[00:09] Hario: I think wakelocks are disabled by design [00:10] why? may i ask [00:11] they can be useful [00:13] Hario: battery saving. Everything gets suspended or killed [00:14] Hario: I think the logterm goal is to expose similar functions through frameworks [00:14] yeah, but say, i want to make download manager or torrent app [00:15] it kills the doesnloads [00:15] *downloads [00:15] Hario: yep. You're supposed to use the frameworks for that [00:15] so, is currently not possible? [00:16] Hario: not without using the framework, or an exploit [00:16] is the framework documented? [00:16] Hario: yes, I think so. I haven't used it myself [00:17] https://developer.ubuntu.com/api/apps/qml/sdk-15.04.1/Ubuntu.DownloadManager/ [00:20] Hario: alternatively, you can use an exploit. I have a syncthing app on the store which doesn't spawn a GUI, so is not controlled by lifecycle management. It is an exploit which will be closed "soon", but still worked when I last checked. Not very user friendly, though [00:22] could it be used for torrents? [00:22] the framework [00:22] wait, you said you haven'r used it [00:22] Hario: I don't think the framework would work for torrents [00:23] Hario: I don't hink a background torrent app would work without the exploit I described [00:34] mcphail: mmm, is the syncthing app open source? i'd like to look at the exploit, probably won't use it anyway, just a bit curious [00:36] Hario: just download the .click file and unzip it. I've just wrapped the upstream syncthing binary - https://uappexplorer.com/app/syncthing.njmcphail [00:37] actually, not sure if you can download directly from there. Sure there is a way, though... [00:38] yeah, can't download [00:40] Hario: one moment - I have a tarball somewhere [00:46] Hario: http://themcphails.uk/stc.tar.bz2 [00:46] oh thanks [00:47] Hario: basically, it is just a shell script to start the service [00:49] so, bash scripts are not suspended? [00:49] Hario: not if they don't create a GUI [00:49] lol [00:50] Hario: but bash scripts need to pack lot of things along with them as they don't have access to all the usual utilities. You'll see I had to pack "sed" in the package [00:53] Hario: but expect the hole to be closed "soon" [00:58] anyway, bedtime here. goodnight [01:06] good night === chriadam|away is now known as chriadam === salem_ is now known as _salem [02:46] Ubuntu touch on nexus 7 2012 === timchen1` is now known as timchen119 [04:41] hey guys [04:41] nice to meet you all [04:42] i want to install ubuntu os on my phone [04:42] let me learn from you [04:42] please [04:42] what's it i need to install ? [04:48] hi [04:49] nice to meet you [05:15] Hi, first you need a machine running ubuntu, even a virtual machine [05:16] And a compatible device, I use a Nexus 4 [05:17] It's all here: https://developer.ubuntu.com/en/start/ubuntu-for-devices/installing-ubuntu-for-devices/ === messenja1 is now known as messenjah === chriadam is now known as chriadam|away === marcusto_ is now known as marcustomlinson [07:28] good morning [07:39] afternoon! [08:06] Hello! Ohh... I'm so looking forward to OTA-8 :-) [08:28] Saviq, hey, have you seen https://bugs.launchpad.net/canonical-devices-system-image/+bug/1494889 ? it has been targeted for ota9 and assigned to alecu. i think it's all on unity8 side though; for us it will be just doc update [08:28] Ubuntu bug 1494889 in Canonical System Image "Play movies directly from Scope" [Wishlist,Confirmed] === zz_Foxmaster is now known as Foxmaster === dandrader is now known as dandrader|afk === chihchun_afk is now known as chihchun === dandrader|afk is now known as dandrader [10:12] Good morning all; happy Thursday, and happy Pizza With The Works Except Anchovies Day! 😃 [10:24] Hello anyone on here? I have a question about wifi and hexadecimal key. [10:28] Hello. is there any working SIP client on Ubuntu Touch? [10:42] i got this in my dmesg: unity-system-co: page allocation failure: order:0, mode:0xd0 http://paste.ubuntu.com/13237531/ anybody knows what it means? === _salem is now known as salem_ === dandrader is now known as dandrader|afk === dandrader_ is now known as dandrader [12:46] guys anyone got answer for this ? [12:46] http://askubuntu.com/questions/697155/how-to-make-a-flashable-zip-of-ubuntu-touch [12:46] I've executed make -j6 [12:46] and it was success full [12:49] Is it possible to get the OTA-7 installed on the Meizu MX4? [12:50] I can't seem to get it from the About->Updates [12:51] slo, https://wiki.ubuntu.com/Touch/ReleaseNotes/OTA-7 if you use the stable channel you should automatically have it [12:51] slo, r6 is OTA7 [12:52] Im on r4 [12:52] and you are on the stable channel ? [12:53] I think so, its all original === chihchun is now known as chihchun_afk === Foxmaster is now known as zz_Foxmaster [13:01] yuep, stable/meizu.en === alan_g is now known as alan_g|lunch === zz_Foxmaster is now known as Foxmaster [13:10] Hi, I've noticed that if my phone is trying to connect to eduroam and is failing/in progress, and then I move away from the hotspot (eg on a bus) it still appears in the list like its still connecting and no network works until i switch WiFi off and on again, has anyone else had this before? === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === hasselmm1 is now known as hasselmm [14:08] tvoss: I've to confess that I'm quite lost in the code [14:08] mardy, hangout? [14:09] mardy, or better: gimme 30 until I have stabilized my wifi connection [14:09] tvoss: I'm in a café for 1-2 hours [14:09] mardy, so no ho for you? [14:10] tvoss: nah... maybe we can talk here in IRC; I guess the source of the confusion is the many files around: [14:10] ahayzen: yes [14:10] tvoss: I spot a pattern with a class having {skeleton,stub,implementation}.cpp files, but I'm not sure of their roles [14:11] mardy, so you have got an interface in general (in include), then you have a stub for the client side, and a skeleton for the service side [14:11] stub and skeleton take care of handling communication, nothing more [14:12] the interesting bits are all in implementation.cpp [14:12] tvoss: skeleton is for D-Bus? [14:12] yup, so is stub [14:12] mcphail, yes to the wifi issue? if so do you know if there is a bug tracking it? [14:13] ahayzen: I was waiting to see if the dbus wifi bug fixed it before filing one [14:13] mcphail, i'm on the latest rc-proposed, are there silo's still to land? [14:14] ahayzen: not sure. There was a fix in a silo which was _about_ to be landed a day or 2 ago [14:14] hmm, it might not be fixed then :-/ [14:14] bug 1480877 [14:14] bug 1480877 in Canonical System Image "Access points' "PropertiesChanged" dbus signals freeze UI on mobile devices" [High,Confirmed] https://launchpad.net/bugs/1480877 [14:15] mcphail, note this is not when going out of range of one that you have *connected* to... [14:15] ahayzen: bug was already marked as "fix released" before ota-7, but it was only a partial fix [14:15] tvoss: mmm... so libubuntu-location-service is to be used both by the client and the service? [14:15] mcphail, this is specifically if it is failing to connect but still trying [14:15] mardy, yup [14:15] mcphail, and then you go out of range, then it seems to just keep trying even though its not there [14:16] tvoss: ok, so I can ignore the stubs as long as I work on the service only [14:16] mardy, yup [14:16] ahayzen: mine doesn't even try to connect to eduroam (as i don't have access to it anyway), but it stays on my wifi access point list even when I leave and i sometimes struggle to get connected to my home wifi without switching off/on [14:16] mcphail, well that the trimming of the list seems to be fixed [14:18] ahayzen: I can't run rc-proposed 9as I need phone for work) but I'll let you know if I still get the problem after the next OTA [14:18] mcphail, yeah, i'll continue playing about :-) maybe there are still some bits to land [14:18] ahayzen: to be honest, i keep the wifi switched off when I leave the house now due to the bug mentioned above [14:19] yeah but i want it to disconnect from eduroam and connect to the bus :') [14:20] * mcphail wonders which circle of hell networkmanager inhabits [14:20] hah [14:21] ahayzen: interesting thing is a lot of people who had the dbus bug had it when near eduroam access points as well. Don't know if it is a simple coincidence [14:21] I suppose eduroam access points are very common [14:21] mcphail, you'll have to ask one of the dark priests like awe or cyphermox [14:21] when you are on a university campus or university town yes they are :-) [14:22] they can tell you what to sacrifice to reach which circle of hell to fix NM [14:22] * awe doesn't know whether to be offended by the work "dark" or the word "priest". ;D [14:22] :D [14:22] awe: if you are going to be one, you're as well also being the other [14:22] mcphail, which "dbus bug" [14:22] mcphail, ;/ [14:22] bug 1480877 [14:22] bug 1480877 in Canonical System Image "Access points' "PropertiesChanged" dbus signals freeze UI on mobile devices" [High,Confirmed] https://launchpad.net/bugs/1480877 [14:23] mcphail, I've been working on it for close to 2 weeks [14:23] we're pretty sure we understand