=== duflu_ is now known as duflu [02:17] popey: you probably have to do some trickery to get a full download of the specific revision instead of an image diff === chihchun_afk is now known as chihchun === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [06:20] pitti, good morning :) qtubuntu-sensors is in landing sheet now, and the Mirv set the cmake branch to approved [06:20] tvoss: good morning -- great to hear! so it should auto-merge into trunk soon? [06:21] pitti, yup [06:21] cool; I'll work on qtubuntu-sensor tests then this week [06:27] nexus 5 [06:28] hello developers, this is imjm [06:29] i want to know about ubuntu for my device [06:30] :( [06:30] pitti, great, that sounds good [06:30] imjm, just ask your questions then [06:34] any link for the status of development for Nexus 5? [06:36] preview release === Will is now known as Guest82215 [06:51] What is Ubuntu Touch like on the Nexus 4 these days? I tried it at one point, but had problems with the keyboard and the UI kept hanging. How are things coming? [06:56] any news about nexus 5? [07:07] Guest82215, I use it as my primary phone these days. Glitches: yes, but very usable [07:15] tvoss, Does bluetooth function? Specifically, can you use a bluetooth keyboard? I sometimes use my phone as a laptop replacement that way. [07:16] Guest82215, I'm not sure tbh, would need to check. However, you can now dual boot android and ubuntu, so easier to keep Ubuntu Touch around on your phone [07:18] I've thought about going the dual boot route, but I only have an 8GB model, so I've been hesitant to do that. But it's probably my best bet at the moment. [07:18] Thanks, tvoss. [07:18] Guest82215, sure, yw :) === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [07:45] good morning === chihchun is now known as chihchun_afk === jodh` is now known as jodh [09:03] tsdgeos: Saviq: hi! I seem to have found out how to reproduce the ubuntu-ui-toolkit segfaults locally.. I wonder if you could help? at least I don't find any existing upstream reports yet. bug #1268507 [09:03] bug 1268507 in qtdeclarative-opensource-src (Ubuntu) "V4 crash if executing a Qt Quick test without /proc mounted" [Critical,New] https://launchpad.net/bugs/1268507 [09:06] Mirv: i can have a quick look [09:07] let me see if i can repro [09:07] Mirv: can i unmount proc on my running system and hope nothing will explode? [09:08] tsdgeos: haha :) just a second, I'm ready for a reboot, I'll test what happens [09:09] tsdgeos: it's possible, via sudo umount -fl /proc - system still works, and I can launch more gnome-terminals but not more firefoxes [09:09] tsdgeos: I can then launch firefox again after sudo mount -t proc proc /proc [09:10] so seems "safe" for short periods [09:10] ok [09:12] Mirv: yeah happens here too [09:14] if builders do their builds in similar procless chroot, that might explain why it isn't gotten locally normally when proc is mounted but is seen in the PPA. [09:16] Mirv: is not even qt [09:16] it's just pthreads [09:16] Mirv: http://paste.ubuntu.com/6743566/ crashes if you don't have proc mounted [09:17] Mirv: basically i'd say not having proc mounted doesn't seem a good idea :D [09:17] that is the code inside Qt that's crashing when proc is not mounted [09:17] i tried valgrind-ing it [09:18] but valgrind doesn't like the fact proc is not mounted either [09:18] :D [09:18] maybe some error checking is missing in that code [09:18] let me see [09:19] eh.. [09:20] right checking the funcions return makes it not crash [09:20] not sure i can still make the Qt code be meaningful (i.e. that code that crashes is what i copied out from the Qt code that is crashing) [09:20] it's not like i invented it from nowhere :D [09:22] I've compiled all the Qt modules themselves including tests at least without /proc, so it seems not too common to have that without error checks [09:22] since I tend to compile them locally on another machine inside chroot and I don't bother mounting the /proc usually [09:24] ok [09:24] i think i can make it not crash easily [09:25] but it seems that is the code used to check how big the stack is [09:25] so may have side effects [09:32] Mirv: yes, running QML stuff needs pthread_getattr_np to succeed that needs /proc/self/maps to be there, don't think there's much we can do here, i don't think i'll be able to convince the Qt guys not having /proc around is a priority for them [09:32] but i can try [09:33] cjwatson: do the builder build environments have /proc mounted? there is a crasher that happens when using pthread without error checking and /proc not mounted, and I wonder how it is on builders [09:34] tsdgeos: if error checking is something usually done in that situation, I'd guess they'd accept a patch at some point even if not a priority [09:34] Mirv: the thing is that i can't see how to trivially add error checking [09:35] they are using the pointer to do nasty stack collection stuff from what i can see [09:35] so they need that value [09:35] trying to see if i can think of something that makes some kind of sense [10:13] Mirv: ok, ended up with https://codereview.qt-project.org/#change,75282 seems to work enough for the unittests, let's see how the review goes [10:17] With last two rev on Nexus 4 wifi doesn't work. I have no dual-boot. what could be the cause? [10:23] WebbyIT, you need the 4.3 radio firmware [10:24] sergiusens, this is weird, I have