**** BEGIN LOGGING AT Tue Nov 04 02:59:57 2008 Nov 04 03:07:32 is cbrake's weekly changelog automated? I'd hope so Nov 04 08:36:02 has anyone used to test sd card read write speed Nov 04 08:36:14 i want to check the speed of read/write Nov 04 09:33:40 bonjour Nov 04 09:47:35 good morning Nov 04 09:59:44 florian: good morning Nov 04 10:22:59 03Michael 'Mickey' Lauer  07org.openembedded.dev * ra52a9c2799 10openembedded.git/packages/images/fso-illume-image.bb: fso-illume-image: new recipe; An fso-compliant bare illume image as a starting point for X-based applications Nov 04 11:37:47 morning Nov 04 11:52:15 03Michael 'Mickey' Lauer  07shared/xorg-7.4-update * ra52a9c2799 10openembedded.git/packages/images/fso-illume-image.bb: fso-illume-image: new recipe; An fso-compliant bare illume image as a starting point for X-based applications Nov 04 11:52:23 03Graeme Gregory  07shared/xorg-7.4-update * r82c5e910ca 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into shared/xorg-7.4-update Nov 04 13:03:06 Crofton: hows your guy doing? Nov 04 13:04:14 just crawled from bed Nov 04 13:04:39 I voted absentee, so I am avoiding the long lines I am reading about in the paper though Nov 04 13:04:45 heh heh Nov 04 13:04:55 you actually going to have a worthwhile turnout this time Nov 04 13:05:05 I think last election you were down to 35%? Nov 04 13:05:09 I vote in Virginia, which in a change, you lot will hear a lot about Nov 04 13:05:19 I forget Nov 04 13:05:59 but yes, election turnout and national embarrassment can be used in the same sentence :) Nov 04 13:06:36 * XorA is personally hoping of Obama, not sure if the world can take Palin Nov 04 13:07:13 * hrw -> lunch Nov 04 13:07:16 um Nov 04 13:07:25 * Crofton -> coffee Nov 04 13:19:44 ;) Nov 04 13:20:14 I am running buildbot on 8 core system and it is nice to see all cores ~100% cpu time occupied Nov 04 13:20:58 ~hail 6h time difference between .pl and .nyc.us Nov 04 13:20:59 * ibot bows down to 6h time difference between .pl and .nyc.us and chants, "I'M NOT WORTHY!!" Nov 04 13:26:17 hrw: big speedup? Nov 04 13:26:45 hrw: I have a 4 core system that I use for compiling and it works fine . I'm wondering if 8 core could speedup it a lot or not Nov 04 13:26:53 hrw: since IO is one of problems Nov 04 13:26:58 otavio: will know when next builds will be done - now lot of time is spent on fetching Nov 04 13:27:09 hrw: oh o Nov 04 13:27:21 hrw: last machine was a dual core or a quad core? Nov 04 13:27:27 hrw: previous machine Nov 04 13:34:03 hrw, you are pleased you get to eat lunch, before I have coffee Nov 04 13:34:25 and I am in .phoeniz.az.us for a few weeks Nov 04 13:34:29 all your coffee are belong to us Nov 04 13:34:38 crappy coffee at that ... Nov 04 13:34:50 * XorA has decent coffee here Nov 04 13:35:07 Apparently, in Virginia you are required to one have one voting machine per 750 voters Nov 04 13:35:13 :-D Nov 04 13:35:14 otavio: this 8core is not mine but client Nov 04 13:35:18 so, if it takes 5 minutes per voter Nov 04 13:35:30 how to does it take 750 people to vote? Nov 04 13:35:38 Answer: 62.5 hours .... Nov 04 13:35:54 stupid computerisation Nov 04 13:35:59 yeah Nov 04 13:36:04 are you going to have hanging chads this year? Nov 04 13:36:16 more like crashed voting machines Nov 04 13:36:24 sounds like a meeting in a mens public bogs :-) Nov 04 13:36:35 forensic analysis of flash cards Nov 04 13:36:41 EMP the voting machines Nov 04 13:38:00 only in certain distrcits .... Nov 04 13:39:28 Crofton: red ones? Nov 04 13:39:35 :) Nov 04 13:39:59 hrw is learning how to "fix" elections Nov 04 13:40:38 what sucks is I have to go to my paying job, so I can't spend the day obsessing Nov 04 13:43:50 :-D Nov 04 13:43:55 election party tonight then! Nov 04 13:44:12 my friends are having one, about 2000 miles from here Nov 04 13:44:30 Crofton: some sucky friends thay are :-D Nov 04 13:44:44 good point, I will relay this to them Nov 04 13:44:54 and you lot will all be in beed when things get exciting Nov 04 13:44:57 Crofton: BTW ever been to Baltimore? Nov 04 13:45:01 yeah Nov 04 13:45:07 opinions? Nov 04 13:45:10 you going? Nov 04 13:45:13 maybe Nov 04 13:45:27 not a place I think of as a tourist destination Nov 04 13:45:35 the inner harbour area is noce Nov 04 13:45:43 aquarium, submarine, a few boats Nov 04 13:54:31 "This is a big day for sure. I wore all blue with red and white striped