**** BEGIN LOGGING AT Wed May 13 02:59:57 2009 May 13 03:09:01 Hey! Anyone here? May 13 04:18:02 03Angus Ainslie  07fso/milestone5.5 * r2ad2bd67d1 10openembedded.git/ (conf/checksums.ini recipes/python/pyring_1.1.8.bb): pyring : ship all the installed files May 13 04:45:04 03Roman I Khimov  07org.openembedded.dev * r2dc3792241 10openembedded.git/recipes/glibc/ (glibc_2.7.bb glibc_2.9.bb): May 13 04:45:04 glibc 2.7/2.9: add march-i686.patch May 13 04:45:04 This patch is there for 2.6.1, but for some reason was not carried on to May 13 04:45:04 2.7 and 2.9. Without it build for i686 machine fails. May 13 04:45:04 Acked-by: Denys Dmytriyenko May 13 07:15:00 good morning May 13 08:12:04 morning May 13 08:14:16 hey May 13 08:17:27 03Marco Cavallini  07xora/angstrom-srcpv * r0cdaf0036a 10openembedded.git/MAINTAINERS: May 13 08:17:27 Added myself in MANTAINERS as KaeilOS mantainer May 13 08:17:27 Signed-off-by: Marco Cavallini May 13 08:17:27 Signed-off-by: Marcin Juszkiewicz May 13 08:17:29 03Peter Chubb  07xora/angstrom-srcpv * rf47a90dc04 10openembedded.git/classes/siteinfo.bbclass: May 13 08:17:32 Allow use of Itanium as build machine May 13 08:17:34 Hi, May 13 08:17:36 I have access to a cluster of Itaniums that I want to use as May 13 08:17:38 build machines for openembedded. This patch enables Itanium as a May 13 08:17:40 build environment. May 13 08:17:42 Signed-off-by: Marcin Juszkiewicz May 13 08:17:44 03Florian Boor  07xora/angstrom-srcpv * r50947a23fc 10openembedded.git/conf/machine/micro2440.conf: micro2440.conf: Prefer an up to date kernel. May 13 08:17:47 03Florian Boor  07xora/angstrom-srcpv * r6349eba5e1 10openembedded.git/conf/machine/topas910.conf: topas910.conf: Enable building images with summary information. May 13 08:17:52 03Florian Boor  07xora/angstrom-srcpv * rb5f722ac2f 10openembedded.git/recipes/gsoko/ (gsoko-0.4.2-gpe6/fix_install.patch gsoko_0.4.2-gpe6.bb): gsoko: Apply fix to suppress strip installing binary. May 13 08:18:00 03Roman I Khimov  07xora/angstrom-srcpv * r2dc3792241 10openembedded.git/recipes/glibc/ (glibc_2.7.bb glibc_2.9.bb): May 13 08:18:03 glibc 2.7/2.9: add march-i686.patch May 13 08:18:05 This patch is there for 2.6.1, but for some reason was not carried on to May 13 08:18:07 2.7 and 2.9. Without it build for i686 machine fails. May 13 08:18:09 Acked-by: Denys Dmytriyenko May 13 08:18:11 03Leon Woestenberg  07xora/angstrom-srcpv * r2531c335e6 10openembedded.git/recipes/linux/linux-rt_2.6.24.bb: linux-rt-2.6.24: Fixed missing defconfig for mpc8315e-rdb. May 13 08:18:14 03Graeme Gregory  07xora/angstrom-srcpv * r055108de5a 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into xora/angstrom-srcpv May 13 08:18:19 03Graeme Gregory  07xora/angstrom-srcpv * r58582919d6 10openembedded.git/recipes/u-boot/u-boot-bug_svn.bb: u-boot-bug_svn.bb : SRCREV -> SRCPV May 13 08:18:24 03Andrea Adami  07xora/angstrom-srcpv * r5d93237c35 10openembedded.git/ (conf/checksums.ini recipes/linux/linux_2.6.29+2.6.30-rc3.bb): linux_2.6.29+2.6.30: jump to RC5 May 13 08:18:27 03Andrea Adami  07xora/angstrom-srcpv * r6e13bf0bd8 10openembedded.git/recipes/maemo4/hildon-thumbnail_0.14.bb: May 13 08:18:32 hildon-thumbnail: fix dependencies from gdk-pixbuf -> gtk+ May 13 08:18:34 - the dep was triggering the build of gdk_pixbuf_2.10.14 May 13 08:18:36 - with the nasty effect of overwriting the .pc files in staging May 13 08:18:38 - (old gtk+ version and wrong target=linux-fb) May 13 08:18:40 03Andrea Adami  07xora/angstrom-srcpv * r9905f8c0a1 10openembedded.git/conf/machine/netbook-pro.conf: netbook-pro.conf: fix bogus 755 permissions on checkout May 13 08:18:45 03Graeme Gregory  07xora/angstrom-srcpv * r6124e4ed1e 10openembedded.git/recipes/llvm/llvm2.6-native_2.5+svnr20090504.bb: llvm2.6-native_2.5+svnr20090504.bb : SRCREV -> SRCPV May 13 08:18:48 03Graeme Gregory  07xora/angstrom-srcpv * r0567523ab9 10openembedded.git/recipes/llvm/llvm2.6_2.5+svnr20090504.bb: llvm2.6_2.5+svnr20090504.bb : SRCREV -> SRCPV May 13 08:18:51 03Marcin Juszkiewicz  07xora/angstrom-srcpv * r71997fc5fe 10openembedded.git/classes/patch.bbclass: patch.bbclass: use hashlib with Python 2.5+ - removes DeprecationWarning May 13 08:18:58 03Graeme Gregory  07xora/angstrom-srcpv * r7c0c3e3457 10openembedded.git/recipes/xorg-lib/pixman_git.bb: pixman_git.bb : SRCREV -> SRCPV May 13 08:19:01 03Florian Boor  07xora/angstrom-srcpv * r0380701311 10openembedded.git/conf/local.conf.sample: local.conf.sample: Add description of BB_NUMBER_THREADS and ad US locale to the example list which is necessary for Angstrom. May 13 08:36:17 good morning May 13 08:45:29 http://xkcd.com/583/ roxx May 13 08:53:48 is bugs.openembedded.net also valid for stable patches? May 13 08:54:08 no May 13 08:54:50 I'm begging for a couple of ACK :-) May 13 08:55:42 links to patchwork please then May 13 08:56:24 that's what I was looking for, so is patchwork and not bugs, where is patchwork? May 13 08:56:46 mckoan: send patch to OEML and it will appear on patchwork.openembedded.net May 13 08:57:22 .org it seems May 13 08:57:24 found : http://patchwork.openembedded.org/patch/510/ May 13 08:59:03 mckoan: acked. when patchwork will grab my mail I can push it May 13 08:59:16 thank you May 13 08:59:43 I have a question about stable, because I'd like to help ACK-ing too May 13 09:00:19 "Testing of changes" means that I have to build using the specific patch before ack-ing it May 13 09:00:44 but with which platform? the ones I used frequently? May 13 09:01:44 and more, if you have a buildbot I suppose that it test only with a few of platforms May 13 09:01:51 I do tests with "at91sam9263ek qemuarm qemux86 vortex86sx beagleboard"/glibc and "atngw100 at91sam9263ek qemuarm qemux86 beagleboard"/uclibc May 13 09:05:04 fine, I knew that, but what if my patch broke a platform not tested by buildbot? May 13 09:05:41 03Koen Kooi  07org.openembedded.dev * rf6862553f0 10openembedded.git/conf/checksums.ini: checksums: add more checksums May 13 09:05:42 03Koen Kooi  07org.openembedded.dev * rb41d60eff8 10openembedded.git/contrib/angstrom/upload-packages.sh: angstrom feed uploader: upload to unstable feed, the 2008 feeds are populated from stable/2009 May 13 09:19:07 good morning May 13 09:21:50 hello May 13 09:22:31 what's the bitbake target to update the package list ? can't remember it May 13 09:22:57 something like: bitbake -c rebuild package-list ... ? May 13 09:23:06 bitbake package-index May 13 09:24:05 thanks hrw May 13 09:27:32 hi florian May 13 09:28:58 morning May 13 09:29:09 hi ant_work May 13 09:38:52 ~curse python in debian May 13 09:38:53 May you be reincarnated as a Windows XP administrator, python in debian ! May 13 09:39:23 ubuntu switched to python2.6, debian still use python2.5 and python2.6 from experimental is no-no for bitbake May 13 09:39:44 unbelievable! May 13 09:40:16 BTW what's wrong with python2.5 ? May 13 09:41:04 mckoan: 2.6 deprecated some stuff and I want to catch them May 13 09:52:38 Who feeld responsible for xserver-kdrive-common? May 13 09:52:41 with Python 2.6 I get complaints about sets in COW.py and importing md5 instead of hashlib in do_patch May 13 09:52:43 eh feels May 13 09:52:45 <_cpo_> i have a question about the SRC_URI variable, if i have file://defconfig for example, oe looks in the ${PN},${PN}-{$PV}, and files directory for this specific file. but the directory can contain subdirectories like ${DISTRO} or ${MACHINE}, how could i extend the fetch feature to alos look for ${IMAGE_NAME} as a directory, so that i can specify diffent config files for different images that i build May 13 09:54:43 _cpo_: AFAIK that's not possible May 13 09:55:01 akheron: in stable/2009? May 13 09:55:31 hrw: yes, though I haven't pulled for a while May 13 09:56:00 akheron: I posted two patches for it to ML May 13 09:56:59 <_cpo_> akheron: ideas for .bbclass files that i could have a look for implementing? May 13 09:57:24 <_cpo_> and the next thing is, i have an image with "inherit image" that always fails the build with initscripts/udev/sysvinit missing May 13 09:57:34 <_cpo_> but those are defined in the image.bbclass file as ?