hrm [13:06] that's actually a good question [13:06] Laney: it's demonstrated here: http://www.youtube.com/watch?feature=player_detailpage&v=ZTdUmlGxVo0#t=1859 [13:07] but it's not just security...something has to manage who gets access to devices when the active user gets switched, and which power policies applies, etc. [13:07] i have no idea if it still works the same way, but if it does not, then my question is: what changed? [13:07] mdeslaur: yeah i realise that now - there is also an element of cooperation [13:08] in order to make the lock screen the greeter, a lot of stuff needs to be modified, it's a big undertaking [13:09] well it's working in xubuntu - for certain values of working [13:14] ali1234: and what is making sure the devices stay active when the greeter comes up as a lock screen, and what revokes permissions when another user is selected? [13:15] i don't know but i know *something* is, because on xubuntu, your audio player stops when the screen saver starts :) [13:16] oh, well, that's the issue [13:16] that's precisely why the greeter needs to be modified, as we _don't_ want devices to be revoked when the greeter comes up [13:17] but surely modifying the greeter won't help.. you'd need to modify pulseaudio and consolekit to handle this use case [13:18] i mean the greeter doesn't know about this stuff, it's just a thing that draws a login window on screen === alan_g is now known as alan_g|lunch [13:19] ali1234: yes, instead of having everything monitor logind, which is monitoring vt switches, logind needs to be controlled by the greeter [13:20] and the greeter will then decide when it's time to revoke devices (ie: when you actually switch to a new user, instead of a simple vt switch) [13:20] yes - assuming you mean the display manager not the actual greeter [13:21] yes [13:21] another use-case is when you want the screen to blank upon locking, which is what normally happens with screensavers/lockers. while we've implemented that in gtk-greeter, it'd be nice if after a mouse-touch it wouldn't again take forever for the screen to go blank... [13:21] although, what if the greeter wants to make sounds? [13:21] and then there's the case of power management [13:22] if I decided I want my laptop to suspend after 10 minutes, and the greeter is my lock screen...the DM now needs to know what my power management preferences are [13:22] yeah, and in a multi-user env, different users might have different powermanager-settings [13:23] well in that case, if multiple users are logged in, who's settings get used? [13:23] exactly ^ [13:23] ochosi: i thnk this whole thing might be more trouble that it's worth [13:23] ali1234: the last user who had an active session [13:24] ali1234: possible. i'm starting to think we need to re-fork gnome-screensaver for xubuntu and have locking within the session and use the greeter only for user-switching [13:24] ali1234: if we modify how the greeter uses the .ui file, we could theoretically even load the greeter's .ui file with gnome-screensaver [13:25] ochosi: yeah i was just thinking that [13:25] that would be pretty cool [13:25] something needs to cover the vt anyway [13:25] mdeslaur: that's the main advantage of using a gtk greeter over some other stuff ;) [13:25] (i mean being able to reuse the .ui file) [13:25] ochosi: or even make gnome-screensaver directly use gtk-greeter like how lightdm does [13:26] ali1234: hm, that sounds nice, but also like a lot of work, no? :) [13:26] no idea [13:26] ochosi: hehe, at the disadvantage of looking like gtk :) [13:26] but yeah, would be the ideal solution [13:26] messing with .ui files sounds harder to me :) [13:26] reusing the greeter binary would work also with unity-greeter [13:27] but they would probably both need to be modified to support it [13:27] ali1234: well it's insecure in the sense that you can never know what the UI file of a greeter contains [13:27] mdeslaur: yeah, although gtk3 really gives you a lot of freedom [13:28] ochosi: yes, I'm merely joking ;) [13:28] ali1234: so basically a liblightdm would have to be written for gnome-screensaver [13:29] mdeslaur: have to admit though, unity-greeter *does* look quite nice... :) [13:29] you could also perhaps modify gnome-screensaver to actually run the greeter by emulating the dm [13:29] yeah that's what i meant [13:30] ali1234: ah, yes, missed that comment [13:30] and it's much better than re-using the .ui file [13:30] which would probably end up being a xubuntu-only solution [13:31] yeah and lightdm-gtk-greeter only as well === alan_g|lunch is now known as alan_g === bregma_ is now known as bregma [14:36] Laney, yay for removing code! === rickspencer3__ is now known as rickspencer3 [15:04] kenvandine: it's to offset my christmas weight gain [15:04] haha [15:05] awesome to consolidate the accountsservice stuff [15:07] yeah, it felt too nasty to write it again [15:07] ugh, uploaded xorg 1.15 rebuild to ppa [15:09] I bet you could reduce the boilerplate