**** BEGIN LOGGING AT Fri Sep 24 10:52:15 2010 Sep 24 10:52:20 03Koen Kooi  07org.openembedded.dev * rdf9bdda99e 10openembedded.git/recipes/networkmanager/networkmanager_0.8.1.bb: networkmanager: add 0.8.1 Sep 24 11:02:17 JDuke128: check if the igep0020 is listed in the kernel recipe to raise DEFAULT_PREFERENCE there Sep 24 11:05:22 woglinde_: that 4.1 issue is because most maintainers never clean up Sep 24 11:07:03 effem same with cups Sep 24 11:07:18 i know Sep 24 11:07:41 cups: will push 1.4.4 soon Sep 24 11:07:51 works fine here Sep 24 11:07:53 on arv5 Sep 24 11:08:00 afaik, i'd say remove them, actually there are more recipes with security vulnerabilities Sep 24 11:08:04 do you think I remove the previous one ? Sep 24 11:08:17 ok Sep 24 11:09:40 e.g. openssl 0.9.8m also has a few of these, mentioned this before, but no one fixed it (and I decided not to touch openssl again) Sep 24 11:09:59 ericben fine Sep 24 11:10:07 some people consider having lots of versions around as a sign of quality :-( Sep 24 11:14:21 omg why does it start building iphone stuff when i switch my opendreambox oe over to gcc 4.5 ?!! Sep 24 11:17:11 florian ,DEFAULT_PREFERENCE where ? Sep 24 11:17:34 florian , what u mean ? Sep 24 11:19:19 JDuke128: you need to set it in the linux bb you are going to use too. just take a look, ususally there is a list on top of the linux...bb Sep 24 11:32:25 CIA-68: ping Sep 24 11:32:28 03Marcin Juszkiewicz  07org.openembedded.dev * r0e3642cacf 10openembedded.git/recipes/gcc/ (6 files): (log message trimmed) Sep 24 11:32:28 gcc: added 2010.09 Linaro release of gcc-4.4 Sep 24 11:32:28 more about release: https://wiki.linaro.org/WorkingGroups/ToolChain Sep 24 11:32:28 This version does not have any OE patches applied (other then Sep 24 11:32:28 zecke-no-host-includes.patch in gcc-cross). Sep 24 11:32:29 Built armv7-a console-image without problems with binutils-cross 2.20 - Sep 24 11:32:30 modified Angstrom distribution was used. Did not booted into is because Sep 24 11:34:28 florian , virtual/kernel whats its real name of .bb file ? Sep 24 11:36:55 03Marcin Juszkiewicz  07org.openembedded.dev * r845625651f 10openembedded.git/recipes/gcc/gcc-4.4-2010.09.inc: gcc: fix INC_PR of Linaro gcc-4.4 Sep 24 11:41:07 JDuke128: grep igep recipes/linux/* Sep 24 11:45:04 ericben , thanks i got 1 file Sep 24 11:45:47 i add there new line ? Sep 24 11:46:08 i add there a new line ? Sep 24 11:46:50 http://pastebin.com/0FThxv7h Sep 24 11:47:49 well you should wait until you can pull from git server in order to get the 2.6.33 recipe Sep 24 11:48:25 or get it through cgit : http://git.openembedded.org/cgit.cgi/openembedded/tree/recipes/linux/linux-igep2_2.6.33.7-0.bb Sep 24 11:48:42 + this dir : http://git.openembedded.org/cgit.cgi/openembedded/tree/recipes/linux/linux-igep2-2.6.33.7-0 Sep 24 11:49:04 git clone http://git.openembedded.org/cgit.cgi/openembedded/tree/recipes/linux/linux-igep2-2.6.33.7-0 Sep 24 11:49:07 like this i do ? Sep 24 11:50:12 why we cant use "git pull" ? Sep 24 11:51:57 warning: remote HEAD refers to nonexistent ref, unable to checkout. Sep 24 11:52:00 waow Sep 24 11:53:28 when will server open ? Sep 24 11:53:32 git Sep 24 11:53:45 jduke when ka6sox is awake Sep 24 11:54:14 :) Sep 24 12:01:17 03Koen Kooi  07org.openembedded.dev * r065ed847e4 10openembedded.git/recipes/gtk-webcore/ (files/waf midori_0.2.8.bb): Sep 24 12:01:17 midori: add 0.2.8 Sep 24 12:01:17 * waf hacks aren't necessary anymore \o/ Sep 24 12:01:27 03Koen Kooi  07org.openembedded.dev * r5d462f7230 10openembedded.git/recipes/webkit/ (5 files in 2 dirs): webkit-gtk: bump SRCREV Sep 24 12:13:46 hmm. gcc and gcc-cross-sdk failed Sep 24 12:24:35 JDuke127: git://git.openembedded.org should work now Sep 24 12:28:03 cbrake you fixed it? Sep 24 12:33:24 woglinde_: yes Sep 24 12:34:41 woglinde_: the service script did not have a full path to git-daemon so it was not starting automatically on re-boot Sep 24 12:35:36 ah Sep 24 12:35:54 hm maybee ka6sox overlooked it yesterday Sep 24 12:41:21 03Klaus Kurzmann  07org.openembedded.dev * r6e76c88872 10openembedded.git/recipes/xinput-calibrator/ (2 files in 2 dirs): Sep 24 12:41:22 pointercal-xinput: add calibration data for nokia900 and bump PR Sep 24 12:41:22 Signed-off-by: Klaus Kurzmann Sep 24 12:41:22 Signed-off-by: Martin Jansa Sep 24 12:51:25 mrmoku, hi Sep 24 12:51:41 mrmoku, how did you get the logs of maemo's xserver? Sep 24 12:53:54 oops wrong channel Sep 24 13:17:42 03Eric Bénard  07org.openembedded.dev * r410035824b 10openembedded.git/classes/update-rc.d.bbclass: (log message trimmed) Sep 24 13:17:43 classes/update-rc.d.bbclass: move update-rc after postinst Sep 24 13:17:43 without this, launching a daemon can fail if postinst is supposed, for Sep 24 13:17:43 example, to fix permissions of directories used by the daemon. Sep 24 13:17:43 Signed-off-by: Eric Bénard Sep 24 13:17:43 Acked-by: Marcin Juszkiewicz Sep 24 13:17:44 Acked-by: Frans Meulenbroeks Sep 24 13:17:45 03Eric Bénard  07org.openembedded.dev * r2c8a897ee7 10openembedded.git/recipes/perl/liburi-perl_1.52.bb: Sep 24 13:17:46 liburi-perl: update to 1.5.5 Sep 24 13:17:46 Signed-off-by: Eric Bénard Sep 24 13:17:47 Acked-by: Frans Meulenbroeks Sep 24 13:17:47 03Eric Bénard  07org.openembedded.dev * r2218504b04 10openembedded.git/recipes/nostromo/ (5 files in 2 dirs): (log