= May 13 09:58:01 _cpo_: ${DISTRO}, ${MACHINE} gets picked based on OVERRIDE... May 13 09:58:17 _cpo_: the order of OVERRIDES is important... May 13 09:58:17 hrw: ok, they look fine May 13 09:58:30 zecke: but you cannot set OVERRIDES in the image recipe, right? May 13 09:58:48 in a way that it gets propagated to other recipes May 13 09:59:01 akheron: right, it is outside the scope :) May 13 09:59:20 akheron: but for things like defconfig.... machine is most of the time stronger than defconfig? May 13 09:59:28 _cpo_: so you'll need separate recipes for separate images May 13 09:59:53 <_cpo_> akheron: that i wantet to get rid of with that solution :) May 13 10:00:52 _cpo_: you can do it quite elegantly, just "require foo.inc" in each recipe May 13 10:01:08 <_cpo_> akheron, thats th ecurrent way May 13 10:01:15 ok May 13 10:02:25 <_cpo_> next thing is, that in image.bbclass the ${IMAGE_VARS} is not used. so my image doe not require ${IMAGE_INITSCRIPTS} or ${IMAGE_DEV_MANAGER} May 13 10:02:36 <_cpo_> but my image has "inherit image" May 13 10:04:22 what do you mean by "not used"? my image.bbclass has RDEPENDS += "${IMAGE_INSTALL} ${IMAGE_VARS}" May 13 10:06:27 <_cpo_> mine, too but when doing a bitbake -g myimage May 13 10:06:52 <_cpo_> neither udev (for IMAGE_DEV_MANAGER) nor initscripts (for IMAGE_INITSCRIPTS) will pop up May 13 10:09:43 you should check log.do_rootfs and run.do_rootfs May 13 10:10:11 <_cpo_> for what? i know that those packages will never be build May 13 10:10:22 <_cpo_> because they dont show up in the dependency file May 13 10:11:38 weird May 13 10:11:49 they should at least be built because they are in RDEPENDS May 13 10:12:06 <_cpo_> yes, should :) May 13 10:12:12 use bitbake -i to peek RDEPENDS in your image May 13 10:12:23 you may have a typo somewhere May 13 10:13:48 i have a question concerning git source urls May 13 10:14:10 i wanted to make a recipe for gstreamer-plugins-base with git sourcedate today May 13 10:14:57 i put it as SRC_URI = "git://anongit.freedesktop.org/git/gstreamer/${PN};date=${SRCDATE};protocol=git \ since the manual command would be git clone git://anongit.freedesktop.org/git/gstreamer/gst-plugins-base but bitbake then tells me this: May 13 10:15:25 http://pastebin.com/d11aa541f May 13 10:19:50 <_cpo_> akheron: thought so, too... gonne tripple check for typos now May 13 10:21:00 an idea how i have to put the uri? somebody? May 13 10:21:30 03Graeme Gregory  07xora/angstrom-srcpv * rd0ec5f1f43 10openembedded.git/recipes/u-boot/u-boot_git.bb: u-boot_git.bb : SRCREV -> SRCPV May 13 10:25:38 mickeyl: good morning May 13 10:26:08 good morning pb__ May 13 10:43:39 fraxinas: seems to me that there's something wrong with the git fetcher May 13 10:59:17 iwe had some issues with the tune-thumb.inc actually defaults to use thumb as are not expected other tune inc files like xscale that includes it May 13 11:00:24 the problem is my build uses thumb even if i cant find a trace in the distro .bb files that should turn it on . May 13 11:00:33 or my own config fiile May 13 11:06:44 xranby: using Angstrom? May 13 11:09:42 yes May 13 11:09:52 gm May 13 11:10:16 XorA|gone: peek openjdk-6-shark ARM_INSTRUCTION_SET responds arm May 13 11:10:59 thebohemian: ping May 13 11:11:10 ant_work: pong May 13 11:11:14 hey :) May 13 11:11:21 it's about java May 13 11:11:38 ant_work: yep :) May 13 11:11:41 I'm building x11-gpe-java-image May 13 11:11:51 now fails against classpath May 13 11:11:59 xranby: we are just about to turn off thumb for all Angstrom builds again May 13 11:12:05 btw there are missing PREFERRED_PROVIDER May 13 11:12:11 ant_work: still the gtk+ thingie asking for a gdk-linux-fb? May 13 11:12:21 no, solved May 13 11:12:28 xranby: after extensive testing it seems there is no binutils/gcc combo that works properly with thumb May 13 11:13:15 ant_work: if this was done with a patch it would be needed for oe-stable as well as .dev May 13 11:13:34 ant_work: what else fails? (pastebin?) May 13 11:13:44 about providers: May 13 11:13:45 XorA|gone: right, my issue are that bitbake defaults to use thumb even if i explicitly sets ARM_INSTRUCTION_SET = "arm" in my local conf. so my thinking are that the tune-thumb.inc file are broken May 13 11:13:47 http://rafb.net/p/qwG7cD59.html May 13 11:13:52 about failure: May 13 11:14:17 http://tinderbox.openembedded.net/builds/150375/ May 13 11:14:55 xranby: your on armv4t or armv5te? May 13 11:15:01 armv5te May 13 11:15:16 xranby: then remove the line in conf/distro/angstrom-2008.1.conf May 13 11:15:50 ant_work: http://wiki.openembedded.net/index.php/Java May 13 11:15:50 the various virtuals are expected to be set in local.conf. if it ok we could provide some defaults ( ?= ) to angstrom May 13 11:16:02 yep May 13 11:18:04 ant_work: odd. "java -classpath antlr.Tool -o ../tools/generated/gnu/classpath/tools/gjdoc/expr" that misses a path between "-classpath" and "antlr.Tool" May 13 11:18:17 ant_work: perhaps some variable name clash .. May 13 11:23:10 thebohemian: I was just checking whether it builds...I'm about *zarro* vs. Java, using just VMs... Congrats for the progresses, btw May 13 11:33:16 ant_work: is this oe-stable or .dev? May 13 11:37:41 03Robert Schuster  07org.openembedded.dev * r8079cec72c 10openembedded.git/recipes/cacao/cacao-native_0.99.4.bb: cacao-native 0.99.4: Set default preference to -1. May 13 11:37:47 03Robert Schuster  07org.openembedded.dev * r49608b5012 10openembedded.git/recipes/llvm/ (4 files in 2 dirs): May 13 11:37:47 llvm2.6: Updated to newer SVN snapshot (includes some patches developed in OE). May 13 11:37:47 llvm2.6-native: Dito. May 13 11:37:49 03Robert Schuster  07org.openembedded.dev * r70daf01d4f 10openembedded.git/recipes/glib-2.0/glib.inc: glib.inc: Packaging workaround for chinook-compat. May 13 12:00:52 thebohemian: .dev May 13 12:07:38 * RP wakes up May 13 12:08:03 yawn :-) May 13 12:14:15 RP: hey...don't you have any fix to backport wrt packaged-staging? May 13 12:26:14 ant_work: ok, I am building classpath for angstrom atm on a amd64 box. I hope I will see this error, too and fix it ASAP May 13 12:28:07 thx May 13 13:01:55 03Florian Boor  07org.openembedded.dev * rf102e3c340 10openembedded.git/recipes/xserver-common/xserver-common_1.24.bb: xserver-common: Add 1.24 May 13 13:01:56 03Florian Boor  07org.openembedded.dev * r0a8cf40b68 10openembedded.git/conf/machine/topas910.conf: topas910.conf: Add cleanmarkers to jffs2 filesystem May 13 13:01:56 03Florian Boor  07org.openembedded.dev * r0be94d1320 10openembedded.git/conf/machine/micro2440.conf: micro2440.conf: Provide display information and create filesystem with summary information. May 13 13:01:57 03Florian Boor  07org.openembedded.dev * rbdc8535060 10openembedded.git/recipes/images/gpe-image.bb: gpe-image: Soem husekeeping, sortign, comments and improve Angstrom compatibility. May 13 13:02:01 03Florian Boor  07org.openembedded.dev * r79f185638a 10openembedded.git/recipes/psplash/ (files/topas910/psplash-default psplash.inc): psplash: Change the splash orientation for the topas910 since the default splash screen won't fit in the native QVGA landscape framebuffer. May 13 13:02:04 03Florian Boor  07org.openembedded.dev * r8f800370f5 10openembedded.git/recipes/xserver-kdrive-common/ (2 files in 2 dirs): xserver-kdrive-common: Run display calibration later to make sure the device node for the touchscreen has been created already so that the touchscreen test succeeds. May 13 13:02:10 03Florian Boor  07org.openembedded.dev * rad7c3aa1f5 10openembedded.git/recipes/images/gpe-image.bb: gpe-image: Add license information and allow a "demo" style distro to log in automatically. May 13 13:02:13 03Florian Boor  07org.openembedded.dev * r7a5357cee6 10openembedded.git/conf/checksums.ini: checksums.ini: Add gpe-mixer 0.50 and xserver-common 0.24 May 13 14:02:54 03Angus Ainslie  07fso/milestone5.5 * rc1f9d1f740 10openembedded.git/recipes/initscripts/ (initscripts-1.0/g_ether.sh initscripts_1.0.bb): initscripts : fixup the g_ether.sh scipt and remove some on openmoko distro May 13 14:07:18 is anybody familiar with using git source uris in packages? May 13 14:09:57 or recipies who they're called nowadays :) May 13 14:10:55 they were always called recipes :) May 13 14:11:09 the directory they were in was poorly named May 13 14:13:57 i like recipe better too :) but still i don't understand why it doesn't check out the git repo correctly there... May 13 14:14:22 i put it as SRC_URI = "git://anongit.freedesktop.org/git/gstreamer/${PN};date=${SRCDATE};protocol=git \ since the manual command would be git clone git://anongit.freedesktop.org/git/gstreamer/gst-plugins-base but bitbake then tells me this: http://pastebin.com/d11aa541f May 13 14:17:53 fraxinas: in general you can use either -v (-vv, -vvv) or -D (-DD, -DDD) to see the used command May 13 14:18:08 ah cool idea, i shall try that May 13 14:20:42 hmmm doesn May 13 14:20:57 oops. doesn't really tell me much. it does fail at the pull command. May 13 14:21:10 when i manually check out git then i don't use a pull at all. i just clone May 13 14:21:24 it did create a file in sources though: -rw-r--r-- 1 andy andy 1977141 11. Dez 11:24 gst-plugins-base_anoncvs.freedesktop.org__20081011.tar.gz May 13 14:21:58 hmm nah can't be... wrong date May 13 14:22:20 raster: http://labs.trolltech.com/page/Projects/Graphics/Kinetic/DeclarativeUI? May 13 14:45:58 ant_work: What kind of fix? May 13 14:46:28 since some months, the packages are rebuilt but.. May 13 14:46:43 packages using package_stagefile_shell() May 13 14:46:51 e.g. zaurus-updater May 13 