**** BEGIN LOGGING AT Tue Jun 23 02:59:58 2015 Jun 23 05:18:25 morning Jun 23 05:41:45 elvispre: morning Jun 23 05:41:48 Merged your PR ;) Jun 23 05:41:54 Working great so far :) Jun 23 05:42:11 It's now up to morphis or someone else to expand the audio service for the other features ;) Jun 23 09:44:51 morning, i have a question about the latest stable of luneos, when i try and add my google account using the accounts app, it doesnt show a complete keyboard, i can therefore not add my account details. Jun 23 09:45:26 brobostigon: What device are you using? Jun 23 09:45:46 Herrie|Veer: nexus 4 Jun 23 09:46:10 OK well that should work. What are you missing exactly? Jun 23 09:46:23 At least for English keyboard layout Jun 23 09:46:33 half the letters, the spacebar and a few other elements. Jun 23 09:46:34 We're still working on improving the others Jun 23 09:46:40 Hmmz Jun 23 09:46:54 Let me try on mine Jun 23 09:47:00 ok, ty. Jun 23 09:48:55 also i have a question about system updates, do what we call ota updates work, ? Jun 23 09:49:04 You're trying Google C+Dav or regular email? Jun 23 09:49:14 the former. Jun 23 09:49:55 brobostigon: We've been recently working on the update mechanism so OTA can work (though manually through Settings app), not sure if that work has been completed yet. Jun 23 09:50:16 ok, :) Jun 23 09:53:44 Lots of work in progress in many areas while also updating core OS parts so we occasionally have some regressions too ;) Jun 23 09:54:16 :) Jun 23 09:54:34 did i rightly observe that keyboard issue/ Jun 23 09:54:38 ? Jun 23 09:56:17 Not sure it doesn't show me any vkb or login screen here. Got very poor data connection + no wifi here in office :'( Jun 23 09:56:38 ah, ok. Jun 23 09:56:46 Ah well it works now Jun 23 09:56:52 But I have latest nightly Jun 23 09:56:56 keyboard included? Jun 23 09:57:03 English layout is fine there it seems Jun 23 09:57:25 ok, i shall have to try the nightly then, and check the keyboard layout. Jun 23 09:58:16 German, French, Russian, Swedish, Danish should also be OK now :) Jun 23 09:58:27 nightly and stable? Jun 23 09:59:33 nightly. Stable only English works. Nightly is not very stable currently due to a segfault issue in Luna Next Cardshell/QT 5.4.1 and/or QT Wayland but that will hopefully be addressed soon. Jun 23 10:00:18 New stable release should be out early each month. But the currently nightly issue is a release blocker so it depends a bit if we're able to solve it on time. Jun 23 10:00:24 ok, in theory, i should try and set the keyboard layout to english again? Jun 23 10:00:32 Yeah Jun 23 10:00:38 Not sure what you're using? Jun 23 10:00:44 ok, i will try that, thank you. Jun 23 10:01:05 the last stable on nexus 4. Jun 23 10:01:14 Language I mean Jun 23 10:01:24 that i am just checking. Jun 23 10:01:48 In latest release I think we only tweaked the English layout, since then I've been tweaking the others as well :) Jun 23 10:04:05 Just press the globe icon and select EN (when you have the globe icon of course) Jun 23 10:16:33 ok, i have account open, and have pressed google c-dav Jun 23 10:17:12 yep, its not showing a complete keyboard. Jun 23 10:17:50 When you go to browser or Just Type you have the same? Jun 23 10:18:24 nope, there its showing a full english keyboard. Jun 23 10:19:27 It works OK for me in the latest nightly so probably a bug we sorted already ;) So with next stable I guess it'll be fine :) Jun 23 10:19:59 Lots of work has been done on the virtual keyboard over the last month so difficult to keep track of every little change/fix Jun 23 10:20:14 ok, thank you, :) Jun 23 10:21:47 Normally I would suggest you to try the nightly but cannot really recommend that currently because of the issue described above. You could still do it though to verify your issue is indeed sorted ;) Jun 23 10:23:53 Herrie|Veer: that issue is present in release as well Jun 23 10:27:57 nizovn: Might be but not in nightly? Jun 23 10:28:22 Latest release is already 3 weeks ago lots has been changed/fixed since ;) Jun 23 10:34:08 i'm a little confused :) so, we are talking about segfault issue, then it's present in both release and nightly