=== duflu_ is now known as duflu === duflu_ is now known as duflu [05:00] not sure who are around today, not the US folks at least, but filed a bug about autopilot tests for the unity(7) stack bug #1219636 [05:00] bug 1219636 in Unity "unity autopilot tests fails to move mouse, leading to infinite test run" [Critical,New] https://launchpad.net/bugs/1219636 [05:01] bregma: ^ (when you're around, it seems that bug is here to stay for now ie. not random) [05:14] Mirv: I'm approaching EOD but I can take a quick look. To confirm, if I had a VM to try and reproduce on I would add the ppa 'ppa:ubuntu-unity/daily-build' and dist-upgrade ? [05:16] veebers: yes, that's the correct PPA [05:16] Mirv: cheers [05:40] Mirv: hey, re: the test and the mouse move, was there a video for that failing test? [05:44] Mirv: Failing me being able to re-produce the bug, there should probably be checks in autopilot for infinite mouse movement loops (I thought there was already) [05:44] if not today, I'll propose something for autopilot adding that [06:19] veebers: no video, I just checked that not even with the one run that ran for 7200 seconds until it self-aborted (http://10.97.0.1:8080/job/autopilot-saucy-daily_release/1483/label=autopilot-intel/) [06:19] (and sorry, some flakiness in Internet connection today [06:25] Mirv: nw, cheers [08:22] seb128: hi [08:22] seb128: what do you think about my reply on the mail? [08:22] mzanetti, hey [08:22] mzanetti, we already have a way to customize defaults, that's https://launchpad.net/ubuntu/+source/ubuntu-settings [08:24] seb128: so we 'd need a second version of this package for touch, right? [08:26] mzanetti, if we want different defaults on touch and ubuntu yes [08:26] mzanetti, but I though the goal was convergence, having one image, same device being a phone and desktop once docked [08:26] seb128: I really need that for the phone [08:27] seb128: so you're saying we should either have libreoffice on the phone's launcher or remove it from the desktop one? [08:27] yes [08:27] what happen to your phone once docked [08:27] do you expect libreoffice to appear then? [08:27] and what about the dialer app? have that in the desktop launchers all the time? [08:28] well, the issue your describe there is not as easy as "default" [08:28] you want to launcher content to change depending on the screen resolution [08:28] and the hardware available [08:28] and the docking status [08:28] the launcher content* [08:29] you don't want libreoffice on your phone launcher [08:29] but once you dock that phone it becomes a desktop [08:29] seb128: no... I don't want the launcher's content to change when the screen resolution changes [08:29] then you want it [08:29] so you converged desktop wouldn't have libreoffice when used in desktop mode? [08:29] converged device* [08:29] seb128: no [08:30] why not? [08:30] that seems buggy [08:30] well, it's there and can launch it, but not in the launcher as pinned app [08:30] unless I pin it myself in there. [08:30] why not? we have it on the desktop today as pinned [08:30] that seems like a desktop regression [08:30] seb128: I only want to ship that config on phones [08:30] there is no "phones" [08:31] there is [08:31] there is a convergence story [08:31] on day your phone is going to be a desktop once docked [08:31] one day* [08:31] yes.... but still we ship it as a phone [08:31] and there we don't want libreoffice in the launcher [08:31] with a desktop... [08:32] if the user goes home, docks it and drags office to the panel. fine [08:32] the desktop should be full featured and similar to our current desktop once used as a desktop [08:32] you are making desktop users second class citizen there [08:32] saying "let's make the desktop experience less good, if they want they can fix it themself by hand" [08:33] no... if you install your pc at home you will have that stuff in there [08:33] but you won't have the phone app [08:33] that's not really convergence [08:33] that's having 2 OS, 2 images [08:33] not 2 images [08:33] that's not what we are aiming for [08:33] just 2 different sets of configs [08:33] default configs [08:33] right, so 2 images [08:33] one with a desktop config and one with