14:47:12 have strange issues (even *not* moving deploy) May 13 14:47:22 ant_work: I'm totally unaware of the issue :/ May 13 14:47:28 worst thing is another! May 13 14:47:47 the dir /cross/bin/armv5te/angstrom-bla May 13 14:47:59 is recreated as /cross/bin/angstrom-blah May 13 14:48:06 without +armv5te' May 13 14:48:19 effect: no crosscompiler found May 13 14:48:50 ant_work: I don't have time to look into something like that atm I'm afraid May 13 14:48:51 btw: I kept whole /deploy May 13 14:49:01 once /pstage was enough May 13 14:49:21 pstage should always be enough May 13 14:50:09 I think you told me already...got different hints recently... May 13 14:50:16 anyway, no changes May 13 14:50:54 RP: there is definitely some obscure issue May 13 14:54:27 03Marco Cavallini  07stable/2009 * rbd4137e5a0 10openembedded.git/conf/distro/kaeilos.conf: May 13 14:54:27 Modified conf/distro/kaeilos.conf May 13 14:54:27 * added IPKG_VARIANT so postinst script can run May 13 14:54:27 * added Prefer bluez4 May 13 14:54:27 Signed-off-by: Marco Cavallini May 13 14:54:29 Acked-by: Marcin Juszkiewicz May 13 14:59:51 03Andrea Adami  07stable/2009 * r5fb7102b11 10openembedded.git/recipes/ipaq-sleep/files/install-fix.patch: May 13 14:59:51 ipaq-sleep: don't strip during install - using coreutils this now breaks (strip: unable to recognise the format of the input file) May 13 14:59:51 Signed-off-by: Koen Kooi May 13 14:59:51 Acked-by: Philip Balister May 13 14:59:54 03Andrea Adami  07stable/2009 * rc59ce9b670 10openembedded.git/recipes/gpe-nmf/gpe-nmf-0.17/fix_makefiles.patch: May 13 14:59:57 gpe-nmf: don't strip during install - this breaks now with new install - strip is done in packaging May 13 14:59:59 Signed-off-by: Koen Kooi May 13 15:00:01 Acked-by: Philip Balister May 13 15:00:03 03Andrea Adami  07stable/2009 * rea9ca72057 10openembedded.git/recipes/gpe-question/files/makefile-fix.patch: May 13 15:00:06 gpe-question: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:00:08 Signed-off-by: Koen Kooi May 13 15:00:10 Acked-by: Philip Balister May 13 15:00:12 03Andrea Adami  07stable/2009 * r2d4900caf2 10openembedded.git/recipes/gpe-lights/ (2 files in 2 dirs): May 13 15:00:15 gpe-lights: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:00:25 Signed-off-by: Koen Kooi May 13 15:00:27 Acked-by: Philip Balister May 13 15:00:29 03Andrea Adami  07stable/2009 * r9e7365b7bd 10openembedded.git/recipes/gpe-othello/gpe-othello-0.2-1/fix-makefiles.patch: May 13 15:00:32 gpe-othello: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:00:34 Signed-off-by: Koen Kooi May 13 15:00:36 Acked-by: Philip Balister May 13 15:00:38 03Andrea Adami  07stable/2009 * r6a35c98da5 10openembedded.git/recipes/gpe-soundbite/files/makefile-fix.patch: May 13 15:00:41 gpe-soundbite: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:00:43 Signed-off-by: Koen Kooi May 13 15:00:45 Acked-by: Philip Balister May 13 15:00:49 03Andrea Adami  07stable/2009 * r7ed4ecddbb 10openembedded.git/recipes/gpe-go/gpe-go-0.05/fix-make.patch: May 13 15:00:58 gpe-go: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:01:00 Signed-off-by: Koen Kooi May 13 15:01:02 Acked-by: Philip Balister May 13 15:01:06 03Andrea Adami  07stable/2009 * rae2810eb77 10openembedded.git/recipes/gpe-calculator/gpe-calculator-0.2/fix_makefile.patch: May 13 15:01:09 gpe-calculator: don't strip during install - this breaks now with new install - strip is done in packaging May 13 15:01:11 Signed-off-by: Koen Kooi May 13 15:01:13 Acked-by: Philip Balister May 13 15:01:15 03Andrea Adami  07stable/2009 * r464925ed13 10openembedded.git/recipes/libdisplaymigration/ (3 files in 2 dirs): May 13 15:01:20 libdisplaymigration: Makefile fix - don't strip on install May 13 15:01:22 Signed-off-by: Koen Kooi May 13 15:01:28 (102 lines omitted) May 13 15:19:34 03Koen Kooi  07org.openembedded.dev * r448b9103e4 10openembedded.git/conf/distro/ (angstrom-2008.1.conf include/angstrom.inc): (log message trimmed) May 13 15:19:34 angstrom 2009.X: toolchain updates: May 13 15:19:34 * make gcc 4.3.3 the default compiler (except for ppc and avr32) May 13 15:19:34 * make binutils 2.18 the default (except for armv7a and avr32) May 13 15:19:34 * disable thumb for armv4t/5te since there's no gcc/binutils combo that works 100% with thumb May 13 15:19:38 * bump distro_pr to force a rebuild May 13 15:19:40 * point feed configs at unstable feed May 13 15:19:42 03Koen Kooi  07org.openembedded.dev * re0ea8fd0f4 10openembedded.git/ (conf/checksums.ini recipes/bluez/bluez4_4.39.bb): bluez: add 4.39 May 13 15:21:40 mickeyl: hi! this was an impressive breakfast :) May 13 15:22:09 indeed May 13 15:23:07 03Florian Boor  07org.openembedded.dev * r45c80d10ad 10openembedded.git/conf/machine/topas910.conf: topas910.conf: Advertize the small NOR flash since everything needs to fit in there for now. May 13 15:23:37 tarrarus, ping May 13 15:24:04 This reminds me that I'm still editing the wiki on one of my machines... May 13 15:32:42 waite1, pong May 13 15:33:02 how's it going Tartrus! May 13 15:33:12 pretty good :) May 13 15:33:31 Hey did all of the patches to make C++ meta-toolchain relocatable make it to mainline yet? May 13 15:33:53 yes, it's in dev and stable May 13 15:34:01 with the caveat about libtool just not being relocatible May 13 15:34:19 Excellent. You will make make my devs happy. May 13 15:34:22 heh May 13 15:34:35 I wonder if MV found it useful :) May 13 15:34:49 heh. That is the second edge of FOSS May 13 15:37:34 we should send MV an invoice :) May 13 15:38:29 that reminds me, i have a few libtool fixes that need to go up. i know of at least one target/circumstance where /usr/lib can still leak into the link May 13 15:38:42 heh May 13 15:38:49 I was wondering when you would commit them May 13 15:43:34 03Graeme Gregory  07org.openembedded.dev * r54d7edcaa7 10openembedded.git/conf/machine/om-gta02.conf: om-gta02.conf : bring the hard work of xf86-video-glamo to om-gta02 in .dev May 13 15:43:45 03Graeme Gregory  07org.openembedded.dev * r97f9aaebb4 10openembedded.git/conf/distro/include/sane-srcrevs.inc: May 13 15:43:45 sane-srcrevs.inc : update xf86-video-glamo revision to latest, cures screen May 13 15:43:45 corruption May 13 15:43:52 i still have trouble with my git checking out recipe but i have some new findings. i added a tag=master to my src_uri line and it actually does create the [branch "master"] section in the ./.git/config upon git clone, however the pull command is still performed without a proper branch parameter and still throws the same error http://pastebin.com/d11aa541f :( how can i convice bb to pull the correct branch? May 13 15:44:42 03Koen Kooi  07xora/angstrom-srcpv * rf6862553f0 10openembedded.git/conf/checksums.ini: checksums: add more checksums May 13 15:44:42 03Koen Kooi  07xora/angstrom-srcpv * rb41d60eff8 10openembedded.git/contrib/angstrom/upload-packages.sh: angstrom feed uploader: upload to unstable feed, the 2008 feeds are populated from stable/2009 May 13 15:44:43 03Robert Schuster  07xora/angstrom-srcpv * r70daf01d4f 10openembedded.git/recipes/glib-2.0/glib.inc: glib.inc: Packaging workaround for chinook-compat. May 13 15:44:45 03Robert Schuster  07xora/angstrom-srcpv * r8079cec72c 10openembedded.git/recipes/cacao/cacao-native_0.99.4.bb: cacao-native 0.99.4: Set default preference to -1. May 13 15:44:48 03Robert Schuster  07xora/angstrom-srcpv * r49608b5012 10openembedded.git/recipes/llvm/ (4 files in 2 dirs): May 13 15:44:51 llvm2.6: Updated to newer SVN snapshot (includes some patches developed in OE). May 13 15:44:53 llvm2.6-native: Dito. May 13 15:44:57 03Florian Boor  07xora/angstrom-srcpv * r7a5357cee6 10openembedded.git/conf/checksums.ini: checksums.ini: Add gpe-mixer 0.50 and xserver-common 0.24 May 13 15:45:00 03Florian Boor  07xora/angstrom-srcpv * r0a8cf40b68 10openembedded.git/conf/machine/topas910.conf: topas910.conf: Add cleanmarkers to jffs2 filesystem May 13 15:45:05 03Florian Boor  07xora/angstrom-srcpv * rf102e3c340 10openembedded.git/recipes/xserver-common/xserver-common_1.24.bb: xserver-common: Add 1.24 May 13 15:45:08 03Florian Boor  07xora/angstrom-srcpv * r0be94d1320 10openembedded.git/conf/machine/micro2440.conf: micro2440.conf: Provide display information and create filesystem with summary information. May 13 15:45:11 03Florian Boor  07xora/angstrom-srcpv * rbdc8535060 10openembedded.git/recipes/images/gpe-image.bb: gpe-image: Soem husekeeping, sortign, comments and improve Angstrom compatibility. May 13 15:45:18 03Florian Boor  07xora/angstrom-srcpv * r8f800370f5 10openembedded.git/recipes/xserver-kdrive-common/ (2 files in 2 dirs): xserver-kdrive-common: Run display calibration later to make sure the device node for the touchscreen has been created already so that the touchscreen test succeeds. May 13 15:45:22 03Koen Kooi  07xora/angstrom-srcpv * r448b9103e4 