**** BEGIN LOGGING AT Tue Sep 08 02:59:58 2015 Sep 08 03:18:41 * DougReeder waves hello Sep 08 04:35:34 DougReeder: Hello :) Sep 08 04:35:40 We just put out new release :) Sep 08 04:36:15 LOt more stable dashboard & banners are working almost like legacy ;) Sep 08 04:36:28 And many more Enyo 1/2 apps work now without breaking stuff completly in the system Sep 08 04:53:49 Your dash calculator should work too ;) Sep 08 04:53:57 Just after app is closed dasb will close too for now Sep 08 04:54:08 That requires some window management rework which Tofe is looking into Sep 08 05:23:59 Herrie, pong Sep 08 05:24:16 Holiday here Sep 08 05:41:41 Herrie: pong Sep 08 05:42:51 morning morphis Sep 08 05:44:04 Morning! Sep 08 05:45:07 morphis: while you're there, I also have a question: why was the window parenting introduced ? Can I get rid of it ? It seems like Legacy didn't have any notion of "main" window... bug I may be wrong here. Sep 08 05:46:15 Tofe: it did as far as I know Sep 08 05:46:39 it at least had the differentiation between windows and child windows Sep 08 05:53:22 morphis: Sep 06 10:00:31 Speaking of which, I'd like to know the rational for the onStageReadyTimeout introduction Sep 08 05:54:01 morphis: JaMa setup changelog access for me and pushed out release late last night Sep 08 05:54:53 Herrie: great Sep 08 05:55:38 Tofe: the onStageReadyTimeout relates to the PalmSystem.stagePreparing/stageReady calls Sep 08 05:55:49 so when an app is preparing the loading animation is shown Sep 08 05:56:05 and once it calls stageReady it becomes visible Sep 08 05:56:21 however to not let it load forever we have the timeout to force it visible after some time Sep 08 05:58:18 morphis: You aware some legacy apps (in HP Catalog even) were using custom fields in appinfo.json? Sep 08 05:58:30 no Sep 08 05:58:39 Found that while stress testing some Enyo apps over the weekend Sep 08 05:59:03 That didn't run for some reason and found that to be the cause it seems :P Sep 08 05:59:42 Seems not to be documented but HP allowed it and it seemed actively used by quite some apps :P Sep 08 05:59:58 For http://issues.webos-ports.org/issues/923 do you know where this IndexedDB is stored? Sep 08 06:04:00 Herrie: somewhere in $HOME Sep 08 06:12:08 morphis: OK Sep 08 07:00:27 morning Sep 08 07:03:28 morphis: ping Sep 08 07:06:23 Herrie|Veer: pong Sep 08 07:06:48 morphis: Herrie|Veer: is that pulseaudio issue blocking Fido upgrade or can we switch and resolve it in follow-up commits? Sep 08 07:07:20 JaMa: I would say we can fix it with follow up updates but before the next release Sep 08 07:07:31 You OK with JaMa switching from Dizzy to Fido once he sorts the issue with PulseAudio on N4? Sep 08 07:08:05 Anything specific that needs testing still? I didn't see anything obviously broken on N4 and TP Sep 08 07:08:20 the next one after Olla or after Olla is out? Sep 08 07:11:10 Hmmz mobile data dropped :S Sep 08 07:11:18 Anyway I pushed Olla out last night Sep 08 07:11:36 So you'd only have to rebase Fido to have the latest commits from the last 4 or so days from Dizzy Sep 08 07:11:47 Should be about 10 or so commits I guess Sep 08 07:12:21 morphis: Anything specific you'd want me to test? I didn't see anything obviously broken on TP and N4 while I was testing except for PulseAudio on N4 Sep 08 07:13:05 I don't have a N7 to test on Sep 08 07:13:09 And should still check emulator Sep 08 07:13:15 Herrie|Laptop: no, just do exploration testing Sep 08 07:13:43 morphis: OK Sep 08 07:13:59 I'll give the emu a go in a bit, downloading image now, just wifi here is a bit flaky so will take about 20-30 