Jun 23 10:36:02 another bug, did a reboot, my google account i added vanished. Jun 23 10:37:14 Herrie|Veer: i found a way of changing the kayboard, with the kayboard open in the borwser, hit the globe button, hit en, openthe keyboard in accounts, and it showed a normal keyboard. Jun 23 10:40:54 nizovn: You're right. Just the vkb bug might be different from segfault one. Jun 23 10:41:21 One the Google account might be related to the segfault once that's sorted we'd need to double check Jun 23 10:42:06 ok, so i should wait and try the next stable now? Jun 23 10:45:25 You can always try nightly to see if that sorts vkb issue Jun 23 10:45:31 brobostigon: maybe you can provide screenshot of broken keyboard? Jun 23 10:46:23 nizovn: that was a problem i was able to fix, as i described above. let me try. Jun 23 10:48:38 and sound and vibration dont seem to work, and calls just seem to get silently rejected, no indication of calls at all. Jun 23 12:02:30 brobostigon: Sound should work. I can play ringtones in Settings app and also Youtube worked for me (though no hw video acceleration yet). For calls: This is currently work in progress, most backend bits are there they just need to be linked together. Jun 23 12:03:40 Herrie|Veer: ok, let me try it again. ah i see, thats good, :) Jun 23 12:05:51 It's still in alpha stage, so expect bugs all over the place. We had to rewrite many things from scratch because what HP open sourced was quite limited. SMS (and data connection when you're not on a MVNO) should work as well :) Jun 23 12:06:26 Fortunately we can piggyback a lot from Jolla/Ubuntu as well since they are using quite the same base as us ;) Jun 23 12:06:36 For many Android related things at least Jun 23 12:07:10 libhybris, yes, i work on the sailfish port for nexus4 as well. Jun 23 12:07:23 :) Jun 23 12:07:36 have, i should say. Jun 23 12:07:38 Yeah Sailfish has a lot bigger team ;) Jun 23 12:07:52 So they progress quicker ;) Jun 23 12:08:22 Vibration is there in the backend I think (if I recall nizovn did that) just isn't hooked up yet. Jun 23 12:08:29 very true, yes Jun 23 12:08:49 ah. Jun 23 12:11:05 which libhybris version is on the nexus4 port? Jun 23 12:11:12 We don't mind help of course ;) Especially on the low level stuff ;) Jun 23 12:11:28 i think i can help with that, :) Jun 23 12:11:49 We have quite some devs who can do QML, Enyo, JS, graphics etc just low level kernel and components stuff is more tricky Jun 23 12:12:45 i agree, my day job is designing car control systems, so i can help there. Jun 23 12:13:19 morphis should be able to answer that when he's around, maybe Tofe also knows. It seems quite old from what I can see Jun 23 12:13:38 ok, Jun 23 12:13:40 All can be found on Github in general http://github.com/webos-ports/ Jun 23 12:13:59 ty, :) Jun 23 12:14:30 All is open source where with Sailfish some things are not available to public I think? Or that's specifically for Jolla where some things are not open source? Jun 23 12:15:18 thats correct, yes. Jun 23 12:16:03 like the libhybris module for geoclue, they havent opened, so sailfish as a port, cant directly use gps for example. Jun 23 12:16:37 Same for QT Lipstick stuff I think Jun 23 12:16:39 even when test_gps returns gps location data. Jun 23 12:20:15 I think nizovn wrote something for us ;) Jun 23 12:20:29 We do have GPS in Testr app I think Jun 23 12:22:37 But only wifi based currently I think there was some limitation Jun 23 12:23:04 Had to do with Geoclue version from what I recall Jun 23 12:23:20 But believe work is done to upgrade to newer Geoclue Jun 23 12:26:06 ah i see. Jun 23 12:27:16 But all our code is 100% open source :) Jun 23 12:27:30 Including build environment etc ;) Jun 23 12:46:31 :) Jun 23 12:50:12 brobostigon: from what I remember, we are not that far from upstream of libhybris Jun 23 12:51:02 Herrie|Veer: gps is not implemented yet Jun 23 12:51:48 nizovn: Testr DOES return a quite accurate location for me though ;) Jun 23 12:51:49 from what i have experienced from libhybris on sailfish, v.11 has issues on the nexus4, compared to v.10.1 Jun 23 12:52:42 We're still on CM 10.1 = Android 4.2.2 for the