**** BEGIN LOGGING AT Tue Aug 27 02:59:58 2013 Aug 27 08:25:10 morning Aug 27 09:22:15 morning Aug 27 09:43:15 JaMa: I am thinking about switch qtwebkit to use upstream webkit as base Aug 27 09:43:25 as I am have several issues with 5.1.0 Aug 27 09:54:01 JaMa: one option would be to have a history-less repository for our purpose on github Aug 27 09:54:11 which we sync with upstream from time to time Aug 27 09:54:55 upstream webkit or upstream qtwebkit repo? Aug 27 09:55:24 upstream webkit Aug 27 09:56:07 I will play with it first but it might be a option we have to take Aug 27 09:59:17 I don't know enough about the differences to have any opinion on this one, but I trust you to do the right thing Aug 27 10:02:00 JaMa: it's more or less a trail and error I am currently doing to get the web application part working as webkit is rather very complex to debug Aug 27 10:02:14 so checking if a newer version fixes the problem is always preferred :) Aug 27 13:09:27 JaMa: ok, upstream webkit compiled fine Aug 27 14:05:42 got a problem with the latest meta-webos-ports repo -> ERROR: Recipe webos-systemd-services is trying to change PV from '1.0+gitrAUTOINC+2ff87680a7' to '1.0+gitrAUTOINC+dc97c754fd'. This will cause do_package_write_* failures since the incorrect data will be used and they will be unable to find the right workdir. Aug 27 14:59:49 panda-z: see https://bugzilla.yoctoproject.org/show_bug.cgi?id=4102 and run that find command to cleanup pkgdata Aug 27 15:00:24 JaMa: thx, will try that Aug 27 15:20:54 hi Aug 27 15:22:16 Garfonso: hey Aug 27 16:14:35 the patch "sync fix for stable upload" creates an endless loop with the carddav connector. :( Aug 27 16:16:20 this seems to be a bug the patch introduces. Interesting... it seems other synergy connector don't use postPutRemoteObjects. Aug 27 16:40:57 Garfonso: I wouldn't bother too much about that? Aug 27 16:41:01 Since it's legacy? Aug 27 16:42:59 No way to work around the patch in your carddav connector? Aug 27 16:47:28 Or we could "fix" the patch so it doesn't break carddav? Aug 27 16:51:29 work around in the connector would be not to implement postPutRemoteModify, which means all local changes will be downloaded from server again, even if not necessary. Aug 27 16:51:35 could be ok.. Aug 27 16:56:50 Or "patch the patch" so it works properly in all cases? Aug 27 17:01:11 morphis: :S Aug 27 17:01:45 It makes me wonder then if there is a bug in the arm JIT module, or something? Aug 27 17:02:02 It is 100% reproducible, I assume? Aug 27 17:02:27 dkirker: yes Aug 27 17:02:39 with latest upstream webkit I had it once webkit starts Aug 27 17:03:01 :S Aug 27 17:03:30 What's the problem morphis? We seems to be missing a part of the convo here :P ? Aug 27 17:04:21 Herrie: I tried... but I failed.. ;) Aug 27 17:07:01 I would need to see all code before I could come up with something... We could do something nasty with the carddav name maybe and make an exception for that.... Nasty but well :S Aug 27 17:10:17 Herrie: yeah. We moved the conversation from another channel. :P Aug 27 17:10:25 Just to confuse you!! :D :D Aug 27 17:10:57 LOL Aug 27 17:11:10 So newer Webkit doesn't solve the issues you had morphis? Aug 27 17:11:19 Not sure what they were to start with anyway ;) Aug 27 17:11:28 Herrie: consequence from the problem is that we will run webkit without JIT until its fixed Aug 27 17:12:31 And that's bad for performance I assume? Aug 27 17:15:35 morning ! Aug 27 17:20:43 morphis: http://webos-ports.org/wiki/Galaxy_Nexus_Hardware_Overview and http://webos-ports.org/wiki/Nexus_7_Hardware_Overview we still need this? If yes for what? I was thinking about doing this somewhat differently with a template maybe so we can list parts per device and the "working/non working/partially working status"? Aug 27 17:23:46 Herrie: It may be presented another way, but I think that having a place where the hardware specs are clearly documented