a phone config [08:33] yes... we weill have multiple images anyway [08:34] that's a fail if we get there imho [08:34] but anyway, that's not for v1 in any case [08:34] we already have multiple images now. 4 for the desktop + 4 for the phone btw [08:34] no, we have 1 for the desktop [08:34] on different archs [08:36] why not ship 1 image, with multiple configs available, including configs for each platform we support, and the config is dynamically chosen by shell at runtime [08:36] mzanetti, well, anyway, I'm pretty sure some people have been asking for dynamic configs key, where the value change according to the form factor [08:37] mzanetti, e.g Saviq asked for gsettings key where you can have a value for phone mode and one for desktop mode and where the value you get changes once you dock the device [08:37] but that's another discussion [08:37] *that* is violating convergence imho [08:37] mzanetti, let's start with the ubuntu-touch-settings [08:38] mzanetti, the problem with those -settings approch, is that they conflicts, you can't install 2 overrides, you have to pick 1 (or 1 win at the end anyway) [08:38] so you that defeats convergence goal as well [08:38] mzanetti, can you send me a list of keys you want changed, I'm going to upload that ubuntu-touch-settings today [08:45] mzanetti: hey, for tablet, since we can have 2 applications open simultaneously and both have "focus", the launcher will then indicate both apps are focused, right? [08:45] seb128: sent you a mail [08:45] greyback: I'd say yes [08:45] mzanetti, thanks [08:45] greyback: are they really both focused? [08:45] mzanetti, no background image change? [08:45] ? [08:46] seb128: which background? [08:46] mzanetti, org.gnome.desktop.background picture-uri [08:46] unity wallpaper image [08:46] greyback: I don't think we need to change that one, do we? [08:46] we currently rely on the code fallback, because the default image is invalid [08:46] err... seb128^ [08:46] mzanetti: not really. Keyboard input only goes to one [08:47] greyback: yeah. that's what I meant [08:47] mzanetti: it does maybe introduce a tiny bit of confusion, as the model can have 2 "focused" applications, but yet there's only one "focusedApplication" [08:48] greyback: not an issue imho... need to state it clears in the api docs tho [08:48] clearly [08:49] mzanetti: yep that'd work. Was considering renaming the model role to "active" just to make clearer divide between the concepts [08:49] seb128: hmm... do I need to know background image thingie today? I guess it would require shipping the image with the other background pictures etc [08:50] mzanetti, no, that's fine, let's start by just the launcher config [08:50] greyback: would be fine with that too. [08:50] seb128: thanks [08:50] yw [08:50] mhr3, hey [08:50] seb128: you will just install some override for it, right? [08:50] mzanetti, correct [08:50] awesome [08:51] mzanetti, same as ubuntu-settings is doing on the desktop [08:51] with a different config [08:51] * mzanetti doesn't really know much about dconf [08:51] don't worry about it, it's going to just work ;-) [08:52] that was my plan :) [08:56] pstolowski: ok, i can repro some of the problem in my testcase and it's LVWPH specific (at least this one) [08:56] pstolowski: having a look [09:02] * greyback has errand to run, back in about an hour [09:02] seb128, morning [09:04] mhr3, hey, how are you? [09:06] seb128, not bad, not bad, but you know... monday... need to kickstart yourself :) [09:06] mhr3, hehe, don't tell me! [09:06] mhr3, I've a bug for you, maybe that can help to kickstart ;-) [09:07] mhr3, my lenses are showing untranslated in unity7 ... is that a known issue? [09:10] seb128, all of them? [09:10] i was actually doing some po fixes in home scope [09:11] so i wonder if i actually broke it [09:11] seb128, or was it not working for a while? [09:12] tsdgeos: ok, thanks for update! [09:15] mhr3, yeah, none is translated it seems, and dunno when that started, I tend to not be too picky about having english on my screen [09:15] mhr3, where are the strings supposed to come from? [09:16] mhr3, e.g "gettext -d unity-lens-applications "Installed"" returns a french string [09:16] but