=== marcoceppi_ is now known as marcoceppi === wavis_ is now known as wavis === wavis_ is now known as wavis === wavis_ is now known as wavis === rsalveti_ is now known as rsalveti === Ursinha__ is now known as Ursinha_ === steev_ is now known as steev === plars-holiday is now known as plars === Michaela is now known as Mikaela === mariogrip_ is now known as mariogrip === bremnes_ is now known as bremnes === Guest28436 is now known as IdleOne === mariogrip_ is now known as mariogrip === alexforsale is now known as elasrofxela === rsalveti_ is now known as rsalveti === chrisc_ is now known as chrisc [08:35] good morning [08:57] Sleep_Walker: there is an app in ubuntu store called SensorStatus. it has some detailed info on the gps [08:57] lpotter: thanks, I already found that one === john-mca` is now known as john-mcaleely [08:58] problem is that it seems that GPS is not always working [08:58] I brought it up be enabling it and rebooting phone [08:59] Sleep_Walker, why do you think the gps is not working? [08:59] mpt, o/ [09:00] because I can't see any dot on map for really long time, I can't see any data in Sensor Status [09:01] really long time means ~1 hour [09:01] Sleep_Walker, are you indoors? [09:01] Sleep_Walker, which device and image are you using? [09:01] mpt, is ` an apostrophe? [09:02] Aquaris E4.5, OTA-8, 20151118.2 [09:02] tvoss: is agps supported? [09:02] I'm not sure which numbers are relevant here [09:02] lpotter, sure, depends on the image and device, though [09:03] 20151118-205525 [09:03] Sleep_Walker, do you have an active data connection? is wifi enabled? [09:03] phone was behind the window [09:03] Sleep_Walker, are you happy to shell into the phone? [09:03] absolutely :) [09:04] Sleep_Walker, mind running sudo ubuntu-location-serviced-cli --bus system --test and pastebinning the output? [09:04] tvoss, no, that’s a backtick. A typographical apostrophe ’ is U+2019 [09:05] * tvoss searches keyboard === psivaa-afk is now known as psivaa === davmor2_ is now known as davmor2 [09:06] tvoss, with the “English (Macintosh)” layout it’s Option + Shift + ]. Not so easy in non-Mac layouts, unfortunately. [09:07] Sleep_Walker, please note that the test may take some time as it tries to get a gps fix three times from cold start (without aiding data) [09:08] OK [09:08] ’ [09:08] mpt, correct? [09:09] I'm preparing WiFi connection to do the test through SSH [09:09] Sleep_Walker, didn't you have one before? [09:09] my test was yesterday in the evening [09:10] and at that time I was on different network [09:10] Sleep_Walker, ah okay [09:13] btw. is there something like xdotool for Mir? [09:13] I miss such tool very much [09:14] Sleep_Walker, nope, on purpose :) That being said, you can certainly come up with your own input method to support the use case [09:15] Sleep_Walker, on purpose as it leverages a testing api in X. We would rather encourage developers to plug into the input method channels for such use-cases [09:16] I'm still too far from ubuntu sdk and ubuntu touch api so any solution already made could save me time :/ [09:20] Sleep_Walker, nothing available, yet [09:22] * mcphail is giving up on OTA8 and trying rc-proposed [09:23] mcphail, which issue(s) are you facing? [09:23] tvoss: awful battery life [09:23] mcphail, do you use tweakgeek? [09:24] tvoss: I did, but I've tried without tweakgeeking anything [09:24] mcphail, okay, do you use the music app routinely? [09:25] is tweakgeek a problem? [09:25] tvoss: not every day. Didn't use it yesterday but still had rapid discharge [09:26] mcphail, okay, did you file a bug? [09:26] tvoss: I filed a bug against the most frequent finding in my syslog (clock app) in case that was the cause of the battery drain [09:26] mcphail, got link? [09:27] https://bugs.launchpad.net/ubuntu-clock-app/+bug/1519543 [09:27] Ubuntu bug 1519543 in Ubuntu Clock App "Syslog full of spam from alarm settings - is this draining battery?" [Undecided,New] [09:29] tvoss, yes, that’s the one [09:29] mpt, :) shift+alt gr+N [09:30] mpt, mp is updtaed [09:30] updated, even [09:32] Sleep_Walker, any update from the test? === Guest74213 is now known as pixel_ === mariogrip_ is now known as mariogrip [09:47] sorry, I was out [09:47] let me check [09:49] http://sprunge.us/ZjVL [09:49] tvoss: ^^ [09:50] test looks good, it would make sense to tell me which data would be interesting in case of GPS failure so I will collect it next time the problem re-appears [09:51] Sleep_Walker, you want to hand me the output of 'sudo dmesg', 'sudo /system/bin/logcat', all /var/log/ubuntu-location* files, and which app you used for testing [09:51] OK, I'll prepare script for such case, thanks [09:52] and btw. - I'd love to use Activity Tracker but when I switch the phone off, it stops logging - is there something I could do about that? [09:52] Hi - anyone here using the telgram app? I'm able to search and add bots using web interface but not the app - am i missing something? === JoseeAntonioR is now known as jose [10:00] Sleep_Walker, the behavior is intentional, we are evaluating if/how we can support such use cases without causing excessive battery drain [10:00] bee_keeper: might not be implemented yet, they went for basic compatibility with the old app, now they can add new features to it with relative ease with the new back end [10:01] bee_keeper: you might want to double check though on #ubuntu-telegram [10:01] tvoss: I understand your point, but I somehow miss the freedom of choice in that [10:01] Sleep_Walker, sure, I can see that point, too. But: integrating both is an art, without impacting the UX for the non-tech-savvy user [10:02] tvoss: there should be switch - Give me full control [on/off] ;) [10:02] hmm, upower -d is not very grep-friendly [10:02] Sleep_Walker, sure, what means full control, though ;) [10:03] Sleep_Walker, so yeah, under investigation, takes time [10:03] davmor2: ok, thanks [10:03] tvoss: thanks for caring about that [10:03] Sleep_Walker, see bug 1514329 [10:03] bug 1514329 in location-service (Ubuntu) "Applications should be able to setup and subscribe to geo fence events" [Undecided,New] https://launchpad.net/bugs/1514329 [10:03] mpt, that's not exactly the same :) [10:04] mpt, it enables another class of use cases, though [10:04] tvoss, not necessarily, but “wake me up for a moment when I hit the fence” would be [10:04] And I doubt geofences would be that useful otherwise [10:05] mpt, well, activity tracker really is about ongoing monitoring of certain aspects of the system. you could "abuse" geofences to implement that monitoring, but I'm not a fan of that idea :) [10:05] mpt: thanks [10:06] I added myself as affected user [10:07] tvoss, according to the store description it’s about monitoring your activity, not the system’s [10:07] Sleep_Walker, you are not really affected by that specific bug, but feel free [10:08] mpt, sure, that involve's monitoring the system's location for example [10:08] I don’t think a geofence is an abusive way to track the system’s location :-P [10:09] mpt, seriously, you are confusing two concepts here [10:21] Saviq, done http://i.imgur.com/kvNKMvO.jpg but now the fake browser's controls are under the top bar, i can't minimize it. [10:22] mzanetti, bug #1438465 ↑? [10:22] bug 1438465 in unity8 (Ubuntu) "A window dragged under the panel can't be dragged out again" [High,Fix released] https://launchpad.net/bugs/1438465 [10:23] Good morning all; happy Monday, and happy Computer Security Day! 😃 [10:23] Saviq, i also get a lot of binding loop detected when i resize the unity8's window file:///home/pixel/unity8a/trunk/qml/Stages/SurfaceContainer.qml:54:5: QML MirSurfaceItem: Binding loop detected for property "surfaceWidth" [10:23] pixel_, isn't it maximized, though? if you click on it? [10:23] Saviq, hmm... not sure [10:24] there have been lots of changes lately... [10:24] yeah... indeed [10:24] so far no issue on the screenshot [10:24] pixel_, ah, browser is fullscreen is all [10:25] Saviq, if i click on it http://i.imgur.com/lLAFuA4.png [10:27] yep, the fake browser is fullscreen [10:27] pixel_, you can drag the launcher in, long-press on the browser icon to get "Close" [10:27] mzanetti, we should have window controls in panel for fullscreen apps, fwiw ↑ [10:27] Saviq, no [10:27] Saviq, fullscreen on a desktop is fullscreen [10:28] press F11 on firefox [10:28] mzanetti, yes, but you can pull the panel n [10:28] in [10:28] mzanetti, I'm not saying it should be always on screen [10:28] but when you pull it down [10:28] pull it down? [10:28] mzanetti, when