=== kalikiana_ is now known as kalikiana [01:03] Hi all... I was looking for documentation on the python application indicators api, but all links seem to be broken [01:03] does anybody know where I can find these references? [01:04] example of a broken link: http://developer.ubuntu.com/resources/technologies/application-indicators/ === chihchun_afk is now known as chihchun [04:13] balloons, looks like Victor beat me to it, you seen the issue he has? https://code.launchpad.net/~nskaggs/music-app/fix-1292044/+merge/211637/comments/516258 [06:55] good morning [07:04] popey: I don't know who to tell but the download option for 64bit mac on the ubuntu site gives a 404 [07:19] balloons: I don't know who to tell but the download option for 64bit mac on the ubuntu site gives a 404 [07:20] popey: the fix for different results on i386 wheter or not the registry allocator is used or not is upstream. Should probably ask for a distropatch to Mirv https://codereview.qt-project.org/#change,83936 [07:22] tsdgeos: awesome.. Mirv ^^ ☻ [07:23] tbh i haven't tried it applies to our 5.2.1 packages [07:23] Mirv: tell me if you want me to do that or not [07:29] tsdgeos: it says no file to patch.. src/qml/jit/qv4regalloc.cpp [07:31] tsdgeos: chaging jit -> compiler, seems to apply thanks [07:31] popey: tsdgeos: LP bug to refer to would be nice [07:32] this is the 4th qtdeclarative patch in pipeline now [07:32] Mirv: sure. [07:36] Mirv: where shall I file it, we only have an upstream bug at the moment [07:37] popey: https://bugs.launchpad.net/ubuntu/+source/qtdeclarative-opensource-src [07:41] Mirv: bug 1312571 [07:41] Launchpad bug 1312571 in qtdeclarative-opensource-src (Ubuntu) "Simple JS has differing results on AMD64/i386" [Undecided,New] https://launchpad.net/bugs/1312571 [07:48] ah, this's the one Roman e-mailed me about too [07:48] I just didn't get an upstream pointer [07:59] hello - I put the UbuntuPhone image #303 on my tablet Nexus10. The scrolling is very slow / jerky / flickering ! [08:04] morning popey, when you've got a minute, could you review https://code.launchpad.net/~dpm/reminders-app/click-everywhere/+merge/217175 ? It's the same thing as for file manager yesterday [08:07] dpm: sure [08:08] awesome, thanks [08:11] dpm: i may be missing a build-dep? In file included from /tmp/click-everywhere/3rdParty/libthrift/Thrift.cpp:20:0: [08:11] /tmp/click-everywhere/3rdParty/libthrift/Thrift.h:48:39: fatal error: boost/utility/enable_if.hpp: No such file or directory #include [08:12] popey, did you add the build deps as described in the MP? If you did, I'll have to figure out if there is any other one missing [08:12] yes [08:12] well, i already have them [08:13] you already had libboost-dev and libssl-dev? [08:13] dpm: let me play more, [08:13] yes, but i skipped straight to 2) to test click-buddy before qtc [08:13] lemme do them in the right order ☻ [08:13] ok [08:17] it looks like it that file comes from libboost-dev, so I'd double check if it got indeed installed: [08:17] # dpkg -S /usr/include/boost/utility/enable_if.hpp [08:17] libboost1.54-dev: /usr/include/boost/utility/enable_if.hpp [08:18] Successfully built package in './com.ubuntu.reminders_0.5.106_armhf.click'. [08:18] 09:18:40: The process "/usr/bin/click" exited normally. [08:18] ^^ from qtc [08:18] 09:18:40: Elapsed time: 01:32. [08:20] and the package works on-device [08:21] and the package builds from click-buddy too [08:22] dpm: not convinced by this... click-buddy is surely only making a click out of something that's already been built by qtc? [08:22] so no compilation happens there if you run 2) right after 1) [08:22] (which won't happen in the DC) [08:24] popey, click-buddy should rebuild the package. I was told yesterday that it cannot use Qt Creator's build directory - and in fact, with the command line args passed as described on the MP, if you don't tell it where the QtC build dir, then there is no way click-buddy can guess [08:25] popey, you can pass click-buddy the --no-clean arg and it should show you the build dir it used [08:25] it took seconds [08:25] there's no way it built it [08:25] paste.ubuntu.com/7328276/ [08:26] paste.ubuntu.com/7328278/ [08:27] can you build it if you cleanly branch that MP? [08:27] using click-buddy alone [08:27] sure, doing a clean checkout now [08:29] it is building now [08:31] hmm [08:31] * popey tries this too [08:32] popey, yes, click-buddy built the code and created the package successfully on a clean checkout [08:32] same here.. how did that work [08:33] is it re-using the armhf chroot by qtc? [08:33] yes, it does use the same chroot [08:33] regarding the build dir, though [08:33] this is news to me [08:34] This all voodoo to me ☻ [08:34] so I'm guessing it's reading some sort of cmake cache and detects that there is a build dir already, and uses the built files from there to create the