**** BEGIN LOGGING AT Tue Oct 22 02:59:58 2013 Oct 22 03:20:59 LarrySteeze: I made the changes you suggested, take a look? Oct 22 03:21:21 http://pastebin.ca/2469757 Oct 22 03:33:36 LarrySteeze: before you check that, i'm fixing a bug or two. Oct 22 03:34:12 ok Oct 22 03:35:18 HaDAk: I'm heading to bed. Tag me if you want me to look over anything. I recommend tagging ka6sox or scoutcamper if you need review before I get back on. Oct 22 03:35:28 thanks LarrySteeze Oct 22 03:35:33 just gotta fix this one bug. :) Oct 22 04:00:08 HaDAk: Heyo. Oddly, it didn't flash on Roy__ for some reason. Oct 22 04:00:15 weird. Oct 22 04:00:21 i blame EricBlade Oct 22 04:00:25 Did you get whatever it was you needed answered? Oct 22 04:00:30 nope Oct 22 04:00:41 What was it? Oct 22 04:00:51 we're trying to figure out if we're allowed to use this image: http://cl.ly/image/2F3Y3g141y2g Oct 22 08:39:35 morning Oct 22 08:54:38 morphis: ping Oct 22 09:58:05 Unrelated: Not to self: Don't buy cheap tools anymore :P Oct 22 09:58:36 Managed to blow up the engine of my drill while trying to remove a tile floor that was put in cement :P Oct 22 09:58:41 Time to get a proper replacement LOL Oct 22 09:58:48 I guess I'll get a proper brand this time ;) Oct 22 13:12:10 Jay the floor is out :) Now the fun part start: Cleaning & rebuilding ;) Oct 22 14:04:16 Herrie: You named your floor Jay? Oct 22 14:07:20 bah. i let my grouper die. now i have to wait a million years for the battery to charge up enough for it to even turn on Oct 22 14:08:31 Tofe|Away: pong Oct 22 15:52:25 Garfonso: ping Oct 22 15:52:36 pong Oct 22 15:54:19 Garfonso: about the console messages: you tried the inspector? Oct 22 16:07:42 oh, you're right... how did that work again. Oct 22 16:14:28 Garfonso: http://www.webos-ports.org/wiki/Luna_Next_Remote_WebApplication_Debugging Oct 22 16:14:39 Garfonso: but I will enable logging to stdout anyway Oct 22 16:16:05 LarrySteeze, ka6sox, scoutcamper|away: can one of you review this for me? test it out and such: http://pastebin.ca/2469927 Oct 22 16:16:59 HaDAk: when you publish this please don't point per default to http://build.webos-ports.org/webos-ports-staging/latest/images/ Oct 22 16:17:13 can I use that on the emulator? adb doesn't recognize a device and browser does not connect to localhost:1122 Oct 22 16:17:46 morphis: right now, it's just meant for testing. if/when it goes out to the public, it won't point to the bleeding-edge feed Oct 22 16:18:23 HaDAk: can't you add a staging switch like --staging and then point to staging? Oct 22 16:18:34 also, since i'm checking for a local image, i'd like to be able to md5sum the image and compare against an md5sum stored online to see if it's the latest version. plus, if i download the image, i'd like to confirm the image matches the one on the server. Oct 22 16:18:37 Garfonso: ah right, you need to enable port forwarding for 5522 in virtualbox Oct 22 16:18:55 HaDAk: sounds like a good idea Oct 22 16:18:57 morphis: yes, i can do that. i was planning on adding switches once it works Oct 22 16:19:01 JaMa|Off: can we do something like this? Oct 22 16:19:08 or, once people have confirmed to me that it works Oct 22 16:19:19 hm? ssh magic? Oct 22 16:19:20 HaDAk: btw. would it be possible to check wether we're supporting the device which is connected or not? Oct 22 16:19:24 Garfonso: no Oct 22 16:19:44 Garfonso: https://forums.virtualbox.org/viewtopic.php?f=5&t=54073 Oct 22 16:19:44 morphis: again, i'd