Nexus 4. For the Touchpad we're on CM 11.0 Jun 23 12:53:10 ok, Jun 23 12:53:35 morphis was working on a CM11 update but we ran into a critical issue there :'( Jun 23 12:53:49 someone interested in upgrading nodejs in our builds to 0.12? Jun 23 12:54:10 Herrie|Veer: yeah, mozilla location service is pretty exact :) Jun 23 12:55:39 JaMa that would either be Garfonso or DougReeder I guess to say something useful about that ;) Jun 23 12:56:50 ok, will keep it in fido for now Jun 23 12:57:20 I'm sure they'll let us know if there's something interesting they'd like to have Jun 23 12:57:38 How is fido going? Any progress with the remaining issues? Jun 23 12:58:05 yes, nodejs build for qemuarm was the last issue detected in last round of builds Jun 23 12:58:44 Did you do some on device testing as well? Or just build testing? Jun 23 12:59:00 build testing only for now Jun 23 12:59:13 qemux86 runtime testing long time ago Jun 23 13:00:05 If you have any images to test I have a Mako and Tenderloin laying around just for this purpose :) Jun 23 13:01:08 ok, please try latest images from http://build.webos-ports.org/luneos-unstable/images/ when you have time, but no promisses on funcionality :) Jun 23 13:01:54 JaMa: I guess I'll get to that tonight :) Jun 23 13:30:44 So far as I know, we haven’t been updating node.js on our own, we’ve been using the one in OpenWebOS. Jun 23 13:31:09 Is there something important in node v0.12 we need? Jun 23 13:32:39 we're not going to get more nodejs updates from OpenWebOS and there is recipe for v0.12 in meta-oe, so we can drop the recipe inherited from OpenWebOS Jun 23 13:32:54 and I've already fixed qemuarm build in meta-oe recipe Jun 23 13:33:19 DougReeder: Since LG is not updating Open webOS in the public for now I think we need to go our own path probably Jun 23 13:33:51 It's morphis' decision ultimately. Some LG birdie might give us a clue where they're at internally.... Jun 23 13:33:55 Yes. I’ve been thinking some other thing were higher priority. Jun 23 13:35:43 For LG or us? Jun 23 13:35:59 For us. Jun 23 13:36:01 JaMa constantly updates various bits as part of OpenEmbedded releases ;) Jun 23 13:36:55 As soon as I get the next release of Serene Notes out, I’d like to add Promise support to webos-lib. Jun 23 13:37:27 Then I’d like to get automated testingt set up, so we don’t get regressions. Jun 23 13:37:27 So we keep updated with "latest and greatest" (stable?) bits for the OS. And we don't run into outdated stuff that quickly like legacy did due to it's large amount of customizations Jun 23 13:37:56 DougReeder: Automated testing would be really good :% Jun 23 13:37:58 :) Jun 23 15:32:37 Anyone by any chance here knows/speaks Hebrew or Arabic? Working on those vkb layouts now just have no clue if I'm doing it OK :P Jun 23 15:37:30 IIRC, there's a dev over in #e who does? Jun 23 15:38:58 #e ? Jun 23 15:39:18 enlightenment's IRC channel Jun 23 15:39:45 I'm trying to confirm who it was right now Jun 23 15:40:00 I'll ask them if they'd mind helping you once I confirm Jun 23 15:44:21 Herrie|Veer: TAsn in #e said he's willing to help you. Jun 23 15:44:49 Herrie|Veer: he should be able to help you with the Hebrew stuff. Don't know about the Arabic. I didn't mention that to him Jun 23 15:45:05 Yeah usually those don't combine well :P Jun 23 15:45:18 I just want to know if my layout makes sense Jun 23 16:40:38 DougReeder: automated testing sounds good Jun 23 16:40:43 for the js bits? Jun 23 16:40:58 That’s what I’d initially focus on. Jun 23 16:41:05 JaMa, DougReeder: testing is really essential when we upgrade nodejs Jun 23 16:41:16 don't want this to break a lot things which might get hard for us to fix Jun 23 16:41:58 I’m trying to decide when auto tests should be run. I’m thinking (for Enyo apps) when tools/deploy.sh is run. Jun 23 16:42:26 And, of course, we need to figure how to tell bitbake to run them. Jun 23 16:43:01 DougReeder: that is possible Jun 23 16:43:20 JaMa: what do you think? Jun 23 16:43:56 I’m thinking we’ll run the tests in PhantomJS. It’s not identical to the LuneOS runtime, but it should be close enough. Jun 23 16:46:25 hm Jun 23 16:46:26 