click [08:34] but that's not a bad thing, as it just speeds up click package creation :) [08:34] well yes [08:42] popey, here's another easy one https://code.launchpad.net/~dpm/ubuntu-filemanager-app/update-readme/+merge/217183 [08:42] it's just an update to the README file [08:42] happroved [08:49] \o/ [08:50] popey, I think we should be good to go do start the machinery to do a file manager upload, and soon after a reminders upload [08:51] mzanetti, would you mind reviewing https://code.launchpad.net/~dpm/reminders-app/click-everywhere/+merge/217175 when you've got a minute? [08:51] dpm: sure [08:51] dpm: will it build in jenkins? [08:52] great, thanks! [08:52] popey, it should do. Jenkins does a similar thing as click-buddy does. Although I don't know what actually triggers the click build, do you? [08:53] I think fginther or someone on *-ci-eng might need to flick the switch to build a click? [08:53] i can trigger the build [08:54] ah, cool [08:55] * popey does that now your stuff is in trunk [08:55] awesome [08:55] it already built [08:55] \o/ [08:55] * popey tests on device [09:07] dpm: hmm... clicking the maintain button in QtCreator for the click chroot opens a terminal here but closes it down again immediately. Did you see that too? [09:07] i didnt [09:07] it opened here [09:08] mzanetti, no, I haven't seen that, it usually works for me, I just tried it again. Perhaps zbenjamin has got an idea what it could be? ^ [09:08] seems to print the usage for click in that terminal. but I can't read it all, it closes down too fast [09:09] mzanetti, is it a recent click chroot with the ubuntu-sdk-14.04 framework? [09:09] yeah, just generated it right now [09:09] Or something else you could do might be to update the chroot [09:10] ok, then in that case it's up to date [09:10] let me update my chroot to see if I can reproduce [09:11] /usr/bin/python /usr/bin/x-terminal-emulator -e click chroot -a armhf -f ubuntu-sdk-14.04 -s trusty maint /bin/bash [09:11] thats what it runs [09:11] for me ☻ [09:12] popey: where did you find this? [09:12] ps aux while it's running [09:13] I updated the chroot, but I cannot reproduce [09:13] hmm... running it manually (without python in the beginning) seems to work fine here too [09:14] mzanetti, perhaps running qtcreator from the command line might give a clue? [09:15] never seen that before .. [09:16] nope... doesn't print anything... [09:16] sounds like the click chroot args are maybe wrong [09:16] yeah. seems so. where can I see what it's trying to use? [09:17] mzanetti: check if it prints a error the general messages tab [09:17] but i guess it does not because the process starts [09:17] zbenjamin: nope. not a single line [09:20] dpm: I left a few comments [09:21] mzanetti, thanks. Re: 12 + configure_file(manifest.json ${CMAKE_CURRENT_BINARY_DIR}/manifest.json) [09:21] I would have preferred to use the .in convention too, but we need to leave it as manifest.json, so that QtC can pick it up [09:21] that's actually the main change of this branch [09:27] mzanetti, I think I addressed your comments. Would you mind having another look at https://code.launchpad.net/~dpm/reminders-app/click-everywhere/+merge/217175 and approving/rejecting as appropriate? [09:32] dpm: btw. the patch for the sorted list performance issue has been accepted upstream [09:33] mzanetti, yeah, I saw that, and Mirv packaged it up, thanks! [09:33] purrfect! [09:43] Mirv, now that U has opened, where do we stand with https://code.launchpad.net/~timo-jyrinki/kubuntu-packaging/qtdeclarative-for-u-series-multiple-fixes/+merge/216821 ? I saw you committed a fix today, is the MP looking good, or do you foresee any complications? [09:44] popey, the reminders' click can also be triggered from Jenkins and tested and uploaded to the store :) [09:46] nik90, when you've got a minute, do you think you could address the last comment on https://code.launchpad.net/~nik90/ubuntu-weather-app/improved-toolbar-icons/+merge/216098 ? I think it should be an easy fix. Thanks! [09:47] dholbach, as a packaging expert, do you think that last comment on https://code.launchpad.net/~nik90/ubuntu-weather-app/improved-toolbar-icons/+merge/216098 makes sense? [09:48] suru-icon-theme is only available on 14.04 and onwards [09:48] dpm, yep, that should work [09:48] cool, thanks for confirming :) [09:57] dpm: testing will show the status, currently known is that it builds. but no silo build to test yet, since U is not really open yet (toolchain updates should be done today) === _salem is now known as salem_ [09:59] ah, I see, thanks Mirv [09:59] dpm: ok === salem_ is now known as _salem [10:17] zbenjamin, I'm trying to set a breakpoint and debug lp:ubuntu-filemanager-app, using the Desktop kit. However, when I press the Debug button in QtC, I get this message about it not being a debug build: http://i.imgur.com/OOKQ7bJ.png - how