**** BEGIN LOGGING AT Tue Sep 06 02:59:58 2016 Sep 06 04:51:07 morning Sep 06 07:09:44 Morning Sep 06 07:09:59 morning Sep 06 07:10:42 Morning! Sep 06 07:11:04 oh, I already said that. Maybe I need some coffee. Sep 06 07:11:58 Tofe: Me too, little one decided 4:45 was a good time to get up :P Sep 06 07:12:08 Luckily I wasn't in bed late yesterday :P Sep 06 07:16:44 Tofe: I'm not sure how to get this pepper flash stuff working :P Sep 06 07:17:05 We have plugins enabled in the webview. I copied the .so to all possible locations but doesn't seem to work :s Sep 06 07:17:11 Nothing in logs etc :s Sep 06 07:17:31 Maybe we need to enable something in the build like for the codecs. I'll dig a bit :P Sep 06 07:19:04 I added the .so to the paths in the Qt docs, to /usr/lib/qt5/plugins/ppapi/ as well, no luck. Sep 06 07:20:21 I don't know, I never looked at the how the plugins are loaded Sep 06 07:21:09 Tofe: Well it determines the path based on OS hence my question yesterday, but when that fails it seems to look in the default plugins path which is /usr/lib/qt5/plugins/ppapi Sep 06 07:21:44 At least that's what I was able to read from the code I linked yesterday Sep 06 07:22:05 ok Sep 06 07:23:03 I took an arm libpepflashplayer.so from a Chomebook image, so that theoretically should work. Sep 06 07:25:57 you could try starting luna-qml-launcher with "--ppapi-flash-path=/usr/lib/qt5/plugins/ppapi/libpepflashplayer.so" Sep 06 07:27:23 Tofe: OK will give that a go tonight... Sep 06 07:27:33 USB port on N4 is flaky though :s Sep 06 07:27:41 Need to take it apart & maybe replace :P Sep 06 07:33:02 embedded_linux.pri has enable_plugins=0 in GYP_CONFIG. I guess I could try to force it to be enabled during a build, similar to what we do for the codecs Sep 06 07:43:38 But I'll try your suggestion first :) Sep 06 07:47:08 Yeah seems we need that in the build. When I check meta-qt5 it seems to do quite some bits based on enabled_plugins Sep 06 07:49:51 ah yes, if that's not enabled, there won't be any pepper plugin Sep 06 07:49:56 I thought it was Sep 06 07:50:49 Tofe: I guess not, but this whole Chromium build is like a black box to me :P Sep 06 07:51:02 I'd need to check the logs but quite sure it's probably not enabled Sep 06 07:51:41 maybe even on purpose, as the first step was to get a good usable qtwebengine before shaking it with flash :p Sep 06 07:51:57 Tofe: Yeah :P Sep 06 07:52:24 I'm not sure the ChromeOS ARM .so will work, but guys @ Rpi forums seem to have success with it :P Sep 06 07:59:34 So it might work for us too ;) Sep 06 08:31:08 Tofe: Seems I might have broken my battery fixing the USB port :s Sep 06 08:31:14 It's so weak :P Sep 06 08:31:26 Nothing compared to what I'm used from Palm batteries LOL Sep 06 08:43:27 Well new battery should be here tomorrow... Doesn't hurt to replace it anyway because it started to bulge a bit already Sep 06 08:44:42 You use them fast :) Sep 06 08:45:40 Tofe: Well this is an old phone, so don't know what the previous owner did with it... Sep 06 08:45:48 Battery is weak though :S Sep 06 08:46:00 It would bend too easily imho :P Sep 06 08:46:40 Followed the ifixit guide and was careful but it still bended due to it sticking so well to the case chasis :P Sep 06 10:11:24 Tofe: Another small PR is in :P Sep 06 12:46:44 JaMa: Where can I find the output of the various build log files from Jenkins again? Sep 06 14:23:31 Herrie|Pre3: you mean during the build or after? Sep 06 14:26:10 JaMa: After. I changed some webengine config and it failed in do_configure Sep 06 14:27:14 Want to check the logfile for it Sep 06 14:29:21 if the workspace still exist you can use workspace link in the jenkins job Sep 06 14:29:31 but if it was removed from tmpfs already then it's gone Sep 06 14:31:49 Where do I find the workspace link? Sep 06 14:32:01 I've been clicking around but no luck so far :P Sep 06 14:32:06 Must be blind LOL Sep 06 14:33:03 http://jenkins.nas-admin.org/view/luneos-testing/job/luneos-testing_grouper/ws/ Sep 06 14:37:21 OK thnx Sep 06 14:40:55 kernel upgrade merged in meta-smartphone as well Sep 06 14:42:04 JaMa: Thnx, I somehow thought that was done already :s Sep 06 14:42:20 Need more coffee... 4:45 is not a good time to get up :P Sep 06 14:42:24 only to the linux repo Sep 06 14:42:39 and you didn't update KERNEL_VERSION in the recipe, not sure if it's important or not Sep 06 14:47:05 JaMa: Also not sure tbh... Sep 06 14:48:02 I checked the output of the CM11 jenkins job and that showed 3.0.101-12915-g7cc80bd so I thought it shouldn't be changed. Sep 06 14:48:08 morphis: ^ Sep 06 15:17:07 Tofe: ping Sep 06 15:46:44 Garfonso: ping Sep 06 17:21:50 Herrie|Pre3: pooong Sep 06 17:27:58 Tofe: I'm trying to address some of the issues in http://issues.webos-ports.org/issues/1097 looking at 4 now Sep 06 17:29:02 I have a way to do it but it's not very proper... I.e. Adding https://github.com/webOS-ports/luna-next-cardshell/blob/master/qml/Notifications/NotificationArea.qml#L154 after L160 Sep 06 