10openembedded.git/conf/distro/ (angstrom-2008.1.conf include/angstrom.inc): (log message trimmed) May 13 15:45:27 angstrom 2009.X: toolchain updates: May 13 15:45:29 * make gcc 4.3.3 the default compiler (except for ppc and avr32) May 13 15:45:31 * make binutils 2.18 the default (except for armv7a and avr32) May 13 15:45:33 * disable thumb for armv4t/5te since there's no gcc/binutils combo that works 100% with thumb May 13 15:45:35 * bump distro_pr to force a rebuild May 13 15:45:37 * point feed configs at unstable feed May 13 15:45:39 03Graeme Gregory  07xora/angstrom-srcpv * r97f9aaebb4 10openembedded.git/conf/distro/include/sane-srcrevs.inc: May 13 15:45:44 sane-srcrevs.inc : update xf86-video-glamo revision to latest, cures screen May 13 15:45:46 corruption May 13 15:45:48 03Graeme Gregory  07xora/angstrom-srcpv * r54d7edcaa7 10openembedded.git/conf/machine/om-gta02.conf: om-gta02.conf : bring the hard work of xf86-video-glamo to om-gta02 in .dev May 13 15:45:51 03Florian Boor  07xora/angstrom-srcpv * r79f185638a 10openembedded.git/recipes/psplash/ (files/topas910/psplash-default psplash.inc): psplash: Change the splash orientation for the topas910 since the default splash screen won't fit in the native QVGA landscape framebuffer. May 13 15:45:59 03Florian Boor  07xora/angstrom-srcpv * rad7c3aa1f5 10openembedded.git/recipes/images/gpe-image.bb: gpe-image: Add license information and allow a "demo" style distro to log in automatically. May 13 15:46:02 03Graeme Gregory  07xora/angstrom-srcpv * rc96a32ba1b 10openembedded.git/conf/distro/include/sane-srcrevs.inc: May 13 15:46:07 sane-srcrevs.inc : update xf86-video-glamo revision to latest, cures screen May 13 15:46:11 corruption May 13 15:46:13 03Koen Kooi  07xora/angstrom-srcpv * re0ea8fd0f4 10openembedded.git/ (conf/checksums.ini recipes/bluez/bluez4_4.39.bb): bluez: add 4.39 May 13 15:46:16 03Florian Boor  07xora/angstrom-srcpv * r45c80d10ad 10openembedded.git/conf/machine/topas910.conf: topas910.conf: Advertize the small NOR flash since everything needs to fit in there for now. May 13 15:46:23 03Graeme Gregory  07xora/angstrom-srcpv * r0ab134424f 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into xora/angstrom-srcpv May 13 15:46:26 03Graeme Gregory  07xora/angstrom-srcpv * r6d5e1d36a2 10openembedded.git/conf/machine/om-gta02.conf: om-gta02.conf : bring the hard work of xf86-video-glamo to om-gta02 in .dev May 13 15:46:29 (4 lines omitted) May 13 15:48:15 okay so my question was successfully flushed far far away ^^ May 13 15:48:25 i hope nobody will mind when i repeat it? May 13 15:48:34 i still have trouble with my git checking out recipe but i have some new findings. i added a tag=master to my src_uri line and it actually does create the [branch "master"] section in the ./.git/config upon git clone, however the pull command is still performed without a proper branch parameter and still throws the same error http://pastebin.com/d11aa541f :( how can i convice bb to pull the correct branch? May 13 15:49:11 hi XorA May 13 15:49:17 hey florian May 13 15:51:09 tartarus, what is the deal with libtool? May 13 15:52:16 ask RP over there :) May 13 15:52:25 Or google libtool sysroot and see what he had to say at some point May 13 15:53:40 * RP hides May 13 15:53:56 someone should teach libtool about sysroots... May 13 15:54:12 reading up now. May 13 15:55:56 libtool should die in a fire May 13 15:56:11 but i guess teaching it about sysroots is good too May 13 15:56:11 indeed May 13 15:56:13 * RP hands kergoth a lighter May 13 15:56:36 hey kergoth, how does MV solve this? :) May 13 15:57:01 don't think we've been bitten by it as of yet May 13 15:57:13 Er, still support relocation, right? May 13 15:57:38 Is OE using libtool 2.x yet? May 13 15:57:44 that depends on the answer to the question 'relocation of what' May 13 15:57:55 If not, thats something it should get from poky as 2.x is a lot nicer May 13 15:58:08 The conversion nearly drove me mad though :/ May 13 15:58:08 looks like 2.2.4 is current May 13 15:58:11 in oe that is May 13 15:58:16 cool May 13 15:58:27 kergoth, well, installation into an arbitrary location May 13 15:58:46 to which my obvious question is "installation of what"? May 13 15:59:00 kergoth: packaged staging I suspect May 13 15:59:05 heh, how radically different is mv6? :) May 13 15:59:22 well, last i checked, oe has no installers, so i have no idea what you're talking about here. May 13 16:00:12 if you're talking about issues resulting from the use of pstage packages when moving tmp, i've fixed some issues there, yes May 13 16:06:16 kergoth, I simply want to install meta-toolchain.tgz into any dir that I want and have it do the right thing, instead of hardcoding it at build time. Every devel I know complains. Some want the toolchain in /opt, some in /usr, some in ~ May 13 16:11:48 *sigh* May 13 16:22:18 03Angus Ainslie  07fso/milestone5.5 * r5003ee8222 10openembedded.git/recipes/initscripts/ (initscripts-1.0/g_ether.sh initscripts_1.0.bb): initscripts : more g_ether tweaks May 13 16:28:45 anybody here who can explain to me how to seta git src_uri correctly? May 13 16:31:08 uhm seta = set up :) May 13 16:36:44 fraxinas, grep the recipes May 13 16:37:53 gnutoo: i've formatted it correctly after looking in the docu and at other recipes but git pull fails still :/ May 13 16:37:54 fraxinas, ah ok checkouting the oe branch fail...I thought you wanted to create a recipe May 13 16:38:18 no gnutoo you were right in the first place, i'm talking about creating a recipe for gst-plugins-base May 13 16:38:28 ah ok May 13 16:38:42 fraxinas, so grep the recipes for git May 13 16:40:03 fraxinas, I'll be back soon May 13 16:40:04 gnutoo i have, but it seems this repo needs something more tricky May 13 16:40:12 ah okay cool May 13 16:50:10 03Koen Kooi  07org.openembedded.dev * r8126c9725a 10openembedded.git/recipes/udev/udev_141.bb: udev 141: continue cleaning out udev legacy stuff May 13 17:03:00 fraxinas, I'm back May 13 17:04:13 fraxinas, what is the URI of your git repo? May 13 17:04:20 s/your/the/ May 13 17:13:06 Hmm May 13 17:13:18 NOTE: The MD5Sums did not match. Wanted: '62749c6cdf28ce31aae335092fa107df' and May 13 17:13:18 Got: 'd41d8cd98f00b204e9800998ecf8427e' May 13 17:13:18 NOTE: Task failed: Checksum of 'http://downloads.sourceforge.net/hdparm/hdparm-6 May 13 17:13:19 .9.tar.gz' failed May 13 17:19:16 SRC_URI = "git://anongit.freedesktop.org/git/gstreamer/${PN};module=${PN};date=${SRCDATE};tag=master;protocol=git " May 13 17:19:36 with the date being SRCDATE = "20090513" and the PN gst-plugins-base May 13 17:22:00 i've noticed too that it doesn't automatically put master as the default branch like stated in the doc. i explicitely have to put it otherwise the ./.git/config doesn't contain that [branch "master"] section which it doesn when i manually clone the repo from the command line May 13 17:22:20 which it _does_ (-n) May 13 17:23:16 fraxinas, maybe it's possible to do it without git config May 13 17:23:43 fraxinas, what is PN already? May 13 17:24:19 filename's gst-plugins-base_0.10.23.1.bb May 13 17:24:32 ok May 13 17:26:44 any way how to get it to work is welcome to me :) May 13 17:29:47 tag=master is not necessary because it's the default according to the manual May 13 17:30:31 I'll look if srcdate is valid in git May 13 17:30:42 becuase git uses hashes if I remember well May 13 17:31:20 yeah i am not sure about the srcdate either May 13 17:31:39 but it doesn't work even when omitting it alltogether May 13 17:32:25 and about tag=master being default i wrote something just earlier May 13 17:32:41 it does make a difference May 13 17:34:38 ah May 13 17:34:45 fraxinas, could you pastebin your recipe May 13 17:34:50 it would be easier May 13 17:35:53 also maybe module is only for svn-like version control system May 13 17:37:52 if i don't put the module then it doesn't clone the git repo May 13 17:37:58 yeah sure here we go wait... May 13 17:39:17 http://pastebin.com/d78a54619 May 13 17:44:02 fraxinas, in the .inc file there is SRC_URI = "http://gstreamer.freedesktop.org/src/${PN}/${PN}-${PV}.tar.bz2" May 13 17:44:14 I'll re-look at your output May 13 17:44:44 yeah that's the _release_ but i overwrite the src_uri because i need a newer git version May 13 17:45:22 are you shure you are overriting it because it's = not ?