**** BEGIN LOGGING AT Mon Sep 08 02:59:58 2014
Sep 08 05:58:11 <morphis> morning
Sep 08 06:30:54 <Herrie> morning
Sep 08 06:31:31 <Herrie> browser PR is taking shape, expect it to be completed later today, just some code cleanup + layout tweaking to do for N4 ;)
Sep 08 08:28:56 <nizovn> morphis: is it possible to check hardware type here: https://github.com/webOS-ports/sleepd/blob/webOS-ports/master/src/pwrevents/machine.c#l121 ? This will reduce flood rate(577) from every 0.5s to 300s. The problem is sleepd written in C, while LunaSettings in C++. Do you think it's good to rewrite MachineGetName to get network name and use it?
Sep 08 08:34:19 <morphis> nizovn: that exactly do you want to change with checking for the hardware type in there?
Sep 08 08:34:56 <morphis> s/that/what/
Sep 08 08:35:17 <nizovn> morphis: MachineCanSleep can return false for emulator or desktop
Sep 08 08:36:10 <morphis> nizovn: but imho we should never come into that case as https://github.com/openwebos/sleepd/blob/master/src/pwrevents/suspend.c#L290
Sep 08 08:36:19 <morphis> on the emulator the display should be always on
Sep 08 08:45:24 <nizovn> ok, but for some reason emulator reaches SuspendRequest. will see why.
Sep 08 08:50:22 <nizovn> seems the problem is that IsDisplayOn is different in openwebos and webos-ports.
Sep 08 08:56:48 <morphis> yeah
Sep 08 09:03:22 <morphis> nizovn: if we disable display timeouts for the emulator (like you did with your PR) it should just work and sleepd should never try to suspend the system
Sep 08 09:04:30 <nizovn> morphis: it goes into SuspendRequest state and floods ls-monitor.
Sep 08 09:34:29 <nizovn> maybe problem somewhere in init, because after turn off, turn on screen, it stops flooding.
Sep 08 09:45:18 <Tofe> morning
Sep 08 10:14:55 <morphis> nizovn: on emulator?
Sep 08 10:15:04 <nizovn> yes
Sep 08 10:15:22 <morphis> as I said, the screen should not go off on the emulator
Sep 08 10:16:13 <morphis> nizovn: so I think doing the check you added in your PR should be enough to get this fixed
Sep 08 10:17:26 <nizovn> seems screen not goes to off, but sleepd goes into suspendrequest state once after boot.(i've tested with PR changes)
Sep 08 10:17:49 <morphis> ok
Sep 08 11:03:02 <nizovn> ok, it gets displayInactive event and decide that screen is off. be back after 2.5 hours.
Sep 08 17:15:04 <Garfonso> not sure what you are talking about, but currently emulator is still unusable for me. Right after boot luna-next seems to crash, or at least does not react to mouse input at all. After a luna-next restart (which solves the issue on "stable" emulator), the lock screen is displayed, reacts to mouse input but can not be unlocked.
Sep 08 17:16:33 <Garfonso> ls-monitor shows call to control/setLockStatus {"status":"unlock"} goes through, but lockscreen stays. Hm. strange.
Sep 08 17:16:57 <Garfonso> and now it's in "suspend flood" again. ;)
Sep 08 17:20:19 <Herrie> Tofe: ping
Sep 08 17:32:12 <nizovn> Garfonso: you are right, this is not luna-next crash, but screen turns off(it reacts on mouse however). You can temporary solve this by turning it on by luna-send. Fix is almost ready.
Sep 08 17:43:53 <Garfonso> what is the luna-send cmd to turn it on?
Sep 08 17:57:49 <nizovn> Garfonso: luna-send -n 1 luna://com.palm.display/control/setState "{\"state\":\"on\"}" If not works, then turn off it before.
Sep 08 18:19:34 <morphis> nizovn: I have two other changes regarding display management
Sep 08 18:20:46 <adam_> ?
Sep 08 18:22:13 <morphis> nizovn: https://github.com/webOS-ports/luna-appmanager/pull/3
Sep 08 18:23:30 <nizovn> morphis: for now i think to enable DNAST mode for emulator. What does it mean?
Sep 08 18:23:35 <nizovn> :)
Sep 08 18:24:19 <morphis> nizovn: it's a blocker for the display
Sep 08 18:24:33 <morphis> not sure if it was used in legacy for the same problem we try to fix here
Sep 08 18:25:35 <morphis> nizovn: and https://github.com/webOS-ports/luna-sysmgr/pull/9
Sep 08 18:25:40 <nizovn> morphis: seems it was used, but emulator was called "desktop" in those times
Sep 08 18:26:06 <morphis> really?
Sep 08 18:26:24 <morphis> nizovn: emulator and desktop should be different even on legacy
Sep 08 18:26:50 <morphis> nizovn: https://github.com/openwebos/luna-sysmgr/blob/master/Src/base/DisplayManager.cpp#L705
Sep 08 18:28:53 <morphis> nizovn: putting that back in might fix the problem we have
Sep 08 18:28:59 <morphis> if you use the settings approach
Sep 08 18:29:10 <morphis> we're using the same when the first use app is running
Sep 08 18:29:14 <morphis> works quite fine
Sep 08 18:29:17 <morphis> display never goes off
Sep 08 18:29:58 <nizovn> morphis: there is a problem while i tested it. lunasettings returns Device there for some reason.