The bigger change is getting people to grok that breaking the tests is like breaking the build. Jun 23 16:47:28 I.e. if a test fails, you need to understand why, and fix either the code or the test. Jun 23 16:47:33 right Jun 23 16:47:43 the best would be really if that is done with std. tools Jun 23 16:47:49 not especially with the LuneOS runtime Jun 23 16:48:02 as that one should be very close or equal to a std. webkit/chrome Jun 23 16:52:25 morphis: yes, auto testing would be awesome Jun 23 16:53:01 fwiw: I've backported my nodejs fix for qemuarm to 0.10 version we're using now, so we can postpone 0.12 to Yocto 1.9 or just apply separately later in fido Jun 23 16:53:21 gtg Jun 23 16:56:14 morphis: suddenly i have free ~2 days. what to do? Jun 23 16:56:14 Tofe: morphis: btw i compiled emulator image, and luna-next crashes on every close of enyo app Jun 23 17:23:34 nizovn: Write the Bluetooth plugin :P ? Jun 23 17:24:03 morphis: For audio-service, what would need to be done in order to be able to set volume for different things like ringtone? Jun 23 17:24:24 Also brobostigon was looking for which version of libhybris we're using... Jun 23 17:35:05 Herrie: :) morphis: it should be service like com.palm.bt? there is no much info about it in google Jun 23 17:39:06 nizovn: It should be a wrapper for connman just like wifi so it's accessible via LS2 Jun 23 17:39:12 Let me find the wifi one Jun 23 17:40:14 https://github.com/webOS-ports/connman-adapter/blob/master/src/wifiservice.cpp Jun 23 17:41:01 Let me search legacy doctors for LS calls Jun 23 17:42:44 it's com.palm.bluetooth Jun 23 18:03:00 I think initially we're OK with turning BT on, scanning for devices and pairing. It will eventually need a lot more but that would be a good start Jun 23 18:03:10 Let me see if I can find the legacy bits refering to this Jun 23 18:10:59 Herrie: I am not sure if we should write a complete service for that Jun 23 18:11:09 could be get really complex to wrap all those bluetooth details Jun 23 18:11:43 basic discovery/connect/disconnect isn't that much but everything else could get complex Jun 23 18:13:51 Herrie, nizovn: so implementing something like https://developer.mozilla.org/en-US/docs/Web/API/Web_Bluetooth_API should be more what we want Jun 23 18:14:17 similar to https://github.com/webOS-ports/luna-webappmanager/blob/master/src/extensions/wifimanager.cpp Jun 23 18:14:37 those APIs are currently only exposed for system apps Jun 23 18:27:08 Herrie, nizovn: problem is that developing such dbus wrappers can be very time and resource (people) consuming Jun 23 18:27:09 morphis: Yeah I guess BT could get complex quickly Jun 23 18:27:26 so sticking with a direct use of those APIs behind a very simple js API is the better approach for us Jun 23 18:27:38 also in terms of compatiblity with others Jun 23 18:28:04 nizovn: you have an N4 or touchpad? Jun 23 18:28:14 nizovn: I'll try to look more into the crashes Jun 23 18:28:18 He has a TP I think Jun 23 18:28:36 ok Jun 23 18:28:46 Herrie: we got bluetooth working on the TP or just on the N4? Jun 23 18:28:51 morphis, Herrie: right, TP Jun 23 18:28:58 morphis: did we change the way we delete surfaces, lately? Jun 23 18:29:08 Tofe: not sure Jun 23 18:29:21 but I doubt we changed anything in the compositor Jun 23 18:29:32 I would more guess that qt has changed somewhere Jun 23 18:29:46 yes, that's my guess too, it's just less easy to track :) Jun 23 18:30:04 yeah Jun 23 18:30:11 nizovn: so bluetooth would be one thing Jun 23 18:30:19 I don't know if we still have an image with older qt Jun 23 18:30:25 once we have basic things in place we can look into support for specific profiles Jun 23 18:31:21 morphis, Tofe: not sure what version of qtwayland we use Jun 23 18:31:40 5.4 with some additional changes Jun 23 18:45:12 * HaDAk is so exhausted Jun 23 19:08:20 Not sure what version of QT JaMa has in unstable, but that one segfaults as well Jun 23 19:08:27 Both TP and N4 boot OK with unstable though :) Jun 23 19:09:07 Herrie: guess its still 5.4.x Jun 23 19:10:16 Herrie: triggered a new testing build Jun 23 19:10:19 it crashes on surface destruction when luna-next