= May 13 17:45:27 I'll check May 13 17:45:28 i've seriously taken in consideration checking out manually and diffing against the release so that i can simply add a giant patch, but that would be greatly uglish May 13 17:45:43 yes it would be uglish May 13 17:45:58 actually i am not sure! May 13 17:46:07 let me simply wipe the src_uri from the inc and retry May 13 17:46:54 retrying May 13 17:47:01 ok May 13 17:48:33 same problem still May 13 17:48:37 so that didn't bother it May 13 17:49:28 fraxinas, parsing... May 13 17:49:54 humm? May 13 17:51:18 fraxinas, I'll test your recipe but it's doing that actually: NOTE: Handling BitBake files: - (1629/6769) [24 %] May 13 17:51:38 tag=master? May 13 17:51:56 branch=master would make more sense May 13 17:52:25 sorry for not spoting that but there were too much wrong things... May 13 17:52:32 okay gnu May 13 17:52:44 fraxinas, see better talking in #oe than in private May 13 17:52:50 hehe :) May 13 17:53:37 XorA: cf. section 9.13.5. of the docu, the git src only has a tag parameter but no branch :( May 13 17:54:17 fraxinas: please fix the documentation then May 13 17:54:20 yes I just saw that May 13 17:54:31 fraxinas, send a patch for the doc in the ml May 13 17:54:38 there should be no module on there either May 13 17:54:56 and probably no srcdate? May 13 17:55:16 fraxinas, yes should be a hash normally May 13 17:55:32 okidok i'll find out the correct code May 13 17:55:38 I think SRCDATE works due to historic reasons May 13 17:55:57 but its frowned upon now May 13 17:56:15 ok May 13 17:57:21 anyhoo, time for StarTrek May 13 17:57:36 yaaay May 13 17:57:40 energize May 13 17:58:16 i still need to take my mom to the movie (mother's day present) May 13 18:01:26 ouch...I have to re-parse... May 13 18:01:47 funny enough... but my latest changes produce the exact same result :) May 13 18:01:59 I didn't update my local.conf of the curent target with the COLLECTIONS like I did with another target May 13 18:02:40 http://pastebin.com/d69a034f8 May 13 18:04:24 ok I'll try that recipe instead May 13 18:06:16 fraxinas, do you know bitbake -i and python? you could try to see what's happening while my computer is parsing May 13 18:06:31 interactive mode you mean May 13 18:06:55 haven't used it in a year May 13 18:07:13 03Cliff Brake  07org.openembedded.dev * r0eba3a4797 10openembedded.git/MAINTAINERS: test edit #3 for new repo May 13 18:07:33 BB>> fetch gst-plugins-base May 13 18:08:11 receiving now but that's what it does always. the clone works, the problem is the pull step May 13 18:09:03 hmm yeah same error message again in the -i mode May 13 18:09:25 i need to type python? May 13 18:09:40 fraxinas, no May 13 18:09:51 fraxinas, I meant using peek etc... May 13 18:09:57 okay because now i'm in EXPERT BB> which i am certainly not ;) May 13 18:10:27 you mean like BB>> peek gst-plugins-base TAG May 13 18:10:27 117686abc55c93a8431fef237b4e2ed8ff43d381 May 13 18:10:41 fraxinas, for python: I meant trying to do print values to see what's happening in the clases or in bitbake May 13 18:10:51 I used it once for debuging something May 13 18:10:56 BB>> peek gst-plugins-base SRC_URI May 13 18:10:56 git://anongit.freedesktop.org/git/gstreamer/gst-plugins-base;tag=117686abc55c93a8431fef237b4e2ed8ff43d381;branch=master;protocol=git file://gst-plugins_configure_skip_shave.patch;patch=1;pnum=0 May 13 18:14:34 ok May 13 18:15:00 not okay ^^ sucky² May 13 18:15:31 git is designed for ingenious people like linus himself ^^ May 13 18:15:50 fraxinas: debug prints to git fetcher May 13 18:15:57 code May 13 18:16:02 find out what's wrong May 13 18:17:52 there is that line: runfetchcmd("git clone -n %s://%s%s%s %s" % (ud.proto, username, ud.host, ud.path, repodir), d) May 13 18:17:58 try to print what it does May 13 18:18:00 it's in : May 13 18:18:12 the clone works May 13 18:18:15 it's the pull which fails May 13 18:18:27 rungitcmd("git pull %s://%s%s" % (proto, host, path),d) May 13 18:18:31 it's missing the branch May 13 18:18:49 why would it pull? May 13 18:18:57 I'd do a checkout after clone -n May 13 18:19:05 it does a pull May 13 18:19:10 fraxinas, ok May 13 18:19:22 git pull git://anongit.freedesktop.org/git/gstreamer/gst-plugins-base failed May 13 18:19:27 http://pastebin.com/d69a034f8 May 13 18:19:27 ok May 13 18:19:40 that's what it decided to do... not what i want it to do :) May 13 18:20:04 hmmm hey May 13 18:20:12 you're specifying both tag and branch May 13 18:20:15 in SRC_URI May 13 18:20:16 why? May 13 18:20:16 i don't checkout nor pull because when i clone with -n then i have the source right there in the directory already May 13 18:20:44 just give one of them and it should work May 13 18:20:46 wanted to be on the safe side and not miss anything that could possibly be necessary May 13 18:20:49 ah okay May 13 18:21:25 the fetcher should be fixed to handle situations where both are given May 13 18:21:48 either to give precedence to the tag or fail with a sane error msg May 13 18:21:54 with a randomizer picking either or? May 13 18:21:55 ChangeLog: - Use git-fetch instead of git-pull removing any need for merges when May 13 18:22:32 i might have an outdated version of bb .. .let's see May 13 18:22:43 BitBake Build Tool Core version 1.4.2, bitbake version 1.4.2 May 13 18:22:48 uuh May 13 18:22:54 akheron: nay - same error still May 13 18:23:02 let me guess: update bb May 13 18:23:08 what's the newest, 1.12 or so? :) May 13 18:23:14 yay May 13 18:23:21 sorry for bothering you with the old crap ^^ May 13 18:23:29 i didn't even think of that May 13 18:23:33 you should use the one shipped with oe May 13 18:23:43 yes you have an old version of bitbake May 13 18:24:00 at least in the stable branch, I assume it's in .dev too May 13 18:24:01 update and see if it work,but at least you have something recent to hack on May 13 18:24:09 with _our_ oe comes 1.4.2 (opendreambox) May 13 18:24:37 what's the current recommended version? May 13 18:24:43 the error msg git gives you is quite new May 13 18:24:51 fraxinas, svn version I think May 13 18:24:56 so probably you just have too new git for bb 1.4 May 13 18:25:03 narf does that mean it's gonna need to do the whole rebuild rosters thing? May 13 18:25:19 Crofton|work: you available this Sat to do DNS changes to switch git to new server? May 13 18:25:57 git version 1.6.1.2 May 13 18:26:21 Crofton|work: cgit is now working, so I'm going to propose a move on the mailing list: http://new.openembedded.org/cgit/ May 13 18:26:22 most likely May 13 18:26:37 any idea what time May 13 18:26:48 Crofton|work: does 2PM suit you? May 13 18:28:39 hi May 13 18:28:47 fraxinas: 1.8.12 is last release May 13 18:29:13 fraxinas: but I do not know how much .dreambox branch would need to be changed May 13 18:29:40 hmm i can't really do that anyways May 13 18:29:44 cbrake: can you kill hrw/* branches there? May 13 18:29:49 because it would fail for all of our image builders May 13 18:29:51 :// May 13 18:30:13 annoying :( so that ugly giant patch thing is the only alternative i guess :( May 13 18:30:23 hrw: checking ... May 13 18:31:03 fraxinas: you can work on backporting git fetcher but it would be hard probably May 13 18:31:39 i'll check how many packages will fail when updating to latest bb May 13 18:31:55 fraxinas: I am afraid that most May 13 18:31:56 mmm...what's that? ERROR: log data follows (/home/embedded/oetmp_openmoko/work/armv4t-angstrom-linux-gnueabi/gst-plugins-base-0.10.23.1 git20090513-r1/temp/log.staging_helper.5665) <- space in the name May 13 18:32:04 shit... why is that? May 13 18:32:13 cbrake, earlier would be better May 13 18:32:20 fraxinas: classes/ conf/ would need adaptations May 13 18:32:21 don't know...lol May 13 18:32:36 crap May 13 18:32:58 but it'll have to be done at some point May 13 18:33:03 fraxinas: .dreambox branch did not tracked .dev ;( May 13 18:33:14 Crofton: I'm busy from 9:30-1:00, but any other time of the day should be fine May 13 18:33:16 fraxinas: merge .dreambox into .dev maybe? May 13 18:33:30 that's out of my responsibility hrw May 13 18:35:11 omg okay error spamming when i try it with 1.8.12 May 13 18:35:13 ouch...re-parsing(I did a ctrl+c when I should not have done one) May 13 18:35:26 ah gotto del cache May 13 18:35:52 still :( May 13 18:35:53 fraxinas, I think you need to change the package version with +something May 13 18:37:46 currently i broke my bb May 13 18:38:14 ERROR: local variable 'bb' referenced before assignment while parsing ..... May 13 18:38:24 manual says: PREFERRED_VERSION_linux-omap1_omap5912osk ?