Sep 08 18:30:11 <morphis> that can be fixed
Sep 08 18:30:53 <morphis> nizovn: https://github.com/webOS-ports/luna-sysmgr-common/blob/webOS-ports/master/Src/base/settings/Settings.cpp#L608
Sep 08 18:30:56 <morphis> is where it is read
Sep 08 18:31:39 <morphis> nizovn: that installs the config https://github.com/webOS-ports/meta-webos-ports/blob/daisy/recipes-webos/luna-sysmgr-conf/luna-sysmgr-conf.bb
Sep 08 18:32:02 <morphis> nizovn: which is here for emulator: https://github.com/webOS-ports/luna-sysmgr/blob/webOS-ports/master/conf/luna-qemux86.conf
Sep 08 18:32:15 <morphis> there might be something wrong with this if it doesn't work
Sep 08 18:32:39 <morphis> nizovn: you can add -t -c to the cmdline for LunaSysMgr in /lib/systemd/system/LunaSysMgr.service
Sep 08 18:32:44 <morphis> that should give you more verbose output
Sep 08 18:32:52 <morphis> and should also mention the hardware type it reads
Sep 08 18:33:09 <nizovn> btw there is emulator line in luna-platform.conf
Sep 08 18:34:18 <morphis> nizovn: on the emulator rootfs?
Sep 08 18:34:23 <nizovn> yes
Sep 08 18:34:49 <morphis> hm
Sep 08 18:34:59 <morphis> then something with the parsing is wrong when it returns device for you at the other end
Sep 08 18:36:16 <morphis> nizovn: have to drop off, can you also look at my two PRs?
Sep 08 18:36:40 <morphis> nizovn: should fix #575
Sep 08 18:39:30 <nizovn> morphis: i'll test them too. not sure they fixing 575, seems more 577
Sep 08 18:43:49 <nizovn> hm.. hope #9 will solve 575 on mako and tenderloin
Sep 08 18:47:58 <HaDAk> hey morphis, are LuneOS install instructions up on the wiki?
Sep 08 18:49:29 <HaDAk> nevermind. i found it.
Sep 08 18:49:30 <HaDAk> :)
Sep 08 19:41:15 <Tofe> Herrie: pong
Sep 08 19:47:20 <Garfonso> uhm.. regression: Parameters are not handed to Custom UI anymore. :(
Sep 08 19:47:54 <GodGinrai> like, Launch Parameters?
Sep 08 19:48:16 <Herrie|Laptop> Tofe: here from laptop
Sep 08 19:48:59 <Garfonso> GodGinrai: right
Sep 08 19:49:46 <Tofe> Herrie|Laptop: well, pong|laptop then :)
Sep 08 19:49:57 <GodGinrai> k
Sep 08 19:50:07 <Herrie|Laptop> LOL
Sep 08 19:50:18 <Herrie|Laptop> Just wondering how I could run some queries in QML while loading the UI
Sep 08 19:50:47 <Herrie|Laptop> I need to pull some data right at the beginning. Now have some onClick when I actually need it but would like to load it when UI is build
Sep 08 19:51:08 <Tofe> Herrie: Component.onCompleted is what you need
Sep 08 19:52:07 <Herrie|Laptop> OK will try that :)
Sep 08 19:52:09 <Tofe> it's available on all types of QML components; just be careful: if you have A that contains B, B will be completed before A is fully completed.
Sep 08 19:52:29 <Tofe> Usually it's used on the root component of the module you want to handle
Sep 08 19:53:07 <Tofe> There are plenty of examples in the cardshell, if you need some.
Sep 08 19:53:49 <Herrie|Laptop> OK
Sep 08 19:54:08 <Tofe> Also, be careful with async requests; but that's nothing new compared to Enyo
Sep 08 19:54:27 <Herrie|Laptop> I guess I might have some of those ;)
Sep 08 19:54:36 * Herrie|Laptop doesn't know Enyo that well :P
Sep 08 19:55:02 * Tofe doesn't know Enyo
Sep 08 19:55:14 <Tofe> (boooo)
Sep 08 19:55:59 <Herrie|Laptop> So what should I take care of with the async ones?
Sep 08 19:56:09 <Herrie|Laptop> I need to have a success and failure call in there?
Sep 08 19:56:24 <Tofe> just that when the UI is ready, you may not yet have your underlying data ready
Sep 08 19:59:14 <Herrie|Laptop> OK yeah have that here it seems ;)
Sep 08 19:59:21 <Herrie|Laptop> But well that's a bug to fix for later :P
Sep 08 20:01:31 <Herrie|Laptop> I guess that causes my "Unable to assign [undefined] to QString"? Any quick way around this?