it's showing in english in the UI [09:16] seb128, home scope [09:18] mhr3, but applications.scope has Domain=unity-lens-applications [09:19] seb128, but what you see is coming from the master scope, and that is defined in home-scope [09:22] mhr3, ok, I guess that's the problem then, those strings are not translated there [09:22] seb128, yey, i just successfully delegated stuff to translator team :) [09:23] lol === jamesh_ is now known as jamesh [10:18] mzanetti, https://launchpad.net/ubuntu/+source/ubuntu-settings/13.10.4 [10:19] mzanetti, just installing unity-touch-settings should give you the launcher defaults [10:19] mzanetti, wget https://launchpad.net/ubuntu/+source/ubuntu-settings/13.10.4/+build/4926146/+files/ubuntu-touch-settings_13.10.4_all.deb if you want to try it === mpt_ is now known as mpt [10:25] pstolowski: ok, i've found the two problems, there was one in LVWPH and another in GenereicScopeView [10:26] the one in GenereicScopeView i'm not sure i've fixed properly yet [10:26] tsdgeos: awesome [10:47] larsu: good morning. [10:56] greyback: what do you think where the desktop file parser should be located? [10:56] greyback: or better: where do you need to use it from? [10:57] mzanetti: I use it in unity-mir, in the application manager. [10:57] mzanetti: separate library with it's own package would be nicest IMO [10:58] probably, yes... [10:58] Wellark: ^ [11:01] greyback, mzanetti: we would not want to introduce gazillion of small libraries, though.. [11:01] on their own packages that is [11:01] greyback: could it live in ubuntu-mir? [11:02] all the other projects I know are splitting their code into gazillions of small libraries right now [11:02] above should be "on their own *source* packages" [11:02] binary packages are ok [11:04] Wellark: it doesn't quite belong to unity-mir, as reading desktop files not a mir specific request. And other tools like ubuntu-software-center might want to be able to read desktop files maybe? [11:06] dednick: hi! [11:09] larsu: hi. just wanted to follow up on that messaging action issue i was having. [11:11] dednick: ah right, I applied that but never commites/pushed. Sorry about that. [11:12] dednick: https://code.launchpad.net/~larsu/indicator-messages/remove-root-parameter/+merge/183426 [11:19] larsu: cool. Anything about why reply isnt working? [11:20] dednick: that's next on my list, currently trying to figure out #1217676 [11:22] larsu: that may be something in the unity8 side. [11:23] larsu: i think the qmenumodels unity8 "flattening" model is resetting when you add-remove messages. [11:23] could be what's causing it [11:25] larsu: a problem which should be sidestepped in the unitymenumodel version. [11:26] dednick: yeah I had the feeling this could be it, but I'm not seeing it with qmenumodel when running on my desktop. [11:27] dednick: I just made sure though, the service definitely doesn't clear the items. Let me try with your branch [11:28] larsu: i'm not sure the clear will work with my branch (i think it may suffer same problem as reply) [11:29] larsu: it will remove the item from the list, but may not actually call into the backend. [11:30] dednick: heh, I guess I'll be looking into that issue first, then :) [11:30] larsu: see what i did there? ;) tricked you into doing that first. haha [11:31] haha, clever ;) [11:33] seb128: I've tested your package. works fine [11:33] mzanetti, great [11:33] seb128: is this already in the image? or when will it land? [11:33] mzanetti, it needs to be seeded [11:33] that means in terms of timeline? [11:33] > a week? [11:33] mzanetti, no, today [11:34] ah. perfect [11:34] I'm pinging ogra to get seeded, then it's going to be on the next image build [11:36] seb128: I'll put you on the reviewers of this: https://code.launchpad.net/~mzanetti/unity8/launcher-defaults-from-dconf/+merge/183420 [11:36] seb128: once the defaults have landed you can approve [11:37] Wellark: I think I'm done: https://code.launchpad.net/~mzanetti/unity8/launcher-defaults-from-dconf/+merge/183420 [11:37] mzanetti, ok [11:40] mzanetti, so resetting the launcher config (we need to do that in system setting) would be basically resetting the gsettings key and calling