**** BEGIN LOGGING AT Tue Oct 29 02:59:58 2013 Oct 29 05:04:44 !list Oct 29 05:05:13 ,list Oct 29 06:38:03 !topic Oct 29 09:32:25 morning all Oct 29 09:56:19 good morning silvio_l_, all Oct 29 09:57:19 morning mckoan, all Oct 29 09:58:47 hi bluelightning, hi mckoan Oct 29 09:59:28 morning silvio_l_ Oct 29 10:51:52 do we have a web page onlayer/version/maintainer anywhere? Oct 29 10:59:05 Crofton|work: maintainership can be recorded per-branch in the layer index Oct 29 10:59:12 which reminds me I need to add dora now... Oct 29 11:22:50 boost.inc now uses packageconfig to decide if it should add python. how do I ... use this? must I set PACKAGECONFIG[python] = "python" Oct 29 11:24:42 tasslehoff: depends where you're doing it from Oct 29 11:25:41 tasslehoff: assuming local.conf or distro config, PACKAGECONFIG_append_pn-boost = " python" Oct 29 11:31:41 bluelightning: thanks. Oct 29 13:24:44 ok, dora branch is now in the layer index Oct 29 16:31:01 khem: is there a schedule for upgrading to gcc4.7 to 4.7.3? http://thread.gmane.org/gmane.comp.handhelds.openembedded.core/36125/focus=36182 Oct 29 16:32:18 if not, how easy is it to upgrade myself? Oct 29 19:33:00 JaMa, hello, Oct 29 19:33:23 I tried experimenting with meta-qt5 & qtwayland in qemuarm. Oct 29 19:33:45 Using weston as compositor. Oct 29 19:34:01 Simple examples work (more or less). Oct 29 19:34:23 cinematixecperience just segfaults. Oct 29 20:38:44 Am I correct in understanding that the update-rc.d.bbclass now uses 3 different ways, in the package scripts, to detect wether it is running on the host or the target ? 1. -z "$D" (on target) 2. -n "$D" (on host) 3. "x$D" == "x" on target. Although 1 and 2 use the same scheme of checking the length of $D, shouldn't be cleaned up somewhat ? Is there a preferred way of checking the host/target thing ? Oct 29 20:41:08 And even "$D" != "" is in there Oct 29 20:45:47 3. is a portability hack encouraged by autoconf thats pointless in our context, we know our shells support -z or -n. so 1. or 2. are the preferred methods, as appropirate Oct 29 20:45:56 afaik. Oct 29 20:46:59 sounds like a janitor task.. Oct 29 20:47:15 the only requirement so far is that if (environemnt) D is defined, we're on a host doing a cross install Oct 29 20:47:20 if it's not defined.. we're on a target Oct 29 20:47:45 indeed Oct 29 20:48:53 ...and whatever is implemented needs to work with whatever shell the user sets /bin/sh to.. Oct 29 20:49:02 bash, dash, zsh? Oct 29 20:53:19 What does it mean that D is defined ? The value of $D is actually used in the options passed to update-rc.d it seems Oct 29 20:53:53 when running at image creation time on the host, D is set to the path to the filesystem Oct 29 20:54:00 when running on the target, it's not set to anything Oct 29 20:55:09 aha Oct 29 20:55:32 lumag_: I haven't tried that on qemu* and I'm not using cinema* thing, do you have backtrace? Oct 29 21:14:12 JaMa, no. I will try running with gdb, but it will take some time. Oct 29 21:14:37 Did you ever try it with weston fbdev backend? Oct 29 21:15:30 update-rc.d postinst still confuses me. If the package already is installed, shouldnt we stop the service in preint(), instead of postinst() ? Oct 29 21:19:20 'maybe' Oct 29 21:19:33 upgrading hasn't been high on anyones agenda (other then perhaps Angstrom) Oct 29 21:28:32 JaMa, do I need to do any special things to generate backtrace, or installing -dbg package would be enough for gdb to pick it up? Oct 29 21:36:45 it should be enough Oct 29 22:23:12 Is it ok to post RFC, request for comments, patches to the oe-devel list ? Not sure I want to send this update-rc.d.bbclass cleanup to patchwork right away Oct 29 22:38:30 oe-core ML not oe-devel Oct 29 22:41:19 ok Oct 29 22:59:54 rtollert: we only backport specific fixes and no plans to upgrade 4.7 Oct 29 23:00:31 rtollert: it should be fairly easy to upgrade recipes its the validation thats the tough part Oct 29 23:26:39 khem: thanks, can you provide any refs to how validation has worked in the past? Oct 29 23:28:55 khem: ("no" and "rtfm" are valid answers :) ) **** ENDING LOGGING AT Wed Oct 30 02:59:58 2013