the underlying cause(s) [14:23] but we're still working on fixing it ( in at least two places ) [14:23] awe: I was hearing one of the current silos has a near-fix [14:23] not quite yet [14:23] awe: when you get it fixed, I will buy you a pint [14:24] so the basic cause is that we have multiple system daemons that are endlessly added DBus match rules to watch NM signals [14:24] awe, i seem to have an issue where, for certain eduroam connections it won't connect so it keeps spinning, then when you move out of range of the access point it still tries to connect to it and blocks other connections (cellular data) until i flip wifi off/on again [14:24] and these processes never cleanup these rules when the APs go away [14:24] ahayzen, that's a different issue [14:25] awe, is that report somewhere? [14:25] or do i need to make a new bug? [14:25] not that I'm aware of [14:25] * mcphail needs to put down his sandwich and get back to work [14:25] I would suggest filing a new bug, however until we fix this perf problem [14:25] it won't see much attention [14:26] awe, cool, against network-manager (ubuntu) or indicator-network (ubuntu) .. or? [14:26] mcphail, also in vivid, the per process match rule limit was increased from 512 [14:26] to 5k [14:26] which means if you have 8 process all adding AccessPoint object match rules [14:26] and they all max out [14:27] the dbus daemon could have ~40k match rules to consider for each message sent on the bus [14:27] bad [14:27] bad [14:27] bad [14:27] one source is code in loc services, the other appears to be in Qt [14:27] why do we allow match rules at all ? [14:27] it's a fundemental dbus mechanism [14:27] it doesn't work without them [14:28] instead of having a proxy process that handles matching once [14:28] read the spec dude [14:28] ;)- [14:28] pfft specs ... :P [14:28] basically this is how process (a) says... I want to see signals from process (b) [14:28] right, but if you flood the bus having a proxy would be better [14:29] ogra_, the bus is the proxy [14:29] so you match once and notify the other listeners instead [14:29] tvoss, yeah, but thats bad design if you can hog it by to many requests [14:29] ogra_, how would the proxy know about the listeners? ;) [14:30] ogra_, not debating that, pub-sub is just difficult in these scenarios [14:30] well, the listeners would have to register with the proxy instead of listening directly on the bus [14:31] ogra_, so that only moves the problem [14:31] it moves it out of the global bus though ... wheer you have other stuff going on that you dont want to slow down [14:32] ogra_, sure, it does not solve the underlying problem, though [14:33] it cleans up traffic [14:33] on one bus, by migrating it to another [14:34] well, i should have said it splits the traffic ... but indeed you are right [14:34] at any rate, nothing we can do about the design right now, it is what it is [14:36] * ogra_ wonders how that design will work with kdbus ... will we saturate the kernel eventually ? :) [14:37] iirc, the match rule matching leverages a bloom filter to speed up [14:37] avoiding the O(n) iteration in a lot of cases [14:37] not all of them, though === Guest42341 is now known as radheuse === dandrader is now known as dandrader|afk [15:01] tvoss: would you mind a very quick look? https://code.launchpad.net/~mardy/location-service/last-known-position/+merge/277358 [15:01] mardy, in 10, finishing an iteration on the trust-store MP [15:02] tvoss: it's the dumb implementation with no accuracy reduction, but I'd like to know if the logic is correct (and in the correct place) [15:02] I only verified that it builds :-) [15:02] mardy, ack === alan_g|lunch is now known as alan_g [15:11] ogra_, we have bugs in our high level dbus bindings that are causing the issue(s) [15:11] in theory, this kind of thing shouldn't happen [15:11] ah [15:11] but dbus offeres a lot of rope [15:11] and in this case [15:12] we've tied ourselves into knots [15:12] so yea, kdbus would choke too [15:12] heh, fun === dandrader|afk is now known as dandrader === Foxmaster is now known as zz_Foxmaster === BOHverkill_ is now known as BOHverkill [15:47] mardy, yup, that looks good [15:47] mardy, but please avoid & [15:48] mardy, you want a weak_ptr in this specific case, to avoid keeping the session instance alive [15:50] tvoss: OK, thanks [15:50] mardy, also updated https://code.launchpad.net/~thomas-voss/trust-store/fix-1504022/+merge/277266 according to your comment === zz_Foxmaster is now known as Foxmaster [15:54] mardy, hang on, need to push one other revision [15:56] mardy, good