= "2.6.18+git" May 13 18:41:57 fraxinas, mine seems to work May 13 18:42:29 fraxinas, I've rev 1179 May 13 18:43:04 bye May 13 18:43:58 gnutoo, what seems to work? May 13 18:44:03 i'm frustrated right now -_- May 13 18:44:08 parsing May 13 18:48:20 (parsing works for me) May 13 18:52:45 i checked out bb-1.8.12 but now everything fails -_- May 13 18:58:33 03Koen Kooi  07org.openembedded.dev * ra6abba0064 10openembedded.git/contrib/angstrom/build-feeds.sh: May 13 18:58:33 angstrom feed builder: allow builder to override machinelists on the commandline May 13 18:58:33 * usage: sh build-feeds.sh machine1 machine2 machineN May 13 19:11:04 Hi, anyone around? May 13 19:13:58 no May 13 19:14:39 03Angus Ainslie  07fso/milestone5.5 * ra6832a2e7c 10openembedded.git/recipes/openmoko-projects/paroli_git.bb: paroli : add paroli-illume recipe May 13 19:20:58 esoren, lol ask your question directly May 13 19:28:57 hi florian May 13 19:29:29 think you could help me with some recipe stuff? May 13 19:31:35 re May 13 19:31:44 taelot: i can try May 13 19:32:28 florian: awesome. I'm just trying to do some basic modifications to a task (task-native-sdk to be precise) May 13 19:33:18 however my additions to RDEPENDS don't seem to get installed, I'm trying to add emacs so I can edit files in a friendlier environment than vi May 13 19:33:38 do I need to do more than add the package to RDEPENDS? May 13 19:35:20 taelot, look at the emacs recipe and check what is packaged? May 13 19:37:04 taelot: did you increase the PR variable so that task-native-sdk is rebuilt at all? May 13 19:37:22 Crofton: what does that get me slash what am I looking for? adding the emacs_22.3 recipe to RDEPENDS isn't enough? May 13 19:38:57 RDEPENDS needs package names :) May 13 19:39:06 not recipe names May 13 19:39:27 try just emacs May 13 19:39:30 but you need to have the DEPENDS as well May 13 19:39:37 yes May 13 19:39:55 DEPENDS to force the build, RDEPENDS to dorce the package install May 13 19:40:28 Crofton: Ah, I see... that makes sense May 13 19:41:35 I'll go try that, thanks all! May 13 19:47:35 hi all May 13 19:48:00 i'm having issues getting "bitbake helloworld" to work on a fresh git checkout on the fso milestone5.5 branch -- can anyone help? May 13 19:48:02 oh well.. I managed to fix a Makefile in oe twice May 13 19:48:42 tonyg: maybe :) May 13 19:48:59 hooray :) May 13 19:49:02 me too! can you help me? :) May 13 19:49:36 I'm having trouble compiling octave because it seems I am missing a fortran compiler May 13 19:49:37 * florian should start an OE consulting company ;) May 13 19:49:37 so: the error is in .../recipes/freesmartphone/frameworkd_git.bb, apparently May 13 19:49:45 bitbake says: May 13 19:50:01 I uploaded the entire log at: http://www.esorensen.com/elog.txt May 13 19:50:05 NOTE: Handling BitBake files: \ (2667/6457) [41 %]NOTE: :EOL while scanning string literal (, line 1) while evaluating: May 13 19:50:05 ${@base_set_filespath([ "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/frameworkd-0.8.5.1+gitr${SRCPV}-r3", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/frameworkd-0.8.5.1+gitr${SRCPV}", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/frameworkd", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/frameworkd-0.8.5.1+gitr${SRCPV}", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmar May 13 19:50:05 tphone/frameworkd", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/files", "/home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone" ], d)} May 13 19:50:07 ERROR: Error in executing: /home/tonyg/openmoko/org.openembedded.dev/recipes/freesmartphone/frameworkd_git.bb May 13 19:50:11 er sorry. May 13 19:50:16 perhaps there's an OE pastebin? May 13 19:50:28 ~pastebin May 13 19:50:30 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste May 13 19:50:36 ta May 13 19:50:51 cbrake, you forgot the TZ info :) May 13 19:51:57 Crofton|work: doh, I thought everyone was EDT :-) May 13 19:52:03 heh May 13 19:52:12 I had to think about what TZ you were in May 13 19:52:28 Crofton|work: thanks, I'll clarify May 13 19:52:53 I did :) May 13 19:53:08 esoren: iic some toolchains don't have fortran indeed. May 13 19:53:28 Its ages ago I need fortran at all... May 13 19:53:34 http://paste.lisp.org/display/80159 May 13 19:53:52 Crofton|work: thanks! May 13 19:54:33 florian, I was thinking I need to edit the gcc files to include fortran and "bitbake -c rebuild gcc" does this make sense? May 13 19:55:09 which branch of git do most people here use? May 13 19:55:44 (and is anyone working with an openmoko gta01) May 13 19:55:49 florian, when I look at gcc-4.1.2.bb I see the following lines: May 13 19:56:11 #Set the fortran bits May 13 19:56:16 # ',fortran' or '', not 'f77' like gcc3 had May 13 19:56:22 FORTRAN = "" May 13 19:57:19 do you think it is as simple as changing the last line to FORTRAN = "fortran" ? May 13 19:57:20 cbrake, Crofton|work: Don't we have UGT here? ;) May 13 19:57:31 heh May 13 19:57:48 this is for server dns change, so relative time would be bad :) May 13 19:58:15 esoren: yes.. mitght be worth a try. its likely not to be tat well tested... May 13 19:59:02 ah... the wiki! May 13 19:59:30 florian on ly git May 13 20:01:17 Crofton|work: ok :) I have an edit window open for a day or so now... May 13 20:01:17 esoren, = ",fortran" May 13 20:01:49 thanks tartarus, I was wondering if that was a typo May 13 20:01:52 Tartarus: do you know why it isn't enabled per default? May 13 20:02:03 florian, one more question. Thanks for your help! Should I try to edit gcc*.bb or gcc-cross*.bb ? May 13 20:02:04 esoren, it should work, or at least tell why it won't in log.do_configure, iirc we ahve the fortran deps as a global May 13 20:02:08 florian, nope May 13 20:02:24 florian, I have a vauge impression there was a specific thing, only being built/cared about on ARM at the time May 13 20:02:33 and a gcc+fortran didn't work on ... something May 13 20:04:10 florian: H; thanks for the ping and the work! May 13 20:04:37 Thanks for your help guys. This gives me something to go off of! :) May 13 20:05:31 hi dth May 13 20:05:43 florian: hi May 13 20:06:20 florian: hmmm gpe, impressive.... May 13 20:08:26 florian: 203 of 500 May 13 20:12:51 does this mean anything to you guys? May 13 20:12:54 configure: error: GMP with MPFR support is required to build fortran May 13 20:14:06 yeah, check that gmp and mpfr are built for the target May 13 20:14:17 I thought those deps were already listed May 13 20:14:19 but perhaps not? May 13 20:15:47 okay May 13 20:24:24 dth: impressive... well, it could need some love :) May 13 20:30:31 florian: do you have any idear how much work is in it? ;-O May 13 20:32:14 dth: gpe? May 13 20:33:24 yes May 13 20:35:37 btw thanks to mwester in #openmoko-cdevel, the bitbake issue i was having has gone away... May 13 20:35:45 ... but now I'm having trouble building libtool-native! May 13 20:35:47 dth: quite a lot... its hard to calculate because a lot of gpe was experimental code. May 13 20:36:00 | configure: error: source directory already configured; run "make distclean" there first May 13 20:36:01 | make[1]: *** [i686-linux-libtool] Error 1 May 13 20:36:18 dth: it is close to 10000 checkins May 13 20:39:26 florian: lot of work May 13 20:39:35 http://paste.lisp.org/display/80163 fwiw May 13 20:41:56 Tartarus, I built MPFR and GMP successfully. I am rebuilding GCC and its at the compile stage. So far so good! Thanks! May 13 20:50:49 03Koen Kooi  07org.openembedded.dev * r5391c3d2f4 10openembedded.git/recipes/gcc/ (gcc-4.3.3.inc gcc-cross_4.3.3.bb gcc_4.3.3.bb): May 13 20:50:49 gcc 4.3.3: fix regression from 4.3.1 that caused fortran to get disabled for linux-gnueabi May 13 20:50:49 * powerpc people are welcome to enable fortran as well :) May 13 20:51:06 heh May 13 21:02:24 wow...they integrated the patches of qemu-arm that were necessary to build emacs into qemu 0.10.x... May 13 21:02:49 I'll clean my emacs-x11 recipe and ask for review May 13 21:07:34 anyone mess with bitbake's eventhandlers much? May 13 21:07:42 wondering about two potential cleanups May 13 21:07:49 1) switch to isinstance() rather than getName May 13 21:08:11 2) bitbake checks if theyr eturn Handled, or anything else. given its the common case, perhaps we should remove all the unneeded 'return NotHandled' fromt he eventhandlers May 13 21:08:15 RP: opinion? May 13 21:09:51 kergoth, RP suggested the best way to bring bitbake stuff to him is use the bitbale list yesterday May 13 21:10:01 this isnt bitbake stuff :) May 13 21:10:05 it's OpenEmbedded class stuff May 13 21:10:12 urg May 13 21:10:22 I need coffee May 13 21:10:38 the event passed into our event handler is a class, kind of like how python does its exceptions. right now they call getName to get the object/class name May 13 21:10:41 which feels.. icky May 13 21:10:55 either the events should be text, or they should be a class hierarchy May 13 21:10:59 but right now its .. both May 13 21:11:02 :) May 13 21:11:41 Crofton|work: if you have a moment, I'd appreciate an opinion on http://thread.gmane.org/gmane.comp.handhelds.openembedded/23680 May 13 21:13:14 basically, you are trying to automatically create a preferred version file for the entire buidl May 13 21:13:21 not just what is in the distro files May 13 21:13:26 so no version floats? May 13 21:13:46 yep. once you've built a given verson of something, that version si preferred until you wipe tmp or manually remove tmp/versions.conf May 13 21:13:58 so your build wont go changing out from underneith you May 13 21:14:21 what if you remove the line from the tmp version May 13 21:14:29 and have bumped the version May 13 21:14:54 basically, you are deliberately chnging one version without chaning everything else May 13 21:15:15 manually remove the preference from tmp/versions.conf? then itll go ahead and switch to the newer, unless your distro or someone has a preference May 13 21:15:22 ok May 13 21:15:32 but you could also just do bitbake foo-1.5 to switch it from 1.3 to 1.5, itll store the new version May 13 21:15:40 what if the distro pref changes? May 13 21:16:01 distro's prefs are ?