fullscreen, you can still go top-edge-swipe for the panel [10:29] mzanetti, we might enable top-edge-push with mouse, too, TBD [10:29] ah, with touch [10:29] Saviq, at least firefox on X11 uses top edge stuff on its own [10:29] mzanetti, yeah, we want it to stop, because it's hell inconsistent everywhere [10:30] mzanetti, that's what I prototyped and showed off back in Dallas [10:30] aanyway, before we get there [10:30] since you can get the panel, it'd be nice for the window controls to be there :) [10:40] hmm, is it my eayes or did the panel color just chane by a small nuance with todays rc-proposed upgrade ? [10:40] *change [10:40] ogra_, it did change [10:40] :) [10:40] ogra_, and the orange line is gone [10:40] it so minimal that i thought its my brain or so :) [10:41] oh, yeah, thats sad, i liked it [10:41] ogra, you can add it back :D (1 line of qml) [10:42] nah, i'm not *that* attached to it :) [10:42] :)) [11:06] mardy, I'm tempted to drop http://bazaar.launchpad.net/~phablet-team/location-service/15.04/view/head:/src/location_service/com/ubuntu/location/time_based_update_policy.cpp from the default engine configuration [11:06] mardy, rationale: I don't think the service is the right place to do the right thing for *all* applications consuming position updates [11:07] mardy, we could probably think about providing example code to carry out low-pass filtering on the client side, though [11:07] mardy, what do you think? === vrruiz_ is now known as rvr === chihchun is now known as chihchun_afk [11:21] my port finally booting but with no graphics [11:22] any idea what to do next? [11:22] https://github.com/alexforsale/ubtouch-dmesg [11:22] tvoss: it might need some adjustment, but I think that the server side should do that kind of filtering [11:23] mardy, I think we are not clever enought for that tbh, we would have to do the heavy lifting of fusing with sensor data [11:23] tvoss: I cannot imagine the apps being more clever, TBH :-) [11:24] mardy, well, not right now, but I'm thinking along the lines of a full-blown navigation app [11:25] tvoss: what problems do you see with the current code? [11:26] tvoss: I also see some problems, but I want to see if we find the same ones :-) [11:27] tvoss: the problem I see is that if the update is coming from the same provider which gave us the last valid update, we should never filter it out [11:28] mardy, that's one issue: the other is the timeout parameter, which is notoriously difficult to choose [11:28] tvoss: so, for example, it's fine to filter out wifi-based results when they are less accurate than the GPS, but if the last update was from the GPS and now we get a more inaccurate update again from the GPS, we should not filter it [11:28] mardy, sure, I agree with that [11:28] mardy, the other bit is: we really should decrease the accuracy of the last update to age it out [11:29] tvoss: and if we move this logic to the client, we should also extend our signals to include some info about the source of the update [11:29] mardy, and that is something I would like to avoid [11:29] tvoss: yep [11:30] tvoss: indeed, I think that if we don't get new updates from the providers, we should still continuing emitting updates to the client, reusing the last known position with decreased accuracy [11:31] tvoss: BTW, the Qt API allows to specify an update interval; we don't have that in the platform API, do we? [11:31] mardy, it's hidden, but we have it [11:33] tvoss: so, back to the original point: I think that that code needs some improvements, and might need to be moved in the session class, if we find a reason for which different apps would want it to behave differently; but I strongly feel that it should stay in the server side [11:35] mardy, okay, so I think a reasonable first step would be to (a.) move to session (not making it configurable right now), (b.) expose enough information to ensure that chains of updates from individual providers are always accepted and propagated [11:35] tvoss: ok [11:35] mardy, (a.) would give us the opportunity to also integrate a timer that samples the reference position at the requested interval (either faster or slower than actual updates) [11:37] tvoss: indeed, as well as obfuscating the real position to apps that shouldn't be