Ubuntu Touch since september I haven't flashed Android since then... [10:25] sergiusens, so, I have to flash Android 4.3 and then reflash Ubuntu Touch? === ondrat is now known as ondra [10:25] just the radio image [10:25] let my try, thanks [10:32] tsdgeos: and first approval already! I'll do a test build with that included. [10:52] Morning all [11:01] didrocks, "TypeError: expected tuple for callback user_data", rings a bell? [11:02] didrocks, and it's python-gi [11:02] -gobject doesn't affect [11:03] pitti, didrocks said you might now something of the above ↑? === chihchun is now known as chihchun_afk [11:04] Saviq: yeah, really looking like the warning/check in latest changelog [11:04] Saviq: do you have a pointer to the code that causes this? [11:04] Saviq: stupid question, do you have a tuple as the callback in your code (even if you don't use it)? [11:04] pitti, http://bazaar.launchpad.net/~unity-team/unity8/trunk/view/head:/tests/autopilot/unity8/shell/emulators/create_interactive_notification.py [11:05] Mirv: yes they have /proc mounted [11:06] didrocks, not sure ↑ that's the code that triggers it [11:06] didrocks, action_callback I think [11:07] Saviq: https://developer-next.gnome.org/libnotify/0.7/NotifyNotification.html#NotifyActionCallback says that the callback just receives one user_data argument [11:08] Saviq: I don't have an off-hand idea, I'm afraid; could you turn this into something small and runnable and file a bug report? [11:10] pitti: it seems it started with latest python-gobject, do you think it's a side-effect of the additional checks or it started to show a real issue in the code itself? [11:11] didrocks: it's more likely to be a regression; but I don't have a firm idea yet [11:11] ok ;) === MacSlow is now known as MacSlow|lunch === zsombi is now known as zsombi|lunch === chihchun_afk is now known as chihchun [11:49] tvoss: I figured out the cmake/gtest/QtPlugins integration now and committed a first sensor API integration test case: http://bazaar.launchpad.net/~pitti/qtubuntu-sensors/integration-tests/revision/46 [11:50] pitti, great, let me have a look [11:50] tvoss: I'll make it more Qtish now and drop some C stuff (mkstemp/write -> QTemporaryFile etc., but this part was copy&pasted from platform-api) === abeato is now known as abeato-afk [11:51] pitti, ack === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [11:58] rsalveti: !! [11:58] wake up mate [11:58] asac, mup him :P [12:00] nah :) [12:00] not important enough [12:00] i think he first checks public comm before checking /msg when getting up, so pinged him here :) === abeato-afk is now known as abeato === chihchun is now known as chihchun_afk === alan_g is now known as alan_g|afk === _salem is now known as salem_ [12:18] can I read D-Bus annotations using QDBus somehow? [12:25] pitti, bug #1268578 [12:25] bug 1268578 in pygobject (Ubuntu) "Notification callback causes exception in gi" [Undecided,New] https://launchpad.net/bugs/1268578 === MacSlow|lunch is now known as MacSlow === alan_g|afk is now known as alan_g [12:51] Saviq: thanks === alan_g is now known as alan_g|lunch [13:10] tvoss: ping [13:11] tvoss: (related to app lifecycle) === zsombi|lunch is now known as zsombi === dandrader is now known as dandrader|afk [13:20] nik90, shoot [13:23] tvoss: when the clock app is running a timer and is sent to the background by navigating to the dash etc, how does clock app signal that the timer is completed to the user? [13:24] tvoss: I read a small part of https://docs.google.com/document/d/1ij8RtPsR_eYMW3mys8Gu1Y2CVFZpjXdMpdIjIGZ1SCA/edit# [13:25] tvoss: where Zygmunt mentions that the clock app has to let the system know to trigger a notification in a certain amount of time which in turn will trigger a snap decision. [13:25] tvoss: am I right? If yes, any idea how the clock app sends a message to the system? [13:27] nik90, first of all: yes, the clock app needs to hand over to the system to complete this task. Second: libnotify should be your friend here. I don't know if we have a qml binding, yet. [13:27] thostr_ is the best person to ask here. [13:28] tvoss: but how does the clock app detect that it has been sent to the background by the user? [13:28] tvoss: is there some sort of platform api for that? [13:28] nik90, sure, your app is being informed of about to be suspended [13:28] nik90, you then have a grace period to clean up/setup notifications and such [13:29] tvoss: ok..do you happen to know the qml api for that? something like onSuspended: create notification ? [13:29] or where I can find the docs for that [13:29] nik90, yup, something like that [13:29] mhall119, can you help nik90 here? [13:30] nik90, sorry, but I rarely use the qml apis myself [13:30] tvoss: no worries === dandrader|afk is now known as dandrader [13:34] kalikiana, tvoss: to confirm, in qtubuntu-sensors we currently only support accelerometer and orientation, no light and proximity, right? [13:35] pitti, correct === alan_g|lunch is now known as alan_g === ara is now known as Guest58013 [14:03] ChickenCutlass: !! [14:04] greyback, ricmm_: hi guys, any progress on getting https://code.launchpad.net/~ricmm/unity-mir/sidestage-reenable/+merge/198489 merged in? [14:17] good day! [14:21] didrocks: who was it that was looking for the qmlscene crashes? We have some more of those showing up again [14:21] plars: earlier, it was Mirv [14:22] any updates