=. once a version is stored in tmp/versions.conf, itll be used over the distro's prefs May 13 21:16:04 03Tom Rini  07org.openembedded.dev * r657105b567 10openembedded.git/recipes/gcc/ (4 files): gcc*4.3.3: Switch to INC_PR, start at r3 May 13 21:17:34 I wonder what the right thing to do in that case is? May 13 21:18:02 also, is this optional, or does this change existing behavior? May 13 21:18:13 it looks like it changes existing behavior May 13 21:18:43 it changes existing behavior within the context of a single tmp/ dir. wiping tmp brings you back to whatever the defaults were May 13 21:19:36 imo deterministic builds are a necessity. if I'm merrily building ncurses 5.3, and upstream adds 5.7, i don't want to switch to 5.7 unless i wipe tmp to start from scratch May 13 21:20:39 although you could argue that wiping tmp and rebuilding it should build the same thing May 13 21:20:51 you can do that too, by copying tmp/versions.conf to conf May 13 21:20:51 basically, it looks like a good thing May 13 21:20:58 I'd argue that that decision should be more explicit May 13 21:21:07 I'm trying to make sure i understand it so I can explain it May 13 21:21:56 my first implementation directly emitted versions.conf into conf/ in the first path in BBPATH, but i think thats *too* automatic, and too noticable a change May 13 21:22:11 yeah May 13 21:22:30 i dunno, I'm open to ideas, I've already played around with 3 ways of doing it May 13 21:22:46 this time i decided to try to make it sane for use by default May 13 21:23:06 remember the .eu crowd is mostly not working by now, so hopefully some of them comment in their am May 13 21:23:10 yeah May 13 21:23:12 * kergoth_ nods May 13 21:23:18 this seems pretty reasonable May 13 21:23:39 apparently its comments need to be more clear though, since clearly they didn't answer all your questions :) May 13 21:24:37 well, I was asking questions to make sure I understood everything May 13 21:24:45 * kergoth_ nods May 13 21:25:23 it's simple, but it takes a second to wrap your head around the implications in daily use May 13 21:25:59 i thought about trying to make it not lock anything down until you actually get a full successful build, but that adds a whole can of worms to the thing May 13 21:26:02 added complexity May 13 21:26:07 still may be worth looking into, though May 13 21:27:53 Tartarus, are you still around? May 13 21:27:59 yeah May 13 21:29:17 Tartarus, does this error mean anything to you? May 13 21:29:20 http://paste.debian.net/36061/ May 13 21:30:14 Other than I guess that's why fortran was disabled, no, sorry May 13 21:30:26 Although... May 13 21:30:57 Yeah, OK May 13 21:31:02 Here's roughly what you need to do May 13 21:31:16 Add 'file://gfortran.patch;patch=1' to gcc-4.1.2.inc May 13 21:31:31 bitbake -c clean gcc && bitbake gcc May 13 21:31:42 See if (a) the patch applies cleanly and (b) if that fixes the error May 13 21:32:34 I think the patch is already being applied. I see it being applied with the rest of the patches. May 13 21:33:14 Hmm May 13 21:33:20 OK, try not applying it then May 13 21:33:24 okay May 13 21:33:25 Maybe it's erronious there May 13 21:41:24 dth1: you could try to convince your router to reconnect at a better time :) May 13 21:41:30 XorA: wb May 13 21:41:34 hey florian May 13 21:44:44 * florian needs some sleep - good night May 13 22:07:34 * Jay7 -> sleep May 13 22:34:04 does anyone know why anonymous python functions are being run both at the end of the parse of a recipe, and once for *every* task it executes for that recipe? May 13 22:34:10 this seems broken May 13 22:35:28 If anyone is wonder: www.fortransucks.com is available to register. May 13 22:46:37 * XorA|gone has finished the rum, now time to sleep May 13 22:49:26 hello May 13 22:49:45 i'm having problems building a sub package May 13 22:49:49 for example bitbake iputils-ping May 13 22:49:54 RROR: Nothing PROVIDES 'iputils-ping' May 13 22:50:04 how do i fix this. May 13 22:50:04 "sub package"? May 13 22:50:18 bitbake accepts a provider on its commandline, not a runtime package name. May 13 22:50:24 okay. May 13 22:50:34 see bitbake --help May 13 22:50:38 how do i specifiy that in side a image ? May 13 22:50:39 and/or the bitbake manual May 13 22:50:42 what do you mean? May 13 22:50:55 IMAGE_INSTALL = 'iputils-ping May 13 22:51:01 yes? May 13 22:51:03 what about it? May 13 22:51:07 it doesn't work. May 13 22:51:32 "it doesn't work" is a useless declaration. it provides absolutely no useful information May 13 22:51:55 how do i tell the image to install iputils-ping ? May 13 22:52:05 it works for kernel-module-* packages May 13 22:52:48 by putting it in your IMAGE_INSTALL, which in turn puts it inot the image's RDEPENDS. bitbake knows how to determine what package emits what RDEPENDS using the recipes PACKAGES vars. May 13 22:53:51 where are you seeting that IMAGE_INSTALL? May 13 22:54:30 it wasn't infact in my IMAGE_INSTALL, it was in a DEPENDS line of a package. May 13 22:54:58 iputils-ping should not be in a DEPENDS unless there's a iputils-ping.bb somewhere. May 13 22:55:20 you probably meant to put it in RDEPENDS May 13 22:55:23 okay, so how do i say package-x depends on iputils being installed ? May 13 22:55:30 iputils-ping May 13 22:56:00 iputils is a provider, DEPENDS contains providers. iputils-ping is an emitted package, RDEPENDS is runtime dependencies, which is where your binary package names go May 13 22:56:15 as i just said above, you probably meant to add iputils-ping to your RDEPENDS instead of your DEPENDS May 13 22:56:22 ah. i thought RDEPENDS was Reverse Depends. May 13 22:56:24 no May 13 22:56:24 runtime May 13 22:56:31 as the various manuals describe May 13 22:57:21 does as long as you know what to look for. May 13 22:57:42 it's not like people check the manual when they see DEPENDS May 13 22:57:52 your inability to read the manual isn't my problem May 13 22:57:52 they just asume it means what it is logically. May 13 22:57:59 it's a manual, not a dictionary May 13 22:58:20 http://bec-systems.com/oe/html/recipes_depenencies.html May 13 22:58:24 there, read. May 13 23:02:48 lol, and this is an irc channel and not a monastery May 13 23:02:58 was that supposed to make sense? May 13 23:03:20 you just think about it. May 13 23:03:23 thanks for your help May 13 23:04:16 mmmmk May 13 23:08:46 I have an architectural question about OE. Why is the Linux Kernel in OE created the way it is? Why not have OE create a regular GIT version of the Linux Kernel so that developers can use GIT to create experimental branches and merge with future Kernel additions? Why not have OE do a kernel git pull when we do a OE git pull? My problem is that it is all too easy to loose any code changes I make to the Linux kernel. The GIT rep May 13 23:12:02 john3909: that's not how bitbake/oe works for *any* recipe, there's nothing specific to the kernel here. we fetch pristine sources, apply patches to it, and build from there May 13 23:12:17 john3909: admittedly that sucks for the app/kernel developer use case, but it's lovely for some of the other needed use cases May 13 23:13:31 john3909: it *is* possible to create a .bb file that sits in the source tree for a package and is built there, but then you need a way to fetch all of these source trees to build them. lets say you did this for every package in oe, and have 3,000 git repositories for them all, what tool would you use to check them all out to do a build? May 13 23:13:52 * kergoth_ has been looking into this source of thing lately, he wants it improved too, but isnt sure as to the best approach May 13 23:14:55 my current feeling is an architectural split. one tool that takes a recipe in your app's source tree and builds and installs it, then there's a second tool, which is a higher level build tool, more along the lines of enterprise level build/release tools and continuous integration tools which would pull things out of source control, build those recipes and run specific operations like packaging against the output/artifacts coming from those recipes May 13 23:15:07 but if that is the best route, i can't see a clean way to transition to it. May 13 23:16:26 kergoth: that is what I'm talking about. I like your ideas. May 13 23:16:38 if you want to play with the idea of building a recipe that sits in a source tree, i think i have a class for it somewhere. you basically point S=${PWD}, set