message trimmed) Sep 24 13:17:48 nostromo: add new recipe for this web server Sep 24 13:17:49 * description taken from its website : Sep 24 13:17:50 nhttpd is a simple, fast and secure HTTP server. Sep 24 13:17:50 It runs as a single process, handling connections with select(2). Sep 24 13:17:51 For CGIs and directory listing it does fork(2). nhttpd has the Sep 24 13:17:51 minimum of HTTP/1.1 and CGI/1.1 implemented. Also supported are; Sep 24 13:18:28 are compiled for the target and thus can't be executed. Sep 24 13:18:28 * tested on armv5te Sep 24 13:18:28 Signed-off-by: Eric Bénard Sep 24 13:18:28 03Eric Bénard  07org.openembedded.dev * rf7c547ede1 10openembedded.git/recipes/ppp/ (9 files in 2 dirs): Sep 24 13:18:28 ppp: add version 2.4.5 Sep 24 13:18:28 * tested on armv5te with RTC and GPRS Sep 24 13:18:29 * DEFAULT_PREFERENCE = "-1" for the moment until it's better tested Sep 24 13:18:30 (18 lines omitted) Sep 24 13:18:33 03Eric Bénard  07org.openembedded.dev * r693cf93d04 10openembedded.git/recipes/gs/gs-fonts-std_8.11.bb: Sep 24 13:18:33 gs: add gs-fonts-std Sep 24 13:18:33 * thsi is one of the font packages gs can use to render ps files Sep 24 13:18:33 Signed-off-by: Eric Bénard Sep 24 13:18:51 03Michael Kurz  07org.openembedded.dev * r2abf3e0501 10openembedded.git/recipes/libesmtp/libesmtp_1.0.4.bb: Sep 24 13:18:51 libesmtp_1.0.4.bb: fix the installed but not shipped in any package warning. Sep 24 13:18:51 * add esmtp-plugins/*.la to -dev package Sep 24 13:18:51 * add esmtp-plugins/*.a to -static package Sep 24 13:18:51 * move libesmtp-config to the -dev package Sep 24 13:18:51 Signed-off-by: Michael Kurz Sep 24 13:18:52 Signed-off-by: Eric Bénard Sep 24 13:18:53 03Eric Bénard  07org.openembedded.dev * r1316a0c0c4 10openembedded.git/recipes/apache2/apache2_2.2.14.bb: Sep 24 13:18:53 apache2: unbreak recipe Sep 24 13:18:54 * tested on armv5te / angstrom Sep 24 13:18:54 Signed-off-by: Eric Bénard Sep 24 13:50:06 03Koen Kooi  07org.openembedded.dev * r855cee4976 10openembedded.git/recipes/networkmanager/networkmanager-openvpn_0.8.1.bb: networkmanager-openvpn: add 0.8.1 Sep 24 13:50:09 03Koen Kooi  07org.openembedded.dev * r0fb9215034 10openembedded.git/recipes/networkmanager/networkmanager_0.8.1.bb: networkmanager: bump PR for ppp Sep 24 14:35:04 morning Sep 24 14:35:29 gm Sep 24 14:44:42 03Chris Larson  07master * r99c324fe73 10bitbake.git/lib/bb/cache.py: Sep 24 14:44:42 bb.cache: fix bug from last commit with loading from cache Sep 24 14:44:42 Signed-off-by: Chris Larson Sep 24 14:50:42 * kergoth mutters Sep 24 14:50:49 03Chris Larson  07master * rf29ceb22b6 10bitbake.git/lib/bb/cache.py: Sep 24 14:50:49 Fix another little bug with undefined var from cache Sep 24 14:50:49 Signed-off-by: Chris Larson Sep 24 14:56:45 sheesh Sep 24 14:56:54 bitbake is using a meg of ram just for FetchData instances Sep 24 14:58:49 hi all Sep 24 15:03:56 grrr it doesn't boot anymore Sep 24 15:03:58 on n900 Sep 24 15:04:23 ahh wait a sec Sep 24 15:04:25 cover open Sep 24 15:04:26 hmmm Sep 24 15:05:14 oops Sep 24 15:05:18 sorry wrong channel Sep 24 15:07:35 hi Sep 24 15:09:37 03Frans Meulenbroeks  07org.openembedded.dev * r2aa46c0ad9 10openembedded.git/recipes/dosfstools/files/ (5 files): Sep 24 15:09:37 dosfstools : moved unused files to obsolete dir Sep 24 15:09:37 Signed-off-by: Frans Meulenbroeks Sep 24 15:22:47 03Koen Kooi  07org.openembedded.dev * re06553979d 10openembedded.git/recipes/speex/ (speex-fpu.inc speex_1.2.bb): speex: make float and asm settings more intelligent Sep 24 16:05:05 03Koen Kooi  07org.openembedded.dev * r4bb71df326 10openembedded.git/recipes/dhcp/ (7 files in 2 dirs): dhcp: add 4.x from poky Sep 24 16:05:16 03Koen Kooi  07org.openembedded.dev * rf5b20e5258 10openembedded.git/conf/distro/include/ (2 files): angstrom: switch to newer networkmanager Sep 24 16:05:17 03Koen Kooi  07org.openembedded.dev * rb7c4a26d9b 10openembedded.git/recipes/networkmanager/ (2 files in 2 dirs): networkmanager 0.8.1: cleanup recipe, fix loopback init with a patch from debian, fix dhclient and iptable paths as well Sep 24 16:21:21 kergoth ping Sep 24 16:22:46 03Frans Meulenbroeks  07org.openembedded.dev * re06e9cd83c 10openembedded.git/recipes/directfb/ (directfb-1.1.1/fix-pkgconfig-cflags.patch directfb_1.1.1.bb): (log message trimmed) Sep 24 16:22:46 directfb-1.1.1: remove Sep 24 16:22:46 code is from 2008, not pinned Sep 24 16:22:46 src tarball is not available any more Sep 24 16:22:46 (see http://www.directfb.org/downloads/Old/) Sep 24 16:22:46 lastest version we have is 1.4.2 Sep 24 16:22:47 Signed-off-by: Frans Meulenbroeks Sep 24 16:23:33 When you install the udev package, does it auto-mount all USB devices automatically? Sep 24 16:24:39 no Sep 24 16:25:10 Hmm.. it seems to, for me. Sep 24 16:25:10 you need udev and usbmount Sep 24 16:26:03 03Frans Meulenbroeks  07org.openembedded.dev * r9b713a003c 10openembedded.git/recipes/directfb/ (6 files in 3 dirs): Sep 24 16:26:03 directfb: remove some old versions Sep 24 16:26:03 Signed-off-by: Frans Meulenbroeks Sep 24 16:26:03 Acked-by: Henning Heinold Sep 24 16:31:21 03Frans Meulenbroeks  07org.openembedded.dev * re86bc83bc2 10openembedded.git/recipes/directfb/files/ (fix-includes.patch fix-pkgconfig-specs.patch): Sep 24 16:31:21 directfb: moved unused files to obsolete dir Sep 24 16:31:21 Signed-off-by: Frans