**** BEGIN LOGGING AT Fri Sep 04 02:59:58 2015 Sep 04 06:32:21 Tofe: Your PR fixes the flashcard issue? Sep 04 07:01:53 Herrie: it should yes Sep 04 07:02:39 And I think I will add a LunaWebView to LuneOS.Components, with all the dialogs preconfigured Sep 04 07:16:03 morning Sep 04 07:25:47 morniing Sep 04 07:28:09 Tofe: morning Sep 04 07:32:29 Tofe not sure what you meant by your LunaWebView comment? Sep 04 07:33:37 Herrie|Veer: I see that today, for the various dialogs of the WebView, we use exactly the same code Sep 04 07:34:09 What I just fixed for flashcard could very well reappear in the browser for instance Sep 04 07:34:54 Tofe: Yeah that's true ;) Sep 04 07:35:13 Would make sense to have it once in a central location ;) Sep 04 07:35:44 yup Sep 04 07:42:56 I'm reworking the spinner a bit still ;) Seems Enyo 2 one is 59x58 px and legacy small one was 32px so will scale it down a bit Sep 04 07:43:32 I think I keep the 59x58 for the large one which was 128x128. Tried to CSS stretch it but looks ugly Sep 04 07:46:50 Herrie|Veer: my idea would be like that: https://github.com/webOS-ports/luneos-components/pull/14 Sep 04 07:47:27 Herrie|Veer: for the spinner, you say we are replacing the spinning qml one, but I didn't see any qml modification... did I misunderstand ? Sep 04 07:47:55 Tofe: No this is Enyo 1.0 Sep 04 07:48:16 CSS no QML ;) Sep 04 07:50:37 See #773 for screenshot of the problem. Been debugging it yesterday seems -webkit-transform rotateZ is somehow broken. (It tried to rotate a fixed image from 0 to 359 degrees in 1s). Enyo 2 simply uses an animated gif instead. So I've backported that to o Sep 04 07:50:37 ur Enyo 1.0 this way we can drop the whole -webkit-transform stuff :P Sep 04 07:55:18 ah ok Sep 04 07:56:39 did you try to just remove the -webkit prefix? css transforms should be pretty standard by now. Sep 04 08:11:17 Garfonso: Yeah tried that to no avail Sep 04 08:11:46 Probably animated gif is less resource hungry too :P Sep 04 08:12:27 Tofe: LuneOSComponent looks good :) Sep 04 08:19:10 Did you test it already? Sep 04 08:21:22 Herrie|Veer: not at all! Sep 04 08:21:34 Herrie|Veer: on the other hand, it's not used either :) Sep 04 08:21:43 Tofe: Yeah :P Sep 04 08:23:16 So in Browser I would just import this in WebView.qml with import LunaWebView 1.0 and then just use LunaWebView instead of WebView? Sep 04 08:23:57 no, you already import LuneOS.Components, so imports are ok Sep 04 08:24:15 Ah OK :) Sep 04 08:24:21 just use LunaWebView instead of WebView, and remove the properties that are already set (and identical) in LunaWebView Sep 04 08:24:51 If you do that, take that opportunity to rename WebView.qml to something else, like BrowserWebView.qml :) Sep 04 08:25:09 Yeah was thinking about that :P Sep 04 08:29:36 Let me try that quickly. Should be straightforward :) Sep 04 08:34:51 Tofe: You might want to add experimental.preferences.fileAccessFromFileURLsAllowed: true and experimental.preferences.javascriptEnabled: true as well Sep 04 08:39:32 Other then that it's working fine :) Sep 04 08:45:10 Herrie|Veer: ok, I'll add that Sep 04 08:46:30 Thnx Sep 04 08:46:44 Let me know when done and I'll merge. I'll PR browser now too Sep 04 08:47:14 Done Sep 04 08:49:02 OK let me merge in a minute and will PR browser Sep 04 08:51:39 OK PR done. Sep 04 08:52:33 Nothing too shocking. For the dummy data I copied the FirstUse LunaService.qml and added the history and bookmarks. Sep 04 08:52:56 Might be good to centralize those as well since we're maintaining them in multiple places now? Sep 04 09:01:57 I mean the test/import stuff we now need to duplicate in all QML apps ;) Sep 04 09:11:46 Yeah I still have to see how we could do that Sep 04 09:12:31 The test stuff is only used on Desktop, so I wouldn't like it to be installed on device as a package Sep 04 09:12:49 But maybe we could create a separate repo, dedicated to that Sep 04 09:14:01 Yeah something more central