on cm-11.0 based ubuntu touch?? [14:22] didrocks: also, I thought the dialer-app crash was supposed to be fixed, at least according to bug #1257844 [14:22] bug 1257844 in dialer-app "dialer app crashes during autopilot tests" [Undecided,Fix released] https://launchpad.net/bugs/1257844 [14:23] didrocks: but http://ci.ubuntu.com/smokeng/trusty/touch/mako/125:20140113:20140107.1/6032/dialer-app-autopilot/ seems to indicate otherwise [14:23] plars: we didn't get any new release of dialer app requested [14:23] hmm, I wonder why he marked it fix-released then?] [14:24] I guess we are waiting a landing ask from bfiller [14:24] plars: well, it's an upstream task [14:24] not dowstream [14:24] true [14:25] plars: this shouldn't have been marked fixed released, we have not figured it out yet [14:25] Mirv: there are a couple of qmlscene crashes in the maguro results from today if you are looking for some still [14:25] bfiller: ack, thanks [14:26] boiko: any progress on this? https://bugs.launchpad.net/ubuntu/+source/dialer-app/+bug/1257844 [14:26] Ubuntu bug 1257844 in dialer-app (Ubuntu) "dialer app crashes during autopilot tests" [High,In progress] [14:28] pitti: is it better from your perspective to implement a workaround in unity8 for the pygi regression (compared to just backout pygi changes?)? whats the real fix? [14:28] asac: if you need to fix it right now, you can add the workaround (it won't hurt after the fix) [14:29] asac: I need to catch simon feltman about this, but he's asleep, so will need to wait until tomorrow [14:30] pitti: so just reverting the pygi isnt feasible technically? point is that unity8 would have to land with a workaround and unity8 is pretty big beast [14:31] we can do that, but from outside perspective i see both options as valid and would choose whatever is less work :) [14:31] (assuming that we solve this problem upstream anyway) [14:31] asac: certainly it's feasible [14:34] pitti: so is the change/workaround in unity8 the right thing to do? if its the right thing anyway, i would go the unity8 route; otherwise only if its more work than a revert upload .. [14:34] * asac still wonders how python is hitting us at all in phone [14:34] thought we dont do python there :/ [14:34] yeah, I thought that, too [14:34] asac: currently testing a reversion of the corresponding commit in pygobject [14:35] cool [14:35] let me also ask in bug why we use python [14:35] or directly here [14:35] asac: autopilot? [14:35] oh :) [14:35] lol [14:35] asac, we obviously ship a lot more python since friday http://people.canonical.com/~ogra/touch-image-stats/20140110.1.changes [14:35] :) [14:35] ogra_: well, not relevant to the current discussion [14:36] asac, system-image uses python all over the place too [14:36] asac: the unity8 revert is risky, we are going to use latest trunk of unity8, and so, way more code [14:36] ogra_: on the device? [14:36] yes [14:36] didrocks: i know. we woul dhave to cherry pick land [14:36] asac: it's from unity8's emulators, so only for tseting [14:36] which we also can do of course [14:36] system-image-cli is all python afaik [14:36] yeah [14:37] shouldn't be actual functionality from the phone [14:37] right. but doesnt matter [14:37] but if this wreaks havoc, I can upload pygobject [14:37] once we have not 100% we cant land anything anymore as it might hide new regressions etc. [14:37] pitti: it would help a lot... we will soon have image == trunk, then it would be easier to land just cherry picks for unity8 and friends [14:37] at least the costs forunity8 workarounds should go down then [14:39] bfiller: nope, I am still trying to find a way to reproduce it [14:40] plars: any ideas how to help boiko reproduce/debug the issue with dialer? [14:40] bfiller, boiko: you don't see it by just running the dialer-app autopilot tests on mako? [14:41] plars: nope, I see another issue, but not this crash that causes the QUbuntu's cannot create application instance message [14:41] boiko: what issue are you seeing? [14:42] plars: so, there is one test for incoming calls, after the call is accepted (and the snap decision is dismissed), dialer-app gets focused again, and the screen turns black [14:42] this one I can reproduce all the time [14:43] boiko: I'm installing 125 locally right now, let me see what I can do [14:45] plars: ok [14:45] ogra_, so no more maguro builds? Effective when? Does this mean I don't need the results of your nested-mode testing on maguro? [14:46] mterry, ? [14:46] mterry, we might drop support for it but i dont see a reason why we should stop building it [14:46] comes with no extra cost [14:51] wow. is it intentional in the latest image that the phone vibrates hard on every tap? [14:51] ah not every tap, but many [14:52] ogra_, sorry, got disconnected. asac's recent mail to ubuntu-phone is what I was talking about [14:52] asac, err, why should we stop doing builds ... we can just stop doing tests, it costs us nothig to leave builds running [14:52] asac, i agree if they start to cause work we should probably drop them [14:53] Saviq, asac: pygobject uploaded [14:53] mterry, sorry, hadn't seen the mail [14:53] (i see it now) [14:53] timp: it is the start of haptic feedback [14:54] asac, didrocks: time for me to chat with my boss about getting an n4 then apparently meh [14:55] rsalveti, seen asacs mail ? [14:55] davmor2: right ;) [14:55] can you still get nexus 