=== chihchun_afk is now known as chihchun === cul_ is now known as cul [05:18] morning touch devs :p === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === psivaa-off is now known as psivaa === chihchun is now known as chihchun_afk [08:09] Good morning all; happy Apollo 11 Day! :-D [08:31] mardy: ping? I have a few questions about online-accounts [08:44] hello, i only want to know if the ubuntu tablet of meizu or bq will have got pen-digitalizator to write in the screen [08:44] i saw the scecs but i didn't found, just 10 points of touch [08:45] and sorry for my "engrish" [09:22] seb128, did you ever see my question about the pkg/abi stuff? [09:23] mhr3, no [09:24] when? [09:24] couple of days ago [09:24] let me find it in logs [09:26] seb128, http://paste.ubuntu.com/7802578/ [09:27] mhr3, yeah, just get it added to the ubuntu-touch seed [09:28] seb128, isn't it a problem that the old version would no longer be buildable? ie the src pkg would be building newer ver? [09:28] yeah, if you do that you need to keep different sources in the archive [09:29] or build old/new abi from the same source [09:29] eh, option #3? [09:36] mhr3, no option 3 [09:37] "don't break abi" :) [09:40] seb128, not what i wanted to hear :/ [09:40] mhr3, yeah, what Laney said is option 3 [09:40] stop changing your interfaces all the time [09:41] * mhr3 looks at libmirserver22 and doesn't feel guilty despite whatever seb128 says [09:41] lol [09:42] at least they don't have clients to support/they transition those with them [09:42] seems it's not your case [09:42] yea, lucky them [09:44] seb128, so what would happen if we added it to the seed and it was no longer buildable? [09:44] mhr3, your new version would stay stucked in utopic-proposed and never reach an image because britney wouldn't let you do that [09:45] seb128, time to migrate to rebecca, britney seems old [09:47] lol [10:35] anybody here familiar with QDbusInterface. The clock app needs to make a call to com.canonical.indicator.datetime and retrieve some properties. However I am having trouble connecting to the dbus interface. [10:36] Here is a code snippet and error output I got http://paste.ubuntu.com/7802826/ === MacSlow is now known as MacSlow|lunch [11:28] seb128, can't we at least have src pkg based based on different series? [11:29] dednick: ping (QDBus help) [11:29] seb128, like lp:foo/abiX [11:29] nik90_: hey. what's up? [11:29] dednick: hey, I am not sure if you remember, but I started writing the custom class for the clock app to get dbus properties from indicator datetime. [11:30] dednick: however I am facing some errors that I am unable to solve, mind giving me a hand if you have time now? [11:30] yup [11:31] dednick: so I have this code at the moment http://paste.ubuntu.com/7803025/ [11:31] dednick: I can confirm that it is able to read the Dbus session bus since I fixed the cmake stuff. However when I run I get the error Service Unknown [11:31] dednick: I verified with d-feet if the interface, path and service names given were correct [11:32] dednick: the entire code is at https://bazaar.launchpad.net/~nik90/ubuntu-clock-app/10-alarm-settings/view/head:/backend/modules/Alarm/Settings/alarmsettings.cpp [11:33] dednick: I get QDBusError("org.freedesktop.DBus.Error.ServiceUnknown", "The name com.canonical.indicator.datetime.AlarmProperties was not provided by any .service files") [11:35] nik90_: firstly, is there a indicator branch that goes with this? [11:35] dednick: I believe charles told me that this has already landed in the indicator side of things. [11:35] dednick: https://bazaar.launchpad.net/~charlesk/indicator-datetime/lp-1318997-export-properties-to-dbus/view/head:/data/com.canonical.indicator.datetime.AlarmProperties.xml [11:36] dednick: MP at https://code.launchpad.net/~charlesk/indicator-datetime/lp-1318997-export-properties-to-dbus/+merge/224743 [11:37] nik90_, your service name is wrong [11:37] nik90_: hang on, i need to update [11:37] that's what the error is saying [11:37] dednick: sure [11:37] mhr3: yeah, but I cross checked the service name [11:37] nik90_, it's without the .Alarm... [11:38] nik90_: yeah, mhr3 is correct. [11:38] "com.canonical.indicator.datetime" is the service name [11:38] what about the object path? Is that still /com/canonical/indicator/datetime/AlarmProperties [11:38] yes [11:39] nik90_: http://paste.ubuntu.com/7803044/ [11:39] I'm confused with the new network settings [11:39] dednick, mhr3: ah thnx [11:39] I have my network listed in "Previous networks", but I don't see a switch to enable it [11:40] dednick, mhr3: that worked :) [11:40] mhr3, well, that's "different source package", and yes you can do that [11:40] mhr3: stop poaching my fixes! ;) [11:40] mhr3, like have lensesv6 and lensesv9 [11:40] seb128, \o/ at least that [11:40] dednick, blame seb128, it took him ages to reply :P [11:41] mhr3, but those can't build any identific binaries, so you need to version the lib/bin/common/etc [11:41] oh ffs [11:41] oh.. disableing and then enabling wifi helped [11:42] seb128, not even if we made sure that the newer ones are always higher version? [11:42] mhr3, no, it doesn't work that way, a binary can only belong to one source [11:42] grrr [11:44] hmm, maybe we could just disable all the other pkgs before releasing new abi? [11:44] nah i guess that wouldn't work [11:47] mhr3, you could stop changing interfaces, rly ;-) [11:47] seb128, tell that to c++ and gcc === alan_g is now known as alan_g|lunch [12:03] huh getting weird apparmor denials on my scope [12:05] jdstrand: any idea why I'm getting this denial? http://paste.ubuntu.com/7803133/ [12:05] using template ubuntu-scope-network === psivaa is now known as psivaa-lunch [12:21] sergiusens, fyi https://launchpad.net/ubuntu/utopic/+source/livecd-rootfs/2.223 === _salem is now known as salem_ [12:24] Hi... recently learned about ubuntu touch... Has ubuntu released an image that will run in any x86 device? [12:26] there is the experimental desktop-next image ... that uses the UI (but none of the backend features a real touch install has) === alan_g|lunch is now known as alan_g [12:45] cwayne: you aren't declaring something correctly somewhere. the items in the zmq directory should be named <"name" from click manifiest>_<"key name under hooks" from scope manifest> [12:45] cwayne: you appear to only have the part after the underscore [12:45] jdstrand: I got it, yep I'd named stuff wrong [12:46] s/scope mainfest/click manifest/ [12:46] ok cool [12:50] mterry, yo [12:50] ogra_, hello! [12:51] mterry, is there any dbus-ish way too find out if the user is capable to lock the screen ? [12:51] what's the word? [12:51] ogra_, capable? [12:51] who isn't allowed to lock the screen? [12:51] i need that for the dev-mode UI (needs to be greyed out until the user is able to lock) [12:52] mterry, i would expect anyone who doesnt have a pw set (or where it is locked) [12:52] ogra_, oh you mean whether a password is required to get into the session? [12:53] i need to know if the user can actually unlock the screen and only then allow that dev mode can be enabled [12:53] ogra_, we do have an EntryIsLocked property [12:54] ogra_, that's false if the user has swipe-to-unlock on and/or an empty password. [12:54] popey: thanks for the quick review :) [12:54] mterry, cool, thanks ... looks like what i need [12:54] ogra_, or true if there is a prompt they have to answer associated with the account [12:54] right [12:54] ogra_, so that's on dest=com.canonical.UnityGreeter, obj=/list, interface=com.canonical.UnityGreeter.List [12:55] cwayne: np [12:55] mterry, cool, thx [12:55] ogra_, you also mentioned not wanting to allow adb if the greeter is currently up? [12:55] for RTM only being able to unlock is the req. [12:55] ok [12:56] Well we have a property for that too [12:56] mdeslaur would like to see adb being stopped on locked state [12:56] no [12:56] i want it to not start if the screen is locked [12:56] but that will break all our testing atm ... so this is post RTM stuff (where we will likely solve it differently by using a key exchange on machine basisi instead) [12:57] if the screen locks while it is connected, that's fine [12:57] right ... i still think we should rather go with a key db and have a key exchange instead ... [12:57] basing on locked/unlocked will cause lots of testing issues [12:57] sure [13:01] cyphermox_: dunno if you've seen but my phone seems to have the radio wedged off.. no matter how I fiddle the airplane switch bug 1342602 [13:01] bug 1342602 in telephony-service (Ubuntu) "Cannot make or receive calls on #133 (and possibly