mins Sep 08 07:33:58 morphis: thanks for the info Sep 08 07:42:25 JaMa: VBox emu is stuck on "Booting the kernel" for me on unstable. Could be my machine. If someone else could test would be good Sep 08 07:51:51 Herrie|Veer: i'll not be able to test until tonight, unfortunately Sep 08 07:52:50 Tofe: no rush, maybe JaMa has chance or nizovn when he's around. I know he uses emulator quite a bit :) Sep 08 14:35:47 morphis: you're right, there was a notion of parent/child in legacy; however, it seems like a card could only have one child card, and then that child card would receive all the events. Looks like a dialog-like card to me, or something like that. I'll keep reading legacy code. Sep 08 14:36:15 Tofe: ok Sep 08 14:51:33 morphis: and then, what do I find ? http://logs.nslu2-linux.org/livelogs/webos-ports/webos-ports.20130201.txt (look for childcard :p ) Sep 08 14:57:05 morphis: in the end, I think we simply are creating far too many child cards. There is no child card, apart if it is explicitely asked at window creation, I'd say Sep 08 14:58:14 I still have to find one concretre example of a card started as a childcard Sep 08 15:00:51 Still, knowing which winId started a new card is a useful information for the window management Sep 08 15:01:53 (but it can be dropped after the card has been put in the correct stack) Sep 08 15:16:31 Herrie|Veer: ping Sep 08 15:17:13 Herrie|Veer: you've pushed one commit to meta-webos-ports/master, I think you wanted it in dizzy Sep 08 15:21:47 JaMa: The new release stuff? I did to both :P Sep 08 15:22:04 Seems unstable qemux86 gets stuck at booting kernel :l( Sep 08 15:25:06 Herrie|Veer: it's booting correctly, but luna-next gets killed with: Could not initialize egl display Sep 08 15:26:18 Herrie|Veer: ok, I've replaced the one in master branch with rebased master Sep 08 15:26:24 nizovn: It's stuck here on my laptop with "booting kernel" or something Sep 08 15:26:28 Herrie|Veer: so fido is also updated to have everything from olla Sep 08 15:26:40 ok, I can look into this one Sep 08 15:26:51 easier than N4 pulseaudio issue as I don't have N4 :) Sep 08 15:27:00 OK I think we're OK to switch if you're able to solve N4 and qemu issue before next release Sep 08 15:27:22 Herrie|Veer: same for me Sep 08 15:27:25 I'll quickly test journalctl when I'm home tonight for TP to make sure nothing sticks out there Sep 08 15:27:36 Then we should be OK to switch Sep 08 15:27:46 QT 5.5 comes after Fido right? Sep 08 15:28:13 I can try to drop pulseaudio-droid upgrade to see if it builds (I think it wont, because that was probably the reason why I did the upgrade) Sep 08 15:28:20 then will need someone to test it on device Sep 08 15:28:26 yes Qt 5.5 after Fido Sep 08 15:28:38 it's already in master branch :) Sep 08 15:29:02 but I got first successful build for qemux86 just few days ago and no runtime testing yet Sep 08 15:29:28 and now I'm switching that to gcc-5.2, so it probably won't build again Sep 08 15:29:46 JaMa: It seems the pulseaudio-droid one could be easy. I recall it was just saying a port wasn't correct or something. Shouldn't be too hard to fix? Sep 08 15:30:10 if you say so :P Sep 08 15:30:47 JaMa: I don't touch kernel and C++ stuff Sep 08 15:31:30 57 commits in pulseaudio-droid upgrade Sep 08 15:32:12 but this one should build with pulseaudio-5.0 as well, so we can test it in dizzy and see if it breaks the pa as weel Sep 08 15:32:34 OK Sep 08 15:32:42 I'll leave that up to you ;) Sep 08 15:32:44 Herrie|Veer: can you try to cherry-pick 79f30ef3563e6371dfeffbd241a37ae740818142 to your dizzy build? Sep 08 15:32:53 I'm happy