**** BEGIN LOGGING AT Tue Jun 07 02:59:58 2016 Jun 07 07:21:45 morninv Jun 07 07:21:53 s/v/g Jun 07 08:22:07 Tofe: ping Jun 07 09:12:34 Seems I might have found the cause of the problem with the windowParams :D Jun 07 09:12:49 Just not 100% sure what's the best way to solve it. Jun 07 09:16:31 Legacy returns PalmSystem.launchParams as an empty string, however on LuneOS we get {} back. var a is set based on this inside the enyo.requiresWindow function. Subsequently there's a check for !a which is true on legacy but false on LuneOS. Jun 07 09:45:46 Herrie|Pre3: pong Jun 07 09:47:13 Tofe: Just writing you an email Jun 07 09:47:38 oh, ok :) Jun 07 09:47:41 Cause for the issue is that PalmSystem.launchParams is {} where it shouldn't be set at all. Jun 07 09:47:50 With some links for code reference Jun 07 09:48:26 ok, then maybe the fix is in webappmanager, and could be straightforward Jun 07 09:51:12 Tofe: OK email sent. Tofe: Yeah I guess so. I could put a nasty workaround in Enyo 1.0 but better to sort at source. Jun 07 09:51:40 Might solve some other issues we see as well like Macaw Enyo callback not working and maybe FilePicker type issue Jun 07 10:06:23 Herrie|Pre3: I see two places to check Jun 07 10:06:25 https://github.com/webOS-ports/luna-next-cardshell/blob/master/qml/LaunchBar/FullLauncher.qml#L354 Jun 07 10:06:46 https://github.com/webOS-ports/luna-next-cardshell/blob/master/qml/LaunchBar/LaunchableAppIcon.qml#L29 Jun 07 10:07:47 and maybe I missed one somewhere else; difficult to say without a debug session :p Jun 07 10:08:11 but that should be easy to test, at least Jun 07 10:08:40 the default value should be "", not "{}" Jun 07 10:09:12 Tofe: Yeah, let me just try that in a minute :) Jun 07 10:09:23 ok :) Jun 07 10:13:40 FullLauncher.qml seems to do the trick :P Jun 07 10:13:50 Let me revert that and try the other one as well. Jun 07 10:14:56 good Jun 07 10:15:04 anyway we should change both Jun 07 10:15:27 LaunchableAppIcon.qml doesn't solve it. Jun 07 10:15:38 I should make it "" there too? Jun 07 10:16:16 yes Jun 07 10:16:36 the default value should refeclt the one of legacy Jun 07 10:19:49 OK I'll PR that and test Macaw Enyo for fun in a bit :P Jun 07 10:19:59 That required quite some debugging lines LOL Jun 07 10:20:37 yeah, js debugging can be cumbersome... Jun 07 10:20:53 but that was a good catch :) Jun 07 10:24:00 Tofe: Yeah especially in Enyo spaghetti LOL :P Jun 07 10:24:33 I also noticed our window name convention is different. Not sure that will cause any issues. Jun 07 10:26:12 name? Jun 07 10:27:18 Yeah the name for the window Jun 07 10:28:00 We have window2 for example while legacy does accountManager_AccountsView_customAccountsUI_crossAppUI Jun 07 10:28:09 I guess the app don't base their logic on that... at least I hope so! Jun 07 10:29:09 I hope so too, just was sometthing I noticed while debugging Jun 07 13:38:33 DougReeder: Seems we have found the cause of the issue :P Jun 07 13:39:05 We were passing {} instead of "" which caused funny things for windowParams in Enyo 1. Jun 07 13:57:44 Garfonso: ^ Should be in the next nightly :) You shouldn't be needing your template workaround anymore and editing accounts might work too :) Jun 07 14:16:13 Herrie: that's great news. Jun 07 14:22:37 Garfonso: It was something extremely small and simple to fix in the end... We were passing "{}" as initial windowParams while legacy did "". A check for the value would become false which lead to them not being populated like they should :P Jun 07 16:35:52 Garfonso: Tracked it back to https://github.com/webOS-ports/enyo-1.0/blob/webOS-ports/master/framework/source/palm/system/windows/manager.js#l171 and l172 Jun 07 16:36:14 Fixed with: https://github.com/webOS-ports/luna-next-cardshell/commit/c02c45b4a25de5709d703959342062996392c22e Jun 07 16:36:18 New nightlies are running now Jun 07 17:24:14 Herrie, one could argue that was a poor design decision by Palm, but you are right that we should match what webOS does here. Jun 07 17:28:55 what, the "" instead of {}? Jun 07 17:33:12 The braces are valid JSON, while the empty string is not. But we should stick with the empty string for compatability. Jun 07 17:34:11 empty string is valid json, is it not? it's the json for a string. Jun 07 17:36:06 DougReederG3: Well both should be fine, but don't want to tinker too much with Enyo 1.0 tbh Jun 07 17:36:16 And in this specific case it caused havoc :P Jun 07 17:38:01 GodGinrai, the empty string is represented in Json by two double quotes. Jun 07 17:40:09 DougReederG3: exactly. Hence, it is not invalid json Jun 07 17:44:43 Yeah, all primatives by themselves are valid JSON Jun 07 17:49:58 I would imagine, though, that parsing that empty string produces undefined or null, which is a ton easier to check for than an object with no keys. Jun 07 17:56:43 Herrie that fix should help in several places; I look forward to finding out what works now. Jun 07 17:59:34 Andolamin: That's what it did as you can see in above link ;) Jun 07 17:59:45 When we passed "{}" this would fail ;) Jun 07 17:59:55 And our windowParams went missing ;) **** ENDING LOGGING AT Wed Jun 08 02:59:59 2016