**** BEGIN LOGGING AT Thu Jun 23 02:59:56 2011 Jun 23 05:00:19 can anyone help me how to change screen resolution with fbdev drivers? Jun 23 07:40:22 hi guys Jun 23 07:40:50 is there a new mirror for the source packages? Jun 23 07:41:00 mirrors.openembedded.org doesn't work Jun 23 07:41:26 03Joel A Fernandes  07org.openembedded.dev * rc28ca0ff20 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Jun 23 07:41:26 linux-omap: Disable UART sleep timeout to avoid serial port problems on wakeup. Jun 23 07:41:26 Signed-off-by: Joel A Fernandes Jun 23 07:41:26 Signed-off-by: Koen Kooi Jun 23 07:44:29 03Koen Kooi  07org.openembedded.dev * r7b10e138c2 10openembedded.git/recipes/linux/linux-omap-psp_2.6.32.bb: Jun 23 07:44:29 linux-omap-psp 2.6.32: bump PR Jun 23 07:44:29 Signed-off-by: Koen Kooi Jun 23 08:35:50 morning all Jun 23 08:36:03 hi, bluelightning Jun 23 09:24:34 good morning Jun 23 09:25:28 hi anarsoul, mckoan Jun 23 09:26:30 hi mckoan Jun 23 09:26:33 hi bl Jun 23 09:27:54 hi woglinde Jun 23 09:28:35 hm some knows where in kernel config -> Device Tree and Open Firmware support Jun 23 09:28:38 is reside? Jun 23 09:29:26 I dont find it this quick with menuconfig Jun 23 09:39:40 woglinde: do you mean FIRMWARE_IN_KERNEL ? Jun 23 09:40:18 mckoan its some menu point where you can enable CONFIG_OF Jun 23 09:40:26 that I need Jun 23 09:41:22 its sourced after mtd Jun 23 09:41:52 woglinde, hi .... still I m facing problem in changing screen resolution with fbdev... Jun 23 09:42:52 how do i change screen resolution with fbdev drivers... Jun 23 09:43:23 fbset Jun 23 09:43:33 man fbset Jun 23 09:44:35 flatten device tree support it is Jun 23 09:44:51 i had use " fbset -xres 800 -yres 600 -vxres 800 -vyres 600 -depth 16 -pixclock 62 -rgba 5/11,6/5,5/0,0/0" Jun 23 09:45:10 but it gives some distorted screen.... Jun 23 09:45:48 it works perfectly only with kernel command line " display=vga1024" Jun 23 09:46:12 woglinde: http://cateee.net/lkddb/web-lkddb/PROC_DEVICETREE.html Jun 23 10:08:52 pb_: can I get your ack on the native/nativesdk parse fix? Jun 23 10:10:30 I was hoping someone more expert than me would ack it. I'll see if I can do that this morning though. Jun 23 10:19:42 Is there any way to use variable expansion in bblayers.conf? Jun 23 10:19:52 * pb_ trying to figure out a way to make oe-core setup less painful Jun 23 10:20:38 pb_: well, variable expansion works there, but bear in mind variables there will get expanded very early Jun 23 10:20:46 or rather variable references Jun 23 10:20:50 if they aren't := Jun 23 10:21:42 so you can't use anything from local.conf in there for example Jun 23 10:23:19 hm. so, if it gets expanded before local.conf, where am I supposed to set the variables from? Jun 23 10:23:58 I guess maybe I can set them in the environment. Jun 23 10:24:15 * pb_ goes to check the secret password to make bitbake not erase the environment Jun 23 10:25:08 ah yes, very good, that seems to work Jun 23 10:25:30 pb_: so you just added something to BB_ENV_EXTRAWHITE? Jun 23 10:25:34 yeah Jun 23 10:46:19 ah, oe-buildenv-internal clobbers BB_ENV_EXTRAWHITE Jun 23 10:46:21 * pb_ stabs it Jun 23 10:53:44 I, for me, dislike env vars Jun 23 10:55:17 but a few must be set, I agree Jun 23 10:59:29 yeah, indeed Jun 23 11:07:21 pb_: you are everywhere... yesterday I stumbled about your first TLS patches for ARM in 2005... Jun 23 11:08:05 pb_: woglinde mentioned you had a problem with two git repos for one source tree. Did you find a solution? I have a simple fetch2/git.py patch to select the 'destdir' for the git checkout