before) mako" [Undecided,New] https://launchpad.net/bugs/1342602 [13:10] Saviq, why the last version of shell has two sections with apps? What the first sections mean? === Tassadar_ is now known as Tassadar === gnu is now known as Guest53452 === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [13:30] AlbertA: hello! :) [13:30] AlbertA: I heard that you're working on the unity-system-compositor fix for the breakage caused by the recent landing [13:31] AlbertA: what's the ETA for a fix? Since if it's not a trivial thing, we would most probably consider a revert [13:37] sil2100: so I think we've identified [13:37] the root cause [13:38] sil2100: I think in about an hour I'll have a fix [13:39] sil2100: basically, we are blocking the QT thread that handles DBus incoming signals/requests [13:40] Excellent [13:46] popey: is that an upgrade or a new install? [13:49] cyphermox_: upgrade, i think it broke when i repeatedly stabbed the flight mode button, but can't be certain [13:50] popey: should I file a bug showing up on home screen against unity or some other project? [13:51] frecel: whats the bug? [13:53] popey: sometimes the home screen doesnt actually update until you touch it, as in if you turn the screen off wait a minute and then turn the screen back on it will be a minute behind and then when you touch it it will refresh and show the correct time [13:57] frecel: i have noticed this also, i believe its a known bug [13:58] sil2100: is the bug frecel is talking about the one you just discussed with AlbertA ? [13:58] popey: no that sounds different [13:59] popey, i think thats rather the one we discussed in several morning meetings now [13:59] yeah, did we file a bug for it? [13:59] where Mirv was checkking if the event blocker is back [13:59] (which doesnt seem the case) [13:59] ok, so new bug? [13:59] balls. [13:59] popey, yeah, i guess so ... and i dont think it is filed yet+ [14:00] what would you file that on? [14:00] heh [14:00] popey, ogra_: yeah... I guess it's the old bug we've been seeing ;/ [14:01] popey, no idea ... it might go even down into the android stack or hybris ... or it might be a simple unity8 thing ... really hard to judge [14:02] ugh. [14:02] this is easily reproducable [14:02] yes [14:03] but hard to find the component at fault === Ursinha is now known as Ursinha-afk [14:05] would it be easier to find the person at fault? :D [14:06] bug 1342742 [14:06] bug 1342742 in unity-mir (Ubuntu) "Screen doesn't update till you touch it" [Undecided,New] https://launchpad.net/bugs/1342742 [14:06] there you go, confirm that ☻ [14:24] bzoltan1: mhr3: I'm trying once again to write a scope from Trusty [14:24] I have an i386 emulator with utopic and the associated kit [14:25] but when I switch to that kit, the only build target I get is "simpletest" [14:25] bzoltan1: also, I think there's a bug in the QtC emulator management [14:26] mhall119: what the bug is about [14:26] mhall119: I think I should backport the latest QtC from Utopic to the PPA. We have fixed tons of things related to the scopes. We have brand new templates and project wizard. [14:27] bzoltan1: I created an emulator called 'utopic-devel', but it shows in QtC as only 'utopic' and it won't start [14:27] running "ubuntu-emulator run utopic-devel" works fine [14:27] mhall119: that could be a real bug [14:28] bzoltan1: wasn't the plan to continuously backport qtc to Trusty? [14:29] popey: are you going to fix the mediaplayer-app icon hide MR? [14:29] cwayne: ping [14:30] mhall119: pong === Ursinha-afk is now known as Ursinha [14:30] mhall119: for the Trusty I apply massive testing and dogfooding ... it takes time [14:30] cwayne: hey man, how have you been developing scopes, are you running topic on your desktop? [14:31] mhall119: i will release a new trusty QtC tomorrow [14:31] ok, thanks bzoltan1 [14:31] mhall119: yep [14:31] cwayne: ok [14:31] mhall119: I do not take risks... I got burned too many times [14:31] mhall119: although it sucks that I have to.. it'd be so much better to be able to do it from trusty [14:31] bzoltan1: do we have enough automated testing around qtc and our plugins? [14:32] cwayne: I agree, which is why I'm sticking to trusty until we get it sorted [14:32] jhodapp: do we have the right incantation? [14:32] mhall119++ [14:33] mhall119: absolutely no... autopilot was not up to the task in the last two years. Now i am testing the latest autopilot from the lp:autopilot and it seems to have major improvements. It can introspect object what it did not see before [14:33] popey: I thought so...we can copy what other .desktop files do in /usr/share/applications/ [14:34] popey: like this: "NoDisplay=true" [14:34] popey: sync-monitor-calendar.desktop uses that [14:35] jhodapp: so remove OnlyShowIn=neverShow and add NoDisplay=true ? [14:35] popey: yes [14:35] ok, 2 min [14:35] great thanks === psivaa-lunch is now known as psivaa [14:36] jhodapp: updated branch, pushed [14:36] popey: all I can suggest right now is to run /usr/share/ofono/scripts/online-modem to get it back online [14:37] popey: awesome [14:37] popey: testing it [14:37] mhall119, tbh never tried the emulator, let me know how that goes [14:38] mhr3: so far it's not [14:38] * mhr3 has to live on bleeding edge [14:38] cyphermox_: that did it, thanks [14:38] I'm going to try re-creating an i386 emulator without hypens in the name to see if I can get further [14:39] popey: yep it's not visible [14:39] i got some of my scopes working in the i386 emulator, so it should work [14:39] sweet jhodapp ! [14:42] popey: added us to the CI sheet [14:42] \o/ [14:47] popey: it's building...will land right after [14:48] seb128, heyo! Did you have time to look at the locking-hash branch? [14:48] will keep an eye out jhodapp [14:48] cool [14:48] ogra_, you mentioned we can't bindmount writable files individually? I see on my device that /etc/init/ssh.override is in my mount output. Am I misreading that? [14:49] mterry, we cant delete bind mounted files ... [14:49] ogra_, but I'm thinking for /etc/shadow and whatnot [14:49] mterry, why do you want to fiddle with that file ? [14:50] ogra_, for the same reason we want libnss-extrausers! [14:50] ogra_, we can avoid the need for that altogether [14:50] but do we want to ? note that this was discussed at the sprint ... [14:50] and the nss-extrausers solution was deemed to be safest [14:51] ogra_, my memory from the sprint was ya'll saying that we couldn't do individual files [14:51] we can [14:51] well.. geeze. Just do that [14:51] :) [14:52] Do we not trust the existing protections for those files? [14:52] not sure we want that ... [14:52] What's the argument against it? [14:53] mdeslaur, would there be security concerns about making /etc/shadow, /etc/group, /etc/passwd writable? [14:53] mterry, hey [14:53] mterry, I was sort of waiting for the security team to do a review ;-) [14:54] mterry, but sure, let me do one from the settings side [14:54] seb128, I'm worried about timing so the more we can front-load the better :) [14:54] mterry: writable by who? [14:54] mdeslaur, bindmounted from the writable filesystem instead of the RO one [14:55] mdeslaur, normal file permissions and such [14:55] * ogra_ sighs [14:56] mterry: I believe all the tools write a temp file beside them and then rename them in place...so I'm not sure you can bind mount those [14:56] ogra_, why am I bumming you out? Am I retreading ground we've already discussed? I don't remember this being seriously considered at the spinrt [14:57] mdeslaur, interesting [14:58] isn't that part of the reason why there was discussion about an different file using nss? [14:58] mterry, it was discusssed at the foundations table in the great ballroom for a bit ... feel free to implement it or not ... just going back and forth in that discussion wont get us forward iirc there were valid reasons to not make the files writable [14:58] mdeslaur, right [14:58] mdeslaur, the prob is that we need the tools to work with nss-extrausers ... passwd, adduser namely ... [14:58] there was also the upgrade scenario, but my memory is a bit fuzzy on that one [14:59] ogra_, OK. There must have been some discussions I wasn't a part of, I thought I was there for those meetings [14:59] ah right, if the groups get changed in the ro image, they won't be represented in the rw version [14:59] popey: np. I'm working on a more permanent fix but it seems there is some disagreeing on how to correct fix it [14:59] mterry, it wasnt actualyl a meeting [14:59] cyphermox_: supply a sledgehammer with every phone [15:00] ogra_, well regardless. I'm caught up [15:00] mdeslaur, yeah, thats