allowed to get an accurate position (in in the future we implement such a policy) [11:37] mardy, yup [11:37] mardy, sounds good. ssweeny might take that one once he finishes the async provider stuff [11:38] tvoss: as for (B), after thinking over your suggestion that we should decrease the accuracy of the last update as it ages, I think it might not be needed [11:39] mardy, well, if the provider delivers faster than expected interval, we sample at a slower pace [11:39] mardy, which we wouldn#t do in the "aging out" scenario [11:40] tvoss: right [11:42] mardy, so I think we want a simple Sampler that is periodically invoked [11:44] tvoss: maybe, but I wouldn't like a wake-up timer... I mean, we shouldn't cause more wake-ups than now [11:44] tvoss: just thinking about battery :-) [11:45] I'm trying to build and run the Unity 8 shell following this tutorial over here: https://unity.ubuntu.com/getinvolved/development/unity8/ but I get an initctl error saying the unity8 process is unknown: http://paste.ubuntu.com/13573548/ Somebody knows how to fix this problem? [11:46] mardy, well, *if* the session is active, we have to satisfy the client requirement [11:47] mardy, in which case we need the sampler [11:47] mardy, it obviously shouldn't run if the session is not active [11:50] tvoss: I mean, suppose that the client wants an update every second, and the GPS gives us an update every 0.97 seconds [11:50] mardy, in which case the GPS is running and the additional cost of just having an additional wakeup is moot :) [11:51] tvoss: OK, s/GPS/wifi or whatever :-) [11:51] mardy, still :) [11:51] tvoss: I mean, we should hand out updates when we have them [11:51] mardy, which is not what the client would expect [11:51] mardy, *if* we support the frequency flag [11:52] Ooh, I just flashed my Nexus 7 with ubuntu-touch/rc-proposed/ubuntu-pd. I like it! What is it? :) [11:52] tvoss: well, the client doesn't really have a say in that, it's just a preference [11:52] mardy, that's not what the qt api suggests, though, iirc [11:52] tvoss: at least the Qt documentation explicitly says that updates could come at any frequency [11:53] okay, that's difficult to me, tbh [11:53] either we support it, or not :) [11:53] Regardless, the Unity 8 theme looks very nice and the window manipulation is nice. The Browser handling touch and mouse different is perfect and I could completely imagine working with the tablet as an impromptu pocket desktop. Which also now I think I know what -pd means. [11:53] in the sense that: if we support the flag we should implement it such that a developer can rely on the frequency [11:53] * tvoss claps :) === Celtor is now known as vitimiti [11:55] And if *that's* what it is, well, I'm sad that the legacy apps icons are only mockups. :) But seriously, there's plenty of polish still needed (as always with these things, that comes last) but it's a giant leap ahead of OTA-8. Very impressive. [11:57] tvoss: I think we should take the desired interval from all the connected the clients, and start computing updates at a frequency acceptable to all (the minimum among them, I guess) [11:58] tvoss: the if the means that we are waking up every second, but we have one client which only cares for updates every 5 seconds, we can let that client skip some updates [11:58] tvoss: but the main point, is that we should have only one internal timer, and not a timer for each client [11:59] mardy, let me think about that. I like the idea of the minimum frequency, which we could also propagate to providers (the gps chipset driver supports setting that value) [12:00] mardy, so how would you force propagation of updates at a later point then? [12:00] mardy, per session, that is [12:00] tvoss: actually, something like internal_frequency = max(1sec, max_common_divisor([freq_client1, ..., fre_client_N])) [12:01] mardy, okay, so you propose to reuse the tick of the system, and update per client every frequency / internal_frequency tick [12:01] correct? [12:01] tvoss: yes [12:01] mardy, where frequency is the client-specific frequency [12:02] yes, so for every session we store the time of the last update (to that session), and the desired frequency [12:03] tvoss: if enough time has passed, we deliver the update, otherwise we don't [12:04] mardy, that