Meulenbroeks Sep 24 16:33:14 ping eicben Sep 24 16:33:18 ups ericben Sep 24 16:33:30 yes woglinde Sep 24 16:33:43 hm has cups its own configure script? Sep 24 16:33:55 what do you mean by configure script ? Sep 24 16:34:02 or why did you patch configure instead of configure.in/ac 4-macros Sep 24 16:35:14 I started from 1.4.3 recipe which is patching configure Sep 24 16:35:26 not good Sep 24 16:35:57 patching configure and Makefile.in is the wrong way Sep 24 16:38:48 ok what is the right way ? fixing the ac4 macros and not having a do_configure and letting autotools' one doing the job ? Sep 24 16:38:57 right Sep 24 16:39:15 in fact? I didn't check if this patch is still necessary, will do it asap Sep 24 16:39:42 it was in 1.4.3 so it went to 1.4.4 ... Sep 24 16:40:51 Hmm, so udev does automatically mount block devices as they are added. Sep 24 16:40:55 Why does it do that? Sep 24 16:41:12 are you really sure Sep 24 16:41:18 did you check your fstab? Sep 24 16:41:20 Yes, it's in local.rules Sep 24 16:41:22 The rule for it Sep 24 16:41:25 hm Sep 24 16:41:30 /etc/udev/rules.d/local.rules Sep 24 16:41:31 remove the rule Sep 24 16:41:39 And it does not do that. Sep 24 16:41:51 I was just wondering why it was decided that udev should do that Sep 24 16:44:12 Mmm.. git's not smart enough to track the packages/->recipes/ path rename in 2009 Sep 24 16:53:00 03Martin Jansa  07org.openembedded.dev * rd33ca8e166 10openembedded.git/ (5 files in 5 dirs): (log message trimmed) Sep 24 16:53:01 EFL: bump SRCREV Sep 24 16:53:01 * python-efl: python EFL bindings are now using autotools instead of setuptools started with r52282 Sep 24 16:53:01 * python-efl: prefix includedir value returned from pkg-config --variable=includedir with STAGING_DIR_HOST Sep 24 16:53:01 * python-efl: don't install .svn files with examples Sep 24 16:53:01 * python-edje: update patch for unicode conversion Sep 24 16:53:01 * elmdentica: jump back to EFL_SRCREV Sep 24 16:53:58 good bye Sep 24 16:55:42 kelvie_, see git log --follow and -M/-C. it works fine for tracking across it, but not at the directory level, at the file level. for dirs, you can just specify both -- e.g. git log -- packages/foo/ recipes/foo/ Sep 24 16:56:22 kergoth, all good now? Sep 24 16:56:41 what is? Sep 24 16:59:50 kergoth, did you manage to peek into the SRC_URI issue ? Sep 24 17:00:52 kergoth`: yeah, I did that Sep 24 17:01:04 It just got added in 2005 Sep 24 17:01:09 Because someone wanted automount, I suppose. Sep 24 17:01:32 heh Sep 24 17:02:09 I guess no one has complained for 5 years, so people do want usb automount by default Sep 24 17:02:23 kergoth: I am trying to package /lib64 so I set libdir=/usr/lib64 and base_libdir=/lib64 but it also trickled into native recipes Sep 24 17:02:27 or don't use udev :-) Sep 24 17:02:47 Ooh, what alternatives are there? Sep 24 17:02:54 kergoth: currently OE is hell bent to only package /usr/lib and /lib only Sep 24 17:02:56 busybox's mdev Sep 24 17:02:56 mdev? Sep 24 17:03:01 Mm Sep 24 17:03:23 kergoth: what I want is change the libdir for target but not for host/native Sep 24 17:04:34 heh, would be nice if that was default behavior Sep 24 17:05:02 btw what is the proper/best way to get the testing-next branch (what to do after clone; git checkout --track -b testing-next does not seem to do the job Sep 24 17:05:35 If I use mdev in an initramfs, do I have to kill it before I mount the rootfs and init starts udev? Sep 24 17:06:07 effeM: with recent git, git checkout testing-next Sep 24 17:06:17 effeM: otherwise git checkout -b testing-next origin/testing-next Sep 24 17:06:24 track is default nowadays Sep 24 17:06:33 kelvie_: if you use it only to create the initial devices (mdev -s) it won't run as a daemon Sep 24 17:06:42 ericben: oh, interesting. Sep 24 17:06:48 I guess udev was a bit overkill here. Sep 24 17:06:55 in fact it don't run as a daemon Sep 24 17:07:14 kergoth ah ok, mixed up things Sep 24 17:07:38 with recent git, if you try to check out a local branch that doesn't exist locally and does exist upstream, it'll set it up for you Sep 24 17:07:41 which is convenient Sep 24 17:08:12 * khem observed that too Sep 24 17:08:19 kergoth do i need to be inside the openembedded dir for checkout (tyring to recreate my autobuilder script as the system had issues and ict nuked the vm) Sep 24 17:08:32 yes you need .git Sep 24 17:08:36 to do all this Sep 24 17:08:37 i'm not sure how you think git would know what repository you're checking out in if you're not in there Sep 24 17:08:39 it's not psychic Sep 24 17:08:48 meh Sep 24 17:08:50 true Sep 24 17:09:19 kergoth: any idea on my quandry Sep 24 17:09:36 not offhand, other than forcibly setting libdir_native to /lib Sep 24 17:09:53 hmmm ok that would work Sep 24 17:09:58 lemme try that Sep 24 17:12:19 kergoth: hmmm native recipes do use libdir and base_libdir Sep 24 17:12:58 look at native.bbclass Sep 24 17:13:08 could be i'm remembering wrong though Sep 24 17:14:54 what could be wrong here? http://pastebin.org/1151646 Sep 24 17:23:24 hm, still confused with git; I did a git checkout testing-next followed by a git pull but then I am again (?) at the master branch. Do I need to do more? this is on our autobuilder and I cannot login to it, only supply a build script) Sep 24 17:23:42 ? Sep 24 17:23:50 git pull doesn't change the branch you're on