4s? [14:55] thanks pitti! [14:55] davmor2: good to have a start :) but I hope in the end it will be a bit more subtle :) [14:55] Laney: no [14:59] davmor2: yes, absolutely! [15:00] davmor2: there is a PO approved for something related... CC me on the mail you send to jfunk [15:00] asac: I'll talk it through with jfunk fisrt [15:01] ok tell him to talk to me :) [15:01] lol [15:02] ogra_: happy to explain to you the details in private conversation later today or tomorrow morning [15:02] ogra_: yup [15:03] * rsalveti checks backlog [15:03] asac, i just think we should keep them building and just drop them from the tools (phablet-flash etc) so people can still effectively fiddle around with them [15:03] i know that you think that. [15:03] :) [15:04] mterry: we still want to support it officially for a couple of weeks [15:04] mterry: do you have any issues with maguro specifically? [15:05] rsalveti, there is a bug in u-s-c that ogra_ found. I have a fix in hand, just needs to be tested. Was curious how much I should bother with it. But sounds like yes [15:06] rsalveti, both, grouper and maguro currently block the nexted mir mode [15:06] *nested [15:06] both require some work [15:06] right, we'll land the grouper fix probably later today [15:06] (hybris and some testing) [15:07] mterry: but yeah, please push the maguro fix forward [15:07] mterry, where are these debs, the MP doesnt have a link [15:08] hm [15:08] that's weird [15:08] did you get them in the mail ? [15:08] (i'm not subscribed to that one) [15:09] ogra_, no [15:09] weird [15:11] rsalveti, any updates on cm-11.0 based ubuntu touch ports? [15:11] or rather merging cm-11.0 framework to ubuntu phablet frameworks. [15:11] annerajb: got the first working image for mako (nexus 4), but still fixing some final issues [15:11] rsalveti, any public repos that i can start looking at? [15:12] annerajb: not yet, will upload it later today [15:12] still a bunch of diffs locally :-) [15:12] allright === vying is now known as Guest39595 [15:13] ogra_, well I guess just build it the old fashioned way. Doesn't take too long [15:14] nik90: what API do you need help with? === dandrader is now known as dandrader|lunch [15:19] tedg, bfiller: i proposed a branch for UAL which constructs the appid from the triplet [15:19] mhall119: the platform API to detect if the clock app is going to be suspended or not [15:20] mhall119: a little backstory..when the clock app is running a timer and is suspended, it is supposed to send the system a message that the timer is running [15:20] mhall119: but first I need to get a signal when the app is going to be suspended [15:21] kenvandine: nice [15:23] kenvandine, Ah, cool. I'll take a look, did you grab all the tests too? [15:23] i added tests === vrruiz_ is now known as rvr [15:26] Cool [15:27] nik90: ok, and where should that signal be? which component? [15:29] mhall119: in the timer page I guess, where if the signal is triggered it checks if the timer is running and if it does then do something special like sending a message etc [15:30] tedg, only think i don't like is it adds a click depends to the lib [15:30] kenvandine, Yeah, that should hopefully switch to a libclick depends here in a little while, which is better. [15:30] good [15:31] it's not enough to meet the need of url-dispatcher though [15:31] i don't think anyway... but it's a start :) [15:33] bzoltan1: what version are we going to call the SDK for 14.04? [15:34] nik90: I mean in the SDK APIs, where would you expect this to be documented? [15:34] mhall119: oh that..let me check [15:36] mhall119: I am guessing under platform services with a title "App Lifecycle" === vrruiz_ is now known as rvr [15:38] boiko: ok, I was able to reproduce it here at home too [15:39] nik90: hmm, I don't know if there is such a component [15:39] in QML anywya === alan_g is now known as alan_g|tea [15:44] cyphermox: bah, after scrapping my MR cus of failing on maguro, now we go and drop maguro support :) [15:44] the what? [15:44] oh? [15:44] you mean on the customized image or on all images? [15:45] cyphermox, https://lists.launchpad.net/ubuntu-phone/msg05889.html [15:47] dholbach: thanks, I hadn't looked yet [15:48] * cyphermox dives right into code, not email in the morning === alan_g|tea is now known as alan_g [15:54] tedg, any insight into why CI is failing for me? It's failing to generate the source package, can't find click.pm, but i didn't make any packaging changes and click-dev is a build dep [15:54] i'd assume this isn't a new problem... [15:54] kenvandine, Hmm, no I thought we'd fixed it. [15:54] kenvandine, Basically the build system can't install new dh hooks. They have to already be there. [15:55] kenvandine, So fginther needs to install the click-dev into the base image. [15:55] ah... fginther ^^ [15:55] Basically it fails on the original dh_clean before it installs the deps. [15:56] maybe this is fallout from the transition they did recently [15:56] kenvandine, He'll probably want a link to your failure :-) [15:56] fginther, https://jenkins.qa.ubuntu.com/job/upstart-app-launch-trusty-amd64-ci/61/console [16:01] dpm: any chance for the call for community help translating today? the dialer-app, messaging-app and address-book-app are almost completely untranslated, seems bad :( [16:01] kenvandine, got it, I'll take a look after my meeting [16:01] fginther, thx === dandrader|lunch