still leaves us with a fuzzy statement to developers, though [12:44] tvoss: why do you think so? we could still be quite precise with timely delivering updates [12:44] mardy, that depends on the setup of all clients [12:45] tvoss: well, indeed the first update they receive could come anytime within the first second, but any later update would come at regular intervals from the first === pixel__ is now known as pi||aw [12:53] My car still can't see my phone via bluetooth, depite running rc-proposed as per https://lists.launchpad.net/ubuntu-phone/msg16864.html . Is there any way to debug this? [12:54] mcphail: https://wiki.ubuntu.com/DebuggingBluetooth [12:56] davmor2: embarrassingly, I just found that link at the bottom of the link I posted ;) [12:56] mcphail: yeah I thought it was in there :) === alan_g is now known as alan_g|lunch [13:15] does anyone have an idea what this means? (from xsession-errors): [13:15] upstart: Failed to spawn application-click (com.ubuntu.terminal_terminal_0.7.140) main process: unable to switch security profile: No such file or directory === m0n5t3r_1 is now known as m0n5t3r [13:16] looks like you are trying to run a click app without having the right environment installed [13:20] it started doing this not long ago, then I flashed rc-proposed 191 an hour ago and stuff worked; then after it rebooted itself once or twice it's doing it again [13:20] define "the right environment" [13:22] it is a phone app [13:22] so you need a phone environment [13:22] with all the lockdowns and security bits [13:22] the only differences from the base thing is /etc/hosts (because no other way to block ads) and /etc/apparmor.d/tunables/global (a line allowing read from /etc/writable/hosts) [13:23] it is a bq aquaris phone, so I guess that counts as a phone environment [13:23] you cant run click apps on a desktop witout a lot of tinkering with apparmor and friends [13:23] not desktop [13:23] phone [13:23] you said .xsession-errors [13:24] there is no X on phoes [13:24] *phones [13:24] ... but there is an .xsession-errors in /home/phablet [13:25] I guess it's a standard thing so unity puts output there [13:25] not really [13:25] ogra_: do you know what ubuntu-touch/rc-proposed/ubuntu-pd is? Does it replace ubuntu-touch/rc-proposed on flo? :) (If not, MultiROM Manager is being weird again.) [13:25] no idea where it comes from or why anything would write to it [13:25] nhaines, ask sil2100 ... no idea what it is [13:26] sil2100! :) [13:26] m0n5t3r, the actual log you are interested in is in ~/.cache/upstart somewhere [13:26] nhaines: hey! ubuntu-pd is the pocket-desktop experimentation channel [13:26] ignore ~/.xession-errors [13:27] sil2100: sounds like just the thing to take up room on my Nexus 7. Unfortunately, Bluetooth is broken on the N7 (in OTA-8 too, not just the pd channel). When can I expect Libertine to show? :) [13:27] nhaines: pocket-desktop is nothing more than ubuntu-touch + libertine and puritine [13:27] sil2100: Then Unity 8 snuck in some really nice window manipulation stuff when I wasn't looking. :) [13:28] nhaines: yeah, the PD effort is being worked on actively ;) [13:29] ogra_: nothing useful shows up in any .cache/upstart log [13:29] just unity8 saying it started the terminal, and then that the terminal exited [13:29] m0n5t3r, and you see that when just tapping on the terminal icon ? [13:29] how do I copy/paste from the terminal app? [13:29] yup [13:29] weird ... [13:29] not only the terminal [13:30] nothing works apart from the browser [13:30] sil2100: it's a giant step up from the rc-proposed stuff. I was quite pleasantly suprised. If LibreOffice Writer ran and Bluetooth worked, I could really use my N7 on the go! :) [13:30] I'm guessing something crashes [13:30] that sounds like you broke your apparmor installation [13:30] gtg, back in a few hours [13:30] did you install any debs, make it writable or anything like that ? [13:30] nope [13:30] wow [13:30] out of diskspace ? [13:31] just a line in apparmor.d/tunables saying "/etc/writable/hosts r" [13:31] sil2100: is there any way I can try out libertine using the -pd channel? [13:31] you edited apparmor.d/tunables ?? [13:31] like ... directly ? [13:32] ... [13:32] * mpt turns off Hotspot and the phone restarts [13:32] Well