is always useful, indeed Aug 27 17:24:30 Tofe: Agree there... That Aug 27 17:24:36 That's what I was thinking as well... Aug 27 17:24:43 I believe CM's device info is too limited Aug 27 17:26:06 See http://webos-ports.org/wiki/Grouper_Info on the right, could expand that with links probably to the actual components somehow? Thinking about how to get that working nicely Aug 27 17:26:15 But on the other hand, we could also rework the template of such hw overview, and have some sort of current status/support for it Aug 27 17:27:46 ah, yes, it may be more clear that way. Maybe also with one line per peripheral, and maybe with a column with some comment Aug 27 17:28:28 That way all the informations of the two hw overviews could fit in, I guess Aug 27 17:31:44 I think I can create a template which has model/type, links, kernel versions supported etc. When you would click the individual component it would give you that info for the component. In case you would click it for the device it would give all components with it's info... Aug 27 17:36:36 looks like a good idea Aug 27 18:16:12 Herrie: sounds good to me Aug 27 18:17:23 Ok hope to finally finish the redo of the remaining pages tomorrow or the day after Aug 27 18:17:38 Have 1 remaining question for the build/install etc Aug 27 18:19:19 I'm not into Android at all. So what is fastboot and how does this fit in with unlocking bootloader? Not really clear for me Aug 27 18:19:42 CM guys have http://wiki.cyanogenmod.org/w/Install_CM_for_grouper for install of a device. Which bits we need, which ones not? Aug 27 18:21:10 morphis: adding "-no-neon" to qt5 didn't work so well... Aug 27 18:21:20 Tofe: hm Aug 27 18:24:23 I succeeded booting with the "old" kernel, but then I get those "Illegal instruction" when I call busybox. Sometimes. For example, sometimes "busybox ps" works, sometimes not. Aug 27 18:24:54 I would say the memory or the kernel is all corrupted Aug 27 18:25:39 I must have missed something. Aug 27 18:32:48 Tofe: you tried a danny rootfs with a new kernel? Aug 27 18:32:53 JaMa: any idea? Aug 27 18:37:33 morphis: no, not tried that Aug 27 18:38:02 to role out it's the kernels fault Aug 27 18:51:19 ah you mean a dylan image Aug 27 18:51:59 not really, we had similar issues e.g. with thumb on armv4t, but I don't remember seeing anyone reporting similar issues with latest toolchain upgrade (e.g. for all hardfp tunes) Aug 27 19:34:44 Tofe: seems like some people are going to use gcc 4.8 with android too: http://forum.xda-developers.com/showthread.php?t=2108705 Aug 27 19:37:18 morphis: yes, I saw that too, I have to find out what's the problem here :) Aug 27 19:37:59 I know such things are quite frustrating Aug 27 20:13:26 Garfonso: seems like you're stealing stefan_schmidt all the work :) Aug 27 20:14:02 I'm sorry. ;) Aug 27 20:14:37 Garfonso: btw. I might need your help with something different soon Aug 27 20:14:52 what is it? Aug 27 20:15:05 wrting a small js based service to support our update procedure Aug 27 20:15:15 or the foundation for this Aug 27 20:16:13 Garfonso: initial thoughts are here http://www.webos-ports.org/wiki/System_Upgrade Aug 27 20:16:36 this still needs a some more thinking about the details Aug 27 20:16:54 ok, will take a look at this and think about it. Aug 27 20:21:04 Garfonso: lets talk about it again later but just that you know :) Aug 27 20:21:13 JaMa: is the llvm fix already in our oe-core branch? Aug 27 20:21:30 morphis: yes (in our meta-oe revision) Aug 27 20:23:24 ah ok Aug 27 20:23:34 searched it in oe-core :) Aug 27 20:23:45 * JaMa starting new build with latest webkit changes and going off Aug 27 20:29:56 JaMa: thanks Aug 27 20:52:31 good night Aug 28 02:34:00 line 262: 2807 Segmentation fault (core dumped) opkg-cl -f $INSTALL_CONF_IPK -o $INSTALL_ROOTFS_IPK --force_postinstall --prefer-arch-to-version update Aug 28 02:34:07 any ideas? **** ENDING LOGGING AT Wed Aug 28 02:59:58 2013