Jun 23 11:08:40 zecke: no, I never did find a solution for my git problem. Luckily one part of my tree is fairly static so I just makde it into a tarball and put that in SRC_URI instead :-} Jun 23 11:09:09 your patch sounds like exactly what I need Jun 23 11:10:03 pb_: let me send a mail Jun 23 11:11:35 rock Jun 23 12:57:31 03Koen Kooi  07org.openembedded.dev * rea41560b2c 10openembedded.git/recipes/ti/gstreamer-ti.inc: Jun 23 12:57:31 gstreamer-ti: hack it to build with libtool sysroot enabled Jun 23 12:57:31 Signed-off-by: Koen Kooi Jun 23 15:22:01 hmmm Jun 23 15:22:14 configure:3055: checking whether we are cross compiling Jun 23 15:22:37 and then It compiles the test and try to run it: Jun 23 15:22:38 configure:3074: ./conftest Jun 23 15:22:38 /lib/ld-linux.so.3: No such file or directory Jun 23 15:22:48 and fails Jun 23 15:22:58 If you meant to cross compile, use `--host'. Jun 23 15:23:20 the package is privoxy Jun 23 15:23:25 *recipe Jun 23 15:24:18 normally it should fails and continues the test right? Jun 23 15:25:07 *tests Jun 23 15:25:20 like find out that we're indeed cross compiling Jun 23 15:36:45 ahhh Jun 23 15:36:47 ./configure --disable-pthread Jun 23 15:36:55 that's how it was called Jun 23 15:55:34 e Jun 23 16:00:29 03Jeff Lance  07master * r0c28fe8887 10openembedded.git/recipes/ti/ (matrix-tui.inc matrix-tui_1.1.bb): Jun 23 16:00:30 matrix-tui: update SRCREV to 42 with makefile.init fixes Jun 23 16:00:30 * Modify to use OE MACHINE name for platforms Jun 23 16:00:30 * Add all Sitara platforms Jun 23 16:00:30 * Update SRCREV to 42 Jun 23 16:00:30 Signed-off-by: Jeff Lance Jun 23 16:00:30 Signed-off-by: Denys Dmytriyenko Jun 23 16:24:52 hi Jun 23 16:24:56 is anyone here= Jun 23 16:24:58 ?* Jun 23 16:35:22 matr0_: there's lots of people here :) Jun 23 16:35:36 could you try to help me? Jun 23 16:35:50 i am trying to build SHR for iphone3g Jun 23 16:36:23 I can try, but I'm no expert on SHR Jun 23 16:36:53 so, building fails with following log http://www.pastie.org/private/k0rrtrjijsvz2diiywv8q Jun 23 16:37:30 ok.. my brain is shutting down again.. val += "foo" will make it '$val foo' =+ will be 'foo $val'.. what is .= ? Jun 23 16:39:00 fray: without space Jun 23 16:39:10 perfect.. and =. is the same as =+ w/o the space? Jun 23 16:39:37 yep http://bitbake.berlios.de/manual/ch02.html#id869030 Jun 23 16:39:59 =. is prepend, it seems Jun 23 16:40:14 perfect.. thanks Jun 23 16:41:35 Does anybody know whether the "no symlinks in the path to your TMPDIR" thing is still actually an issue? Jun 23 16:41:42 it is Jun 23 16:42:00 can you remind me what it is that breaks? it's some autoconf thing, right? Jun 23 16:42:02 there are a lot of makefiles, configure scripts, etc that do processing based on space seperation Jun 23 16:42:13 primarily autoconf, but not just autoconf.. Jun 23 16:42:22 er, what does space separation have to do with it? Jun 23 16:43:17 in shell..: val = "/foo bar" ; list = "/bin /sbin $val" ; for each in list ; do ... ; done Jun 23 16:43:35 that is -really- common behavior Jun 23 16:43:43 well, sure. but how do symlinks break that? Jun 23 16:43:57 sorry.. I was reading that wrong.. Jun 23 16:44:02 * fray needs sleep.. Jun 23 16:44:15 symlinks are because things do readlink and attempt to rationalize canonical paths.. Jun 23 16:44:22 is it still a problem, I'm not sure.. Jun 23 16:44:34 a lot of things have been corrected over the years.. but I -think- that autoconf may still use readlink for things Jun 23 16:44:47 I guess I should just try it and see. Jun 23 16:45:13 ya.. it usually