calls wl_client_destroy Jun 23 19:10:51 Tofe: null pointer or invalid address? Jun 23 19:11:23 invalid address Jun 23 19:14:56 let me track in which order we destroy things (surface & client) Jun 23 19:17:27 mmh I'm a bit surprised Jun 23 19:18:08 the scenario being: open webbrowser, open new tab, then kill new tab, and then kill first tab (which was the last card opened) Jun 23 19:18:31 killing the new tab doesn't result in destroying the shell surface Jun 23 19:19:00 but killing the last tab results in destroying two shell surfaces, the second of those being invalid Jun 23 19:20:19 However we are called on SurfaceDying, so something has indeed be destroyed Jun 23 19:20:33 Stay put, I'll bring more details Jun 23 19:24:57 morphis: Seems Tizen guys are using BT web api with ConnMan? Jun 23 19:28:11 Herrie: where did you find that? Jun 23 19:28:51 Herrie: looks like you're right: https://github.com/crosswalk-project/tizen-extensions-crosswalk/tree/master/src/bluetooth Jun 23 19:29:07 good starting point for us Jun 23 19:29:59 https://developer.tizen.org/dev-guide/2.3.0/org.tizen.web.apireference/html/device_api/mobile/tizen/bluetooth.html Jun 23 19:38:59 morphis I was searching for Bluetooth web api and connman on Google Jun 23 19:39:09 I thought somebody else might have done some heavy lifting already ;) Jun 23 19:39:29 Herrie: connman doesn't do much with bluetooth Jun 23 19:39:36 just takes a role when using PAN profile Jun 23 19:39:44 and for power management Jun 23 19:40:07 OK but looks Tizen guys wrote stuff for BlueZ 4 and 5 ;) Jun 23 19:40:26 That should be pretty re-usable right? Jun 23 19:55:31 morphis: Is there something brobostigon can help with, he's been working on Sailfish with N4 on libhybris ;) Jun 23 19:58:41 morphis: I think we have a bug in QtWayland, which doesn't seem to respect to the wayland protocol rule saying that the compositor must call wl_shell_surface_destroy when the corresponding surface is destroyed Jun 23 20:03:35 Tofe: Probably not what we're looking for but came across this: https://github.com/qtproject/qtwayland/commit/77a7900aaa790f7bb5e0105f38dbd539f04e219f Jun 23 20:03:58 Or this one: https://github.com/qtproject/qtwayland/commit/ca803d44c8231ff6262221a8d66c1401165f1171 Jun 23 20:05:05 Herrie: on sailfish and wants now to work on Lune too? Jun 23 20:06:14 Tofe or:https://github.com/qtproject/qtwayland/commit/6ac99af0fa3cdf334b59dd3f85803872e7a4194f Jun 23 20:06:20 morphis: Yeah ;) Jun 23 20:12:27 Herrie: hm, does he want to do real development work? Jun 23 20:12:32 brobostigon: ping Jun 23 20:12:57 morphis: I think so ;) He seems to be more for low level stuff ;) Jun 23 20:12:57 morphis: pong Jun 23 20:13:02 We need some help there :P Jun 23 20:15:21 morphis: looks like my guess was wrong, after a discussion with giucam Jun 23 20:18:31 Tofe: Any ideas for a fix already? Jun 23 20:21:17 Herrie: not yet no Jun 23 20:22:21 brobostigon: Herrie said you want something to work on for LuneOS Jun 23 20:22:35 Tofe: why? Jun 23 20:24:17 morphis: i used webos many years ago on my old palm, and have been sorting out some old devices in my colection and that old palm reminded me, i need to more reseatch, and found luneos in the process. Jun 23 20:24:37 ah :) Jun 23 20:24:54 brobostigon: generally we're very similar to sailfish Jun 23 20:25:37 morphis: i have done my research, similarities like hybris. Jun 23 20:25:44 yeah Jun 23 20:27:39 morphis: because there's no wl_shell_surface_destroy request yet in wayland, simply Jun 23 20:28:22 morphis: but I'm now trying with valgrind, which does funny stuff on top of qt+wayland+hybris... Jun 23 20:31:22 Tofe: :D Jun 23 20:34:19 ok, valgrind can't start luna-next properly, I give up for tonight :) Jun 23 20:43:47 * brobostigon starts git'ting code from github. Jun 23 21:13:48 brobostigon: Ubuntu 12.04 x64 works best but I think newer versions might work too ;) Jun 23 21:14:06 You seen: http://webos-ports.org/wiki/Build_for_Mako ? Jun 23 21:14:20 i run debian, so ir shouldnt be a problem. Jun 23 21:14:30 ty. :) Jun 23 21:14:54 Yeah I think it will be fine Jun 23 21:15:34 :) **** ENDING LOGGING AT Wed Jun 24 02:59:58 2015