would be good Sep 04 09:14:18 Because now it's required to update each QML app when you make a change :P Sep 04 09:14:34 Should I put something in the bugtracker for it? Sep 04 09:14:48 No rush to change it. Sep 04 09:14:55 But would be good for longer term Sep 04 09:15:12 yes, that's a good idea Sep 04 09:15:26 Do we have the rights to create a new repo in webos-ports ? Sep 04 09:15:45 I guess not Sep 04 09:15:59 Nopr Sep 04 09:16:06 JaMa or morphis could do Sep 04 09:16:10 yep Sep 04 09:16:16 Any suggestions for a name? Sep 04 09:16:53 luna-qml-stubs ? Sep 04 09:17:21 Sounds like a good one ;) Sep 04 09:17:25 morphis: ping Sep 04 09:19:32 Tofe: You see chance to update luna-webappmanager to use LunaWebView? Sep 04 09:21:31 Herrie|Veer: pong Sep 04 09:22:04 Can you create a luna-qml-stub repo for us and give me and Tofe access? Sep 04 09:22:47 This will be used for the mock data/settings in QML apps on desktop (so LS2 calls etc). Currently each app has their own and it's tedious to update them all ;) Sep 04 09:35:23 why not putting that luneos-components? Sep 04 09:36:52 s/that/that into/ Sep 04 09:37:03 morphis: It's only used on Desktop so we don't really want to deploy onto device Sep 04 09:38:37 no problem Sep 04 09:39:05 just create a stubs folder in there and put things in Sep 04 09:39:24 And we can exclude that in the recipe somehow then? Sep 04 09:39:25 as long as you don't put that into the qmake configuration nothing will be installed Sep 04 09:40:48 Tofe: Is that workable for you? Sep 04 09:41:56 morphis: Tofe's in process of moving the WebView into luneos-components so we can have a shared one across browser and luna-webappmanager. I assume you're happy with that? Sep 04 09:43:32 sounds like a good idea, however the WebView used in luna-webappmanager is quite more advanced Sep 04 09:43:43 so as long as that doesn't introduce any regressions I am fine Sep 04 09:45:01 morphis: currently I've just put in there what is common to both. It still inherits from WebView, so users of that component can customize it as they want Sep 04 09:45:23 Herrie|Veer: Creating a stubs folder is fine for me too Sep 04 09:45:48 Tofe: great! Sep 04 09:46:44 morphis: do you see any drawback of putting flickable=true by default for all apps ? Sep 04 09:46:50 morphis: Yeah it allows us to make features and bugfixes in browser and luna-webappmanager at the same time :) Sep 04 09:47:55 Tofe: yes Sep 04 09:48:05 that causes problems for the legacy enyo apps Sep 04 09:48:21 you can overscroll them Sep 04 09:48:23 morphis: they don't have any viewport html tag ? Sep 04 09:49:01 I doubt that Sep 04 09:49:08 and if that works it would be a good thing to add Sep 04 09:49:10 Couldn't we make it conditional? I.e. Enyo 1 = false, Enyo 2 = true Sep 04 09:49:22 Herrie|Veer: would be too easy :) Sep 04 09:49:49 Problem is, with flickable=false WebView ignore the viewport, and I begin to think that it causes problems with gesture events that are not sent to the app, like for Maps Sep 04 09:50:39 Though it's not entirely clear so far Sep 04 12:09:15 Herrie|Veer: ok for merging spinner? Sep 04 12:10:28 Tofe: I think I did already :P Sep 04 12:10:42 I tested it so should work :) Sep 04 12:15:36 ok! Sep 04 12:17:36 Did you see my question on luna-webappmanager? Sep 04 12:19:25 Herrie|Veer: I'll update it Sep 04 12:19:45 I'll do that when at home, otherwise I can't test it Sep 04 12:23:03 Tofe: True ;) Sep 04 12:31:15 Tofe: I can do the PR now you can review. Should be really straight forward and can kick off a nightly :P Sep 04 12:35:18 Tofe: PR is in :P Sep 04 12:37:17 morphis said: LGTM ;) Sep 04 12:37:33 Well it's a no-brainer not much I could break there :P Sep 04 12:53:07 morphis, Tofe: we also can try to set flickable always to true, and control flick interaction by setting "interactive" property of flickable part of webview(base qml "class"). this should be card-specific Sep 04 12:53:33 nizovn: not sure if this works as the "flickable" one for the webview is a bit special Sep 04 12:59:58 Tofe: Nightly running Sep 04 13:00:23 I got a ton of Enyo 1 apps on my N4 so will see if I can still break stuff with this :) Sep 04 13:03:04 morphis: problem is, some Enyo2 apps like Maps are displayed incorrectly if flickable=false, because the viewport isn't taken into account Sep 04 13:03:26 So it looks like we have to find a third way :) Sep 04 13:12:25 Tofe: feel free to try it first to revert that setting for the Webview Sep 04 13:12:31 maybe there is another good way around this Sep 04 13:13:31 We also still have double taps being registered in Enyo 1 apps Sep 04 13:13:39 Not sure where that comes from Sep 04 13:42:48 scoutcamper|away: ping Sep 04 14:20:23 What example of Enyo1 app do we have around here? Sep 04 14:21:48 Herrie|Veer: note that the flickable option of appinfo.js is now correctly parsed, if you want to play around. Of course there's still the issue that it is a global static variable in WebView, but for testing one app it's ok Sep 04 14:22:51 Tofe: seems just type and email are enyo1 Sep 04 14:23:13 Tofe: Tweaks is maybe most obvious Sep 04 14:23:28 Go to Luna Tweaks Sep 04 14:23:59 Then just click a dropdown Sep 04 14:24:08 Often it appears and disappears etc Sep 04 14:24:38 Or when you press Luna Restart it will show "Yes/No" and it will automatically press No Sep 04 14:26:04 That's what happens with flickable=true? Sep 04 14:26:12 In Accounts app (last tab) when you press "Add Account" (I already have my email in there) it goes to AppleID option automatically. Sep 04 14:26:28 Tofe: No this is with current settings. No tweaks Sep 04 14:26:45 oh, ok. So it's worse with flickable=true? Sep 04 14:26:57 I haven't tested it. Sep 04 14:27:20 Just might be related somehow or at least how touch input it passed to Enyo 1 apps Sep 04 14:27:22 ok. We should try flickable=true and adding a viewport tag in index.html Sep 04 14:27:36 (just to see, as we can't do that for other legacy apps) Sep 04 14:28:02 I'll play around with this, already began yesterday evening Sep 04 14:28:09 OK so adding flickable=true works OK now? Sep 04 14:28:13 yes Sep 04 14:28:32 For Enyo 2 apps for example and also for Trello web app? Sep 04 14:28:41 For example what you can test is Memos Sep 04 14:29:25 for Enyo2 apps, it seems to be fine. Maps UI is now correctly shown (it doesn't go over the right side), for instance Sep 04 14:36:03 Add about 10 memos. You couldn't scroll down Sep 04 14:40:25 ok Sep 04 15:16:53 7.<,lp[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[ Sep 04 15:16:53 [[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[[7/1/-[[[[[[[[[[[[[[[[[[[[[[[[[[[777777777777777777777777777777777777777777777777777777777777777777718 Sep 04 15:21:04 Nizovn: Your cat ate your keyboard :P ? Sep 04 15:28:48 oops, sorry, forgot to lock keyboard. yes, it was a cat. btw she managed to enter magic sysrq key combination once, so cats are fairly dangerous. :) Sep 04 15:31:13 LOL Sep 04 15:42:28 cats are evil Sep 04 16:04:19 morphis: ping Sep 04 16:19:57 I've pushed some fixes for Qt 5.5 compatibility to master-next branches in few components and also to jansa/master branch in meta-webos-ports Sep 04 16:20:04 please check it if you're interested in 5.5 Sep 04 16:20:52 I have few issues with newer NodeJS as well, so for now I'm ignoring keymanager and 3 additional node-* packages, someone with NodeJS knowledge would be really helpful to look at them Sep 04 16:39:19 Herrie|Veer: pong Sep 04 16:58:24 Herrie|Veer: I've now PR'ed the move of our browser and cardshell test stubs to luneos-components Sep 04 16:58:43 It didn't seem to break anything for me, both on windows and linux Sep 04 17:21:10 Tofe: Nice :) Sep 04 17:22:08 morphis: Tofe recently reworked banner and dashboards to act correctly. I was thinking about the notifications for new SMS. You have any ideas how to implement these? Sep 