Sep 08 20:03:19 <Tofe> "undefined" looks like a name resolution error to me
Sep 08 20:03:27 <Herrie|Laptop> I'm calling a function that does a luna-call to DB8 and then populates the values
Sep 08 20:03:38 <Herrie|Laptop> Since it's async it's probably not done yet ;)
Sep 08 20:03:47 <Tofe> mmh
Sep 08 20:03:55 <Tofe> do you have already the code on gh ?
Sep 08 20:04:22 <Herrie|Laptop> Let me push that now :P
Sep 08 20:09:50 <Herrie|Laptop> https://github.com/Herrie82/org.webosports.app.browser/commit/cc6ad910f3f3653804eb83d52418d20478acc0d4
Sep 08 20:10:06 <Herrie|Laptop> Need to add some stub stuff sometime ;)
Sep 08 20:10:23 <Herrie|Laptop> Not very useful to test in QT Creator like this without mock backend DB8 stuff
Sep 08 20:11:50 <Herrie|Laptop> I'm sure a lot of things could be done more optimal but well :p
Sep 08 20:12:14 <Herrie|Laptop> It works more or less for most things :P
Sep 08 20:20:08 <Herrie|Laptop> Tofe: Implemented some of your suggestions (URL suggestions rows lower, full screen on N4, made panel for downloads, history + bookmarks fullscreen on N4 too). It could use some nice transformation effect as well I guess but well :p
Sep 08 20:20:32 <Tofe> Herrie|Laptop: where would that error be, btw ?
Sep 08 20:23:37 <Herrie|Laptop> main.qml line 440: json: getJSONData()
Sep 08 20:24:09 <Herrie|Laptop> and NavigationBar.qml line 769: json: getURLHistory()
Sep 08 20:29:03 <Tofe> well, there's an easy way
Sep 08 20:29:44 <Tofe> First, declare my*Data as string, not variant, because you know it will be a string
Sep 08 20:29:53 <Tofe> (isn't it ?)
Sep 08 20:30:29 <Herrie|Laptop> Well it's a JSON ;)
Sep 08 20:30:31 <Tofe> assign a default value, i.e. property string myBookMarkData: '{}'
Sep 08 20:30:33 <Herrie|Laptop> So i thought variant?
Sep 08 20:30:39 <Herrie|Laptop> Or can do string there?
Sep 08 20:31:13 <Tofe> it looks like you assign the payload to it, but without doing a JSON.parse, so you'll get a string, not an object
Sep 08 20:32:12 <Herrie|Laptop> I do a parse ;)
Sep 08 20:32:23 <Herrie|Laptop> var defbrows = JSON.parse(message.payload)
Sep 08 20:32:46 <Herrie|Laptop> Ah that's for the other ;)
Sep 08 20:32:51 <Tofe> yes
Sep 08 20:34:23 <Tofe> for urlHistory*Data, same remarks
Sep 08 20:35:43 <Herrie|Laptop> I should JSON.parse them I guess?
Sep 08 20:36:05 <Tofe> depends on the "json" property of the JSONListModel, let me see
Sep 08 20:36:44 <Tofe> string it is
Sep 08 20:37:03 <Tofe> so no parse
Sep 08 20:38:07 <Tofe> variant properties should be avoided when possible, because it can't be optimized and slows down a bit the app
Sep 08 20:39:45 <Herrie|Laptop> ok
Sep 08 20:39:56 <Herrie|Laptop> so all strings and initialize with '{}'
Sep 08 20:40:03 <Tofe> yes
Sep 08 20:41:22 <Herrie|Laptop> That works
Sep 08 20:41:27 <Tofe> :)
Sep 08 20:42:39 <Tofe> also, your two qml files begin to be pretty big; maybe at some point in time you'll want to refactor them a bit -- but that can wait, as long as you don't feel the need for that
Sep 08 20:43:15 <Herrie|Laptop> Yeah saw that as well
Sep 08 20:44:30 <Herrie|Laptop> Need to see how :P QML n00b
Sep 08 20:45:57 <Tofe> oh, that's pretty easy: you take a set of consistent lines (for example, a delegate definition), cut&paste in another file; then, you see what data you are missing in the standalone delegate, and put that as properties of the delegate, which you feed in from you previous big file.
Sep 08 20:57:10 <Herrie|Laptop> Ah ok
Sep 08 21:09:09 <Herrie|Laptop> morphis: PR is in for the browser
Sep 08 21:10:03 <Herrie|Laptop> P.s. Should the messaging app be called org.webosports.app.messaging instead of org.webosports.messaging in order to be compliant with our naming convention for the other apps?
Sep 08 21:42:06 <Tofe> Herrie: merged!
Sep 08 21:43:29 <Herrie> Tofe: thnx fixed the few things you mentioned
Sep 08 21:43:39 <Herrie> Will update the readme with what's done and what needs doing still
Sep 08 21:43:49 <Herrie> But it's an improvement so far for sure
Sep 08 21:43:56 <Herrie> Need some help with that FocusScope I guess :P
Sep 08 21:44:02 <Herrie> Tried but failed :P
**** ENDING LOGGING AT Tue Sep 09 02:59:59 2014