works.. but don't be surprised if the target of the symlink gets embedded into things.. Jun 23 16:45:20 and it usually breaks in core-utils with GNUmakefile recursion error... :) which is quite misleading Jun 23 16:45:29 historically it was because autoconf would get confused and give you "this source directory has already been configured, need to run distclean" errors, but I think that was in autoconf2.13 days and it might have gotten better since then. Jun 23 16:45:44 denix: ah, heh Jun 23 16:46:05 where I've seen it the most is someone does: ln -s /target1 .../foo/tmp -- does a build Jun 23 16:46:14 they then mv /target1 to /target2 .. fix the symlink.. Jun 23 16:46:22 and things no longer work.. because of the readlink issue Jun 23 16:47:27 okay. well, if people are doing that and it works until they do the "mv", I guess that means that symlinks per se are no longer a problem Jun 23 16:47:41 the behavior I have seen is outside of OE.. Jun 23 16:47:42 modulo core-utils perhaps... :-} Jun 23 16:48:01 ah, I see Jun 23 19:50:06 has anyone used stock OE to build for gumstix? Jun 23 19:56:26 hmm Jun 23 19:56:36 this seems like a openembedded + virtual machine problem Jun 23 20:08:01 aaaaanyone use OE in virtualbox? Jun 23 20:08:32 err virtualbox + Ubuntu to be specific Jun 23 20:08:43 cause I see many errors Jun 23 20:08:44 MANY Jun 23 20:11:42 Did you follow the general install this stuff on your distro wiki page? Jun 23 20:12:18 YEAH Jun 23 20:12:21 err Jun 23 20:12:24 sorry about caps Jun 23 20:12:42 I've done this before on my laptop, but thats ubuntu directly Jun 23 20:13:01 I remember having trouble with VM + OE a while ago, was wondering if they are still problematic Jun 23 20:13:33 I do tons in VMs all day Jun 23 20:13:39 Ubuntu/RHEL, but backed by VMware Jun 23 20:13:45 So, what specifically are you seeing fail? Jun 23 20:13:48 http://pastebin.com/7nU7kEgh --does that give any clues? Jun 23 20:14:04 its just a shitton of those Jun 23 20:14:10 Any local changes? Jun 23 20:14:31 nope I just got OE, setup local.conf and did a bitbake helloworld Jun 23 20:14:49 I specify MACHINE in the command line Jun 23 20:15:09 and set it in BB_ENV_EXTRAWHITE ? Jun 23 20:16:02 http://pastebin.com/ZbJi0Kjv -- bigger dump, but yeah, just repetition of that same error Jun 23 20:16:12 set what? MACHINE? Jun 23 20:16:47 I've never come across "BB_ENV_EXTRAWHITE" Jun 23 20:18:06 lostincake2: env vars are filtered out. if you set machine in the env rather than via local.conf, you *must* add it to BB_ENV_EXTRAWHITE or it won't be available to bitbake Jun 23 20:20:19 did this change recently? Jun 23 20:20:23 or semi-recently Jun 23 20:20:26 kergoth: is there limitation in amend.inc class that it cannot amend SRCREV in recipes? Jun 23 20:20:35 lostincake2: no Jun 23 20:20:40 lostincake2: couple years ago at least Jun 23 20:20:41 denix: not that i'm aware of Jun 23 20:21:02 kergoth: I mostly ask cause I still do MACHINE=beagleboard bitbake ... on my laptop at home Jun 23 20:21:17 although that's the dev branch, does that work differently? Jun 23 20:21:20 lostincake2: MACHINE is in the white list Jun 23 20:21:56 on dev? or on regular OE too Jun 23 20:21:56 ? Jun 23 20:22:15 cause I have regular OE and MACHINE is apparently being filtered out Jun 23 20:22:27 I am mostly curious, I don't really NEED to specify through command line Jun 23 20:22:31 03Jeff Lance  07master * r94514d1986 10openembedded.git/recipes/ti/ (3 files): (log message trimmed) Jun 23 20:22:31 matrix-gui: update SRCREV to revision 175 Jun 23 20:22:31 * Updated crypto scripts