that dbus command? [11:40] seb128: yep [11:41] seb128: it might slightly change once saviq is back, but I'll keep you updated [11:41] mzanetti, thanks === hikiko is now known as hikiko|lunch [11:51] dednick: your branch doesn't build for me: dh_install: unity8 missing files (usr/share/unity8/unity/indicators/*), aborting [11:51] dednick: the build directory contains a working unity8 binary, but running that doesn't show the messaging menu at al [11:51] *all [11:52] larsu: um. give me a sec [11:53] dednick: oh, it does! It's the icon that is missing ) [11:53] larsu: :) [11:55] larsu: the dh_install should be fixed now. pull latest [12:00] dednick: thanks [12:01] om26er: hi [12:02] mzanetti, hello [12:02] om26er: just had a jenkins failure on the phone again but recalled that we already had seen this: https://jenkins.qa.ubuntu.com/job/generic-mediumtests-runner-maguro/402/console [12:02] mzanetti, yep, internet problem it seems [12:02] om26er: I think we should use network over usb [12:03] we could check with other guys what they are using to connect to the internet in the lab. e.g. in utah. I directly call nmcli they might be doing something different since I never heard any problem from them [12:04] mzanetti, I was hoping that change would get merged in phablet-tools but didn't seem to happen, I do know there was a branch for that but never merged [12:04] om26er: well... wifi is still wifi... for tests I'd recommend going through usb [12:04] om26er: unless explicitly testing wifi [12:05] om26er: this should explain how that works: losca.blogspot.com/2013/05/network-from-laptop-to-android-device.html [12:05] mzanetti, I'll try that locally here and see how it goes [12:06] om26er: actually you don't even need all that stuff... just set the PC as the default gateway on the phone and enable masquerading and ip_forward on the PC [12:06] mzanetti, translate that into command please :) [12:07] om26er: first of all you need to make sure TCP/IP through adb === alan_g is now known as alan_g|lunch [12:08] om26er: adb forward tcp:2222 tcp:22 [12:08] for example [12:09] ah no... that won't work [12:09] it only does port 22 obviously [12:09] we need all of them [12:09] Mirv: ping [12:10] om26er: I guess you need all that howto. [12:11] mzanetti, right, experimenting with that an a bit of the web [12:11] om26er: let me know if you have troubles. also, Mirv is the one that wrote this howto [12:11] there is this post as well http://qa.ubuntu.com/2013/07/24/internet-over-usb-on-ubuntu-touch-devices/ [12:12] om26er: yeah... looks mostly the same, except Mirv uses more low-level commands [12:12] dednick: hm, I'm not even seeing HeroMessage.onItemRemoved being called... [12:12] dednick: well, if "print('..');" does what I hope it does ;) [12:13] larsu: erm. give me a sec [12:13] mzanetti, regarding Mirv script isn't that giving a static IP which will rather result in a IP conflict between devices on the same network [12:14] om26er: yeah, well. I guess needs to be adjusted a little for our szenario. [12:15] larsu: i dont suppose you ship that script to add messages on desktop with the indicator? [12:16] dednick: no, mine is in a constant state of change. But you can use bfiller's, it's attach to bug #1217676 [12:16] bug 1217676 in unity8 (Ubuntu) "performing action on menu incorrectly clears and reloads menu" [High,Confirmed] https://launchpad.net/bugs/1217676 [12:16] greyback: pong [12:16] dednick: you might need to change the desktop files in there to other apps if you don't have the phone apps installed [12:17] larsu: ok. thanks [12:18] Mirv: hey. I was looking into the qtubuntu fail to build on x86. I see that the contents of qtbase/src/platformsupport/eglconvenience are not being compiled into qtbase for x86, yet the header files are being shipped. I think eglConvenience is useful on desktop too. [12:18] larsu: mine is outputting. does your message not dissapear? [12:18] larsu: try using use console.log("...") [12:21] dednick: same problem. I also don't see the Activate() call on the bus... [12:21] dednick: the message disappears [12:22] larsu: if the message dissapears then the onItemRemoved should be called. The Activate might not be going