**** BEGIN LOGGING AT Sat Nov 04 03:00:02 2017 Nov 04 07:27:12 Morning! Nov 04 07:27:20 orning! Nov 04 07:27:26 +M Nov 04 07:42:23 Morning! Nov 04 07:42:40 660 minutes. That's 11 hours, isn't it? Nov 04 07:43:27 That's "a lot" :) Nov 04 07:43:33 My build failed on (only) cross-localedef-native do_compile in the end. Nov 04 07:43:45 oh, why so ? Nov 04 07:43:55 Well, I was asleep for most of it :-) Nov 04 07:46:42 "glibc/locale/xlocale.h:27:16: error: redefinition of 'struct __locale_struct' Nov 04 07:47:39 weird Nov 04 07:48:03 Well, I am trying an unsupported build OS. Nov 04 07:48:41 ah yes, true Nov 04 07:56:02 I saw this one on mailing list Nov 04 07:57:43 Rebuilding that glibc component should solve it Nov 04 07:57:51 At least if I recall correctly Nov 04 07:58:14 Just build again? (It has often worked in the past, I must admit.) Nov 04 07:59:07 People weren't too happy with the glibc-locale recipe if I recall correctly and wanted to rewrite it from scratch Nov 04 07:59:43 I'm guessing that nobody could understand it :-) Nov 04 08:15:49 elvispre: Yeah some of these recipes get complicated :P Nov 04 08:18:31 Tofe: I realized that we can specify bits per target, so could introduce HaDAk's LuneOS wallpaper for some of our targets :) Nov 04 08:19:34 Herrie|Pre3: good idea :) Nov 04 08:55:39 Tofe: You happy with the PR's or testing? Nov 04 08:55:58 quite happy :) Nov 04 08:56:00 please go on Nov 04 08:57:10 I cross referenced all files that I deleted in our code and there are no references to it, so we should be good for those Nov 04 09:00:19 Herrie: it should be fine Nov 04 09:19:11 Tofe: OK merged, will update some recipes in a bit Nov 04 09:19:35 ok Nov 04 09:49:18 Herrie: testing changed to rocko ? Nov 04 09:49:49 I thought this wasn't ready Nov 04 09:52:41 "ParseError at /media/LuneOS/testing/webos-ports/meta-webos-ports/meta-luneos/recipes-webos/luna-init/luna-init.bb:42: unparsed line: '}' " ? Nov 04 09:58:47 followed by "cannot stat '/media/LuneOS/testing/webos-ports/tmp-glibc/work/all-webos-linux/luna-init/2.0.1-10+gitAUTOINC+972a96ef8e-r0/git/files/conf/CustomerCareNumber.txt'" Nov 04 10:00:51 ... and cust-preferences Nov 04 10:03:43 Tofe: That's what I meant with updating recipes ;) Nov 04 10:03:58 Tofe: We didn't move to Rocko afaik? Nov 04 10:04:08 ah maybe it's just meta-qt5 Nov 04 10:06:39 Tofe: Luna-init is fixed Nov 04 10:06:55 You have anything pending to PR shortly or can start a testing build? Nov 04 10:07:07 nope, please start Nov 04 10:15:55 Herrie: I just installed the latest ubports, and they have bluetooth working smoothly there; I'll try to find out what's different there Nov 04 10:33:11 Tofe: Could be something small in BT stack Nov 04 10:33:29 Tofe: Hammerhead build already finished :) Nov 04 10:33:40 mine too Nov 04 10:36:38 Tofe: If you want I can diff the bluetooth folder between our kernel and ubports? Nov 04 10:37:19 why not; but it may be something else Nov 04 10:38:01 Well we have UBPorts N4 kernel Nov 04 10:38:04 That works fine Nov 04 10:38:09 But other one for N5 ;) Nov 04 10:38:12 That's not UBPorts Nov 04 10:38:21 Not sure where that originated from ;) Nov 04 10:38:27 So might have some diffs here and there Nov 04 10:38:34 yep Nov 04 10:38:37 I can also do the whole kernel diff but that might be quite some Nov 04 10:38:58 no no, no need; and I'm not even sure it's in the kernel anyway Nov 04 10:42:24 their hciattach seems a bit custom Nov 04 10:42:34 like the one from sailfishos Nov 04 10:44:31 You might want to ping morphis Nov 04 10:44:34 He did most of the BT bits Nov 04 10:44:48 He might have some thoughts Nov 04 10:44:55 As to why some things were done Nov 04 10:51:29 ah wait, I think I understand the issue; hciattach didn't actually load the firmware Nov 04 10:53:25 Tofe: That for sure doesn't help :P Nov 04 10:53:27 we definitly need the updated hciattach Nov 04 10:54:16 let's just take bluez5 from mer-core, eh ? Nov 04 11:15:19 Tofe: Why not, just wonder why these changes wouldn't be upstreamed Nov 04 11:35:47 Just I'm a bit wary of the large number of changes that mer does to standard components like oFono, BlueZ and ConnMan Nov 04 11:35:56 There