even further [15:09] for example by subclassing [15:10] yeah [15:11] Laney, i had one nitpick in your background branch [15:11] ya, I've fixed it [15:11] cool [15:11] thx :) [15:11] * Laney commits [15:11] you meant in the header file, right? [15:12] yeah [15:12] it's up now [15:12] thanks for the reviews [15:12] np [15:30] hey hey hey [15:30] it's meeting time [15:30] qengho, Sweetshark, mlankhorst, Laney, tkamppeter, desrt_, attente, larsu: hello ;-) [15:30] g'day mate [15:30] happy newyear! [15:31] oh, first meeting of the year \o/ === desrt_ is now known as desrt [15:31] hey [15:31] happy new year ! [15:31] I hope everyone had good holidays and a nice start of year [15:32] * kenvandine waves [15:32] it's too cold! [15:32] ok, let's get started (I expect the summary to be short, most of us are just back since yesterday) [15:32] it started with uploading x1.15 to ppa and lts-saucy to archive, can't be a good year. :P [15:32] desrt, would be nice if you could give us some cold and snow back! [15:32] seb128: please... take it. [15:32] I just went for some exercice, 13°C out there [15:32] that's wrong [15:33] seb128: summer is coming! [15:33] should be some -10°C with snow [15:33] larsu, :/ [15:33] get started [15:33] ok, let's get started [15:33] qengho, hey [15:33] Hi all. [15:33] ;P [15:33] * Holidays, family, et c. [15:33] * TODO: Working on High-DPI chromium and multitouch chromium. [15:33] * Speaking of multitouch: DONE: gnome-bluetooth pairing small fix. [15:33] EOF [15:33] qengho, where is the new gnome-bluetooth fix? needs sponsoring? [15:34] moo? [15:34] It will need it, yes. I'll ping someone soon. [15:34] feel free to bug me about any bluetooth, especially if you want help testing :) [15:34] cyphermox: hey... i have a bluetooth problem... [15:35] lol [15:35] desrt: after the meeting :) [15:35] you said 'any bluetooth', right? :) [15:35] cyphermox, hide! [15:35] cyphermox, happy new year btw ;-) [15:35] happy new year :) [15:35] Sorry I tricked you, cyphermox. [15:35] get him to test if it happens under 14.04 [15:35] qengho, thanks, just subscribe ubuntu-sponsors to the bug (or ping cyphermox since he offered to look at ità [15:35] ) [15:36] also fwd it upstream! [15:36] Sweetshark, hey, back from holidays yet (I don't think I saw you wrote anything since I'm back) [15:36] cyphermox, speaking about fwd upstream, I think you still need to do a version of that gnome-control-center ssid fix rebased on trunk [15:37] indeed, I do [15:37] canonicaladmin doesn't have Sweetshark off [15:37] but he doesn't seem around so... [15:37] mlankhorst, your turn! [15:37] xorg 1.15 upload to ppa (ready for testing with open source drivers and nvidia-177/304), lots of sru verifications, lts-saucy pushing to archive [15:37] various bugs [15:37] ^D [15:38] mlankhorst, do we aim at 1.15 for the LTS? [15:38] yes [15:38] is that version stable yet? [15:38] 1.15.0 has been released [15:38] was there a discussion about the update (including the Mir team)? [15:38] ppa:canonical-x/x-staging for testing [15:38] seems to work with mir [15:38] and what's the state of fglrx? [15:38] that's blocking xorg 1.15 atm [15:39] ok [15:39] I guess you are going to do a proper call for testing on ubuntu-devel@ in any case before aiming at the archive, right? ;-) [15:39] I doubt many people will, but in theory it's ready [15:40] well, an email there would at least assure everyone is aware of the update coming [15:40] mlankhorst, thanks [15:41] Laney, hey [15:41] hallo [15:41] • Last year! system-settings tests are failing on the device. Made some headway but they're not working yet: https://lists.launchpad.net/ubuntu-phone/msg05683.html [15:41] • Last year! Some system-settings reviews including attente's icu one (should get that released) [15:41] • Make a private library to factor out the several places in which we handle AS in system-settings, in preparation for https://bugs.launchpad.net/ubuntu/+source/telephony-service/+bug/1265528 [15:41] Launchpad bug 1265528 in ubuntu-system-settings (Ubuntu) "Share ring tone with greeter via AccountsService" [Undecided,New] [15:41] • Update libgdata gnome-online-accounts in Debian & sync [15:41] done done done [15:41] hey desrt, how come you have to use a different annotation for default string values? [15:41] is it because of the gvariant text format? [15:41] Laney, are the tests still not happy on the device? I saw your email in my backlog/read through them, but I'm not sure I got all the details [15:41] no [15:41] try that branch I gave [15:41] ok [15:41] you need help? [15:42] that'd be nice [15:42] I found it quite hard to get advice [15:42] did you try mardy's suggestion? [15:42] yep [15:42] I posted a later reply about it [15:42] ok [15:42] I'm going to look at that again after the meeting [15:42] ty [15:42] I guess holidays didn't help to get replies [15:42] Laney, thanks [15:42] It'll take me a while to get up to speed on it again anyway [15:42] tkamppeter, hey [15:42]