is now known as dandrader [16:04] hm... since when does "adb poweroff" or "sudo shutdown -h now" no longer really shutdown a N10, but always cause it to reboot? [16:04] did the method for shutting a device down change recently? [16:05] elopio, hey, I just filed bug #1268640 (related to some code you manually merged in lp:ubuntu-calendar-app apparently), would you mind confirming it? [16:05] bug 1268640 in Ubuntu Calendar App "Autopilot tests (try to) stop maliit-server" [Undecided,New] https://launchpad.net/bugs/1268640 [16:05] try: sudo poweroff -f [16:05] ogra_, good lord... ok I'll try :) [16:05] (not sure if anything changed on N10, but that should alsways power off a device) [16:05] ogra_, but why would a simple command like poweroff need the force-option? :) [16:05] oSoMoN: can it wait ~1 hour? [16:06] elopio, sure, no urgency [16:06] ok, I'll be back with you. [16:06] MacSlow, it tells the kernel to power off immediately [16:06] (no shutdown process) [16:07] ogra_, ok [16:07] ogra_, thx [16:13] dpm: hello! [16:14] cwayne, sorry, on a call, I'll get to it tomorrow morning, this time for real! [16:14] hey sil2100 [16:14] dpm: https://code.launchpad.net/~sil2100/account-plugin-evernote/more_minor_fixes/+merge/201450 <- these are the final small fixes to packaging that are needed before we can release [16:15] dpm: after those are in, I'll test and publish ;) [16:17] sil2100, nice and easy to review, makes sense, approved. Are you taking care of the merge? I see you already did the latest merge. Otherwise I can look at it after the call [16:18] plars: how did you reproduce it? just running phablet-test-run? or anything different I have to do? [16:24] dpm: all is ok, thanks for approving! [16:24] super, thanks [16:31] oSoMoN: that's not my code, at least not while being awake. [16:31] I might be blamed because I touched everything to fix pep8. [16:31] but I've been working with the osk last week, so I can probably help you. [16:32] Hello there I am trying to use libstartup-notification in my cmakelists file and it can not find it did the name change in 14.04 or was it replaced by something ? thanks [16:33] I also can not get autopilot to be found by cmake :( [16:33] elopio, well I don’t have time to work on this myself, was just pointing out the problem and hoping you could acknowledge it [16:34] oSoMoN: yes, I agree with everything you say on the bug. I confirmed it. [16:34] I'll try to make some time to take a deeper look, but this week it's going to be hard. [16:34] elopio, cool, thanks, hopefully balloons can work on it [16:35] what I'll do this week is the text field emulator, hopefully with nice support using the OSK emulator. [16:36] so it will be a lot better for tests, they won't have to pay a lot of attention to the OSK. [16:36] * balloons wonders what is up [16:37] tvoss: do you happen to know a cmake+Qt project where a test uses signals/slots? [16:37] I'm currently fighting my way through those, trying to connect a sensor's readingChanged() to a slot in my test code [16:37] oh, this is so weird. This is my branch, merged as rev 168: https://code.launchpad.net/~elopio/ubuntu-calendar-app/autopilot-1.4/+merge/192631 [16:38] pitti, got a pastebin for me? === tvoss is now known as tvoss|test === tvoss|test is now known as tvoss [16:38] pitti, I am not him but unity8 does that [16:38] it has none of that code related to maliit. I suppose they added some code and merged it all together. [16:38] pitti, got a pastebin for me? [16:38] tvoss: if I just send this to a "normal" method, it never gets called; so apparently I need the Q_OBJECT+Q_SLOTS+moc stuff [16:38] tvoss: yes, sec [16:38] MyExHatesMeButMy: ah, good [16:38] pitti, you need automoc enabled [16:38] $ cat obj/tests/test_sensor_api_automoc.cpp/* This file is autogenerated, do not edit*/ [16:38] enum some_compilers { need_more_than_nothing }; [16:38] pitti, or you use QMetaObject::invoke [16:38] tvoss: that gets built [16:39] tvoss: http://paste.ubuntu.com/6745521/ [16:39] tvoss: I just added the Q_OBJECT and Q_SLOTS bits; without those, it compiles but the slot never gets called [16:40] yup, Q_OBJECT AND Q_SLOTS are needed unfortunately [16:41] pitti, do you know what is up with autopilot not getting founf by cmake in my app ? in cmakelists.txt ; include(autopilot) declare_autopilot_test(directions directions.shell ${CMAKE_SOURCE_DIR}/tests/autopilot/) [16:41] cmake says that it can not find autopilot :( [16:42] MyExHatesMeButMy: I'm afraid I've never seen autopilot being used like that [16:42] thanks for looking [16:42] MyExHatesMeButMy: sure you don't mean add_subdirectory() instead? [16:42] it is stange it use to work [16:43] pitti, I have that else-where [16:44] do you think that I should use find_package ? maybe === ricardodrosales is now known as ricdros [16:46] MyExHatesMeButMy: no, no; autopilot isn't a cmake/C thing, it's just a command that your test suite calls [16:46] << python n00b and knows nothing about autopilot just trying to add some tests to my app [16:47] fginther, can you help MyExHatesMeButMy === gatox is now known as gatox_lunch [16:47] fginther, he has got some issues with adding autopilot tests [16:47] * MyExHatesMeButMy goes back to the documentation but is lurking === dandrader_ is now known as dandrader [16:48] tvoss, sure, I'm in the middle of a meething, but will get back to MyExHatesMeButMy soon [16:48] cool fginther just ping me when you get some free time [16:48] I love this IRC channel ! [16:54] fginther, thank you, very much appreciated === dandrader is now known as dandrader|afk [17:00] MyExHatesMeButMy, I don't know much about cmake, so I might not be real useful but... autopilot is installed to /usr/bin/autopilot with the python-autopilot package. It's just an executable python script [17:02] I don't suppose anyone has gotten ubuntu touch running on the chumby? [17:03] may be incompatible, idk [17:03] thanks fginther yeah I myself know nothing about autopilot and am just learning. But it seems like all apps have it and I guess that My app should also . Only thing is I wish it was c++ lol [17:03] mhall119: it could be 1.1 or 2.0 ... or we can stick to the 1.0 with updates [17:07] MyExHatesMeButMy, lp:autopilot-gtk executes autopilot tests from cmake, you might give that a look [17:07] thanks fginther === dandrader|afk is now known as dandrader [17:26] Will Sidestage now work on 7" tablets ? given we wont officially support the Nexus 10 ? [17:27] bzoltan1: when will you decide on which of those it will be? [17:27] om26er, the new N7 has a fullHD resolution [17:27] so it should work there [17:28] ogra_, but I thought it was kept limited to certain screen size. Since lets say nexus 5 is a fullHD phone but clearly it wont have sidestage [17:28] iirc it was bound to resolution [17:29] I thought it was based on gridunit size [17:29] \o/ consensus [17:30] that is, how many gridunits wide the screen is, not how many pixels there are in a gridunit [17:31] why not just use QScreen ? [17:31] does not work on mir ? [17:31] or does qt-mir have its own class for screen info ? [17:32] http://qt-project.org/doc/qt-5.0/qtgui/qscreen.html#physicalSize-prop [17:34] anyone else in here using GPS in there app ? It takes forever for it to work. meaning that I have to wait like 15 minutes for it to start working on the phone. [17:35] shuts on and off like crazy also [17:39] bzoltan1: what is this? find /home/mhall/ -name *img [17:40] it grinds HDD every time I launch the SDK [17:40] example of code that takes 15 minutes for the GPS to kick in http://pastebin.com/RpCf2hxJ [17:40] mhall119, that looks for anything with the name *.img [17:40] MyExHatesMeButMy: I know that, I just don't know *why* [17:40] maybe find /home/mhall/ -name "*img" [17:40] and since I have a lot of files under ~/, it takes a long time [17:40] popey, https://myapps.developer.ubuntu.com/dev/click-apps/118/ [17:41] kenvandine, the missing cli-common-dev issue should be resolved. I've rebuilt https://code.launchpad.net/~ken-vandine/upstart-app-launch/appid_from_triplet/+merge/201433 [17:41] mhall119, what is it looking for a emulator ? [17:41] sergiusens: ack [17:42] fginther, thx [17:43] mhall119, I would figure that Qdir and Qfile would take care of all of that and would not run external command [17:43] sergiusens: approved [17:43] MyExHatesMeButMy: I have no idea, I just want it to stop :) [17:43] but who knows maybe that would slower but I doubt it. [17:43] bzoltan1, what is it for ? [17:44] bzoltan1, and why not QDir and QFile to find it ? === gatox_lunch is now known as gatox [18:03] hey mardy, just saw ssweeny's MR finally went through! Should we setup a landing ask, or should we wait til all 3 click-related MR's go through? [18:07] pmcgowan: https://plus.google.com/u/0/106250186908628485422/posts/WejtdUeFsBC === dandrader is now known as dandrader|bbl === nerochiaro is now known as nerochiaro|afk [18:13] mhall119, great, I expected someone would [18:14] pmcgowan: it seems (reading older posts in that community) that they're stuck needing 4.4 support too, what's the status of that? [18:15] mhall119, very soon, working in the lab kind of thing [18:16] * annerajb anner can't wait for 4.4 to start on LG G2 Series [18:21] pmcgowan: is there anything we can release to people like them who are trying to port to a device that needs it? [18:23] mhall119, rsalveti would have the latest, I think its close [18:24] mhall119: pmcgowan: in progress, hope to be done by the end of this week [18:26] rsalveti: can you let that G+ community (linked above) know when it's available for them to try with their port? [18:26] mhall119: sure, everybody will know :-) === VargaD_ is now known as VargaD [18:39] plars: so, any suggestions on how to reproduce the dialer-app's crash on smoke tests? [18:42] boiko: I ran the provision script like we run in the lab, which really just combines all the steps you need like phablet-flash, the network setup, etc [18:42] boiko: then I just ran the jenkins.sh script to reboot the device and run the dialer-app tests [18:43] plars: where can I get this jenkins.sh script? [18:43] boiko: are you unlocking by hand and running autopilot directly I guess? [18:43] boiko: lp:ubuntu-test-cases/touch [18:43] plars: yep, launching the tests using phablet-test-run on desktop [18:43] boiko: let me try it like that also [18:44] plars: thanks for the link, I will check it out [18:49] boiko: ok, I got it by running phablet-test-run dialer_app also [18:49] hmm [18:50] boiko: dialer-app version I have is 0.1+14.04.20131209-0ubuntu1 - same as you? [18:51] plars: same version [18:52] plars: what output you get when running the tests? [18:52] boiko: none, just that it ran 6 tests in 72.400s and OK [18:52] boiko: I'm didn't run verbose there [18:53] boiko: in the lab, you can see all the output clicking on the tests - the tests all pass though === ricardodrosales is now known as ricdros [18:53] boiko: so perhaps it's just crashing at the end? [18:53] plars: no idea :/ [18:54] boiko: it does just end up at a black screen at the end - or is that just the other issue you described? [18:54] plars: that might be the other issue I described [18:54] plars: in the output it says it is due to a known Mir crash [18:57] boiko: I'm watching the directory a bit more closely this time [18:57] for the crash file [18:58] boiko: does seem to show up right at the end [19:04] boiko: with -v, and watching for the file to show up, it looks like it happens toward the end of the test_incoming test, around the time it hangs up. I see the incoming call, and hear the ringtone start to play before it hits hangup and that's about the time I see the crash file show up [19:12] plars: so, the screen turning black happens right after the snap decision is dismissed (when clicking accept), and the focus get back to the app [19:33] boiko: any chance that's all related to the crash we're getting? [19:33] boiko: the timing of it is certainly in line [19:35] plars: maybe, but I'm not totally sure yet [19:35] plars: just testing something with real calls before switching back to debugging this === dandrader|bbl is now known as dandrader [20:23] slangasek, sergiusens, beuno, popey: I changed click-sync.py from lp:click-sync into: get-all-clicks.py as: http://paste.ubuntu.com/6746725/ real quick [20:24] lool: ObNamingBikeshed: clearly this should be named get-ALL-the-clicks.py [20:24] ack [20:25] lol [20:26] get-_ALL_-the-clicks.py please [20:26] http://paste.ubuntu.com/6746744/ [20:26] 401 here [20:26] popey, you need oauth credentials ;-) [20:27] bah [20:27] * popey files a bug against lool [20:28] how do I get that? === salem_ is now known as _salem [20:28] hello everyone [20:28] popey, lool, nessita here can help you get those credentials [20:29] (if you give her context :)) [20:29] may I please have some context? [20:29] sergiusens: the auth credentials didn't work :-( [20:29] I get a 401 too [20:30] nessita: hey, trying to get some credentials to pull all free click apps from the app store with a local client [20:30] nessita: apparently we have one such setup already with a format like http://paste.ubuntu.com/6746701/ [20:30] lool: are you hitting staging or prod? [20:30] lool, hmm, well you just need to generate some; let me try again; they may have accidentally expired [20:31] lool, sergiusens: that dict of credentials has empty strings, which are not valid (I guess you already found out that much :-)) [20:31] nessita, he wants those fields filled with your help ;-) [20:31] sergiusens: will generate some [20:31] lool: you need to grab credentials from the phone itself, or from an desktop box? [20:31] nessita: desktop [20:32] but I think sergiusens has pointed out a script that might help [20:32] let me try that [20:32] lool: staging or prod? [20:32] so I provide the proper command [20:32] prod [20:33] ok, building the curl query [20:34] sergiusens: bah trying to install the click-toolbelt thing I ended up with errors in easyinstall of "stevedore" [20:35] lool, just wait for the magic curl line [20:36] oauth is black magic to me [20:37] hum, network hiccup [20:37] lool: replace email and password with values from an existing account https://pastebin.canonical.com/102880/ [20:37] nessita1: thanks! [20:38] lool: hum, let me confirm the url [20:38] I tried it IRL and got an error [20:38] it says staging [20:38] sergiusens: remove it! :-D [20:38] of course my password has a single quote [20:40]

There's no page with this address in the Ubuntu One service. Check that you entered the address correctly and try again.

[20:42] lool: yeah, missed a /oauth suffix. New command, IRL tested: https://pastebin.canonical.com/102882/ === nessita1 is now known as nessita [20:44] lool: did it work? [20:44] nessita: looking at your new URL [20:45] if you are considering using your canonical account it will fail requiring 2fa, let me know and you can send the next otp via the query as well [20:45] "message": "2-factor authentication required.", [20:45] "code": "TWOFACTOR_REQUIRED", [20:45] nessita: Yup, OTP would be nice :-) [20:46] lool: add "otp": "123456" to the json encoded dict sent as data in the POST [20:47] of course 123456 has to be the next valid otp for you [20:47] worked, thanks [20:47] anytime [20:48] lool: is this something you need to automate or is this procedure enough for your needs? [20:50] nessita: it's fine for now; thanks! [20:51] slangasek: so with https://pastebin.canonical.com/102882/ + "otp": "123456" in the request, I get credentials that I filter down to these http://paste.ubuntu.com/6746701/ [20:51] slangasek: then this get-all-clicks.py script http://paste.ubuntu.com/6746884/ works with: mkdir d && ./get-all-clicks.py --credentials-file ~/.click-credentials d [20:52] feel free to rename it to get-\*-clicks.py [20:52] lool: no way to get these without having to pass passwords on the commandline? [20:53] now i get TypeError: __init__() got an unexpected keyword argument 'openid' [20:53] lool: seems to not be working for me, anyway [20:53] * lool ran this on his laptop in a "sh" instance to avoid shell history [20:54] popey: you have to filter down to the list in