need a list online to pull from Oct 22 16:19:55 HaDAk: you can hard code the devices Oct 22 16:19:58 for now Oct 22 16:20:00 for now, sure. Oct 22 16:20:15 oh.. didn't now virtual box can do that... Oct 22 16:20:18 should be possible to determine the device type with adb shell when booted within android Oct 22 16:20:23 Garfonso: me too :) Oct 22 16:26:28 hm.. still no luck. I now forwarded 1122 to 1122. That's the port the log output says.. but still browser won't display anything. Is it really enough to start the app? Or do I have to do something special there, too? Oct 22 16:29:09 Garfonso: ah yes, you need to start WebAppMgr with --debug Oct 22 16:29:31 yes, that's what I do. Oct 22 16:30:10 after that I see some output in the console and it tells me to connect a browser to http://127.0.0.1:1122 Oct 22 16:30:17 so I think that should be fine... Oct 22 16:30:50 I added Host:1122, Guest:1122 to the forwarding list in virtualbox (and restarted before I tried again). Oct 22 16:31:00 but still the browser can't connect. :( Oct 22 16:33:02 hm Oct 22 16:34:57 Garfonso: you need to have an app open Oct 22 16:35:59 Garfonso: hm, I think the problem is that it's listen 127.0.0.1 Oct 22 16:36:40 yes, probably. Oct 22 16:37:58 yes, ssh also listens on the eth interface... and that's working. Hm. Oct 22 16:39:32 Garfonso: can you try something like: QTWEBKIT_INSPECTOR_SERVER=10.0.2.15:1122 WebAppMgr --debug Oct 22 16:39:39 where the ip is the ip of the vbox network interface Oct 22 16:39:57 ok, thats work fine here Oct 22 16:41:41 yes, that's working. thanks :) Oct 22 16:41:53 Garfonso: can you add that to the documentation? Oct 22 16:42:04 damn... it says debug.html. Where does that come form? Oct 22 16:42:06 ok Oct 22 16:42:08 Garfonso: I will add the port forwarding as default Oct 22 16:52:22 ok, i found a way to get the model via adb Oct 22 16:52:36 it's a bit roundabout, but it works Oct 22 16:52:57 adb shell dumpsys|grep grouper | head -n 1 Oct 22 16:54:49 HaDAk: use adb shell getprop ro.product.device Oct 22 16:55:03 well. i guess that might work :| Oct 22 16:55:15 :) Oct 22 16:55:27 :) Oct 22 17:04:28 morphis: ping² Oct 22 17:05:05 it seems the update app can't load neither enyo.js nor app.js Oct 22 17:05:10 Tofe: pong Oct 22 17:05:18 Garfonso: sounds bad Oct 22 17:05:54 morphis: I couldn't make maliit appear on the image I downloaded, nor on the qemux86 image I built Oct 22 17:06:16 Tofe: hm Oct 22 17:06:28 Tofe: you can check wether the maliit-server process is running? Oct 22 17:06:32 morphis: what was your exact scenario, from scratch ? Oct 22 17:06:34 yes sure Oct 22 17:06:48 Tofe: lets see first if everything is correctly installed Oct 22 17:07:14 Tofe: I need the output of: Oct 22 17:07:17 - opkg files qtbase | grep maliit Oct 22 17:07:24 ok Oct 22 17:07:36 - ls /usr/lib/qt5/plugins/platforminputcontexts/ Oct 22 17:08:23 I will use the emulator, because on gnexus it seemed to be that the reticle issue was still there, which was weird. But maybe I was a bit tired or smthg Oct 22 17:09:12 opkg files qtbase | grep maliit --> nothing Oct 22 17:09:46 (ls /usr/lib/qt5/plugins/platforminputcontexts/ --> only libibusplatforminputcontextplugin.so Oct 22 17:10:13 looks like I don't have maliit plugged in ? Oct 22 17:11:47 I'll now check gnexus Oct 22 17:12:15 while i'm flashing the hell out of this tablet, imma try out ubuntu touch Oct 22 17:12:21 can't hurt to test the competitors, eh? Oct 22 17:12:31 Tofe: hm Oct 22 