can I set up a debug build in QtC? [10:18] dpm: did you add -DCMAKE_BUILD_TYPE=debug? [10:19] nope, I guess that answers my question :) [10:19] dpm: you can clone your buildconfig and name it Debug, so you can switch between them when needed [10:19] yeah, that's exactly what I did :) [10:19] testing it now [10:28] zbenjamin, that works well, thanks. Would it make sense to create a Default and Debug build config? Right now it's not too obvious how to debug, and it'd make it easier for developers to be able to start debugging straight away rather than creating new configs [10:29] dpm: we could do that yes [10:34] ok, reported it as bug 1312644 [10:34] Launchpad bug 1312644 in qtcreator-plugin-ubuntu "Make it easier to debug C++ apps with a Debug build config" [Undecided,New] https://launchpad.net/bugs/1312644 [11:43] dpm, so, continuing the conversation here, if I want to run music-app on the 14.04 desktop, the right service is mediascanner, not -2 [11:43] ara, correct [11:43] dpm, OK, thanks! [11:44] ara, if you install the music-app package from the core apps PPA it should pull the right dependencies for you. Did you install that, or did you try to run the app from source? [11:44] dpm, from universe, and it didn't install the dependencies [11:45] dpm, sorry, from the PPA [11:45] hm, weird [11:45] dpm, and it didn't install the service [11:45] * dpm looks at debian/control [11:45] (nor it did run it) [11:45] dpm, I was planning to file a bug [11:48] ara, sounds like a valid one if you want to file it. I see that music-app depends on grilo-plugins-0.2-mediascanner, but that does not seem to be enough to pull in mediascanner [11:49] dpm, https://bugs.launchpad.net/music-app/+bug/1312684 [11:49] Launchpad bug 1312684 in Ubuntu Music App "[desktop] Music should install and run automatically mediascanner when installing/run" [Undecided,New] [11:56] dpm, is music-app using gstreamer 0.10? [11:57] ara, I know it works with 0.10, but I'm pretty certain it does with 1.0 too. It's not a feature of music-app, though, rather of Qt [11:57] mmm, it does not work for me with 1.0 :) [11:57] on touch 1.0, on desktop 0.10 [11:57] ah, thanks sergiusens [11:57] ah, that's why [11:58] dpm, but mediascanner does find mp3 with only 1.0 installed [11:58] we for qtmultimedia-opensource-src to support 1.0; but that brought in an incomplete port for video [11:58] we could afford to break that in touch, but not desktop [11:58] that is going away with media-hub [11:59] that problem, although this cycle is the cycle for media-hub on desktop [12:03] ara, commented on your bug. Feel free to review the proposed branch too :) [12:06] done [12:08] dpm, thanks, do you want me to file a separate bug for the second point? [12:09] ara, it would probably be best, yes, but we should probably ask Satoris if he's around first [12:10] I can't see him or jamesh online [12:10] dpm: ara on my laptop mediascanner _is_ launched, and I didnt do that manually [12:10] perhaps it only launches on new login? [12:11] popey, that might be it [12:11] popey, why didn't this fix the mediascanner dependency? https://bugs.launchpad.net/music-app/+bug/1241403 [12:11] Launchpad bug 1241403 in mediascanner "Music app's grilo plugin package needs to depend on mediascanner" [Undecided,New] [12:12] oh, that should have definitely have fixed it [12:12] dpm, it looks like the grilo plugin hasn't been rebuilt? [12:13] ahayzen, indeed, it seems to be out of date [12:13] could undo the change you did and get qtgrilo rebuilt? [12:13] let me see [12:13] popey, looks like jenkins disproved dpm's MP [12:13] good old jenkins for once does one thing right :P [12:14] hah [12:14] ☻ [12:15] dpm, popey, although the point about it autorunning i'm less sure about [12:16] dpm, popey, currently you have to run mediascanner-service ~/Music after you have installed it [12:18] Is there anything special I need to do to open the Reminders App in the SDK? Maybe I'm missing something, but I'm not seeing a *.qmlproject file like there is in many of the other apps.. [12:18] gerlowskija: open the cmakelists.txt file [12:19] gerlowskija, check out the README file too, it should tell you everything you need to know about it [12:20] gerlowskija, it's a mixed C++ and QML app, so it's a CMake project in Qt Creator, not a pure QML (.qmlproject) one [12:20] ahayzen, popey, it's weird, it seems I've got the latest grilo package installed (0.0.20130610-0ubuntu3), but it does not have the mediascanner dependency -> http://bazaar.launchpad.net/~music-app-dev/qtgrilo/trunk/changes/15?start_revid=15 [12:21] dpm, magic [12:21] aha! [12:21] ahayzen, popey look at r14 on that link :) [12:22] dpm, that is 20130610...Victor's branch was 2013-11-14 [12:22] seems like the mediascanner dep addition got reverted [12:22] ahayzen, yeah, Victor did the change on r11, but it got reverted on r14 [12:22] dpm, ah [12:23] dpm, popey: Ok, makes sense. Thanks guys. [12:24] Mirv, do you remember why you dropped the mediascanner dependency on http://bazaar.launchpad.net/~music-app-dev/qtgrilo/trunk/revision/14 ? We'll need to re-add it, as otherwise there is no music scanning going on [12:24] dpm: I didn't drop it, I just synced to the branch what had already been uploaded to archives [12:25] popey: Also, when I was looking around for info about the Reminders App, there's a wiki page linked to from a few different places that doesn't seem to exist: https://wiki.ubuntu.com/Touch/CoreApps/RemindersApp. Idk, if it was moved or deleted intentionally, just wanted to mention it in case it hadn't been brought to your attn. [12:26] dpm: so this upload in September https://launchpad.net/ubuntu/saucy/+source/qtgrilo/0.0.20130610-0ubuntu2 didn't include that line. [12:26] Mirv, right, but the dependency effectively was dropped, so I'm wondering whether there was a specific reason. Can we just re-add it? [12:27] dpm: it does not look like it ever was in a released version. so I don't think there was other reason than unability to do new uploads if debian/ is out-of-sync with archives. [12:27] gerlowskija, thanks. That should read https://wiki.ubuntu.com/Touch/CoreApps/Reminders (without the *App) [12:28] gerlowskija, where did you find the broken links? [12:28] dpm: so feel free to add it back, but just check it's really enabled when it's added there in the build dependencies (so something checks during build there's mediascanner and then adds dependency to runtime deps?) [12:29] thanks Mirv, but I'm not sure I follow the thing about doing an extra check for dependencies. Can't I just add it to the list of deps in debian/control and that should be it? [12:30] dpm: https://launchpad.net/reminders-app and https://blueprints.launchpad.net/ubuntu-phone-commons/+spec/reminders-app-development [12:32] fixed in both places, thanks gerlowskija! [12:43] gerlowskija, let us know how you're getting on with getting Reminders up and running, we welcome any help, so if you're interested, please feel free to pick up any of the open bugs at https://bugs.launchpad.net/reminders-app [12:45] Thanks dpm, that's actually what I'm working on. I grabbed a bitesize one to start off with (https://bugs.launchpad.net/reminders-app/+bug/1289347) and then we'll see how it goes! [12:45] Launchpad bug 1289347 in Ubuntu Reminders app "[desktop] No icon available for "Add notebook", "Add note" and "Add reminder"" [Medium,Confirmed] [12:45] gerlowskija, awesome! Let us know if you need any help [12:46] dpm: I mean, in the diff you quote, it doesn't depend on mediascanner from runtime dependencies, only from build time dependencies. so if the mediascanner is installed during build, did it automatically affect the build so that mediascanner is used? [12:47] dpm: but sure if it works so that during it build it checks if mediascanner is installed and acts accordingly, then that's the way to go. [12:48] Mirv, good point. I don't think it needs it during the build, I think it should be in the runtime deps [12:58] dpm: I'm getting this error in the SDK when I try to start the app (I'm on trusty desktop btw): [12:59] Reminders File: qml/reminders.qml does not exist at any of the standard paths [12:59] Looks a lot like this: https://bugs.launchpad.net/reminders-app/+bug/1303763 [12:59] Launchpad bug 1303763 in Ubuntu Reminders app "File: qml/reminders.qml does not exist at any of the standard paths!" [Critical,Fix committed] [12:59] gerlowskija, just a sec, I've noticed this same error on File Manager and I'm fixing it now [12:59] it seems to affect the desktop only [13:08] gerlowskija, I think that should get you running: http://paste.ubuntu.com/7329706/ [13:33] dpm, I have opened a new bug about running mediascanner automatically on music-app start up [13:33] https://bugs.launchpad.net/music-app/+bug/1312739 [13:33] Launchpad bug 1312739 in Ubuntu Music App " [desktop] Music should run automatically mediascanner on start up" [Undecided,New] [13:42] ok, thanks ara [13:54] balloons, here's one for you: I've noticed file manager did not run on the desktop, and as part of the fix I changed the installation path from ubuntu-filemanager-app to just filemanager. Tests run well on my desktop, but they fail on Jenkins: https://code.launchpad.net/~dpm/ubuntu-filemanager-app/fix-desktop-run/+merge/217239 [13:54] I can see why from http://91.189.93.70:8080/job/generic-mediumtests-trusty/2214/testReport/junit/ubuntu_filemanager_app.tests.test_filemanager/TestFolderListPage/test_show_file_properties_with_mouse_/ [13:54] 13:38:00.923 INFO testcase:258 - Attempting to launch application '/usr/bin/filemanager' with arguments '-q /usr/share/ubuntu-filemanager-app/qml/ubuntu-filemanager-app.qml' as a normal process [13:55] it should be launching /usr/share/filemanager/qml/ubuntu-filemanager-app.qml [13:55] do you know if this command line is hardcoded anywhere in Jenkins? [13:56] I don't quite get why all tests pass on my desktop but they fail in Jenkins [14:04] DanChapman has so what happens when you try and build trojita? [14:06] popey i've just noticed there is the UbuntuSDK for armhf kit but it's saying it's sdk__13.10_saucy_default is that right? i'm on trusty? [14:10] mine's 14.04 kit here. [14:20] popey, ok so i'm getting http://pastebin.ubuntu.com/7330100/ where do i add modules to the CMAKE_MODULE_PATH [14:20] * DanChapman needs to figure out why it says saucy 13.10 aswell [14:23] DanChapman: in qtc, tools -> options -> Build & Run.. you can manage your kits there [14:29] dpm, there are 3 ways of launching things via autopilot [14:29] deb, click, and local [14:30] you need to make sure all three launch correctly. Jenkins uses the deb method, while you are likely launching via the local install on the desktop [14:30] ok [14:30] the phone of course uses click [14:30] where are these 3 ways of launching defined? [14:32] dpm: ideally in the ui toolkit :-P at least elopio once commented he would move it there; today in setup http://bazaar.launchpad.net/~ubuntu-filemanager-dev/ubuntu-filemanager-app/trunk/view/head:/tests/autopilot/ubuntu_filemanager_app/tests/__init__.py#L63 [14:32] as an example [14:33] aha! Thanks sergiusens [14:33] dpm: as you can see, first thing that evals to true is used [14:34] dpm: I hope we drop the debs this cycle with the advent of unity8 :-) [14:36] \o/ [14:36] yes, it's a real pain to maintain support for both click and .deb on core apps [14:37] deb is already pretty much dead. [14:42] hah, you wish [14:42] 14.04 has 5 years support friends! ☻ [14:43] popey: but the core apps aren't in main nor the archives, right? [14:43] no, but in a ppa that PES expect to continue working [14:50] wha?! [15:03] balloons: not necessarily the ppa we currently have, but _a_ PPA [15:16] dpm: Belated thanks for the "qml/reminders.qml not found" workaround earlier. Is that something that should be committed, or was that just due to some quirk on my machine? [15:18] gerlowskija, it should be committed, I just noticed it now. Feel free to submit a fix [15:19] sure. [15:21] dpm, you still need help landing https://code.launchpad.net/~dpm/ubuntu-filemanager-app/fix-desktop-run/+merge/217239 [15:21] ? [15:22] balloons, I think that now I know where to look I should manage, thanks! But I'm done with core apps today, so it might take a while for me to come back to it [15:23] if you don't care then I'll just finish it [15:23] I'd like to push fm to the store and this shouldn't be lingering [15:23] I assume popey also wants to do the same :-) [15:24] dpm ^^ [15:24] already done [15:24] file manager is updated in the store, reminders needs to be [15:24] balloons, popey is always one step ahead of us ;) [15:24] popey, the tests passed fine for you? :-) [15:24] yup [15:24] awesome... [15:24] file manager ones did [15:24] not tried reminders yet [15:24] on my list [15:25] for reminders, I need elopio to finish this: https://code.launchpad.net/~elopio/reminders-app/test_go_to_accounts2/+merge/214163 [15:25] he needs one small tweak to finish.. it works fine [15:25] balloons, yeah, if you want to fix filemanager on the desktop, feel free to branch off from my branch and submit a new MP or if you point me to your new branch I'll merge from there [15:26] balloons: when do you think that'll happen? [15:26] since popey has what he needs, I'll leave it.. simpler that way [15:26] I do? [15:26] popey, it's a 3 character change.. I'm hoping he does it right now :-) [15:27] feel free to poke me if you need any buttons pressing [15:27] I'm good at that [15:27] the crucial part is knowing which ones of course. less good at that bit [15:27] hehe.. reminders will definitely land. I'll work an mp to fix the build and merge his mp for him if he doesn't get to it [15:28] i'll push to the store and ping you sometime after your dinner I'd guess :-) [15:32] dpm: not sure if you're still around, but wanted to give you a pointer to https://code.launchpad.net/~gerlowskija/reminders-app/app-dir-standard-path/+merge/217262 [15:40] gerlowskija, approved, thanks a lot! [15:40] You did all the work haha, I just put my name on it. [15:41] a team effort, then :) [15:41] balloons, which build needs fixing? [15:48] dpm, reminders has the same issue as fm.. I'll do the same fix you did to it [15:49] balloons, which issue? The not running on desktop one? It seems to work for me, although I'm not sure if the package in the PPA is a recent one [15:51] dpm, no reminders also does not build https://bugs.launchpad.net/ubuntu-filemanager-app/+bug/1308280 [15:51] Launchpad bug 1308280 in Ubuntu File Manager App "Can't build with phablet-tools" [High,Fix committed] [15:51] does the QML RotationSensor