B=${S}/objdir-${HOST_SYS} or something, override do_fetch, do_unpack, do_patch to do nothing, and make sure TMPDIR/staging are pointing somewhere common to all of the things you're building, since you still need a shared area for linking and stuff May 13 23:17:03 i *always* wanted bitbake to function well from a source tree to help the app/kernel dev use case, but it's never really gotten any love, sadly May 13 23:17:13 I'm hoping i can talk mvista into letting me do some work on such a thing May 13 23:18:18 kergoth: I look forward to your work as this will help a lot. May 13 23:20:15 i have a lot of crazy ideas floating around right now, but haven't quite worked out how to actually make any of it happen. I'd like to avoid trying to do a rewrite, because such things rarely pan out the way you plan, and disrupt things May 13 23:23:26 I also like the idea of building from a source dir. I'm right now rebuilding a kernel after menuconfig. Would be nice to just recompile it in place, but I hear that the way to do a reconfig is create a patch and add it to the kernel.bb... which seems like too much for a reconfig,rebuild cycle May 13 23:23:44 right, indeed May 13 23:23:53 you can implement easier wayts to manipulate the config though May 13 23:24:03 for example, if the defconfig is loaded via a file:// uri May 13 23:24:11 you can manipulate FILESPATHBASE to override it via your build dir May 13 23:24:15 kergoth: would it not be possible to add additional stages to deal with the special case of GIT repositories? Something like a git-reset and a git-pull to get a pristine version of the code, then continue with the patch stage, etc. May 13 23:24:20 FILESPATHBASE =. "${TOPDIR}/files" May 13 23:24:39 cp path/to/kernel/recipedir/linux-1.2.34/defconfig files/linux/ May 13 23:24:43 assuming the PN is linux, there May 13 23:24:47 then itll prefer the one in files/linux May 13 23:24:50 kergoth: after all, there aren't that many GIT repositories in OE. May 13 23:25:13 then, you can make a bitbake task like devshell that uatomaitcally copies the newly altered defconfig from your menuconfig into the override location May 13 23:26:45 john3909: it's conceivable you could implement some altered classes to just clone the once, in a shared location outside of tmp, and operate there, rather than following the usual process.. someone posted such a thing on the list last week i think, in a thread where we got to talking about this sort of thing. it's still suboptimal since the recipe and stuff isn't in the source tree, but it'd improve your use case, I'm sure May 13 23:27:35 john3909: http://thread.gmane.org/gmane.comp.handhelds.openembedded/23403 May 13 23:27:43 kergoth: Thanks, I'll take a look. May 13 23:27:44 esoren, no luck? May 13 23:27:46 john3909: http://thread.gmane.org/gmane.comp.handhelds.openembedded/23403/focus=23412 specifically May 13 23:27:58 john3909: i think roman's post there might be exactly what you want May 13 23:28:18 john3909: i still think we need some bigger changes eventually, but whatever works for now :) May 13 23:28:22 Tartarus, no. May 13 23:28:54 Tartarus, I don't suppose you have any more ideas? May 13 23:29:05 use 4.2.4? :) May 13 23:29:35 john3909: let me know if you come up with any ideas on how to improve matters for you, beyond these workarounds. I'm trying to get some good dialogs about longer term bitbake/oe changes going May 13 23:29:47 I'll try that. May 13 23:30:45 kergoth: I'll surely do that. Thanks again. May 13 23:33:48 kergoth_: thanks for the reconfig suggestion. Are the oe defconfigs documented? I don't see "defconfig" in the UserManual May 13 23:34:20 Tartarus, Can you give me an idea how tell the toolchain to default to a different version of gcc? May 13 23:34:48 I don't know why it is choosing 4.1.2 instead of one of the higher versions May 13 23:35:08 what i described about filespath isn't in any way specific to defconfigs. FILESPATH is how bitbake finds file:// files, and FILESPATHBASE is the lowest piece of its construction, it enables you to easily override any file:// file without messing up the collection May 13 23:35:47 I see May 13 23:39:40 esoren, what DISTRO? May 13 23:41:18 I am developing for a Gumstix system. I found an include file with lots of lines like "PREFERRED_VERSION_gumstix-kernel = "2.6.21" " May 13 23:41:36 I am planning on adding one for gcc. Does that sound right? May 13 23:41:59 there's one already in conf/distro/gumstix bet May 13 23:42:00 must go May 13 23:42:32 thanks for the help May 13 23:45:46 esoren, which gumstix/ May 13 23:46:18 Crofton|work, verdex pro xm4 May 13 23:46:51 the OE that is supplied by gumstix is old May 13 23:47:03 yeah I've noticed :( May 13 23:47:47 badger gumstix to pay to have it updates so they support all product from one version of OE :) May 13 23:48:41 Tartarus, the conf/distro/gumstix file isn't there. But I did find a comment in the gumstix.inc file that explains it. May 13 23:48:53 # The following *should* be in a distro.conf file # We include them here to avoid creating a new distro May 13 23:51:09 esoren, you might try building against .dev May 13 23:51:41 but I'm not sure how much stuff we copied back from the gumstix copy May 13 23:52:19 we would welcome patches to make dev work for old style gumstix :) May 13 23:53:55 what is the full url for the .dev ? May 13 23:54:03 I might just can this and start over soon May 13 23:55:08 Actually, if I recall correctly I was using the .dev when I was developing for the Gumstix Overo May 13 23:55:28 And I had far fewer problems. May 13 23:56:53 yeah Overo is very close to dev May 13 23:57:03 sakoman updates from .dev there May 13 23:57:08 git has made this easier May 13 23:57:27 there is progress, it is just hard for the older product May 13 23:59:39 I really liked developing on the Overo. But the hardware wasn't suitable for my application. :( May 14 00:00:16 why not? May 14 00:00:42 I am using the onboard sound to record hydrophones underwater. May 14 00:00:51 try .dev and building for the gumstix May 14 00:00:59 The summit board for the Overo is very noisy. May 14 00:01:01 what is wrong with overo for that? May 14 00:01:04 ah May 14 00:01:11 but the older ones are not? May 14 00:01:28 The verdex board is extremely clean. May 14 00:01:57 Sakoman actually helped me when I was trying to figure that out. May 14 00:02:50 He's a great guy May 14 00:02:55 I should note that the Overo itself probably ins't noisy. May 14 00:03:16 I suspect it is fine for typical voice apps May 14 00:03:33 could you use some form of usb sound card May 14 00:03:48 I was looking into that May 14 00:04:01 but I am trying to keep the form factor and power at a minimum May 14 00:04:14 these sensors are being deployed underwater for 2-3 days at a time. May 14 00:04:24 try Angstrom and the gumstix machine in .dev May 14 00:04:27 cool May 14 00:04:58 yeah, we have loads of work to do on the omap3 kernel to get the PM stuff working well :( May 14 00:05:51 you are suing a verdex mb? May 14 00:05:57 yep May 14 00:06:14 there is a chance it will work May 14 00:06:39 openembedded/conf/machine/gumstix-verdex.conf May 14 00:07:51 it sounds like an awesome project :) May 14 00:08:00 thanks! May 14 00:08:25 I'm having a lot of fun with it, but I am rapidly approaching some deadlines with a lot of work ahead of me. :) May 14 00:08:36 Good thing there are 24 hours in a day! May 14 00:08:38 yeah May 14 00:08:44 I know how that is May 14 00:09:10 if you have anything that wouild make it easier, I can commit to .dev May 14 00:09:40 I am going to try going the .dev route on my laptop right now May 14 00:09:57 that way I won't contaminate anything I have working already. May 14 00:09:58 yeah, it helps to have a few machies to run builds on :) May 14 00:10:02 right May 14 00:10:50 when I first downloaded the files I used: May 14 00:10:52 svn co https://gumstix.svn.sourceforge.net/svnroot/gumstix/trunk gumstix-oe May 14 00:11:08 what should I use to get the .dev files? May 14 00:12:34 Tartarus, I know you are gone, but I just thought I'd let you know the 4.2.4 build failed with the same error. :( May 14 00:13:30 http://wiki.openembedded.net/index.php/Getting_started May 14 00:13:55 thanks May 14 00:14:14 gl May 14 00:44:40 what's the magic to get a debuggable binary? Looks like the stripped info is saved to PN/some/path/.debug/binary_name May 14 00:45:25 how to rejoin debug part to the stripped executable? May 14 00:52:37 hoj, few min.. May 14 00:53:46 k, though a bus might carry me away ~ :) May 14 00:55:29 heh May 14 00:56:31 so, no need to combine May 14 00:56:37 gdb handles them split just fine May 14 00:56:50 just need to 'set sysroot /path/to/extrated/image" May 14 00:57:00 or set solib-absolute-prefix same/value May 14 00:57:05 now back to my kid May 14 00:57:15 thanks **** ENDING LOGGING AT Thu May 14 02:59:57 2009