http://paste.ubuntu.com/6746701/ [20:54] slangasek: which part doesn't work for you? [20:54] lool: "message": "Provided email/password is not correct.", [20:54] "code": "INVALID_CREDENTIALS", [20:55] slangasek: that's with your login.u.c credentials? [20:55] yes [20:55] I had to escape my password as it had special chars [20:55] ah, thanks [20:55] hope we don't kill the store, all beating it up like this ☻ [20:55] slangasek: BTW there is no actual <> around the email or password [20:55] popey: if we do it's a good test [20:56] lool: yeah, I didn't put any there [20:56] alan@deep-thought:~/all_the_clicks$ ls -l *.click | wc -l [20:56] 26 [20:56] working here [20:56] popey, lool, I welcome the scalability testing ;) [20:56] it's about 54M of downlaod [20:56] 50M of that is one click [20:56] the bitcoin one [20:57] popey: ah note that I dont see UK-only apps [20:57] I do see the FR-only apps though [20:57] :-P [20:57] beuno pointed out that this is geofiltered [20:57] be interested to know how many we each get [20:58] 97 [20:58] * lool quickly uploads 10 webapps [20:58] for 10 French websites [20:58] * popey uploads webapps depicting french presidents in compromising situations [20:59] slangasek: I guessed you used an actual 2fa OTP code? [20:59] yes [20:59] slangasek: no idea what else could be in play; perhaps nessita can help [20:59] dodgy characters in password? [20:59] no comment [20:59] :P [21:00] (there are no escaping issues in my command ;) [21:02] * lool goes off for dinner [21:03] lool, popey: so instead of having all of us fight with credentials, does one of you want to post the batch where I can grab them? [21:04] slangasek: sure [21:06] * lool uploads [21:07] http://people.canonical.com/~alan/click/ [21:07] 100 total [21:09] done [21:09] popey: ta [21:09] np [21:10] bah [21:10] my upload bandwitdh sucks! :-) [21:15] popey: 100 total? http://people.canonical.com/~alan/click/click_list is 98 lines long [21:16] my bad [21:16] i was doing ls ☻ [21:16] didn't realise there were not just click files in there [21:16] alan@deep-thought:~/all_the_clicks$ ls -l *.click | wc -l [21:16] 98 [21:31] lool, add a space in front of the command and it won't go into history with bash at least [21:34] where's the right place to get qt5.2, e.g. as a build-dep for ubuntu-ui-toolkit? [21:37] barry, probably one of the qt5 edgers ppas [21:38] I think Mirv sent out and email wrt [21:38] sergiusens: cool, thanks. i'll check the archives [21:44] barry: Apparently https://launchpad.net/~ubuntu-sdk-team/+archive/ppa but it looks like "ubuntu-ui-toolkit 3 days ago Failed to build: amd64 armhf i386" might be problemmatic [21:45] genii: yeah, that's the one i'm trying to build locally. looking [22:12] Hello [22:12] I just saw that development on Ubuntu Touch for tablets is going to be focused only on the 2013 Nexus 7 [22:13] About how long will it be until it's possible to get a build of Ubuntu Touch for it? [22:20] slangasek, popey lool there are no armhf packages here: http://people.canonical.com/~alan/click/click_list [22:21] it's not 'all clicks' per se :-) [22:22] the query itself doesn't return them either [22:22] sergiusens: yep i do have 9apps, i only see 8 of them [22:23] sergiusens: well, er, that's decidedly unhelpful for what I'm trying to do then [22:23] slangasek, I imagined it was about that; reason for checking [22:25] beuno, is there a query that we can ru that would return the list of arch specific packages? [22:25] sergiusens, you can filter by arch, yes [22:25] let me find the query parameter... [22:26] or a !all :-) [22:29] maybe JamesTait is around? [22:33] slangasek: any particular package you're expecting to see? [22:33] popey: no, I don't know anything about what is or isn't arch-dependent in the app store [22:34] oh okay [22:34] bitcoin is an arch dependent one [22:34] and its not in https://search.apps.ubuntu.com/api/v1/search?q= [22:34] but it's the arch-dep ones that I need to look at :) [22:36] sergiusens, https://search.apps.staging.ubuntu.com/api/v1/search?q=architecture:armhf [22:36] https://search.apps.staging.ubuntu.com/api/v1/search?q=architecture:i386 [22:36] and https://search.apps.staging.ubuntu.com/api/v1/search?q=architecture:amd64 [22:36] would be !all, I think [22:37] ty [22:37] slangasek: ok, modded script, am getting all the armhf ones too, will upload those to the same place [22:37] popey: cheers :) [22:38] On the GalaxyNexus running Ubuntu Touch, the camera app can only take pictures (not video)... Is this a UI issue or is video recording not supported by the underlying platform / libraries? [22:51] 156 [22:51] thats more like it [22:54] popey, \o/ [22:54] popey, for the sake of it; have you seen any amd64 or i386 packages? [22:55] sergiusens: not tried.. lemme see [22:56] slangasek: lots in http://people.canonical.com/~alan/click/ now [22:56] popey, just in case you haven't noticed; make sure the staging part is removed :-) [22:57] i did [23:00] popey, no i386 or amd64 :-) [23:00] not many armhf either [23:00] http://paste.ubuntu.com/6747470/ [23:00] slangasek, ^^ [23:00] some seem improperly added as such [23:01] yeah, none [23:02] lool, I'd say it's safe to call for a recompile for framework-14.04 [23:02] or whatever it's going to be called [23:07] popey: thanks, snarfing [23:49] rsalveti, any luck with uploading cm-11.0 hardware libs? === hb is now known as Guest42571 [23:54] weird