Sep 24 17:24:03 it merges upstream into the current branch after fetching upstream, that's it Sep 24 17:24:08 kergoth that is what I thought Sep 24 17:24:56 this is what the script does: Sep 24 17:24:58 git clone git://git.openembedded.org/openembedded Sep 24 17:24:58 (cd openembedded; git checkout testing-next) Sep 24 17:25:09 followed by (cd openembedded;git pul Sep 24 17:25:23 (the first two lines are only done if openembedded does not exist) Sep 24 17:25:52 03Martin Jansa  07org.openembedded.dev * rda43f76972 10openembedded.git/recipes/shr/ (5 files): Sep 24 17:25:52 e-wm-theme-illume: PACKAGE_ARCH all for all SHR themes Sep 24 17:25:52 Signed-off-by: Martin Jansa Sep 24 17:29:07 03Martin Jansa  07org.openembedded.dev * r6fe2aa105a 10openembedded.git/recipes/shr/ (e-wm-theme-illume-gry_git.bb e-wm-theme-illume-neo_git.bb): e-wm-theme-illume-(gry|neo): bump SRCREV for last fix Sep 24 17:29:40 hm, the checkout failed: Sep 24 17:29:41 + cd openembedded + git checkout testing-next error: pathspec 'testing-next' did not match any file(s) known to git. Sep 24 17:31:32 eFfeM: git branch -a | grep testing-next Sep 24 17:32:01 effeM: like i said, that only works with very recent git :) Sep 24 17:32:05 do the -b version instead Sep 24 17:32:17 will do Sep 24 17:43:25 03Frans Meulenbroeks  07org.openembedded.dev * rbab1c136c9 10openembedded.git/recipes/cdparanoia/cdparanoia/ (Makefile.patch fixes.patch): Sep 24 17:43:25 cdparanoia : moved unused files to obsolete dir Sep 24 17:43:25 Signed-off-by: Frans Meulenbroeks Sep 24 17:49:18 kergoth: is use of layout_* vars not encouraged ? Sep 24 17:50:18 i .. can't remember. i think maybe they became unnecessary with the switch to sysroot from staging Sep 24 17:50:59 kergoth: but libdir is used in both native and target recipes Sep 24 17:51:29 kergoth: I wished we have libdir_native and libdir_target Sep 24 17:51:51 we have prefix_native -- just add a libdir_native too Sep 24 17:52:02 lib should be the only other path that may differ, otehr than prefix, i would think Sep 24 17:52:20 libdir_native=${prefix_native}/lib Sep 24 17:53:00 yeah Sep 24 17:53:11 then libdir = ${libdir_native} in native.bbclass Sep 24 17:53:18 right Sep 24 17:53:21 lemme try that Sep 24 17:53:22 we really should get a _target override set up at some point Sep 24 17:56:59 base_libdir_native is needed too Sep 24 17:57:15 ah, indeed Sep 24 17:57:26 clearly not many people have worked on multilib stuff Sep 24 17:57:34 right Sep 24 17:58:19 actually first step I am trying to use a different multilib dir although full scale multilibbing will be a later job Sep 24 17:58:32 * kergoth nods Sep 24 18:00:36 once I can make OE to choose between lib and lib64 Sep 24 18:00:43 that will be half the battle Sep 24 18:01:04 things get really "fun" when you start to look at multilib setups that need more than two configurations Sep 24 18:01:06 :) Sep 24 18:03:10 yeah mips has 3 Sep 24 18:03:24 I think it would be that much we could support Sep 24 18:03:37 more than that will become inhumane Sep 24 18:04:26 itd be cool if we could automagically create recipe variants to actually *build* different tools for all available multilib configurations, rather than just the core C lib, etc... but bitbake doesn't support that, unless you create a class per variant -- lib64.bbclass, etc... Sep 24 18:05:33 you mean to build recipe 3 times itself or the tools to support it build that way Sep 24 18:05:49 yeah. well, having the variants doesn't mean you actually have to build them all Sep 24 18:05:56 but you could pick and choose which of which to add to your image Sep 24 18:06:11 ugh that would be fantabolous Sep 24 18:06:22 ncurses, ncurses-lib64, etc... Sep 24 18:06:36 right Sep 24 18:06:49 makes more sense for x86 type of systems Sep 24 18:06:54 but OE is ARM infested Sep 24 18:07:16 ERROR: Task 3 (/home/kraj/work/oe/openembedded/recipes/stage-manager/stagemanager-native_0.0.1.bb, do_populate_sysroot) failed with exit code '1' Sep 24 18:07:17 you could use bbclassextend for it, its just you'd need one class per configuration, which is rather ugly Sep 24 18:07:21 /scratch/oe/work/i686-linux/stagemanager-native-0.0.1-r13/temp/run.sysroot_stage_all.2337: line 513: autotools_stage_dir: command not found Sep 24 18:08:12 autotools_stage_dir where is it Sep 24 18:14:44 03Martin Jansa  07org.openembedded.dev * rb5e7497030 10openembedded.git/recipes/freesmartphone/libfreesmartphone-glib_git.bb: Sep 24 18:14:44 libfreesmartphone-glib: bump SRCREV Sep 24 18:14:44 Signed-off-by: Martin Jansa Sep 24 18:14:47 03Martin Jansa  07org.openembedded.dev * r6294fb831a 10openembedded.git/recipes/openmoko-3rdparty/mokosuite2_git.bb: Sep 24 18:14:47 mokosuite2: bump PR after libfreesmartphone-glib change Sep 24 18:14:47 Signed-off-by: Martin Jansa Sep 24 18:15:33 kergoth: staging.bbclass line 76 Sep 24 18:15:50 kergoth: why is autotools_stage_dir used instead of sysroot_stage_dir Sep 24 18:16:04 seems a typo Sep 24 18:22:39 dunno, probably is Sep 24 18:25:07 I think it is Sep 24 18:25:12 I am sending a patch Sep 24 18:25:18 now stagemanager is happy Sep 24 18:32:25 03Koen Kooi  07org.openembedded.dev * r266656047a 10openembedded.git/recipes/gnome/epiphany_2.30.2.bb: epiphany: bump PR for webkit change Sep 24 18:34:18 03Koen Kooi  07org.openembedded.dev * r68db203261 10openembedded.git/recipes/speex/speex_1.2.bb: speex: disable asm for arm since too many toolchains trip up Sep 24 18:49:04 03Khem Raj  07master * raf5d0dbecc 10openembedded.git/classes/staging.bbclass: Sep 24 18:49:04 staging.bbclass: Use sysroot_stage_dir instead of autotools_stage_dir Sep 24 18:49:04 Some packages like package-stagemanager-native dont use autoools Sep 24 18:49:04 and autotools_stage_dir can not be found for such recipes. Sep 24 18:49:04 Signed-off-by: Khem Raj Sep 24 18:49:04 Acked-by: Chris Larson Sep 24 18:49:47 03Khem Raj  07master * rcf18d5677a 10openembedded.git/recipes/eglibc/eglibc.inc: (log message trimmed) Sep 24 18:49:47 eglibc.inc: Transfer PARALLEL_MAKE to PARALLELMFLAGS Sep 24 18:49:47 * Currently we are passing -jN flags to eglibc makefiles Sep 24 18:49:47 however eglibc uses a make variable called PARALLELMFLAGS Sep 24 18:49:47 to pass information about parallel build to sub-makes Sep 24 18:49:48 * Currently make install ran without parallelism but now Sep 24 18:49:49 it should have parallel building flags passed to do_install Sep 24 19:01:40 khem: make install doesn't normally get -j in any recipes Sep 24 19:05:15 kergoth: ok any reason for that Sep 24 19:05:38 kergoth: eglibc builds improve if we do parallel make install Sep 24 19:05:54 far, far more buildsystems fail with -j at install than they do at compile Sep 24 19:06:07 and the benefit is much less in most cases Sep 24 19:06:39 I see Sep 24 19:07:04 eglibc could be excused Sep 24 19:07:20 03Roman I Khimov  07org.openembedded.dev * r5c658a893e 10openembedded.git/recipes/hostap/ (hostap-daemon-0.7.3/defconfig hostap-daemon_0.7.3.bb): Sep 24 19:07:21 hostap-daemon: add version 0.7.3 Sep 24 19:07:21 Signed-off-by: Roman I Khimov Sep 24 19:07:30 for one its huge time and two its install works in parallel Sep 24 19:07:31 03Roman I Khimov  07org.openembedded.dev * rb495b6f66a 10openembedded.git/recipes/groff/groff_1.20.1.bb: Sep 24 19:07:31 groff: fix groff-native build Sep 24 19:07:31 Fails at install stage in doc/ with missing images otherwise. Sep 24 19:07:31 Signed-off-by: Roman I Khimov Sep 24 19:10:29 khem: yeah, i could see it being used on a case by case basis, just wanted to make sure you knew Sep 24 19:10:40 I thought so Sep 24 19:10:51 there might be a reason Sep 24 19:41:12 couchdb -- some very interesting stuff: http://www.youtube.com/watch?v=ESDBM9-U804 Sep 24 19:41:19 so who is going to add it to OE :-) Sep 24 19:46:46 khem: your parrallel install of eglibc fails for me Sep 24 19:47:23 it makes form "make PARALLELMFLAGS=-j 4" this: "make -r PARALLELMFLAGS="-j"" Sep 24 19:47:42 and then complains: make[1]: *** No rule to make target `4'. Stop. Sep 24 19:59:42 I'm just trying to make a recipe that downloads a tarball, applies a patch, and builds it, but my patch isn't applying. If I apply the patch manually, it works fine, so I must be doing something incredibly stupid in my recipe Sep 24 20:00:49 recipe is here: http://pastebin.com/UmYC27G6 patch is here: http://pastebin.com/Mfc6pWCY Sep 24 20:02:51 jluxenberg: there is no dir in the patch, is that ok ? Sep 24 20:02:54 try appending ;patch=1 to your patch's SRC_URI Sep 24 20:04:57 eFfeM: have tried with and without a directory. according the docs, patches are applied with striplevel=1 by default, so the first item in the path shouldn't matter Sep 24 20:05:09 jconnolly: what does patch=1 do? Sep 24 20:05:23 jluxenberg: that corresponds to the patch -p# param you use Sep 24 20:05:33 jluxenberg: lets bitbake know that it is a patch Sep 24 20:05:46 jluxenberg: only needed in stable, dev will take care without it Sep 24 20:05:54 as long as it end in .patch or .diff Sep 24 20:06:10 jluxenberg: where did you put the patch? Sep 24 20:06:39 stefan_schmidt: I put the patch in user.collection/recipes/uvccapture. Same location as the .bb file for this recipe Sep 24 20:06:57 jluxenberg: thats the problem Sep 24 20:07:04 oh? Sep 24 20:07:09 make a dir called files in it and put it there Sep 24 20:08:15 no dice Sep 24 20:08:23 jluxenberg: bitbake searches either in a dir called files or a dir called like the package name for the file Sep 24 20:08:27 jluxenberg: dice? Sep 24 20:08:28 the build process is finding the patch file just fine, its just failing to apply it Sep 24 20:08:35 er, that didn't work Sep 24 20:09:03 is there a way to get bitbake to show me the command line its using to apply the patch? Sep 24 20:09:04 jluxenberg: so it does find it? Sep 24 20:09:18 jluxenberg: then it is indeed a problem with the striplevel Sep 24 20:09:42 * stefan_schmidt is confused that bitbake finds the patch directly in the recipe dir. Maybe I never tested it. :) Sep 24 20:10:04 the tarball unpacks into a directory uvccapture-0.5/*. is that possibly a problem? Sep 24 20:10:08 jluxenberg: is the file you want to patch in the top level dir? Sep 24 20:10:34 it's in uvccapture-0.5/v4l2uvc.g Sep 24 20:10:41 er, uvccapture-0.5/v4l2uvc.h Sep 24 20:10:54 hm, that should be ok Sep 24 20:11:06 and if I apply the patch manually it works Sep 24 20:11:22 have you checked if the patch is in the workdir ? Sep 24 20:11:30 jluxenberg: whats the patch commandline when doing it manually? Sep 24 20:11:32 and have you checked temp/do_patch* Sep 24 20:14:13 hmm, interesting, there are two directories in my tmp/work/... directory: http://pastebin.com/77Fq8EJr Sep 24 20:14:27 