work? I thought it did at some point, but I'm not getting readingChanged events on mako/r303 [15:52] I saw you update it, but did you push something to fix it? [15:52] hah, you did ;-) [15:52] balloons, it does build now. I fixed it this morning. That's the click package that popey intends to upload [15:52] filemanager is already uploaded to the store [15:52] you didn't link it.. but I should have checked [15:53] rev 163 [15:53] popey, yes, we were talking about reminders [15:53] good stuff [15:53] oh, sorry ☹ [15:53] balloons, I linked the branch indeed :) [15:54] mind you, the wrong one :) [15:54] actually, no, it was the right one and you even reviewed it :) [15:55] lol [15:55] no matter, I'll get elopio [15:55] summary: so we're looking good now for those two :) [15:55] 's merged and do the mini fix as another mp [15:55] balloons, cool, looking forward to seeing more reminders tests! [15:55] yep.. that's why I want to land it :-0 [15:55] balloons, did u see Victor's comment on your MP? [15:56] ahayzen, I did.. kind of odd, I haven't looked into it yet, but I will in a second [15:56] ahayzen, did it/does it work for you? [15:57] balloons, i'll try it in a minute just literally submitted two bits of coursework lol [15:57] ok guys, calling it a day and starting the weekend, see you all next week, have a good one! [15:58] dpm, cheers, thanks for fixing the builds! [15:58] np, cheers! [16:09] balloons: hello! [16:09] elopio, :-) [16:09] everything is all set, I approved your mp [16:09] balloons: I thought that the rename of the click packaga was something you did on your branch. [16:09] elopio, it was, so I just merged both [16:09] oh, cool. Thanks. [16:09] I wasn't planning on merging my branch, that's why I wanted you to do it [16:10] anyways, we're good now ;-) [16:10] balloons: if my click-buddy still doesn't work to provision it, where should I report the bug? [16:10] reminders or click-buddy? [16:10] reminders [16:12] ok, I'll give it a try in a few. [16:12] mardy: I need more help from you with accounts. Do you have time? [16:23] 33 [16:23] oops [16:30] popey, could you verify this calendar bug? It seems as though something in a recent revision caused the app to look zoomed in when pushing to the device: https://bugs.launchpad.net/ubuntu-calendar-app/+bug/1312480 [16:30] Launchpad bug 1312480 in Ubuntu Calendar App "Calendar app from trunk installed on the phone seems to run in wide mode" [Undecided,New] [16:35] vthompson: how are you running it from qtc? [16:36] popey, no deploying to the device via click-buddy. I suppose you might be able to verify via qtc by resizing [16:38] popey, actually resizing seems to work just fine. [16:40] ah okay, will test with click-buddy [16:41] vthompson: confirmed [16:43] popey, ok, I might try to figure out what happened as the bug makes it almost impossible to test any future fix on the device [16:47] vthompson: and you're sure it's not a calendar change? [16:47] very little has changed in the image in a week [16:47] hey vthompson, just looking at the feedback you left on my mp.. I'm not getting that issue [16:48] but, I do wonder if you need to push the toolkit to your device again. That's my guess [16:49] adb push /usr/lib/python3/dist-packages/ubuntuuitoolkit /home/phablet/autopilot/ubuntuuitoolkit [16:49] popey, I assume it's a calendar app change that caused it... [16:50] balloons, how does that normally get updated? [16:51] vthompson, phablet-test-setup should be doing it [16:51] balloons, ok [16:52] I need to bug sergio again, or maybe just propose a fix.. I've seen other dependencies needing to be pushed too === gatox is now known as gatox_lunch [16:55] balloons, /usr/lib/python3/dist-packages/ubuntuuitoolkit doesn't exist on my system [16:55] that'd probably explain it [16:57] well adb push /usr/lib/python2.7/dist-packages/ubuntuuitoolkit /home/phablet/autopilot/ubuntuuitoolkit would push the py2 version, which you have [16:57] the desktop runs ap via python2, but the default on phone is python3 [16:58] you just need to push the updated package to the device.. it's the same whether py2 or py3 (since the toolkit is compatible with both) === jhodapp is now known as jhodapp|walk [16:59] balloons, I installed music-app-autopilot which installs ubuntu-ui-toolkit-autopilot. The directory seems to be there now. I'll retest [17:01] balloons, so what is "phablet-test-setup"? I don't see such a script [17:02] phablet-click-test-setup [17:02] phablet-click-test-setup, sorry :-) [17:03] popey, balloons ah. What should be calling phablet-click-test-setup? I've never explicitly ran it before. [17:03] Does anyone know if this wiki page on Platform Services is up to date? https://wiki.ubuntu.com/Touch/CoreApps/Blockers [17:03] gerlowskija: its not, wassup? [17:05] vthompson: paste.ubuntu.com/7331155/ [17:06] popey, ok. I had assumed --provision in click-buddy