to go [15:56] tvoss: OK. BTW, why are you searching the filename as a substring, instead of adding ".desktop" and looking for a match? [15:57] mardy, that's essentially the same, isn't it? [15:59] tvoss: actually, you are also comparing the name 338-343 of the diff, so my question is what are the lines 329-336 for [16:01] mardy, those mimick http://bazaar.launchpad.net/~system-settings-touch/ubuntu-system-settings/trunk/view/head:/plugins/security-privacy/trust-store-model.cpp#L94 following [16:08] Elleo: hey, have you seen a maliit bug where tapping a key on the osk causes the character to appear twice in the pre-edit? [16:09] attente: I've seen that on X11, haven't had a chance to dig into it though [16:09] attente: is it happening on phone devices for you too? [16:10] Elleo: no, just X11. i tried checking out an older source tree, but it still exists. so i'm thinking maybe it's an ubuntu-keyboard plugin bug [16:10] attente: yeah, quite possibly [16:11] attente: might be something different in mouse handling to touch handling [16:11] as I know we have our own custom touch area stuff for handling multitouch [16:11] so could be something goes wrong there with mice with respect to press/release [16:11] Elleo: ok, thanks for the tips. i'll investigate further [16:12] attente: cool, let me know what you find [16:12] will do [16:23] mzanetti: hey, curious. I setup my authenticator for use with github and it seems to work fine, but when I open authenticator, the github account has a timer and keeps cycling through. I guess this is intended? [16:23] jdstrand, heh, yes [16:24] jdstrand, there are 2 ways for OTP, HOTP (as sso.ubuntu.com) or TOTP (like github) [16:24] one is counter based, the other time interval based [16:24] 90% of services I know out there use TOTP [16:24] in face sso.ubuntu.com is the only one I have that uses HOTP [16:24] fact === Guest27111 is now known as alpha_ [16:25] interesting. thanks! :) [16:51] tvoss: approved [16:51] mardy, ack and thx === hasselmm1 is now known as hasselmm [18:20] hey all. anyone got any tips on how to proceed with an aquaris 4.5 stuck on boot on the bq logo on white background? this happened when rebooting on flashing the ubuntu-touch/devel-proposed/krillin.en-proposed -channel. any other options to get into recovery than waiting for the battery to end? [18:22] jaywink: why did you use the image instead of factory reset? [18:23] lotuspsychje, I just wanted to try the bleeding edge stuff, not do a reset to stable [18:23] (yes I know the risks :)) [18:24] ok np just asking [18:25] ok it seems holding power for a looooooooooong time booted. didn't just try long enough. booting to recovery now, hope that works ;) [18:28] jaywink: idle here if you cant get it fixxed [18:28] jaywink: devs will wake up at other times for sure [18:29] sure thanks. just flashing a ubuntu-touch/devel/krillin.en in fastboot [18:40] OK I guess the problem was this in both flashes (last one failed too): Failed to enter Recovery .. found this AU: http://askubuntu.com/questions/602035/how-do-i-use-ubuntu-device-flash-with-the-bq-aquaris-e4-5-and-aquaris-e5 ... if a recovery image is always needed for production phones (it's installing now once I tried with that), should probably say that in https://developer.ubuntu.com/en/start/ubuntu-for-devices/installing- [18:40] ubuntu-for-devices/ page? === alan_g is now known as alan_g|EOD [19:05] how (not) to do mobile phone security... http://arstechnica.com/tech-policy/2015/11/everyone-blames-someone-else-as-classified-military-smartphones-lack-patches/ [19:06] what does make them "military" ? the camouflage rubber bumper ` [19:06] ? [19:06] ogra_: an extra delay in pushing security fixes :p [19:06] ah :) [19:08] google writes patch, sends them to manufacturer, manufacturer tests, sends to carrier, carrier tests, sends to DoD, DoD tests, patch gets deployed (several months after the vulnerability was published) [19:08] all that provided your phone isn't too old [19:08] you think the DoD has carriers involved ? [19:08] ogra_: apparently [19:08] crazy [19:09] i mean ... that alone is already ... ummm ... [19:09] they designed their own phone, then dropped the project when they discovered it would have been very outdated by release time and cost 4.5k / phone [19:09] heh [19:10] so then they had to scramble to get other phones, and went to the carriers for that === salem_ is now known as _salem [21:21] hi === Elimin8r is now known as Elimin8er [22:11] Ok, seems that now that Im on WiFi that I can update my phone... === Foxmaster is now known as zz_Foxmaster