out because the action doesnt exist. [12:22] greyback: ok. the configuration hasn't changed from 5.0.x, but I wonder if that's a necessary evil from using -opengl desktop instead of -opengl es2 in configuration [12:23] larsu: did you ./build after you made the qml change? [12:23] the plugins need to be built as they get copied into the build folder during compile. [12:24] dednick: ya, I just noticed that (I'm in a build area and thought I wouldn't need to rebuild in there) [12:25] Mirv: quite possible. Things that need the qeglConvenience stuff are QPA plugins (like qtubuntu), which are platform specific. You cannot enable both GL types in the build? [12:26] greyback: the -opengl switch at least is either or, no possibility of defining both [12:27] larsu: still not getting the log statements? you're not greping for the output are you? [12:27] katie: hi [12:27] dednick: I'm getting them now that I'm editing the right files :) [12:27] larsu: ah :) [12:29] katie: can you please give an answer here when you have a minute: https://code.launchpad.net/~mhall119/unity8/finish-tour/+merge/183037 [12:30] Mirv: I'm very curious how/why this changed since 5.0. But it must be a bug that the header files for it are being shipped [12:33] greyback: yeah. that might be a bug indeed, upstream though, as --fail-missing makes sure that anything upstream installs gets into the packages [12:36] mzanetti, sure.. just looking now :) === hikiko|lunch is now known as hikiko [12:42] dednick: found the issue (UnityMenuModel::activate() doesn't look at action-namespace). Making a fix now. [12:43] dednick: do you have an idea why it takes 10-20s to start unity8? [12:43] Mirv: ok well then all qtubuntu can do to work-around it is include it's own copy of those eglConvenience cpp files. Which isn't a nice solution really. Any other ideas? [12:45] dednick: what do I need to test your indicators.accesspoint-design branch on the phone? it's just empty here [12:49] greyback: not much. there's some -eglfs switch that we don't define but is enabled by default, and we install egl/es headers also on x86. but the build logs say that eglfs support is still not enabled on x86 - I'm not sure why, but it might be simply because of the -opengl desktop which the automatically disables it. [12:49] buildlog https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta2/+build/4922638/+files/buildlog_ubuntu-saucy-amd64.qtbase-opensource-src_5.1.1%2Bdfsg-2ubuntu1%7Esaucy1%7Etest5_UPLOADING.txt.gz search for "EGLFS" [12:50] (install egl/es headers = build-depend on ubuntu's egl/es packages also on x86, so they'd be available during build) [12:52] Mirv: if egl isn't being included, then yeah the eglfs (that's an example QPA for Qt to render to fullscreen egl surface) won't be built. [12:53] Mirv: though I'm still puzzled that the headers are installed, as QT_CONFIG does not contain "egl" and so those headers should be left out. (see qtbase/src/platformsupport/eglconvenience/eglconvenience.pri) === alan_g|lunch is now known as alan_g [12:56] larsu: nope. dont know why. doesnt on my box [12:57] dednick: this works for me, can you try as well? lp:~larsu/qmenumodel/use-full-action-name [12:57] mzanetti: er. have you got some old indicator files on your machine/device? [12:57] dednick: gotta run for a bit, will be back in 15 min or so [12:58] mzanetti: ok, will give it a try [12:58] dednick: dunno... I flashed the phone on friday evening [12:59] mzanetti: is indicator-messages-service running? [12:59] mzanetti, the issue that you reported is actually something different than we were seeing in the past. the logs say, Error: NetworkManager is not running. [12:59] mzanetti: sorry, indicator-network-servce [12:59] previously that has not been an issue. probably we should increase the delay to 90seconds for the device to settle [13:00] mzanetti: arg. indicator-network-service [13:00] dednick: it works normally... if I run_on_device your branch I get file:///home/phablet/shell/builddir/plugins/Unity/Indicators/IndicatorPage.qml:111:5: Type MenuItemFactory unavailable [13:00] om26er: never use sleeps :D [13:00] mzanetti: reason given for that? [13:00] om26er: otherwise it'll fail when we'd