04 17:22:41 Tofe: At the gym now will be home in 30 or so mins and flash latest nightly to test the "broken" Enyo 1 apps. Sep 04 17:22:47 Will also check the PR Sep 04 17:25:25 ok Sep 04 17:26:34 JaMa: I guess that would be Garfonso or EricBlade or maybe DougReeder Sep 04 17:29:09 Herrie: I never remarked, is it normal that the Memos app now has a much smaller font and graphics ? Sep 04 17:29:37 Ah no, it's tied to the flickable=true thingy Sep 04 17:29:55 Herrie: flickable=true doesn't solve the scrolling issue on Memos Sep 04 17:30:04 It just doesn't scroll. Sep 04 17:57:39 Tofe: Could be memos is just broken Sep 04 17:58:00 Dunno Sep 04 18:00:45 Tofe: On test/imports/LuneOS/Service/LunaService.qml: Better take the one from Browser since it has the FIrstUse + Browser dummy data as well Sep 04 18:03:19 Oh, didn't I? Sep 04 18:03:35 I don't think so ;) Sep 04 18:03:40 ok, will fix that Sep 04 18:03:59 Also on browser.qmlproject: + importPaths: [ "../luneos-components/modules", "../luneos-components/test/imports", "qml/test/imports" ] Sep 04 18:04:08 Shouldn't we drop qml/test/imports? Sep 04 18:04:34 no, there is still a little import specific to the c++ part of the browser it seems Sep 04 18:04:44 (browserutils) Sep 04 18:04:57 Ah ok Sep 04 18:05:06 Yes nizovn wrote those for icons etc Sep 04 18:10:24 Updated Sep 04 18:10:48 We already began to diverge, it was time to put that in a common place Sep 04 18:24:53 Tofe: Yes that's why I mentioned it ;) Sep 04 18:25:04 Also FirstUse and Phone would need to be udpated for this I think Sep 04 18:28:57 Ah, they also have their own copy of the imports ? Sep 04 18:29:17 Yeah ;) Sep 04 18:29:31 So it was even more needed Sep 04 18:29:38 Not sure about FingerTerm ;0 Sep 04 18:29:39 ;) Sep 04 18:29:55 I think FingerTerm might not use them Sep 04 18:30:12 FirstUse uses it for font sizes + mock data for timezones, countries etc Sep 04 18:30:26 And Phone: Well morphis know that better Sep 04 18:37:41 Tofe: FlashCards lite app now runs :D Sep 04 18:37:48 great Sep 04 18:40:29 I'll try Apollo for the bug that we have outstanding Sep 04 18:40:40 Is for DB size so should have been closed long time ago Sep 04 18:40:47 And have some other apps that were broken before Sep 04 18:40:56 See what those do ;) Sep 04 18:47:15 OK all 3 that didn't work before now work :D Sep 04 18:47:31 Great! Sep 04 18:48:02 On my end, I think that we are not really going to use flickable=true Sep 04 18:48:07 Losungen, HP Twynergy and the FlashCards Sep 04 18:50:07 Spinner also working nicely now :D Sep 04 18:50:15 Things come together well it seems ;D Sep 04 18:50:19 Small bits but still :D Sep 04 18:50:33 Stability & legacy compatibility has been improved a lot over the past weeks ;) Sep 04 18:53:05 yes, it's a good thing to see Sep 04 18:54:27 + lots of small old bugs gone too ;) Sep 04 18:55:34 Do you know why the scaling of the Maps app is wrong ? Sep 04 18:57:09 No? Sep 04 18:57:18 It's really alpha 72ka said Sep 04 18:57:31 He just wanted to get it out to repo before his 2nd child would be born Sep 04 18:57:41 Because he would have little time next half year Sep 04 18:57:50 Seeing how alpha it is it works quite well Sep 04 18:57:57 But I know he has high standards Sep 04 18:58:07 Oh, ok, I thought it was more than alpha Sep 04 18:58:16 then forget about my question :) Sep 04 18:58:17 No really alpha he said ;) Sep 04 18:58:36 I made a few improvements to get it working on the browser etc Sep 04 18:58:44 But 99,9% is his code ;) Sep 04 18:59:02 I'm stress testing Enyo 1 apps now Sep 04 18:59:42 We should definitely keep flickable=false, in the end. In that mode, WebView ignores the html viewport and set its own, corresponding to the size of the WebView item and to the zoomFactor that we set by API in order to have the apps looking as on legacy Sep 04 18:59:51 morphis: ^ Sep 04 19:00:47 If we set flickable=true, we only get issues: the scaling is the one