17:13:50 Tofe: can you try with http://files.gravedo.de/maliit-framework-qt5_0.99.0+git0+dbc0403f32-r0.5_i586.ipk in the emulator? Oct 22 17:14:11 yeah. It launches... again errors in system-ext. Hm. This library has some issues... had to remove CrossAppUI.js to get it working. Oct 22 17:15:31 where do the sysbus files go? Oct 22 17:18:06 morphis: it's now installed, I'll reboot Oct 22 17:18:51 should the id of the service be: org.webosports.service.update? Oct 22 17:19:43 morphis: maliit-server runs Oct 22 17:20:25 I got it ! Oct 22 17:21:10 I see the keyboard (grey and violet), on the right side of the screen, badly oriented. Oct 22 17:21:23 Tofe wins. Oct 22 17:23:34 and... it won't go away. Ever. Oct 22 17:24:44 ah, I can swipe it out Oct 22 17:25:06 The launcher is dead. Oct 22 17:25:33 whoops Oct 22 17:26:29 Tofe: :) Oct 22 17:26:45 Tofe: will try to find why the plugin isn't included by default Oct 22 17:26:53 morphis: I think I understand: when swiping out the keyboard, the overlay is still there, but transparent. It catches all the mouse interaction Oct 22 17:28:21 morphis: typing smthg in the keyboard actually works Oct 22 17:28:50 I just tried with the launcher, I typed "Test", and it found the Tester app :) Oct 22 17:28:59 yeah!!! Oct 22 17:29:10 morphis: do you know if it's possible to change the serial number of a device? Oct 22 17:29:11 Tofe is my hero! Oct 22 17:29:14 excellent! Oct 22 17:29:22 You'll make me blush ! Oct 22 17:29:22 Tofe: you are the best! :D Oct 22 17:29:23 * scoutcamper gives Tofe a high-five Oct 22 17:29:31 * HaDAk showers Tofe in confetti Oct 22 17:29:55 * Tofe cries out of emotion Oct 22 17:30:08 ok, let's see if ubuntu touch boots... Oct 22 17:30:14 * HaDAk crosses his fingers Oct 22 17:30:42 morphis: where could I configure the orientation of maliit ? Oct 22 17:31:22 works :) Oct 22 17:31:31 btw, I used my tofe/dashbord-work branch; this kind-of validates it, even if it doesn't test the dashboard part Oct 22 17:31:36 internet at school is too damn slow(downloading emulator while cloning down build.git) i am being throttled down to 30kbps :( Oct 22 17:32:11 so I'll push it as a PR Oct 22 17:33:11 believe it or not, webos looks to be more stable than ubuntu touch Oct 22 17:33:11 lol Oct 22 17:33:46 oh ? it's still in a bad state ? Oct 22 17:33:56 it's pretty rough. Oct 22 17:34:15 and slooooooooow Oct 22 17:34:28 and wifi keeps dying Oct 22 17:34:30 on which config are you testing it ? Oct 22 17:34:46 i'm on grouper. using the latest dev build. Oct 22 17:35:13 a grouper isn't that slow, as a device... I guess they haven't begun the optimization work yet. At least I hope so. Oct 22 17:35:31 i tapped settings, and the screen flickered, then blanked out for several seconds Oct 22 17:35:48 it's pretty bad XD Oct 22 17:36:18 hm.. something is wrong with the service, now. Strange... did work before. But that was on the old image. Oct 22 17:36:48 currently I get: "MojoService: no activity passed in, creating checkUpdate_0", which triggers some timeout, then endless log output and that's it. Oct 22 17:43:54 hm. it works if I cope the files also to the public folders... probably I did something wrong there? Oct 22 17:47:36 morphis: maybe you already notice, but there are some files that are not found by maliit-server (some qml, and some svg), though it doesn't seem to be fatal Oct 22 17:56:19 what's the mechanism to grab a screenshot off the device, right now? Oct 22 17:56:27 swipe right