**** BEGIN LOGGING AT Sun Mar 15 02:59:59 2015 Mar 15 09:22:30 Herrie: let me debug that a bit in the coming days Mar 15 09:22:46 morphis: Not sure this is the most correct solution but will probably work at least for hiding the popup for the low battery when at 5% and charging kicks in: Mar 15 09:22:49 https://github.com/webOS-ports/luna-systemui/compare/webOS-ports/master...Herrie82:webOS-ports/master Mar 15 09:24:50 Basically added an onResponse for the "chargerStatusQuery" call and moved the bit of the code that was checking for the response for that call to the handleChargerStatus instead of handlePowerNotifications Mar 15 09:25:56 Herrie: not sure that this works for all cases as there seems to be no subscription for charagerStatusQuery Mar 15 09:26:23 and the method doesn't seem to support that Mar 15 09:26:34 Yeah it seems to do an initial call only Mar 15 09:27:03 right Mar 15 09:27:11 The code the way it was didn't work anyway. This might work better but not sure Mar 15 09:27:25 can you revert what you did and change usbdockstatus to USBDockStatus Mar 15 09:27:27 Need to find a quick way to drain my N4 :P Mar 15 09:27:39 https://github.com/webOS-ports/luna-systemui/blob/webOS-ports/master/data/PowerdService.js#L29 in that line Mar 15 09:27:53 the signal afaik is named USBDockStatus and not lower case variant of it Mar 15 09:29:56 Tried the calls on TP with 3.x but seem to get stuck so cannot compare Mar 15 09:40:04 morphis: AFAIK that shouldn't matter because of https://github.com/webOS-ports/luna-systemui/blob/webOS-ports/master/data/PowerdService.js#L46 ? Mar 15 09:40:16 L29 is just the service name Mar 15 09:40:25 L46 sets the parameters for the call Mar 15 09:42:49 ah right Mar 15 09:45:46 Just I don't see how the current code would ever get the response for the chargerStatusQuery, because as far as I can see it can't ;) Mar 15 09:46:57 Because https://github.com/webOS-ports/luna-systemui/blob/webOS-ports/master/data/PowerdService.js#L114 will never have a value when you call batteryStatus or usbdockstatus from what I can see Mar 15 15:00:32 morphis: do you know what the following message would mean ? Mar 15 15:00:46 I get it when trying to start my QOrientationSensor sensor Mar 15 15:01:27 I just don't know how to create a sensor manager :) Mar 15 15:04:15 I've got sensorfwd running, that's for sure, but I don't know how the connection works between this one and my QSensor Mar 15 15:20:35 morphis: could it be that the dbus communication with sensorfwd isn't allowed or something like that? Mar 15 15:49:35 Ah, got it! I was not yet in the event loop, so the sensor called the DBus connection too soon and failed Mar 15 15:49:39 now it works ! :) Mar 15 15:58:11 morphis: these two PRs I just made may need more testing. For instance, the windows are not automatically resized horizontally, which results in a weird display. And the vkbd doesn't adapt its width. Mar 15 16:13:40 Tofe: 2? Cardshell and ??? Mar 15 16:13:44 Does this work on N4? Mar 15 16:13:47 Or only on TP? Mar 15 16:14:07 Herrie: only tested on N4 so far Mar 15 16:14:53 2 PRs are cardshell and hwcomposer QPA Mar 15 16:15:15 Ah I guess for hwcomposer QPA I need to wait on a build then :P Mar 15 16:15:34 Since I have no clue how to build & push this to N4 ;) Mar 15 16:15:41 yup :) you want an ipk by mail ? ;) Mar 15 16:15:59 That would be great ;) Mar 15 16:16:12 I've built it against my yesterday build Mar 15 16:16:12 LNC I can do myself since it's just pushing the QML ;) Mar 15 16:16:21 should be fine with a recent build Mar 15 16:16:22 I have latest image on N4 Mar 15 16:16:28 ok, let's try Mar 15 16:16:50 Was just playing around a bit with luna-systemui for the 5% battery popup to disappear. I think I got that working now Mar 15 16:17:00 Just the notifications don't overwrite each other yet Mar 15 16:17:06 For 20%, 15%, 10% etc Mar 15 16:19:09 Herrie i fixed the yesterday and tomorrow problem pulled in to your calender Mar 15 16:20:53 Herrie: you'll see that we definitely need to work out the height of alerts, it's ugly in landscape on N4 Mar 15 16:22:17 Tofe; OK rebooting N4 now Mar 15 16:22:23 cobblemonster: Saw thnx Mar 15 16:22:31 oh you just needed to restart the cardshell :) Mar 15 16:22:34 Will pick the fixes and push to master Mar 15 16:22:41 Tofe; Oh well it's quick enough ;) Mar 15 16:22:56 ok Mar 15 16:23:25 I'm happy with the boot speed on N4 Mar 15 16:23:32 Herrie: it should work as soon as you see the boot logo Mar 15 16:23:35 It really