handled by stgraber by chowning stuff (or by some implementation to phardcode UID/GID values at build time) === vying is now known as Guest70785 [15:00] does the Ubuntu touch support an accelerometer driver, such as lis3lv02d [15:00] popey: too costly [15:01] ogra_: no, I mean if a group gets added to /etc/group in a package update, the rw image's /etc/group won't get it [15:01] ogra_, did you ever get a hold of stgraber/infinity to discuss how they manage their extrausers setup? [15:01] unless someone writes some merge logic [15:01] mdeslaur, debs arent supported :) [15:01] if you use them you are on your own [15:03] * mterry is just nervous that we are a day away from freeze and we don't seem to understand the details of how to enable extrausers like we need to [15:03] ogra_: so you're telling me the ro images aren't generated by canonical from debs? [15:03] * ogra_ is nervous that we are a day away from freeze and he doesnt get dev mode done because he discusses already planned features over and over [15:04] ogra_, ok, sorry man [15:04] mdeslaur, they are and for changes at build time stgraber works on a solution [15:04] ogra_: and how would you merge that back into the rw /etc/group file? [15:04] which is on the device and not in the image [15:05] mdeslaur, we dont, we dont have a rw file [15:05] mdeslaur, well that's why we probably aren't planning to do rw [15:05] mdeslaur, that was my what-if question [15:05] mdeslaur, not a current plan [15:05] mdeslaur: the idea is that /etc/group is read-only, /etc/writable/group is writable and we use nss-extrausers to look at both files and merge the output at the nss layer [15:05] stgraber: right, which is why we can't just make /etc/group writable like mterry wants [15:06] Right, among other reasons apparently [15:07] mdeslaur, exactly [15:07] stgraber, any idea what we do with "passwd" ? it operates through pam and wont update the /etc/writable/shadow file [15:07] in case you want to set a passwd [15:09] ogra_: I suspect this may be more of a slangasek question seeing how he maintains and contributes to pam upstream :) [15:09] ogra_: there's however a rather ugly but working alternative we could use [15:10] ogra_: basically mount /etc/passwd /etc/group /etc/shadow and /etc/gshadow to /etc/readonly/ and then mount /etc/writable/ over /etc/ [15:10] ogra_: which then makes it so that /etc/passwd contains your local writable entries and /etc/readonly/passwd contains the rest [15:11] making any tool that directly changes /etc/passwd /etc/shadow ... "just work" [15:11] stgraber, hmm and leaving extrausers for the ro files ? [15:11] stgraber, we were just talking about that -- apparently tools that modify them do it by writing new file in /etc then moving it, which makes that plan tough [15:11] ogra_: right [15:12] popey: MR is landing [15:12] charles: ping [15:12] nik90_, pong [15:12] mterry: ah yeah, that's always a bit annoying, we've had the same problem with systemd and ended up having to patch it... as we can't make /etc itself writable for obvious reasons and anything short of doing that just fails [15:13] charles: I got reading the alarm settings from the dbus working :) [15:13] mterry: I suspect the best way there would be to patch pam (or whatever does the edit) to attempt to create the temp file, if that fails, attempt an in-place edit instead [15:13] charles: I wanted to talk to you about the alarm volume though. Is it dependant on the phone volume? [15:13] mdeslaur, ^ how feasible is that from a security pov? [15:13] nik90_, \o/ [15:14] charles: Would setting alarm volume 50 while the phone volume is 0 (silent) still ring the alarm with the alarm volume? [15:14] charles: https://bugs.launchpad.net/ubuntu-clock-app/+bug/1337917 [15:15] Ubuntu bug 1337917 in Ubuntu Clock App "Alarm does not ring when phone is muted" [High,Confirmed] [15:16] nik90_, brb [15:16] charles: ok [15:18] is excessive battery drain on nexus4's a known issue? === dandrader is now known as dandrader|lunch [15:18] I left it on my Qi charging pad last night and woke up to 50% battery [15:18] i haven't seen that for a while, but it can depend whats open [15:19] jcastro, my mako survives a day of moderate usage [15:19] stgraber, mterry: in-place edits on those files is a disaster waiting to happen. Can't