uvccapture-0.5 and uvccapture-0.5-1.0 Sep 24 20:14:31 could that be the problem? Sep 24 20:15:02 yes Sep 24 20:15:11 your ${S} is wrong Sep 24 20:15:32 eFfeM: how do I set it, and what do I set it to? Sep 24 20:15:41 eFfeM: 0.5 I'm guessing? Sep 24 20:15:53 S = ... Sep 24 20:15:56 see other files Sep 24 20:16:08 or, perhaps better, change the name of your recipe Sep 24 20:16:35 change the name of the recipe to what? Sep 24 20:16:49 in which dir are your files? and what is the name of your recipe Sep 24 20:17:03 Heinervdm change PARALLEL_MAKE= -j4 in local.conf Sep 24 20:17:29 recipe is uvccapture-0.5.bb Sep 24 20:17:41 files are in uvccapture-0.5/ Sep 24 20:17:48 khem: that's working Sep 24 20:18:28 alright, I changed the name of the recipe to uvccapture_0.5 and it works Sep 24 20:18:30 thanks all! Sep 24 20:19:03 yw (was about to sayt that, your pastebin has the name with the _ in it Sep 24 20:19:24 jluxenberg: ah, the beloved underscore as separator between package name and version :) Sep 24 20:20:30 Heinervdm: Can you try this platch plz http://pastebin.com/wz2PRb30 Sep 24 20:22:39 Heinervdm: I think -j4 or -j 4 both should be accepted Sep 24 20:23:36 khem: i think so too, i used "-j 4" for over a year till now :) Sep 24 20:23:42 will try the patch Sep 24 20:23:48 ok Sep 24 20:26:29 khem: I switched to qemuarm/minimal to check how bitbake world goes there, but it still comes not further then 13000 or so. Sep 24 20:26:40 strange Sep 24 20:26:45 khem: will start another -k world to give you a tinderbox URL, mom Sep 24 20:26:46 where do u get stuck Sep 24 20:27:24 chillispot has the wrong checksum, glibc, net6, and many others Sep 24 20:27:37 Just fetching latest HEAD and restart it to point you to tinderbox Sep 24 20:27:57 I dont know who is pulling in glibc Sep 24 20:28:01 when we have eglibc Sep 24 20:28:35 yeah, I wondered as well :) Sep 24 20:28:48 * khem fires bitbake -g Sep 24 20:28:55 There should be nothing strange in my local.conf, but let me check Sep 24 20:29:01 hey guys, which package contains the typical C library headers (stdio.h, etc?); or: is there a way I can search for a particular file and find which package installs it? (sort of like yum's what-provides) Sep 24 20:29:19 joshc_: eglibc/glibc/uclibc Sep 24 20:29:46 beware only one of them is installed on a given system Sep 24 20:30:17 I dont know if we have a package browser you might check angtrom they mihgt have Sep 24 20:30:23 khem: sure, hmm. I would have assumed then, that they would be installed...I wonder why that wasn't the case Sep 24 20:30:25 khem: nothing strange I could spot besides some ANGSTROM_BLACKLIST I used earlier, but that should not pull in glibc :) Sep 24 20:30:56 joshc_: eglibc-dev Sep 24 20:31:06 stefan_schmidt: ok Sep 24 20:31:15 so someone broke parsing world Sep 24 20:31:23 ERROR: Nothing PROVIDES 'clutter' Sep 24 20:31:23 03Frans Meulenbroeks  07org.openembedded.dev * rf860c47aa8 10openembedded.git/recipes/navit/files/navit.xml-so.patch: Sep 24 20:31:23 navit : moved unused files to obsolete dir Sep 24 20:31:23 Signed-off-by: Frans Meulenbroeks Sep 24 20:32:18 khem: unfortunately, my distro is glibc-based. I'm assuming that will installing eglibc-dev will cause some problems :( Sep 24 20:32:31 sure install glibc-dev Sep 24 20:32:39 khem: the latst try looked like this: http://tinderbox.openembedded.net/builds/87900/ Sep 24 20:32:45 firefox and friends as well Sep 24 20:34:01 khem: ok, parsing done. New stuff will pop up here: http://tinderbox.openembedded.net/builds/88040/ Sep 24 20:35:17 hm, i'm at bitbake head (that is head of three hours ago) and I still get NOTE: Out of date cache found, rebuilding... Sep 24 20:35:17 on every build (no running numbers though) Sep 24 20:35:46 hmm Sep 24 20:37:00 khem: with that patch works Sep 24 20:37:31 ah, that reminds me. khem are you building with bitbake head or 1.10? Sep 24 20:37:42 03Frans Meulenbroeks  07org.openembedded.dev * r8e38088ef1 10openembedded.git/recipes/numptyphysics/numptyphysics_svn.bb: Sep 24 20:37:43 numptyphysics_svn: changed proto=tthps to proto=http Sep 24 20:37:43 it would not fetch without that Sep 24 20:37:43 Signed-off-by: Frans Meulenbroeks Sep 24 20:37:44 03Frans Meulenbroeks  07org.openembedded.dev * r055bf46e9d 10openembedded.git/recipes/numptyphysics/numptyphysics/gcc43.diff: Sep 24 20:37:45 numptyphysics : moved unused files to obsolete dir Sep 24 20:37:45 Signed-off-by: Frans Meulenbroeks Sep 24 20:37:51 eFfeM: see if it goes away if you checkout 085e66f9c14123ea2c0f1e34f7737cf77071f86d instad of master Sep 24 20:38:02 haven't seen that, but *shrugs* Sep 24 20:39:51 will try that, but got http://www.pastebin.ca/1948356 Sep 24 20:41:16 yes, so? Sep 24 20:41:32 thats exactly what it should do, you're checking out a specific commit, not a branch Sep 24 20:41:33 tried it, did not help Sep 24 20:42:30 03Frans Meulenbroeks  07org.openembedded.dev * rd996fe1d33 10openembedded.git/recipes/netbase/netbase/busybox.patch: Sep 24 20:42:30 netbase : moved unused files to obsolete dir Sep 24 20:42:30 Signed-off-by: Frans Meulenbroeks Sep 24 20:44:03 Heinervdm: OK Sep 24 20:44:21 03Frans Meulenbroeks  07org.openembedded.dev * rd75c48bf8e 10openembedded.git/recipes/neon/files/no-func-checks.patch: Sep 24 20:44:22 neon : moved unused files to obsolete dir Sep 24 20:44:22 Signed-off-by: Frans Meulenbroeks