to test drive stuff Sep 08 15:32:53 and see if it's broken the same? Sep 08 15:32:57 Let me check Sep 08 15:34:42 other option is to upgrade it even more, there is newer version and one commit looks a bit promissing.. Sep 08 15:35:25 Well it can't be more broken, so why not try? Sep 08 15:35:44 Or at least not much more broken ;) Sep 08 15:41:28 JaMa how do I find above commit? Sep 08 15:41:41 Still git n00b Sep 08 15:44:38 Ah found it :) Sep 08 15:45:07 You want me to cherry pick that in dizzy? Sep 08 15:46:13 Herrie|Veer: yes please and after the test Sep 08 15:46:27 you can cherry-pick ca3d5b8 on top of that which will upgrade it even more Sep 08 15:46:32 and possibly fix it :) Sep 08 15:48:41 Ok so I cherry pick to dizzy do new testing build, check it on mako & tenderloin and then pick ca3d5b8 and rinse & repeat? Sep 08 15:50:52 Herrie|Veer: don't forget to lather Sep 08 16:00:04 Herrie|Veer: yes and you can probably test only on mako Sep 08 16:00:14 Herrie|Veer: if the audio was working on tenderloin Sep 08 16:03:53 OK testing build running Sep 08 16:15:59 JaMa: Where can I find ca3d5b8? Sep 08 16:16:11 If you have github url would be helpful Sep 08 16:18:15 Herrie|Veer: it's in jansa/master branch Sep 08 16:18:32 but if you call just "git remote update && git cherry-pick ca3d5b8" you should get it Sep 08 16:23:45 JaMa: I mainly use gh web and windows gui client :P Sep 08 16:47:12 JaMa: build failure... Sep 08 16:47:16 http://jenkins.nas-admin.org/job/luneos-testing_mako/lastBuild/console Sep 08 16:57:50 Tofe: ping Sep 08 16:57:55 pong Sep 08 16:58:17 You have any idea how I can monitor activity so where for http://issues.webos-ports.org/issues/923 the DB is created? Sep 08 16:58:45 ls2-monitor ? Sep 08 17:11:02 ah I thought you will test it in local build Sep 08 17:11:16 I'll drop it and create new build for you with unstable Sep 08 17:11:44 JaMa need to setup my local build env Sep 08 17:12:36 I usually don't touch anything that requires a build (i.e. usually only do QML and Enyo stuff) Sep 08 17:15:14 For the Testr on the TP, opening a new window doesn't work? Sep 08 17:25:28 I think I found an example of a "childcard" Sep 08 17:26:44 in the diagnostic app, start the "Compass" test; it will popup a dialog on top of the current UI. Could it be that one? Sep 08 18:12:07 Herrie, morphis: https://github.com/webOS-ports/luna-webappmanager/pull/32 has been updated. The previous fix attempt has been wiped out, and now there is no parent/child relation between the various cards. However, I kept the notion of a main window. Sep 08 18:12:34 It seems to behave correctly w.r.t. Testr, multiple windows and dashboards. Sep 08 18:14:41 I will eventually reintroduce a child window for the "childcard" window type, if we encounter it. But it might as well be that this window type was only used internally by the LSM UI... Sep 08 18:39:28 Tofe: On TP I noticed that the home button cards and maximizes on a single press Sep 08 18:39:35 Might be something small Sep 08 18:39:44 elvispre: Timezone issues should be gone ;) Sep 08 18:40:32 Herrie: Yeah, I saw them mentioned in the change log for the new release. Sep 08 18:41:39 Maybe I can remove some of the special cases from the Settings display code if they're a bit more regular now. Sep 08 18:43:04 elvispre: They should be ;) Sep 08 18:43:38 I will revisit them then :-) Sep 08 18:43:58 This one is still wrong: https://github.com/webOS-ports/luna-init/blob/webOS-ports/master/src/ext-timezones.json Sep 08 18:44:20 But I believe we don't use it and a new one is generated on the fly by gen-ext-timezones.py Sep 08 18:45:11 