that‘s weird [13:32] thats not the place to make something writable [13:32] mount / -o remount,rw; vi tunables, add line, save; mount / -o remount,ro; reboot [13:32] yeah, that broke all your apprmor profiles [13:32] mpt: that seems like an extremely dramatic way to disable hotspot access. :) [13:32] why? [13:32] nhaines: best if you reach out bregma or kgunn regarding the usage of the ubuntu-pd channel, they're the ones working on the libertine/puritine effort [13:32] it's correct [13:32] sil2100: <3 [13:33] They would know best regarding that ;) [13:33] m0n5t3r, you would have to regenerate them somehow [13:33] doesn't it do that on reboot? [13:33] m0n5t3r, and hacking apparmor is not the way to make something writable [13:33] nope [13:33] it is a way to make /etc/writable/hosts readable [13:33] you want to edit /etc7system-image/writable-paths [13:33] huh ? [13:33] really gtg [13:33] /etc/hosts is definitely readable [13:34] else nothing would work :) [13:34] nhaines: [13:34] (and is a link to the file in 7etc7writable) [13:34] Hello guys [13:34] nhaines: https://docs.google.com/document/d/1gfLy7cBHHwZXD0t74tqYpHvw60W5nc_qJoYUiU4A6mo/edit [13:34] how to install ubuntu touch on nexus 7 2012? [13:35] if you want to try [13:35] how to install ubuntu touch on nexus 7 2012? [13:35] vishnudev: it is not available for Nexus 7 2012. [13:36] kgunn: fantastic! I'll be playing with this a *lot* before Ubucon Summit and SCALE. [13:36] * popey wonders why his laptop now has ofono installed [13:37] well, failed to install [13:38] kgunn: aww, I don't have access to chinstrap.canonical.com. :) [13:38] sorry about that.... [13:39] https://bugs.launchpad.net/ubuntu/+source/ofono/+bug/1521199 [13:39] Ubuntu bug 1521199 in ofono (Ubuntu) "ofono fails to install on xenail desktop" [Undecided,New] [13:40] nhaines can i build? [13:40] vishnudev: no; there are no graphics drivers available for the hardware. [13:42] oh [13:42] but how the old builds working? [13:42] Ancient version of Ubuntu with a different kernel. [13:42] till saucy ? [13:42] ok [13:43] and the point is they did not really work due to the driver [13:43] which is why the device was completely dropped ... the driver issues were not fixable [13:44] kgunn: any guesses as to when ubuntu-pd will include puritine? [13:46] nhaines: hopefully before years end...if not, early in the new year [13:47] I am trying to isntall the old saucy in my nexus 7 2012 [13:47] for what ? [13:47] nothing will run [13:48] can i just flash saucy-preinstalled-touch-armhf.zip and saucy-preinstalled-touch-armel+grouper.zip [13:48] I just want a device to try developing apps [13:48] you surely can ... but you wont be able to install any apps and the majority oif the preinstalled ones are rater in demo state [13:49] kgunn: Thanks! I'll try to bum a puritine click off of someone later, when I'm supposed to be awake. :) [13:50] hmm [13:50] its really bad [13:50] In the meantime, ubuntu-pd is pretty clearly in a development state, but it's a giant leap ahead of the OTA. It's really impressive. [13:50] So good job to all involved! [13:51] I was trying to make a ubuntu touch build for nexus 6 [13:51] it failed and now bought a n old nexus 6 2012 in hope of installing ubuntu touch [13:51] :( [13:52] nexus 6 ? [13:52] you mean 7 [13:52] nexus 6 === alan_g|lunch is now known as alan_g [13:53] nexus 7 2012 [13:53] ah [13:53] well, you should have asked here first ... [13:53] then you would have been told to buy a 2013 model [13:54] yeah [13:54] (which has completely different HW) [13:54] I dint get 2013 here :( [13:56] hm [14:05] vishnudev: even if you got the old ubuntu image flashed onto a grouper, it would be so unusable that you wouldn't be able to test any apps on it anyway. and you especially couldn't test any new apps which require newer frameworks [14:05] you could try the x86 emulator instead [14:07] ok [14:44] mzanetti: when you get a chance could you test https://code.launchpad.net/~michael-sheldon/ubuntu-keyboard/fix-1451554/+merge/278468 (available in silo 2) and double check it fixes the mouse issues with the OSK in pocket desktop mode? [14:45] Elleo, sure [14:46] mzanetti: great, thanks [14:53] awe, one for you? :) https://bugs.launchpad.net/ubuntu/+source/ofono/+bug/1521199 [14:53] Ubuntu