Sep 24 20:44:22 stefan_schmidt: I build with master Sep 24 20:44:46 kergoth, it looks like it just spits out the message, but does not actually rebuild the cache, no 1234/7052 or so Sep 24 20:46:59 can someone try bitbake -g clutter Sep 24 20:47:08 and if it succeeds Sep 24 20:47:42 khem: ok, will switch to master to mimic your env Sep 24 20:48:05 03Khem Raj  07master * r750a7a6a70 10openembedded.git/recipes/eglibc/eglibc.inc: Sep 24 20:48:05 eglibc.inc: Use quotes to cover -jN and -j N Sep 24 20:48:05 Thanks Heinervdm for testing. Sep 24 20:48:05 Signed-off-by: Khem Raj Sep 24 20:49:57 khem: -g clutter is in progress Sep 24 20:50:12 stefan_schmidt: latest master of OE ? Sep 24 20:50:25 khem: master of OE and btibake Sep 24 20:50:29 ERROR: Nothing PROVIDES 'clutter' Sep 24 20:50:33 thats what I get Sep 24 20:50:49 minimal/efikamx Sep 24 20:51:38 anyone got a e-CAM32_OMAP_GSTIX ? im looking for the .patch distributed with the hardware Sep 24 20:51:53 khem: minimal/qemuarm has no problem Sep 24 20:52:30 also anyone available for hire to help me create board-overo-camera.c and patch board-overo.c? Sep 24 20:53:57 khem, should that work for minimal/armv5 ? Sep 24 20:55:20 eFfeM: its a new problem for me Sep 24 20:55:40 mang: you want to use OE on it Sep 24 20:55:46 mang: don't know how many gumstix devs are around here Sep 24 20:56:18 kergoth this time, I get an out of date cache and it indeed starts to rebuild (although I didn't really do anything special afaik) Sep 24 20:56:23 stefan_schmidt: indeed it works for qemuarm Sep 24 20:56:29 lemme try it for beagleboard Sep 24 20:56:42 khem: so just a armvt5 vs armvt5e problem? Sep 24 20:56:57 stefan_schmidt: efikamx is armv7 Sep 24 20:57:05 same as beagleboard Sep 24 20:57:07 khem, might try time permitting, but almost bedtime here Sep 24 20:57:12 khem: oh, ups :) Sep 24 20:57:16 eFfeM: no problem Sep 24 20:57:38 eFfeM: you should get that pedalling beer bar fixed Sep 24 20:57:46 :-) Sep 24 20:57:46 I heard it has gone kaput Sep 24 20:58:07 yes, they found it too wide to qualify as a bicycle Sep 24 20:58:22 bte. now I have all Linaro armv7 flashing optimization patches Sep 24 20:58:26 ported into OE Sep 24 20:59:04 yes OE.. gitorious/gumstix-oe with linux-omap-psp-2.6.32 Sep 24 20:59:59 btw no testing-next results from me this week, we had issues with our autobuilder, now we're running on an older version but just found out it does not have chrpath installed and I have no access to it, so can't install it Sep 24 21:00:14 re Sep 24 21:00:29 ah COMPATIBLE_MACHINE Sep 24 21:01:42 infact if something DEPENDS on a recipe which has COMPATIBLE_MACHINE Sep 24 21:01:57 the depender should inherit COMPATIBLE_MACHINE Sep 24 21:02:04 kergoth_: what do u think Sep 24 21:02:26 I think that's too implicit, personally Sep 24 21:02:34 it could easily lead to things happening that you don't expect Sep 24 21:02:52 we could spit out notes Sep 24 21:02:57 if someone isn't incompatible with the current machine and depends on something which is, something is broken and should be fixed Sep 24 21:03:01 so people know whats happening Sep 24 21:03:09 and you'll see an unsatisfied dependency which should tell you Sep 24 21:03:19 it was briefly discussed on the ML wrt clutter Sep 24 21:04:23 choices are either to have CM also in recipes that depend on it, or let them fail (which is nasty with bb world), or inherit Sep 24 21:04:43 no, those aren't the choices Sep 24 21:05:02 letting them fail isn't an option. if you depend on a recipe that isn't buildable for this machine, your recipe *is broken* Sep 24 21:05:07 either it shouldnt depend on it, or it should also be CM Sep 24 21:06:14 i could see adding a bb.msg.debug to base.bbclass, indicating that this recipe isn't buildable due to incompatibility with X machine, for determining *why* this thing you think should be available isnt Sep 24 21:06:33 but i'd argue against inheriting CM -- also because bitbake has no knowledge of CM Sep 24 21:06:46 CM is entirely in the metadata, it raises an exception, bitbake doesn't know *why* it was raised Sep 24 21:07:52 kergoth I think I found the cause of my "out of cache" message, apparently my cache was out of date, but I did a bitbake -b, that spits out the message but does not rebuild the cache Sep 24 21:08:52 ah, that makes sense -- it can't rebuild the cache with -b, it can only reparse that one recipe Sep 24 21:09:28 kergoth_: yeah spitting info about the cause would be great Sep 24 21:09:33 mystery solved :-) we might want to improve the error message Sep 24 21:09:46 khem: my main concern with doing that is bitbake's parsing is way, way too verbose today Sep 24 21:09:57 to the point that adding a debug message will likely be lost in the crowd when using -D Sep 24 21:10:07 eFfeM: sometimes we are dumb too :) Sep 24 21:10:08 too many bb.msg.debug's in bitbake in general Sep 24 21:10:48 kergoth: thats true as well I think we should rework on what bb outputs Sep 24 21:10:53 and curb the crap Sep 24 21:11:08 a lot of them were likely useful while *writing* the code Sep 24 21:11:12 but isn't really useful to anyone at all now Sep 24 21:12:07 he kergoth Sep 24 21:12:12 did you fix bitbake? Sep 24 21:12:32 yeah Sep 24 21:12:34 heh is their peace in the world Sep 24 21:12:35 two fixes went in Sep 24 21:12:40 should be fine now Sep 24 21:12:43 sorry about