require the timeout to be 91 seconds [13:01] mzanetti: UnityActionMenu not a type? [13:01] om26er: can't you do a while [ `pinc -c 1 ubuntu.com` -ne 0 ]; do sleep 1; done? [13:01] or rather UnityMenuAction . i seem to be half asleep today [13:01] s/pinc/ping/ [13:02] dednick: http://paste.kde.org/pef088624 [13:03] mzanetti: you need to upgrade your qmenumodel. and i need to update dependency version. [13:03] om26er: obviously adding some safety mechanism... like wait at max 2 minutes, otherwise fail the test suite [13:04] dednick: ok. flashing my phone now... that should update it, right? [13:04] mzanetti: yeah [13:04] um. possibly [13:04] mzanetti, the problem that we see with nmcli is that it won't work even if we tried after 2 minutes. if the internet worked the first try it will keep working if it did not we'd actually need to retry with nmcli [13:05] that's my finding. I have logged into devices with the problematic internet after the failed jobs and the internet is still not working there unless tried to be connected with nmcli again [13:05] mzanetti: you flash with pending? [13:05] dednick: yes [13:05] mzanetti: then yes. [13:10] mzanetti, so now we can no longer show those running apps thumbnails when running unity8 on the desktop? [13:10] dandrader: huh? [13:11] mzanetti, it used to be possible to launch the fake phone app and then have its fake thumbnail showing up in the recents/running section of the applications dash [13:12] dandrader: I'm not really sure why you ask me :) [13:13] dandrader: did I break it? [13:13] aah. because lp:unity-api doesn't have the screenshot property any more... now I get it [13:13] dandrader: greyback asked me to remove that. he said it belongs somewhere else [13:14] iirc [13:14] I would not use the term "break". maybe it's not supposed to do it anymore. just asking you because I had to ping someone :) [13:14] dednick: and the corresponding mr: https://code.launchpad.net/~larsu/qmenumodel/use-full-action-name/+merge/183453 [13:15] mzanetti: dandrader for unity-mir, we'll be using an image source to grab app screenshots. But for compatibility we'll add support for the ApplicationImage component that's currently being used. [13:16] greyback, not talking about that. I'm talking about the ability to play around with launching and closing fake apps when running on the desktop [13:16] dandrader: possible something broke in the fake ApplicationImage plugin maybe?? Since it's on your desktop only [13:16] larsu: did reply work for you? [13:17] or is this only clear related at the moment? [13:18] dandrader: that still works [13:18] dandrader: ecept that for dialer-app, address-book-app and messages-app we don't have fake apps [13:18] except [13:20] mzanetti, with "./run --fake"? I just get dummy dashes with that one [13:21] dandrader: hmm... seems to have broken there... I never use --fake [13:22] dednick: hm maybe I should have checked. It doesn't :) [13:24] ah, it still works with plain "./run" [13:26] larsu: huh. i think i may just be an idiot. [13:26] dednick: problem is, x-canonical-message-actions doesn't get the "indiator." prefix [13:26] larsu: nevermind. apparently i'm smarted than i think :) [13:27] dednick: hm? It still doesn't work, right? [13:27] larsu: no. i just realised something that i thought wasnt, was. [13:29] larsu: ah. it's finding a "." though, which i guess is causing it to assume it's part of a group [13:29] dandrader: I think the reason is that there are no fake apps for "Tile.0" for example [13:29] dednick: I don't understand. What's finding a '.'? [13:30] dandrader: changing the fake backend to return for example "telephony-app.dekstop" for every tile should fix it I think [13:30] larsu: sorry, nevermind. i'm talking rubbish [13:31] dednick: standup [13:31] dednick: hehe, okay. The problem is that a UnityMenuAction is not directly associated with a GtkMenuTrackerItem (the only thing that knows which namespace the action is coming from) [13:31] dednick: I'm trying to find a halfway elegant solution... [13:36] dednick: model.fullActionNameForItem(index, name)? [13:36] man that's ugly... [13:36] * larsu is starting to think UnityMenuAction should have been a non-qml-creatable