we make passwd use the rw files? [15:19] and i charge it via Qi on my nightstand [15:19] as an option [15:19] mdeslaur, thats what we'll do [15:19] popey, ok I had some stuff open, will try it again [15:19] mdeslaur, see stgraber's explanation above [15:20] mdeslaur, but even with the rw files passwd wants to back the original up [15:20] before editing [15:23] stgraber, separately, how does nsswitch handle merging the different databases? Like, if I wanted to add phablet to a group also defined in /etc/group, can I have lines for that group in both and they get merged? [15:28] mterry: I don't know [15:29] stgraber, ok. That's actually not 100% necessary -- we can function without adding phablet to nopasswdlogin [15:30] my guess would be that nss will go through the databases in order and return the first matching entry [15:30] yeah [15:30] which would only be a problem if you had a group in the read-only file which contained members and to which you want to add extra members but IIRC we don't have any of those in our default install (just empty groups) [15:30] stgraber, I also suspect that. Would be nice (for this use case) if those lines were indexed by user instead of group [15:30] thats most likely why the order of entries in nsswitch.conf matters [15:31] stgraber, well we do have that -- nopasswdlogin -- but we can survive without it by setting the user's password to blank [15:32] So yeah, that means that the only big question for libnss-extrausers is how to handle passwd changing the password [15:34] by flipping the files [15:34] ogra_, and making passwd edit in-place, but mdeslaur seemed to hate that [15:35] mterry, well, or work around the temp file creation differently [15:35] of course, you'll hit a race and corrupt the file and the user will no longer be able to log in [15:35] mdeslaur, yeah ... /me not liking [15:35] mdeslaur, we can presumably create the files in a folder besides /etc if we do some patching? [15:36] yeah, that would be my suggestion [15:36] have a dedicated writable dir for the tempfile creation [15:37] needs to be on the same partition if you want to mv it into place [15:37] so you can't bind mount it [15:42] mdeslaur, stop being difficult! ;) [15:42] * mterry needs lunch brain food [15:42] lol [15:43] mdeslaur, can we just go back to storing a crypt hash of the password in ~/.unity8-greeter-demo? so much easier :) [15:43] sure, phablet/phablet was awesome [15:45] lets just drop all that user crap ... make everything rw and the user being root [15:49] mdeslaur, well to be fair, a password hash in the home directory isn't nearly as bad as phablet:phablet [15:50] a user readable shadow file ? [15:51] * ogra_ guesses that is just as bad :) === gatox is now known as gatox_lunch [15:55] nik90_, I don't know how the alarms' sound setting will interact with the overall volume level -- rsalveti just landed (or is landing?) the PulseAudio code so that different roles (e.g. "alarm") can have different volumes [15:55] nik90_, iiuc, for example that's so an alarm can go off at the right volume even if the phone is in silent mode [15:56] charles: iiuc? [15:56] rsalveti, is that correct? [15:56] nik90_, (if I understand correctly) [15:56] :) [15:56] charles: ok [15:56] rsalveti, also while we're on the topic, what does indicator-datetime need to change to set the "alarm" role for the sounds it's playing? [15:57] rsalveti: is there somewhere this is being tracked? [15:57] ogra_, only for the user's own password [15:58] ogra_, which admittedly isn't great, but for Touch apps shouldn't be able to read it [15:58] true [15:58] charles: how is indicator-datetime playing the sounds...does it use a Qt object? [15:58] jhodapp, it's using gstreamer [15:59] charles: right, ok...so yeah it'll be using pulsesink then [16:00] charles: at the moment, I can read/write to the alarm volume, duration. Is there anything else I need to do? For instance should I also use PropertiesChanged signal that you provided? At the moment only the clock app can change these settings. [16:01] nik90_, listening for PropertiesChanged & updating clock-app's state accordingly would be The Right Thing if you've got time to do it [16:02] charles: alrite I will do that as well then. [16:02] nik90_, I agree it's not a hard goal for RTM but if we don't do it