for am3715 Jun 23 20:22:31 * New Camera and multimedia applications for ARM only parts Jun 23 20:22:31 * Modified power scripts to demonstrate standby Jun 23 20:22:31 * Soft Ap application Jun 23 20:22:31 * fix 3D Chameleon man script Jun 23 20:23:21 lostincake2: check your setup scripts - something there must whitelist it... Jun 23 20:27:00 hmm k Jun 23 20:27:07 atleast I know what's up heh Jun 23 20:27:15 thanks much Jun 23 20:27:52 denix? its ALIVE? Jun 23 20:30:08 03Paul Menzel  07master * r7858948b41 10openembedded.git/recipes/pyside/libshiboken_1.0.2.bb: (log message trimmed) Jun 23 20:30:08 libshiboken: Do not override `DEPENDS` Jun 23 20:30:08 This is a fix up for commit 1b9260e7 [1]. Jun 23 20:30:08 commit 1b9260e7a6c756d6938821f013a41c371065980a Jun 23 20:30:08 Author: Paul Menzel Jun 23 20:30:08 Date: Sun Jun 19 11:45:37 2011 +0200 Jun 23 20:30:08 shiboken: Add `apiextrator-native` to `DEPENDS` Jun 23 20:30:16 03Paul Menzel  07master * r99cad4f2f2 10openembedded.git/recipes/freesmartphone/libfso-qt-embedded_git.bb: (log message trimmed) Jun 23 20:30:16 libfso-qt-embedded: Do not override `DEPENDS` Jun 23 20:30:16 Using `=` instead of `+=` overrides the dependencies declared in `libfso-qt.inc`. Jun 23 20:30:16 Task configure therefore fails with the following errors. Jun 23 20:30:17 1. `fso-specs` is not found. Jun 23 20:30:17 checking for SPECS... no Jun 23 20:30:17 configure: error: Package requirements (fso-specs >= 2011.06.01.1) were not met: Jun 23 20:32:17 denix: if you don't mind, any help with this: http://pastebin.com/vk5CFtg3 ? ... still errors I havent seen before :/ Jun 23 20:50:53 ant__, hi, I sent a new patch addressing your comment Jun 23 20:52:03 sounds sane, LZMA_COMPRESSION_LEVEL is unused (only in zaurus.inc) Jun 23 20:52:57 stil, for creating cpio.lzma I think -9 is overkill Jun 23 20:53:16 (I have no probs overriding it, though) Jun 23 20:54:04 anyone have an idea why curl-nativesdk would fail to find the C compiler? Jun 23 20:54:44 GNUtoo: in fact the patch could be renamed like 'unbreak lzma compression' Jun 23 20:55:39 so I resend once more Jun 23 20:55:51 ? Jun 23 20:56:07 at some point I will abandon if it takes too much time Jun 23 20:56:25 up to you, now the title is imprecise ;) Jun 23 20:56:27 anyone have any ideas for me VM problemos? Jun 23 20:56:48 ( http://pastebin.com/vk5CFtg3 ) Jun 23 20:57:27 GNUtoo: being this is a bugfix I think you can commit it with just one ack Jun 23 20:57:39 ok Jun 23 20:57:51 did I have your ack? Jun 23 20:58:10 pls repost last time :) Jun 23 20:58:44 ok Jun 23 20:58:55 I just change the commit message? Jun 23 20:59:09 I tested yesterday's version Jun 23 20:59:39 I suppose the change is trivial enough Jun 23 21:17:45 ant__: around Jun 23 21:17:54 ant__: you have latest gentoo Jun 23 21:17:58 I suppose Jun 23 21:21:05 khem: not latest, a stable profile Jun 23 21:21:11 ok Jun 23 21:21:23 I need someone with latest and greatest Jun 23 21:21:35 toolchain? Jun 23 21:21:39 are there prebuilt images around Jun 23 21:21:43 ant__: no Jun 23 21:21:51 ant__: I need a given package Jun 23 21:22:12 you can unmask one by one Jun 23 21:22:52 which package btw? Jun 23 21:22:58 libtirpc Jun 23 21:24:42 libtirpc-0.2.2-r1 is hanging low Jun 23 21:28:32 yes thatone Jun 23 21:28:36 can you build it Jun 23 21:28:41 and give me the binaries Jun 23 21:28:46 I only need the .so file Jun 23 21:28:57 khem ping? Jun 23 21:28:57 o well that may not be it Jun 23 21:29:01 ka6sox-farfarawa: yes Jun 23 21:29:05 hiya... Jun 23 21:29:17 we shut down 2 things that we suspect of