type [13:40] larsu: without knowing any further details, the would suggest that your sispicions are correct :) [13:40] ...the name would... [13:41] hehe [13:44] mzanetti, "Dash/DashApps.qml:54: TypeError: Object Categories(0x1b3eb30) has no method 'overrideResults'" <- I think that's likely the culprit. it seems the fake scopes implementation didn't get updated to the latest design/api/changes [13:45] dandrader: yeah... we _REALLY_ should have all APIs in lp:unity-api [13:45] that would prevent such things from happening [13:48] larsu: i'm not going to dignify that with a response [13:49] It's a noun :) you can create it! [13:50] dednick: fair enough, but then we need a function that turns actions names into properly namespaced ones [13:50] larsu: add the namespace if it doesnt have it. [13:51] dednick: how? UnityMenuAction is not associated with an item... [13:51] larsu: hm [13:52] larsu: why doesnt it have the namespace on it in the first place? isn't that like "a requirement" [13:52] to be part of the group. [13:53] lunch! === dednick is now known as dednick|lunch [14:00] pstolowski: added you to 4 reviews that fix the stuff we found out on friday [14:04] tsdgeos: yay, you nailed it down? [14:06] i can't find anything obviously wrong at least [14:07] pstolowski: i'll send an email later explaining the problem as i understand it [14:07] in case there's something left to fix [14:07] whoever needs to fix it has a bit of head start [14:07] tsdgeos: awesome, thanks. I'll test your fixes in a minute [14:07] mzanetti: https://code.launchpad.net/~mzanetti/unity8/launcher-defaults-from-dconf/+merge/183420 ? [14:08] tsdgeos: yep [14:10] tsdgeos: on the desktop it should pick up your existing launcher config (or use the default one with libreoffice etc) [14:10] tsdgeos: on the phone, additionally we would override the default launcher entries with this dconf override: https://launchpad.net/ubuntu/+source/ubuntu-settings/13.10.4/+build/4926146/+files/ubuntu-touch-settings_13.10.4_all.deb [14:11] ij [14:11] ok [14:11] mzanetti: if (appIds.count() << m_storedApps.count()) { ¿? [14:11] my math logic exploded [14:11] lol [14:12] * mzanetti wonders that this actually compiles [14:12] number shiftleft number [14:12] gives number [14:12] you can "if" a number [14:13] there should be a warning like when you do a if ( foo = bar) [14:14] tsdgeos: fixed [14:16] mzanetti: i don't understand that if [14:17] tsdgeos: if the old stored list is bigger than the new one... [14:17] which means, items where removed [14:17] then do a cleanup of the itemCache [14:17] isn't it possible that someone sets a larger list that is not a superset of the existing one? [14:18] tsdgeos: yes it is... but also not really critical I'd say [14:18] ok, i mean you can just remove the if [14:20] mzanetti: you probably want to comment/remove the qdebug [14:21] tsdgeos: narf... did I commit another one? === dandrader is now known as dandrader|afk [14:22] tsdgeos: removed === dandrader|afk is now known as dandrader === alan_g is now known as alan_g|tea === dednick|lunch is now known as dednick === alan_g|tea is now known as alan_g [14:59] mzanetti, question for you [14:59] mzanetti, what qt type would "<[{'defaults' : }]>" be? [15:00] QList [15:00] seb128: ^ [15:00] aka QVariantList [15:01] I think [15:01] dednick_: no, namespaces would be worthless if the actions would have to carry them [15:01] mzanetti, I tried to do that: http://paste.ubuntu.com/6055372/ [15:01] mzanetti, but I get [15:01] "(QMap(("defaults", QVariant(bool, true) ) ) ) [15:01] QDBusMarshaller: type `QList' (1458) is not registered with D-BUS." [15:01] mzanetti, when I run my code [15:02] mzanetti, oh, I guess I need to register the type [15:02] seb128: hmm... I don't think you have to... [15:02] larsu: right, but it's not the action, it's the xCanonicalMessageActions. [15:02] dednick_: well, they are in the same action group as the original action, so the same namespacing rules should apply [15:03] dednick_: s/original/standard [15:04] dednick_: I see three possible solutions: (1) add model.getAction(index, name) [15:04] (2) add model.getActionName(index, name) [15:04] seb128: qdbus_cast>(variant.value