now it'll be a bug ticket later as soon as another app starts poking with the alarms :-) [16:02] charles: :-) [16:02] charles: the clock app should be exclusive :P [16:04] jhodapp, ok. Do you know what properties datetime should set there for the alarms, or is that an rsalveti question? [16:05] charles: it's rsalveti since it's dealing with pulse...I'm not exactly sure how you'll get the gstreamer pulsesink to select which output stream to be on [16:06] jhodapp, np; thanks for the info [16:07] charles: looking at the pulsesink docs on freedesktop.org, it does look like there is a stream-properties property for the sink [16:07] charles: that might possibly allow you to select your stream [16:07] charles: I assume you're using playbin to play the alarm sound? [16:08] jhodapp, right [16:08] Laney, i see you have a uss MP related to autopilot tests and updates, do you think that would fix http://paste.ubuntu.com/7804110/ [16:09] kenvandine: nope, those tests are buggy [16:09] charles: yeah, take a look at the pulsesink source code to see what types of properties it'll take in that GstStructure...because you should be able to get a pointer to the pulsesink and set that property then from your code [16:09] wait [16:09] misread [16:09] kenvandine: umm I've not seen that one [16:09] did you look at what this test does? [16:10] i'm sure that failure isn't related to my branch [16:10] it's surprising that it fails [16:10] not yet, it wasn't in anything i touched... so looked to see if it was known [16:10] and saw your branches [16:10] hmm actually I did fix some bugs with 'self.about_page' [16:10] maybe just pre-req on my branch and see if it works after that :) [16:11] I'll see if it fails in CI first [16:11] it fails on my device [16:11] haven't had a CI run on that one yet though so it could be terrible in itself [16:11] the problem was that a recent change was making AP sometimes miss clicking on the about button [16:12] if an update notification comes in and moves it down so that the coordinates change after AP has decided it knows what they are [16:12] nik90_: did that apparmor rule help? [16:13] pitti, I've got someone translating the phone into Korean and he's finding that there is no Korean available for selection in system settings, which is probably because Korean didn't make the cut. I seem to remember that the packages are still on the archive, just not preinstalled [16:13] jdstrand: yeah it did :) [16:13] ok, I'll update the policy [16:13] pitti, so he should be able to just go into RW mode and install the Korean touch langpack, right? [16:13] jdstrand: thnx [16:15] dednick: hey, I am able to now read/write into the dbus. Thnx for your help. Do you have any sample code where I can track the dbus signal to know when the property has changed in dbus? [16:15] dednick: would I need to use connect() as shown in https://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/plugins/Powerd/Powerd.cpp ? === chihchun_afk is now known as chihchun [16:20] dpm: no, we don't build them at all, and include all available packs on the phone [16:22] pitti, oh, but it seems I saw a language-pack-touch-ko package on the archive. Perhaps it's indeed preinstalled already? Let me check if I see Korean myself on system settings... [16:22] nik90_: yep, thats the one [16:24] dednick: will give it a try [16:25] dpm: oh, wait -- I think inclusion of new ones requires rebuilding ubuntu-touch-meta [16:25] dpm: so yes, apt-get installing it is fine for testing [16:26] dpm: I can rebuild -meta now, hten the next image will have it === dandrader|lunch is now known as dandrader [16:27] pitti, that'd be cool, Korean is one of the languages we'd like to have good coverage for, and we've got enthusiastic translators wanting to contribute :-) [16:27] charles: is there a way to change the value in d-feet so that I can check if the clock app correctly updates its state? [16:28] nik90_, you could do it that way, even easier would be using dconf-editor and editing /com/canonical/indicator/datetime [16:29] charles: ah sweet [16:29] nik90_, open up dconf-editor, navigate down the tree on the left-hand-side to indicator-datetime, and then click on the property on the right-hand-side that you want to change [16:30] thnx [16:33] dpm: hm, it was rebuilt three days ago [16:33]