causing that issue. Jun 23 21:29:24 ok Jun 23 21:29:30 lets see how long it goes before it crashes. Jun 23 21:29:36 but eventually it needs a compleat overhaul. Jun 23 21:29:48 melo ? Jun 23 21:29:51 or just apache Jun 23 21:29:53 yes Jun 23 21:29:57 hmm Jun 23 21:29:58 all of Melo Jun 23 21:30:01 its 32bit Jun 23 21:30:11 and hw is 64bit Jun 23 21:30:15 correct Jun 23 21:30:20 ok Jun 23 21:30:28 upgrade to 11.04 ? Jun 23 21:30:34 that could be a daunting task Jun 23 21:30:35 no...not LTS Jun 23 21:30:40 khem: http://tinderbox.dev.gentoo.org/default-linux/ Jun 23 21:30:44 ant__, v5 sent Jun 23 21:30:48 10.04 LTS AMD_64 Jun 23 21:31:20 khem: is in net-libs Jun 23 21:34:13 ant__: yeah figured but its built with older glibc Jun 23 21:34:24 I need one that built with 2.14 Jun 23 21:34:27 ok Jun 23 21:34:30 [ebuild N ] net-libs/libtirpc-0.2.2-r1 USE="-kerberos -static-libs" Jun 23 21:35:02 sys-libs/glibc-2.12.2 Jun 23 21:35:03 ka6sox-farfarawa: sounds ok Jun 23 21:35:04 ^^ Jun 23 21:35:37 ant__: yeah the problem I am looking for is only with 2.14 Jun 23 21:40:10 khem: is still unstable http://bugs.gentoo.org/show_bug.cgi?id=370409 Jun 23 21:40:24 but maybe this helps http://old.nabble.com/glibc-2.14-and-changes-in-rpc-support-(libtirpc)-td31826446.html Jun 23 21:47:49 ant__: that Jun 23 21:47:51 I know Jun 23 21:48:02 problem is I want to look at the binary of libtirpc Jun 23 21:48:04 so Jun 23 21:48:14 built with glibc 2.14 Jun 23 21:48:31 since I am seeing that it misses the crypt_des code Jun 23 22:02:49 khem: even latest snapshots have 2.12 Jun 23 22:03:03 I'm sorry... :/ Jun 23 22:03:15 I guess Jun 23 22:03:16 n Jun 23 22:03:17 p Jun 23 22:33:15 thanks a lot for the ack!!!!!! Jun 23 22:39:30 03Otavio Salvador  07org.openembedded.dev * rc382a9ce64 10openembedded.git/ (conf/bitbake.conf recipes/linux/linux-kexecboot.inc): (log message trimmed) Jun 23 22:39:30 conf/bitbake: unbreak and improve lzma image compression handling and adapt linux-kexecboot.inc Jun 23 22:39:30 We now use xz-native to generate lzma images Jun 23 22:39:30 Kexecboot.inc needed to be adjusted in order to reflect the dependency Jun 23 22:39:30 change from lzma-native to xz-native Jun 23 22:39:31 The initramfs.cpio.gz were tested with linux-kexecboot on nokia900. Jun 23 22:39:32 Signed-off-by: Otavio Salvador Jun 23 22:43:19 Tartarus: ping Jun 23 22:43:46 pong Jun 23 22:43:52 hi there Jun 23 22:44:00 about -       kerneldir=${STAGING_KERNEL_DIR} Jun 23 22:44:00 +       kerneldir=${D}/kernel Jun 23 22:44:17 how does it reflect on klibc? Jun 23 22:44:55 would you sinplify MACH_STAGING_KERNEL_DIR = "${STAGING_DIR}/${MACHINE_ARCH}${TARGET_VENDOR}-${TARGET_OS}/kernel" ? Jun 23 22:46:58 (btw debian builds klibc with linux-libc-headers now afaik) Jun 23 22:51:25 on armv7 klibc works fine, but on armv6 it failed, I don't remember the failures tough Jun 23 22:51:28 why would curl(-nativesdk) try to use x86_64-linux-gcc when my compiler is x86_64-unknown-linux-gnu-gcc? it's resulting in a quite annoying "C compiler cannot create executables" error. Jun 23 22:51:48 Tartarus: I don't know how bad is depending on kernel as we do now. For the initramfs use you need a kernel anyway Jun 23 22:54:57 GNUtoo: afaik there are some armv6 and armv6 VFP Jun 23 22:55:22 it was a vfp issue and it had vfp Jun 23 22:55:28 I can try on novfp if you want Jun 23 22:55:28 :p Jun 23 22:55:41 OK Jun 23 22:55:58 ant__: The answer is that you need to place things inside of D where D=STAGING_DIR Jun 23 22:55:59 some lib was with vfp and something was without Jun 23 22:56:01 Otherwise