of the html viewport (often 1.0), and the font looks to tiny. Sep 04 19:01:04 It doesn't solve the scrolling in Memos, nor the zooming in Maps. Sep 04 19:02:21 Trello app behaves exactly the same. Sep 04 19:04:31 However, I discovered that my PinchArea in CardGroupDelegate.qml does interfere with app interaction. I've commented it out on device, but I'll try to make it smarter in a real commit. Sep 04 19:05:12 Tofe: Those are all very useful discoveries :D Sep 04 19:07:09 Tofe I think I loaded about 50 apps now onto my N4 Sep 04 19:07:12 Testing them now Sep 04 19:07:45 Tofe: without flickable=true we can't scroll webapps like pivotce, if it is set, than can Sep 04 19:08:52 ah, yes, quite true, I didn't think of that Sep 04 19:09:13 Well we can use it for the apps where needed right? Just not globally Sep 04 19:09:49 I hope so Sep 04 19:12:30 Tofe: So far so good ;) Sep 04 19:12:35 All apps loaded so far ;) Sep 04 19:12:43 And I've tested quite some already Sep 04 19:15:44 OK found 1 that didn't work Sep 04 19:15:46 Rebooting now Sep 04 19:16:47 Lots of resolution issues on N4 but that's to be expected if they were developed for TP Sep 04 19:19:46 Tested 46 app of which 45 ran ;) Sep 04 19:19:59 I call that great success :D Sep 04 19:21:39 yup :) Sep 04 19:22:24 I need to test a bit more but it's looking really good! Sep 04 19:23:16 interesting what if we put webview with flickable=false in scrollview? Sep 04 19:24:43 Tofe: merged your 3 PR's Sep 04 19:24:52 Do you want me to do FirstUse? Sep 04 19:25:28 Herrie: yes please, I'm in the middle of WebKit's code, and I don't feel the courage to do something else right now :) Sep 04 19:26:44 :D Sep 04 19:26:54 morphis: ping Sep 04 19:27:44 Tofe: While you're at WebKit stuff you might have a glimpse at http://issues.webos-ports.org/issues/923 which DougReeder raised about DB's not surviving app close ;) Sep 04 19:28:53 Ah let me just test it first Sep 04 19:28:59 Maybe it solved itself recently Sep 04 19:31:34 Tofe: No still there. Sep 04 19:31:40 But I saw this in the log: Sep 04 15:29:05 mako mojodb-luna[732]: [] [pmlog] DB8 MOJ_SERVICE_WARNING {"sender":"com.palm.eventreporter.LunaAppManager","method":"put","payload":{"objects":[{"_kind":"com.palm.contextupload:1","appid":"com.hominidsoftware.searchablenotes","event":"install"}]},"error":"kind not registered: 'com.palm.contextupload:1'","reqErr":-3970} Sep 04 19:31:40 ok Sep 04 19:31:50 That might explain something? Sep 04 19:31:58 seems related Sep 04 19:32:03 Let me check legacy doctor for com.palm.contextupload Sep 04 19:32:07 ok Sep 04 19:33:26 Tofe: https://github.com/openwebos/luna-sysmgr/blob/1393f0af5dd8d9f0e9cc79627f4acb226f8c8d45/Src/base/EventReporter.cpp#L67 Sep 04 19:33:34 Seems some error reporter stuff or something? Sep 04 19:34:27 Tofe we seem to have /etc/palm/activities/com.palm.contextupload/contextupload.json with https://bpaste.net/show/c2cb6fe1919c Sep 04 19:35:04 I guess you can focus on WebKit for now ;) Sep 04 19:35:07 I'll add this to the bug Sep 04 19:35:10 Might be more useful Sep 04 19:37:27 Might be more of a Garfonso or morphis thing to look into & sort Sep 04 19:46:34 Do we actually need the whole eventreporter or maybe we can dump it altogether? Sep 04 19:46:38 morphis: ^ Sep 04 19:55:31 Tofe: https://github.com/webOS-ports/org.webosports.app.firstuse/pull/20 Sep 04 21:02:38 Tofe: new nightly running, let's hope nothing breaks :P Sep 04 21:06:13 Seems mako built correctly... It's now doing Tenderloin, afterwards image should be synced Sep 04 21:06:19 Will test it to see if all is OK Sep 04 21:10:14 Herrie, Tofe, Zap Photoshare is Enyo 1, if that helps. Sep 04 21:18:33 DougReeder: We have some examples ;) Sep 04 21:18:41 DougReeder: You any good with nodejs? Sep 04 21:31:19 DougReeder: You any good with NodeJS? Sep 04 21:42:31 Tofe: So far so good :) Sep 04 21:42:36 Will test some more apps :D **** ENDING LOGGING AT Sat Sep 05 02:59:58 2015