bug 1521199 in ofono (Ubuntu) "ofono fails to install on xenial desktop" [Undecided,New] [15:01] popey, gee thanks [15:01] :D [15:01] we need to figure out why ofono is being installed by default on the desktop [15:01] boiko, ^^ [15:01] i added the aptitude why to it [15:01] it shows why it's installed [15:01] k [15:01] is it by default or did popey just install some autopilot [15:02] see bug :) [15:02] it *shouldn't* be default [15:02] it was a telepathy-related lib in the past that caused this [15:02] ( pretty sure ) [15:02] ah, the indicator [15:02] yeah [15:02] the network indicator? [15:02] pete-woods, ^^ [15:03] indicator-network yes [15:03] yeah, it is even a direct dep [15:03] not a recommends or some such [15:05] awe: network indicator is not included in the desktop [15:05] only on the phone [15:05] but it does, indeed depend on ofono [15:07] pete-woods, are you sure that's true for xenial? [15:07] https://bugs.launchpad.net/ubuntu/+source/ofono/+bug/1521199 [15:07] Ubuntu bug 1521199 in ofono (Ubuntu) "ofono fails to install on xenial desktop" [Undecided,New] [15:07] awe: pretty sure? unless it was newly seeded for some reason [15:07] awe: it's not even in main, so if it gets into the desktop image, then that's a problem === bremnes_ is now known as bremnes [15:16] i assume popey has ubuntu-sdk-libs installed ... which depends on qtdeclarative5-ubuntu-push-plugin ... which depends on libconnectivity-qt1 which then depends on indicator-network [15:17] i do [15:17] that would make sense [15:17] it's set up not to start on the desktop (in the upstart session config) [15:18] well, it might be fine, t7he prob is ofono after all === lotuspsychje_ is now known as lotuspsychje === Ursinha_ is now known as Ursinha [15:51] jdstrand_, when I run an app on my device via ubuntu-app-launch, I get a message in dmesg which implies the app is unconfined (it should not be) - what does this mean? It's a bit scary. [Mon Nov 30 15:31:32 2015] type=1400 audit(1448897492.317:88): apparmor="STATUS" operation="profile_load" profile="unconfined" name="com.ubuntu.sudoku_sudoku_1.1.392" pid=19539 comm="apparmor_parser" [15:51] popey: that audit message is saying that an unconfined process performed an AppArmor profile load [15:52] tyhicks, not sure I follow :) [15:53] popey: its just logging that a profile got loaded, and the confinement information of the task that loaded the profile [15:54] popey: after the app is running, do a `ps auxZ | grep sudoku` and the first column will be the AppArmor profile that is confining the process [15:54] in this case, it is the apparmor_parser that is unconfined, not the app run via ubuntu-app-launch [15:57] ahh [15:57] makes sense, thanks chaps! [16:02] tedg, is there something like ubuntu-app-launch which shows stats for startup of a specific app. Like, I launch calculator then run this thing which tells me how long it took to start up fully? [16:03] tedg, or is that not measured? [16:04] maybe ubuntu-app-watch with some very verbose option? [16:04] popey: It's not measured on a live system. You can enable the LTTng tracepoints and figure it out though. [16:05] popey: But LTTng is blocked by apparmor by default. [16:05] ah okay. [16:05] popey: So you need to enable it and rebuild the apparmor profiles. [16:05] so on a clean phone - flash - enabled LTTng - rebuild apparmor profiles - install app to be tested - run app... [16:05] and gather data from ? === balloons_ is now known as balloons [16:06] (or is this documented somewhere?) [16:06] popey: http://gould.cx/ted/blog/Measuring_Upstart_App_Launch_using_LTTng [16:07] Wow, that blog post is almost two years old. [16:07] They grow up so fast [16:07] oooh! [16:08] thanks tedg [16:09] popey: to save you some time-- adjust these: /usr/share/apparmor/easyprof/templates/ubuntu/1*/* to change this line: 'deny /{,var/}run/shm/lttng-ust-* r,' to be: '/{,var/}run/shm/lttng-ust-* r,' [16:09] popey: then run 'sudo aa-clickhook -f' [16:11] Hi [16:12] Does anybody know how to disable the smartscopes (Amazon, Weather, etc.) in Ubuntu Touch? [16:13] BTW, popey I think that brendand was working on something with app startup. Asked me about this recently. You guys might sync. [16:13] mwu: just don't use them === jdstrand_ is now known as jdstrand [16:14]