Oct 22 17:56:33 then you'll get a png in /tmp Oct 22 17:56:40 cool. Oct 22 17:57:05 glad to see the improved gesture area in recent builds. it looks phenominal. Oct 22 17:57:18 :) wide enough ? Oct 22 17:57:26 it's perfect. Oct 22 17:57:29 cool ! Oct 22 17:58:28 i'm tempted to not even finish the other pngs i was putting together for it. i don't think they matter anymore. Oct 22 17:58:30 i love it the way it is. Oct 22 17:58:46 :) Oct 22 17:59:43 trying to streamline this install process a bit. ubuntu's touch installer has a custom recovery image based off CWM that lets you push autodeploy.zip and reboot to install. Oct 22 17:59:59 hm.. the call to applicationManager/launch comes back with success, but app is not launched. This did work in old emulator. Oct 22 18:00:03 is there a similar mechanism for CWM? Oct 22 18:00:05 morphis: between out nemo-keyboard and the one in http://bazaar.launchpad.net/~phablet-team/phablet-extras/maliit-plugins/view/head:/nemo-keyboard/nemo-keyboard.qml , which one is the most recent ? Oct 22 18:00:09 our* Oct 22 18:00:26 the one from launchpad seems much cleaner Oct 22 18:06:00 morphis: I renamed the service to org.webosports.service.update. That is more in line with other system services. Is that ok? Oct 22 18:06:31 morphis: do you build ipk files? Do you need the packageinfo.json for that? If not I would delete it. Oct 22 18:11:53 morphis: I think the Nemo keyboard from Ubuntu has a lot of fixes; unfortunately they put some "import Ubuntu.Components" here and there, mainly for their DPI management Oct 22 18:14:09 Garfonso: no we don't need the packageinfo.json Oct 22 18:14:26 Tofe: we can use the ubuntu keyboard if we want Oct 22 18:15:05 Tofe: https://launchpad.net/ubuntu-keyboard Oct 22 18:15:23 Tofe: the svg warnings should be fixed already Oct 22 18:15:57 ok Oct 22 18:18:32 morphis: well, for that one I'm not sure, when I see in the main qml file: Oct 22 18:18:45 // !!!IMPORTANT!!! Heavy modifications here might break the assumptions made by Oct 22 18:18:46 // UbuntuApplicationApiWrapper code (e.g. object names and what the items do/contain). Oct 22 18:19:22 applicationManager/launch does strange things.. Oct 22 18:19:53 http://issues.webos-ports.org/issues/445 Oct 22 18:20:29 and the app won't get parameters, or I am reading them wrong (works in legacy webOS, though) Oct 22 18:31:22 morphis: the ubuntu keyboard seems to be a QML + C++ code, but the main UI is in QML, so it should work fine for us. I can try to adapt it to remove the Ubuntu DPI bits, so that we can give it a shot. Oct 22 18:31:34 Garfonso: yeah, I've found that one today too Oct 22 18:31:42 Garfonso: and I've already a fix for it Oct 22 18:31:52 ah. ok. Oct 22 18:31:58 Tofe: ok, should be worth a try Oct 22 18:32:03 Tofe: but can we do that after the release? Oct 22 18:36:13 Tofe: merged https://github.com/webOS-ports/luna-next-cardshell/pull/31 Oct 22 18:37:51 morphis: yes, sure. I'll fix the orientation issue for the nemo-keyboard, that'll make it already look great. Oct 22 18:38:27 Tofe: btw. I don't get it up with your PR merged Oct 22 18:38:54 oh. on the emulator ? I didn't try it yet on the gnexus Oct 22 18:39:41 on the N7 Oct 22 18:39:49 ah, ok, weird Oct 22 18:39:59 I think the orientation detection is quite screwed up anyway, so I'll fix that first we'll retry Oct 22 18:40:50 Tofe: I get it up once after a fresh restart Oct 22 18:40:57 open just type, press the input field Oct 22 18:41:10 keyboard shows