sstate will miss it Jun 23 22:56:08 but I remember very vaguely Jun 23 22:56:32 ant__: Or are you thinking something else? Jun 23 22:56:40 At the end of the day, STAGING_KERNEL_DIR is still there Jun 23 22:56:45 It's just covered by sstate now Jun 23 22:57:01 Tartarus: I think that now smthg is broken (pstage contains unexpanded paths) Jun 23 22:57:15 probably the TOOLCHAIN bits Jun 23 22:57:47 ant__: The short answer is, work things out such that do_install shoves things inside of ${D} Jun 23 22:57:56 Then use a sysroot_stage_all_append bit like the kernel does Jun 23 22:58:00 I aim to have klibc in oe-core so we need to polish the recipe Jun 23 22:58:19 Er, not that Jun 23 22:58:20 sec Jun 23 22:59:03 sysroot_stage_all_append() { Jun 23 22:59:03 sysroot_stage_dir ${D}/kernel ${SYSROOT_DESTDIR}/kernel Jun 23 22:59:04 } Jun 23 22:59:11 is what does it in kernel.bbclass to then fix things up right Jun 23 23:19:20 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rd5805488d2 10openembedded.git/conf/machine/nokia900.conf: Jun 23 23:19:20 nokia900.conf: add support for kexecboot Jun 23 23:19:20 The nokia bootloader(nolo) expects a zImage, so we produce Jun 23 23:19:20 a kexecboot zImage. Jun 23 23:19:20 Signed-off-by: Denis 'GNUtoo' Carikli Jun 23 23:24:15 git pull --rebase Jun 23 23:24:44 * ant__ closes some terms Jun 23 23:24:55 03Andrea Adami  07org.openembedded.dev * r5a0e0b735b 10openembedded.git/conf/machine/include/zaurus.inc: Jun 23 23:24:55 zaurus.inc: remove orphan LZMA_COMPRESSION_LEVEL Jun 23 23:24:55 * cpio.lzma are built using xz-utils Jun 23 23:24:56 * add -e to achieve a slightly better compression Jun 23 23:24:56 * keep -7 preset to be able to decompress on 32MB devices Jun 23 23:24:56 Signed-off-by: Andrea Adami Jun 23 23:26:35 bluelightning: what about the fix for bitbake master and oe-dev? I tested it here too. Jun 23 23:26:58 ant__: I was waiting for an ack on v2 of the patch Jun 23 23:27:13 ah, or "" Jun 23 23:27:56 well I could just commit it, I'm not sure anyone would object at this point Jun 23 23:28:48 I tested that 2nd version, acked now Jun 23 23:35:34 03Pau Espin Pedrol  07org.openembedded.dev * r45081d1878 10openembedded.git/recipes/openmoko-3rdparty/emtooth2_svn.bb: Jun 23 23:35:34 emtooth2: bump SRCREV to 143 Jun 23 23:35:34 Signed-off-by: Pau Espin Pedrol Jun 23 23:35:34 Acked-by: Denis 'GNUtoo' Carikli Jun 23 23:35:45 03Pau Espin Pedrol  07org.openembedded.dev * r2c06ce205c 10openembedded.git/recipes/e17/enjoy_svn.bb: (log message trimmed) Jun 23 23:35:45 enjoy: add alsa and autodetect gst modules. Jun 23 23:35:45 Without autodetect, enjoy segfaults. Without alsa, no sound is played. Jun 23 23:35:45 That happens because emotion needs them to play as shown in its source code: Jun 23 23:35:46 emotion_gstreamer_pipeline.c: sink = gst_element_factory_make("autoaudiosink", "audiosink"); Jun 23 23:35:46 Signed-off-by: Pau Espin Pedrol Jun 23 23:35:47 Acked-by: Paul Menzel Jun 23 23:37:47 thx. gn Jun 23 23:40:28 03Paul Eggleton  07master * r6f3334a9a1 10openembedded.git/classes/ (native.bbclass nativesdk.bbclass): Jun 23 23:40:28 classes/native*.bbclass: fix error during parse with bitbake master Jun 23 23:40:28 Fixes "AttributeError: 'NoneType' object has no attribute 'split'" during Jun 23 23:40:28 parsing with bitbake master. We should not be calling explode_deps with Jun 23 23:40:28 None as the argument, so ensure the value isn't None. Jun 23 23:40:28 Signed-off-by: Paul Eggleton Jun 23 23:40:29 Acked-by: Andrea Adami Jun 24 02:36:09 Hi all **** ENDING LOGGING AT Fri Jun 24 02:59:57 2011