Delete it and see what breaks? Sep 08 18:47:00 Yeah that might be an option :P Sep 08 18:47:54 I tried building myself a new image just now but Pulse Audio won't build with the -Wpedantic flag. Is everything all experimental again already? Sep 08 18:49:45 elvispre: which image did you build ? unstable ? testing ? testing should be alright, I thin Sep 08 18:49:47 k Sep 08 18:52:00 Tofe: testing for mako. I'm starting from a couple of weeks old build and doing a make update on top of it. Perhaps I'll try a clean one. Sep 08 18:53:06 well, the failure on pulseaudio reminds me of the latest changes Herrie & JaMa did for unstable Sep 08 18:57:44 Tofe: Hmm. It is a long time since I chose the variant. I think I said "testing". It doesn't matter. I can always download a pre-built image when I get stuck like this. Sep 08 18:58:21 Herrie: ^ did you apply something related to pulseaudio in testing? Sep 08 19:02:57 elvispre: ah yes I can see the patch to pulseaudio was also applied to dizzy in meta-wop in order to test a bit for the next release, but it failed Sep 08 19:03:47 elvispre: you can simply checkout b00bc3e5d9a78e06386b9802b415ac083e26e52a in meta-webos-ports, that should fix your build for today Sep 08 19:04:13 Tofe: Oh well. No big deal for me... Oh cool! Thanks. Sep 08 19:05:35 Tofe: Yeah to test pulseaudio... Sep 08 19:08:15 Herrie: no pb, if that help JaMa for converging on fido, it's worth it Sep 08 19:12:11 Tofe: Yeah Sep 08 19:39:16 Tofe: Sep 08 21:38:35 mako LunaWebAppManager[1524]: https://minego.net/macawtwitter/?create=0.4291283031925559&oauth_token=0LD8PgAAAAAAO1clAAABT652K9E&oauth_verifier=3G6NJffhNhgRKo1dKflB2heZrjkKlntk:44: JS ERROR: SyntaxError: DOM Exception 12: An invalid or illegal string was specified. Sep 08 19:39:40 I get this after Macaw wants to return after I authorize access on Twitter Sep 08 19:39:45 This used to work before Sep 08 19:39:59 This is not tested with your latest PR yet Sep 08 19:43:53 Tofe: YOu have IPK for me to test your latest PR? Sep 08 20:08:25 Herrie: eventually yes Sep 08 20:09:06 Herrie: https://dl.dropboxusercontent.com/u/4679068/luna-next/luna-webappmanager_0.2.0-8%2Bgit2%2Bd458383343-r0.9_armv7a-vfp-neon.ipk Sep 08 21:08:04 Herrie: Tofe: there is fix in python-tz-native to fix build luna-init with new tzdata Sep 08 21:08:18 Herrie: Tofe: not sure if it's related to issue you discussed earlier :) Sep 08 21:13:26 JaMa: There's a flag in gen-ext-timezones.py to not guess stuff ;) Sep 08 21:13:30 That worked it seems Sep 08 21:13:49 JaMa: https://github.com/webOS-ports/luna-init/commit/f5563f1af67fbfe33f9c386b95ae8f2524450e5a Sep 08 22:23:11 Tofe, morphis, legacy had a concept of a “main” window for an app. Mojo and Enyo1 handled that in different ways. I don’t know if the non-main windows were “children” of the main window. Sep 08 22:27:53 The main window could be a card or a non-visible window. Sep 08 22:28:46 If the primary window was closed while other windows were open, another window would become primary. Sep 08 22:48:57 Herrie: https://github.com/webOS-ports/meta-webos-ports/commit/ec74da5adee11f97876dcc05da35858c851ebd70 Sep 08 22:49:10 Herrie: this is the tz-native upgrade and fix for newer tzdata Sep 09 00:57:06 Initial tests of PopupCalc under Olla have the dashboard & popup not appearing when the app is launched, but appearing when the app is re-maximized. Sep 09 01:00:08 The system popup (containing the calculator) is not the specified height, but that’s not new. Sep 09 01:00:30 Banner messages dissapear as they ought. Sep 09 01:00:35 that’s progress! **** ENDING LOGGING AT Wed Sep 09 02:59:59 2015