17:30:19 But it should actually only do it on a proper launch of the app, so I ended up adding it to https://github.com/webOS-ports/luna-next-cardshell/blob/master/qml/LaunchBar/Launcher.qml#L151 for when the call is successful, however that would mean I'd have to pass in an additional parameter besides to id and params Sep 06 17:30:40 Which would only be for this call and we call launchApplication in a number of places. Sep 06 17:30:56 Wondering if you have a proper suggestion how to do this. Sep 06 17:31:38 Add some property to launcherItem and set that instead? Sep 06 17:34:22 one moment :) Sep 06 17:38:16 What I've done is replaced the undefined in L151 with handleLaunchAppSuccess and let handleLaunchAppSuccess do the closing of the notification. Sep 06 17:42:51 Personnally, what I would do is introduce a third argument to launchApplication, which would be a callback success function Sep 06 17:43:42 we could then simply pass that to the ls2 call Sep 06 17:44:19 if the caller doesn't give that argument, it'll be undefined, so we are back on the precedent behavior Sep 06 17:44:55 and then it's dead simple to implement the success callback in NotificationArea, because we have all the informations to close the notification Sep 06 17:45:59 ah, wait, that's exactly what you did, isn't it ? :p Sep 06 17:46:21 (I don't even have the excuse of having been up at 4:30am...) Sep 06 17:46:28 Tofe: That's what I did... Just wasn't sure how QML deals with the 3rd one if we don't provide it Sep 06 17:46:57 that's good ol' javascript, so it's undefined if it's not there Sep 06 17:47:09 If you say it's OK and will work then I'm fine with it :-) Sep 06 17:47:37 Just I know c++ always expects all if I'm correct... Sep 06 17:47:37 I think it'll work Sep 06 17:47:55 yes, c++ signatures are strict Sep 06 17:48:02 but js is sloppy on all the sides :) Sep 06 17:48:17 OK will test it a bit. I guess should be straight forward. Sep 06 17:49:12 Another one I'd need to update is the call in the chatthreader so it doesn't create multiple notifications but updates where possible. I can just push stuff to the db kind for that right? Sep 06 17:49:39 The chatthreader is a simple watch on the db that gets triggered from what I recall. Sep 06 17:49:56 ah yes, right Sep 06 17:50:27 for that one, we have to somehow recall what was the id of the created notification, to be able to replace it Sep 06 17:50:29 It might be a bit tricky in some cases but will check how legacy does it. Sep 06 17:50:52 Yeah well when it creates the notification it returns the I'd ;-) Sep 06 17:51:09 So we have that in the future.result ;-) Sep 06 17:52:24 so then it should be ok I guess Sep 06 17:52:28 I know what it should do when there are 2 messages from the same sender, just would need to check what it does when there are multiple senders. Sep 06 17:53:27 Tofe: Well it adds a count, so that might be a little tricky bit shouldn't be rocket science. I.e. when you have 5 messages from me you have "Herrie (5)" Sep 06 17:54:03 It's chatthreader who sets the "(5)" I guess ? Sep 06 17:55:38 Tofe: Yeah it should. The thing is that it can send the first notification which is then either opened or not. When it's opened it shouldn't set the (2) etc, so that might require a bit of fiddling with how the new messages are processed bit shouldn't be Sep 06 17:55:38 too hard Sep 06 17:56:23 mmh Sep 06 17:57:32 I.e. to keep track of which id's were created and which ones haven't been closed etc. Sep 06 17:58:05 yes, we'll have to investigate how to do that Sep 06 18:01:34 I'll dig around some legacy code for inspiration :-) Sep 06 18:01:47 Will let you know when I need some help :-P Sep 06 18:02:43 Could be we need the notification plugin in luna-next to be extended a bit and that's c++. Shouldn't be rocket science by any means though. Might even be able to tackle it myself :-P Sep 06 18:06:04 ok :) Sep 06 18:07:05 I kicked off a build this afternoon on Jenkins wir Sep 06 18:07:45 with the enable_plugins. Charging other TP to test. This one has no space left :s Sep 06 18:34:24 Should we wait for the new Flash version for Linux ? :p Sep 06 18:35:19 ahaha Sep 06 18:55:01 Tofe: I need to check, I think we cannot supply the .so from the ChromeOS images due to license/copyright, but that won't stop users to put it themselves :P Sep 06 18:55:08 The official one we could include probably :p Sep 06 18:55:41 The version I have is 22.0.0.209-r1 Sep 06 18:55:44 So pretty new :D Sep 06 18:55:52 yup, not bad! Sep 06 18:56:17 Anyway I'm not a huge fan of Flash; but it some users need it, then why not Sep 06 18:56:21 if* Sep 06 19:01:11 Tofe: Nobody is, but it's still being used. At least it's slightly better then Silverlight :P Sep 06 19:01:37 WideVine in 5.7 will be more interesting for Netflix etc :-) Sep 06 19:02:33 Getting Flash working in 5.6 paves the way a bit for WideVine too for the future ;-) Sep 06 19:13:30 ka6sox: ping Sep 07 01:15:26 Herrie: Ping **** ENDING LOGGING AT Wed Sep 07 02:59:58 2016