did that [17:06] provision does the dbus thing only [17:06] popey: I thinking of taking a look at https://bugs.launchpad.net/reminders-app/+bug/1273311. The description says it's waiting on the Notifications stuff to land. (I clearly don't know anything about the notifications stuff, so I was just going off the wiki ;-p) [17:06] Launchpad bug 1273311 in Ubuntu Reminders app "Add Alarm API integration in reminders" [Medium,Triaged] [17:06] vthompson: oh, also i do an adb shell rm /home/phablet/autopilot to clean up [17:07] gerlowskija: so it's possible to create alarms, both the clock and calendar do that, via EDS, that has landed. whether they trigger or not is another matter. so yes, part of it could well be implmented now, maybe use clock/calendar as reference? === zoopster_ is now known as zoopster [17:08] ok, I'll take a look, thanks for the pointer. [17:09] popey, thanks! [17:09] balloons, she's running now. [17:09] vthompson, awesome.. [17:10] \o/ [17:10] * popey goes afk for a bit [17:11] vthompson, I agree.. provisioning should do this [17:13] balloons, also shouldn't the ubuntuuitoolkit come from the phone and not the development box? [17:13] What if the dev box isn't running the latest? [17:14] vthompson, it the setup step grabs the dependency from source [17:14] I was just telling you to push from your box :-) [17:14] ah. :) [17:15] i'm going to propose a merge.. that should get some action on landing this so click-buddy does it ;-) [17:16] balloons, cool that'd be awesome [17:16] sometimes doing something poorly gets others to fix it :-) [17:17] but perhaps they liked what I propose [17:17] elopio: sorry, I was away; please ping me back on Monday (but at least a couple of hours earlier :-) ) [17:22] mardy: I will. [17:23] balloons: I was trying to be extra-clever with the evernote account: [17:23] https://code.launchpad.net/~elopio/reminders-app/test_with_account/+merge/217171 [17:24] but that doesn't work, I think it's missing the authorization that's done through the evernote website. [17:24] elopio, ohh account time? :-) [17:24] if we don't find a way without the UI, we'll be blocked as there's no nice way to introspect the evernote webview. [17:24] elopio, so we should chat quickly about that.. I did it by faking oachineuth, including the cookie on the local ma [17:25] *machine [17:25] balloons: that's what I wanted to ask mardy, I didn't know you already did it :D [17:25] I still have the mp for that. however, evernote also allows use of a token, which we have a test version of [17:25] * elopio wants. How do I do that? [17:25] elopio, I think we're better off doing the token [17:26] yes, it's the only way I see. [17:26] the oauth hackery was indeed a lot of hackery [17:26] elopio, let me give you the token [17:27] balloons: I have a token [17:27] I just don't know how to tell online accounts to use it. [17:27] elopio, ohh do we still have to use online accounts.. it's been so long since I looked at this [17:27] mmm, maybe not. [17:27] here's what I remember doing.. and let me find the branch [17:27] I don't know. [17:28] I used a test oauth account I created with evernote.. I then manually generated an auth token and cookie on my local box [17:28] then, inside the test, I had it use the same account, and re-setup the token and cookie [17:29] I spoke with the online accounts guys and there is no other way it seems to do it.. [17:29] sounds good. I can create an evernote account and not validate it. I have the token. [17:29] with the token, we should be able to avoid online accounts completely I think [17:30] because evernote allows you to use the a key (sorry I keep calling it token, which is confusing with oauth) instead of doing oauth [17:32] elopio, perhaps it's easier if you just read the site: http://dev.evernote.com/doc/start/python.php [17:32] look under authentication === gatox_lunch is now known as gatox [17:32] bah, they call it a developer token :-) [17:33] balloons: yes, I read about it. What I don't get is how to tell the reminders app to use that token. [17:33] elopio, your mp isn't merge.. I think you need to rebase with trunk: https://code.launchpad.net/~elopio/reminders-app/test_go_to_accounts2/+merge/214163 [17:34] here's the old work I think https://code.launchpad.net/~nskaggs/reminders-app/oauth-ap/+merge/204107 [17:34] balloons: ok, merged with trunk and pushed. === jhodapp|walk is now known as jhodapp [17:37] elopio, I think that's where I ended up as well.. Since the app uses oauth, it was unclear if it was possible for us to use just the token [17:38] so, looking at my old hackery, I see I basically recreated the cookie, signon, and secrets db online accounts uses, rather than invoking it [17:39] balloons: I was hoping for a nice method of gi.repository.Account to link the account I added with a token, so the validation is not needed. [17:39] I have