flies compared to legacy 2.x and 3.x Mar 15 16:23:41 TOfe: already have the first bug ;) Mar 15 16:23:47 there are plenty Mar 15 16:24:00 hopefully it's quite obvious bugs Mar 15 16:24:03 Rotating phone counter clockwise does a 270 degrees right rotation ;) Mar 15 16:24:11 instead of 90 degrees left ;) Mar 15 16:24:11 yes, for instance :p Mar 15 16:24:24 I didn't optimize the rotation path Mar 15 16:25:22 But it's fun to play with Mar 15 16:25:52 Looks quite good already ;) Mar 15 16:25:58 Some stuff needs tweaking but that's fine Mar 15 16:26:01 Ah also I think the sensor plugin doesn't survive a cardshell restart Mar 15 16:26:12 It's a good start Mar 15 16:26:22 Yes, I'm quite happy it's not a disaster Mar 15 16:26:23 I guess we could put the rotation lock in quite easily now? Mar 15 16:26:41 Just Type layout gets squeezed for example, but that's minor Mar 15 16:26:53 Is it exposed on LS2? Mar 15 16:27:03 Good that VKB now can have 2 layouts going forward ;) Mar 15 16:27:14 Because current TP layout is UGLY and not useful :P Mar 15 16:27:54 yeah, it needs a lot of tweaking in landscape, both on N4 and on TP Mar 15 16:28:44 Herrie: for card squeezes, you just need to maximize it and turn it back as card, and it will update to the correct size Mar 15 16:29:06 With LS2 exposed you mean if you can do a LS2 call to activate/deactivate rotation lock? Mar 15 16:29:21 Herrie: yes, and to read it Mar 15 16:29:54 because then, for actually locking rotation, it's just a matter of not listening to events in OrientationHelper.qml Mar 15 16:31:41 Anyway. I think we need to keep screen rotation under a tweak setting at the beginning Mar 15 16:32:01 or something like that Mar 15 16:32:40 It's not really ready yet, but we need to have it on device to test it and converge Mar 15 16:34:03 morphis: what we can do for the moment is merge the hwcomposer QPA PR, and wait a bit for the cardshell PR Mar 15 16:39:01 btw orientation keeps working if turn off screen? Mar 15 16:39:17 ah, good question, let me try Mar 15 16:39:32 yes, it survives the blanking Mar 15 16:40:09 nizovn: but it doesn't survive a luna-next termination Mar 15 16:40:41 Maybe I'm not closing the sensor properly, or something else is happening Mar 15 16:42:54 Tofe: thanks, seems we need new hypothesis Mar 15 16:44:13 nizovn: no problem. From my point of view, the blanking doesn't so much affect sensorfw as the closing of the sensor handlers Mar 15 16:52:15 Tofe: Seems it was in OWO Luna-SysMgr at least to read the value: https://github.com/openwebos/luna-sysmgr/blob/master/Src/base/settings/Preferences.cpp#L434 Mar 15 16:52:23 But seems it's not in ours or moved elsewhere Mar 15 16:53:18 Ok Mar 15 16:53:23 well, that'll be all for me for this week-end :) Mar 15 16:55:54 I guess nizovn could help with the rotation lock, or morphis might have some ideas from a design point of view ;) Mar 15 16:56:05 It seems it simply stores a true or false value Mar 15 16:56:40 So when you rotate 90, 180, 270 degrees, then lock it, it should remember where it was locked and put the value to true ;) Mar 15 17:04:35 Herrie: really, the rotation lock won't be an issue :) There's nothing to remember, we just have to not listen to orientation events Mar 15 17:05:15 does Wayland use xrandr? Mar 15 17:07:36 Tofe: Well it remembers something after a reboot still Mar 15 17:07:57 If you locked it up side down and moved back it's still upside down on a TP after reboot if I'm correct Mar 15 17:09:13 Just rebooting TP to confirm now Mar 15 17:14:01 Yeah it remembers where it was locked after a reboot Mar 15 17:14:07 My screen is now upside down :P Mar 15 20:40:44 ka6sox: no, but there's an equivalent functionality, wlrandr http://www.phoronix.com/scan.php?page=news_item&px=MTMyMzI Mar 15 20:41:32 Herrie: ok, didn't think about that Mar 15 20:51:21 Tofe: I didn't really figure out how legacy kept this value Mar 15 20:51:28 Maybe morphis knows Mar 15 20:51:41 I'm not that good in C++ reading :P Mar 15 20:56:50 Herrie: personnally I would keep it besides the rotation lock preference Mar 15 20:57:51 or just one value, like an enum: { NoLock, Top, Bottom, Right, Left } Mar 15 21:16:29 Tofe: That might cause some compatibility issues with legacy apps because it's a public api? Mar 15 22:10:30 Tofe: You seen http://doc.qt.io/qt-5/qml-qtquick-rotationanimator.html ? Mar 15 22:10:42 Seems to support clockwise and counterclockwise Mar 15 22:10:48 Might be a good approach? **** ENDING LOGGING AT Mon Mar 16 02:59:58 2015