up but the above it is a black screen Oct 22 18:41:18 swipe up doesn't work then Oct 22 19:25:14 Tofe: still here? Oct 22 19:26:47 morphis: was having dinner Oct 22 19:27:02 Tofe: ah :) Oct 22 19:28:35 morphis: the black screen must be the keyboard background; I bring up the overlay and at the same time I progressively set opacity to 1 Oct 22 19:28:47 hm Oct 22 19:29:13 did you try to swipe out the keyboard itself ? Oct 22 19:29:18 yes Oct 22 19:29:21 didn't worked Oct 22 19:29:27 ah, that's weird. Oct 22 19:30:16 ok, I'll work on the keyboard orientation to get it nicer, maybe it will help with that pb too Oct 22 20:07:11 Tofe: ok Oct 22 20:12:04 morphis: we have two problems I think: 1, the overlay window size if of the size of the wrapped window; in this case, it wants to take the whole screen. 2, the MInputMethodQuick.appOrientation value is wrong, at least in the emulator case. Oct 22 20:15:58 for point 2., I guess we can fix it quite easily. For point 1, it needs further reflexion: how to set the overlay window size to the right value ? Oct 22 20:18:30 morphis: can you try to put this https://dl.dropboxusercontent.com/u/4679068/luna-next/nemo-keyboard.qml in /usr/lib/maliit/plugins/nemo-keyboard.qml ? you should see the keyboard better on the n7 at least Oct 22 20:18:58 but it won't solve point 1 Oct 22 20:21:39 Tofe: you're not calling changeSize for the overlay window? Oct 22 20:22:04 Tofe: and what is the difference for the new keyboard? Oct 22 20:22:33 morphis: I'm calling this: " window.changeSize(Qt.size(windowManager.defaultWindowWidth, window.height));" Oct 22 20:22:51 I can't do better than that: I don't know which height it wants to have Oct 22 20:23:29 morphis: the difference is simply that I hardcoded the orientation to be 0 degree ;) Oct 22 20:23:51 I can do much better using QtQuick.Window module Oct 22 20:24:31 but I wanted to see first if that helped solving the issue of the wrong orientation on the N7, as a start Oct 22 20:29:30 Tofe: anyway the input method should afaik automatically use the correct window size Oct 22 20:31:16 but here, the correct window size would be the keyboard layout size Oct 22 20:32:53 and the latter is build anyway using MInputMethodQuick.screenWidth and MInputMethodQuick.screenHeight, which suggests that the screen size is actually used Oct 22 20:33:05 (but I guess it's the window size) Oct 22 20:34:11 Tofe: can you check with http://qt-project.org/doc/qt-5.0/qtqml/qml-qt.html#inputMethod-prop , so Qt.inputMethod.keyboardRectangle? Oct 22 20:36:08 interesting Oct 22 20:38:27 Tofe: I am still getting the keyboard up only once Oct 22 20:38:59 ah but yes, me too Oct 22 20:39:20 that is because we don't get the information when the keyboard is slided down Oct 22 20:39:31 ah ok Oct 22 20:39:40 Qt.inputMethod.visible doesn't help here? Oct 22 20:39:49 or the window visible property? Oct 22 20:40:47 I *thought* I did hook correctly on the window visibilty property Oct 22 20:41:11 but obviously, that was not sufficient or incorrect Oct 22 20:41:53 I don't know if maliit's window is set to visible=false in our case Oct 22 20:42:06 Qt.inputMethod.keyboardRectangle.height seems to be zero Oct 22 20:42:07 hm Oct 22 20:46:23 time to sleep a bit Oct 22 20:46:43 that was a very good day anyway :) Oct 22 20:47:04 gn8 Oct 22 20:47:10 for me too Oct 22 20:47:11 gn8 Oct 22 20:47:19 Tofe: for sure, one step closer now :) Oct 22 20:47:30 :) **** ENDING LOGGING AT Wed Oct 23 02:59:58 2013