no idea if that's doable. I'll leave it for now, and on Monday I'll ask mardy if we have to hack the dbs, or there's an alternate way. [17:40] elopio, when we asked back in Jan, we couldn't do it.. heck, I believe I have a bug on it. let me see [17:40] remember him telling us about account-console? I still have the log :-) [17:41] https://bugs.launchpad.net/account-plugins/+bug/1270264 [17:41] Launchpad bug 1270264 in Online Accounts: Account plugins "Accounts Console autopilot support" [Undecided,New] [17:43] balloons: yes, what I'm doing on my credentials.py is basically what account-console does. [17:46] but I was thinking that what session.process does is to get the token and link it to the account, so reminders doesn't have to go to the evernote page to validate every time. [17:47] that's the step that fails in my branch. So we could skip it and do the link ourselfs. [17:47] but I really don't know what I'm talking about... [17:48] elopio, as far as I understand, this is what needs to happen [17:48] 1) you need an account on evernote's server [17:48] 2) you need to go through a verification on evernote's server (oauth handshake) to allow access for an app to your account [17:49] 3) Once complete, all api calls from that app just work (if you have the token and cookie) [17:49] 4) the access lasts by default for 1 year but can be revoked / managed / increased in evernote [17:50] So, if you do part 2 manually and then pass those keys around to wherever you need to run the test it works until you revoke or the access expires from evernote [17:51] yes, I get it. But the way you pass the manually generated token to the app is throught the database. [17:51] right? [17:56] You need a physical file (the cookie) if I remember correctly. The db stuff is so online accounts has what it needs [17:57] ok, I'll try adding the cookie to what I have now. [17:59] elopio, I basically did the auth and monitored everything that happened.. the network traffic, file writing, db updates, etc [17:59] pretty fun black hat stuff [17:59] :) [17:59] if you want I can try taking the old branch and using it against what you have now.. or you can have at it [18:01] so do you think this is the way we have to go / should go? There's nothing in my mind that can fake the authentication on evernote's servers, so we'll always be doing some of this cooie passing [18:06] balloons: I think that we should use the developer token. I don't know if this is the easiest and cleanest way. [18:07] and at some point, we should use the production servers. This won't work and we will need to instrospect the webview, so it's good to start thinking about it too. [20:07] I'm building something that uses gtk, and I'm calling printf [20:07] but it's not showing up on the screen [20:07] is there anyway to attach a terminal to the output stream of an arbitrary process in order to see what is being printed [21:03] SuperNoeMan: do you have a newline at the end of your string? the stdio functions may buffer their output, adding a newline is a good way to encourage it to happen faster.. [21:03] sarnold: I'll just write to a file [21:04] SuperNoeMan: don't forget the fflush() if you want the output to happen 'quickly' [21:08] sarnold: one question [21:08] while writing to a file i get a seg fault [21:09] why might the following code segfault in a checkbutton callback: [21:09] FILE * fp = fopen("/test.txt", "w"); [21:09] fprintf(fp, "happened"); [21:09] fflush(fp); [21:09] fclose(fp); [21:10] SuperNoeMan: your application does not have privileges to open the file /test.txt -- / is writable only by root -- and so you get a NULL pointer back from fopen(). [21:10] SuperNoeMan: when you try to fprintf() to a NULL filepointer, you get the segv [21:10] oh silly me [21:10] ok... is home writable? [21:11] SuperNoeMan: probably your app is confined in a way that forbids writing to the user's home directory. try just taking off the '/', I'm curious if the current working directory is writable -- I half expect it is.. [21:11] well I don't know exactly where the ... ok [21:13] cool [21:14] SuperNoeMan: did fopen("test.txt", "w"); work? [21:14] yes [21:14] nice [21:14] It printed to a file [21:14] SuperNoeMan: it's not exactly perfect but this wiki page ought to go a long way to explaining the confinement for apps: https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement [21:15] (you'll probably want to skim it heavily, at most :) [21:15] nice, so now my checkbox function callback works [21:15] I need to call python from C... I wonder what hte best way to do that is... [21:49] how can I edit an autogen or configure script to include new dependencies; I need a new include directory, and new library to link to, and I also need to specify that I have a package requirement === jhodapp is now known as jhodapp|afk [22:12] Is there an analogue of sni-qt for qt5?