that Sep 24 21:13:02 the bugs were seen only when using a populated cache, and i was testing from scratch each time :) Sep 24 21:13:32 yes Sep 24 21:13:43 I wondered when I run it the second time Sep 24 21:15:22 bitbake needs unit tests so bad :( Sep 24 21:15:56 kergoth, most of the time OE + bitbake work really well Sep 24 21:16:04 kergoth_: I agree Sep 24 21:16:24 Crofton|work: thats because dev are awesome Sep 24 21:16:27 hehe Sep 24 21:16:31 :) Sep 24 21:17:01 I would not be able to touch gcc if it did not have testsuite Sep 24 21:17:11 kergoth hm for the fetcher is it a good idea when taring from git or svn or hg to exclude the .svn .git and .hg dirs? Sep 24 21:17:21 yes Sep 24 21:17:28 the gcc-svn.tar would be a lot smaller Sep 24 21:17:36 woglinde: yes indeed Sep 24 21:17:43 I thought so many times Sep 24 21:17:53 we should use the export option of scms Sep 24 21:18:14 khem yeah thats a good idea too Sep 24 21:18:25 svn export would be ideal Sep 24 21:18:28 for gcc case Sep 24 21:18:50 i really want to reimplement / merge fetch/unpack into a fetching module / task which *caches* what it downloads into DL_DIR -- not primary destination, just intermediate. as a result, the git one could do a local bare clone into the cache, and clone from there into WORKDIR Sep 24 21:19:09 one of these days i need to try that and see if it sucks Sep 24 21:19:46 * kergoth_ wanders off to get some caffeine Sep 24 21:20:02 exporting sounds like a step in the right direction Sep 24 21:20:25 kergoth_: I will try it for svn Sep 24 21:20:27 lets see Sep 24 21:25:03 03Frans Meulenbroeks  07org.openembedded.dev * r2b81d87d81 10openembedded.git/recipes/transmission/files/webupload.patch: Sep 24 21:25:03 transmission : moved unused files to obsolete dir Sep 24 21:25:03 Signed-off-by: Frans Meulenbroeks Sep 24 21:25:15 vi ok, calling it a day, ca & have fun! Sep 24 21:26:42 hi all Sep 24 21:27:07 virtual/egl is omap3 specific Sep 24 21:27:12 so same story Sep 24 21:27:36 does anybody knows how 2 resolve this: http://pastebin.com/kdML0QyX Sep 24 21:41:15 In the git repo where is bitbake? Sep 24 21:41:37 in git diff, what do these lines do? Sep 24 21:41:39 index 26386a9..9e825cb 100644 Sep 24 21:41:58 w|zzy: its in its own repo: http://cgit.openembedded.org/cgit.cgi/bitbake/ Sep 24 21:42:18 if its creating a new file, can i just cnp them into a patch? Sep 24 21:42:46 ahhh Sep 24 21:42:47 silly me! Sep 24 21:45:14 The getting started part is quite abiguous. later on in the document it talks about bitbake from git but there isnt a git repo in there. Sep 24 21:45:18 Not to worry. Sep 24 21:59:33 good nite Sep 24 22:14:09 03Koen Kooi  07org.openembedded.dev * r86a323f947 10openembedded.git/recipes/gimp/gimp_2.6.8.bb: gimp: bump PR for webkit change Sep 24 22:22:36 kergoth_: I just posted a patch for removing SCM dirs from tars Sep 24 22:22:53 kergoth_: only catch is it will not work with non GNU tar programs Sep 24 22:23:15 I could have used --exclude but I preferred this as it was much better Sep 24 22:24:12 is bitbake intended to be GNU tar free Sep 24 22:25:08 nag damnit! openssl native 64bit is failing to compile Sep 24 23:02:56 khem: Updating to bitbake head worked wonders. At task 18000 now. Seems -k world had some problems with 1.10 Sep 24 23:03:19 Now it keeps going like it should Sep 24 23:03:53 stefan_schmidt: good Sep 24 23:08:23 stefan_schmidt: if you get beyond 58000 that will be awsome Sep 24 23:08:42 * stefan_schmidt crosses fingers :) Sep 24 23:08:49 18362 of 63626 Sep 24 23:08:56 still some way to go Sep 24 23:09:03 burn cores burn ;) Sep 24 23:09:46 will let you know tomorrow where it stands Sep 24 23:09:48 night all Sep 24 23:09:55 gn Sep 24 23:16:15 Which mailing list is for merge requests? openembedded-devel@lists.openembedded.org? Sep 24 23:16:35 Or do I just send patches? Sep 24 23:16:43 how many patches? Sep 24 23:16:45 3 Sep 24 23:16:53 hmm maybe just send patches Sep 24 23:17:01 to the dev list Sep 24 23:17:02 To the -devel list? Sep 24 23:17:02 so anybody knows how to solve locale problem? Sep 24 23:17:05 Alrighty Sep 24 23:17:54 IgorK: does the ipk exist Sep 24 23:20:15 Do I need to subscribe to the mailing list to post? Sep 24 23:21:04 khem: "fiind . -name ipk"? Sep 24 23:24:19 khem: yes, IPK does exists Sep 24 23:24:47 i wonder then why opkg does not find it Sep 24 23:47:21 i dont know. anybody else? ;-) Sep 25 00:03:07 anyone having a problem with openssl-native building on an x86_64 system? http://tinderbox.openembedded.net/packages/openssl-native/ Sep 25 00:03:11 Seems like im not the only one Sep 25 00:03:17 But how can i tell what his problem is? Sep 25 00:21:41 khem: u still here? the path to ipk is "./openembedded/temp/deploy/glibc/ipk". Is this whats expected? Sep 25 00:30:25 and the "ipk" apparently is a directory and not a file Sep 25 00:50:49 Anyone know off the top of their head how to get ahold of the dependency graph before it gets converted to dot? Sep 25 01:41:48 i tried hmc6352 patch, i see it under /sys/bus/i2c/drivers but without a device link Sep 25 01:42:40 its very simple code http://onegrandcircle.com/~justin/hmc6352.patch, wondering how to start debugging Sep 25 02:42:39 has anyone used ruby on openembedded? Sep 25 02:45:16 I'm having trouble getting the "socket" module "required" **** ENDING LOGGING AT Sat Sep 25 02:59:57 2010