shouldn't be that much need for it I would say Nov 04 12:28:16 fixed ! Nov 04 12:28:40 and, like often, it was a very stupid bug Nov 04 12:29:09 it was looking for BCM4335C0 in /etc/firmware... case sensitive. Nov 04 12:38:41 though /firmware is vfat, one might argue our mount options are incorrect Nov 04 12:42:15 Tofe: So no Mer BlueZ5? Nov 04 12:42:21 nope, no need Nov 04 12:46:04 :) Nov 04 12:46:38 I think my patch of bluez5 will be more reliable than the vfat mount option Nov 04 13:11:08 Herrie: https://github.com/webOS-ports/meta-webos-ports/pull/262 will fix bluetooth Nov 04 13:37:54 Tofe: Nice :) Nov 04 18:01:45 HaDAk: We've finally merged your wallpaper stuff ;) Only when cleaning up some of the luna-bits realized that we can specify some bits per device. So on the device where you wallpaper doesn't look good we can keep the butterfly for now ;) Nov 04 18:22:15 Herrie: I'm struggling a bit with the handling of Bluez5 from Nemo.DBus, so I'm trying out bluez5-qt's fork from Mer Nov 04 19:02:32 well, it does work, so why not Nov 04 19:12:40 Morning! Nov 04 19:16:59 Tofe: Weird Nemo.DBus was acting up Nov 04 19:17:07 YOu'd expect dBUS to be pretty straight forward Nov 04 19:17:16 But if bluez5-qt works why not Nov 04 19:17:21 MartinHov: Morning! Nov 04 19:17:37 FYI - I downloaded latest testing image qemux86 version 500 and this download was very slow (it took 15-20 mins) while else it took me 1 minute. Nov 04 19:18:22 I think my internet connection is good and stable .. not sure why this took soooo loooong .. Nov 04 19:18:45 the image is good and works as expected = no change to previous Nov 04 19:20:47 Herrie: it's not Nemo.DBus, it's my implementation which doesn't use it properly :p Nov 04 19:21:29 kde's bluez5-qt also exists for Archlinux, so for me it's the same, I can also test on desktop :) Nov 04 19:21:46 The main work was to setup the recipe Nov 04 19:51:04 Tofe: Yeah recipes are usually the hard part ;) Nov 04 19:59:32 Tofe: NOTE: Running task 3654 of 7569 (/home/jenkins/workspace/luneos-testing/webos-ports/openembedded-core/meta/recipes-core/systemd/systemd-serialgetty.bb:do_rm_work) Nov 04 19:59:32 ERROR: bluez5-5.43-r0 do_patch: Command Error: 'quilt --quiltrc /home/jenkins/workspace/luneos-testing/webos-ports/tmp-glibc/work/cortexa7hf-neon-vfpv4-webos-linux-gnueabi/bluez5/5.43-r0/recipe-sysroot-native/etc/quiltrc push' exited with 0 Output: ... Hunk #2 FAILED at 366. Nov 04 19:59:50 This was for rpi3 Nov 04 20:00:06 I guess this was after I merged your PR when it started to build the pi3 Nov 04 20:00:21 The Mako, Hammerhead, Tenderloin were OK but those finished before your PR I guess Nov 04 20:19:24 Tofe: It works on my local build for Hammerhead Nov 04 20:19:33 So might be something specific for RPi3, checking that locally now Nov 04 20:19:38 But iwll take a while probably :P Nov 04 20:20:39 Herrie: I'm not surprised: Rpi3 has tons of specific stuff for bluez Nov 04 20:29:00 Tofe: Yeah they probably have a conflicting patch in their layer Nov 04 20:30:32 Tofe: This one I guess: https://github.com/agherzan/meta-raspberrypi/blob/pyro/recipes-connectivity/bluez5/bluez5/0002-bcm43xx-The-UART-speed-must-be-reset-after-the-firmw.patch Nov 04 20:31:07 yup Nov 04 20:32:23 well, it would be sufficient to remove a part of my patch Nov 04 20:33:36 just removing https://github.com/webOS-ports/meta-webos-ports/blob/pyro/meta-luneos/recipes-connectivity/bluez5/bluez5/0001-Case-insensitive-firmware-name.patch#L20 and below Nov 04 20:40:36 Tofe: Yeah it's merely cosmetics/logging anyway right? Nov 04 20:40:43 right Nov 04 20:42:12 Tofe: YOu seen http://www.bluez.org/ Nov 04 20:42:17 We're on 5.43 Nov 04 20:42:31 Not sure if there's anything in 5.44, 5.45, 5.46 or 5.47 worth upgrading for? Nov 04 20:42:56 well, right now it works, so let's keep it like that until we've got something :) Nov 04 20:43:27 OK Nov 04 20:43:44 Just not sure there were any specific bugs, nothing that rings a bell for me Nov 04 20:45:16 I didn't remark anything obvious on my side Nov 04 21:04:04 OK well based on changelog it are things we wouldn't need immediately anyway ;) **** ENDING LOGGING AT Sun Nov 05 03:00:02 2017