tie for good luck. Aubrey vetoed the red shoes." Nov 04 13:55:02 blue, red? what are these wierd colours :-) Nov 04 13:55:24 I hope he gets a picture ... Nov 04 14:10:17 florian, fyi: http://community.buglabs.net/kgilmer/posts/3-OpenEmbedded-Tools-for-Eclipse Nov 04 14:12:47 kgilmer: That's pretty cool indeed! Nov 04 14:13:38 thx! Nov 04 14:14:13 Eclipse is quite common at companies... Nov 04 14:14:56 Doing the same with Anjuta should be quite easy... iirc OH has some plugin for Poky already. Nov 04 14:19:31 yeah, i've played with it somewhat. Nov 04 14:19:36 it's offers tighter integration Nov 04 14:19:57 CDT is a little freaky afaik. we're still getting our hands around being able to customize a CDT project properly. Nov 04 14:20:22 I don't think the Anjuta plugin offers bitbake recipe editing though... Nov 04 14:21:19 Anjuta Poky plugin is for software development with Poky Nov 04 14:21:25 not for writing recipes Nov 04 15:01:34 hey all Nov 04 15:02:09 moo Nov 04 15:07:18 yo zecke Nov 04 15:07:20 yo kergoth Nov 04 15:40:02 hmm Nov 04 15:43:33 indeed Nov 04 16:03:08 python 2.6's multiprocessing module is cute Nov 04 16:03:22 even though its just wrappers around pipes and shm, still cute Nov 04 16:03:37 meh Nov 04 16:04:48 hi Nov 04 16:12:56 I'm having some problem building glibc 2.7-r2 for mipsel (angstrom-2008.1) Nov 04 16:14:02 I know what the problem is, a patch that needs to be removed from binutils-2.18-r3 (patch 77 from the binutils patchkit) Nov 04 16:14:28 but i cant find the patch in OE Nov 04 16:14:53 does oe make use of patchkit? Nov 04 16:16:20 apparently not Nov 04 16:16:27 which patchkit? Nov 04 16:23:32 zecke: sorry the patch kit was from gentoo, thought it was from gnu.org Nov 04 16:25:39 trying to go over from gentoo to OE angstrom Nov 04 16:29:12 03Andreas Oberritter  07org.openembedded.dreambox * rb7befdc6d5 10openembedded.git/packages/enigma2/enigma2.bb: enigma2: fetch source from svn Nov 04 16:29:13 03ghost  07org.openembedded.dreambox * r4732495c21 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git.opendreambox.org:git/opendreambox-1.5 into org.openembedded.dreambox Nov 04 16:29:13 03ghost  07org.openembedded.dreambox * r93eba2ba8d 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.opendreambox.org/git/fraxinas/opendreambox-1.5 into org.openembedded.dreambox Nov 04 16:29:15 03Andreas Oberritter  07org.openembedded.dreambox * ra3066324e0 10openembedded.git/: Nov 04 16:29:17 Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Nov 04 16:29:19 Conflicts: Nov 04 16:29:21 packages/enigma2/enigma2.bb Nov 04 16:29:23 03ghost  07org.openembedded.dreambox * r52bc5c4b9b 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git.opendreambox.org:git/opendreambox-1.5 into org.openembedded.dreambox Nov 04 16:29:26 03ghost  07org.openembedded.dreambox * rae7fbcd75e 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git.opendreambox.org:git/opendreambox-1.5 into org.openembedded.dreambox Nov 04 16:29:31 03ghost  07org.openembedded.dreambox * rd7ff3b24ab 10openembedded.git/packages/enigma2/enigma2.bb: enigma2.bb: add gst-plugin-neonhttpsrc to RDEPENDS Nov 04 16:29:34 03ghost  07org.openembedded.dreambox * r46e46fa4e6 10openembedded.git/packages/gstreamer/gst-plugins-good_0.10.11.bb: gst-plugins-good-0.10.11.bb: incremet revision because of new flac version Nov 04 16:29:38 03ghost  07org.openembedded.dreambox * r2bec29b77a 10openembedded.git/packages/flac/ (4 files in 2 dirs): update to flac 1.1.4 (needed for new gst-plugins-good) Nov 04 16:30:10 hmm, should make the notifier just exclude the merge commits. we already see the commits the merge pulled in, its unlikely we need to see the merge commit here Nov 04 16:36:22 yeah, indeed. they're already being excluded from the weekly summary for that reason. Nov 04 17:12:12 hi Nov 04 17:13:00 anyone have any idea why building x11-gpe-image gives me a rootfs where su isnt suid root? Nov 04 17:13:24 it used to do it occasionally but now does it all the time Nov 04 17:34:24 03Graeme Gregory  07shared/xorg-7.4-update * r6cdd131127 10openembedded.git/packages/xorg-driver/xorg-driver-common.inc: Nov 04 17:34:24 xorg-driver-common.inc : fix the header file detections now we are using Nov 04 17:34:24 sysroot. Seems to actually work in use, but maybe there is a better way! Nov 04 17:34:26 03Graeme Gregory  07shared/xorg-7.4-update * rb64d3974fb 10openembedded.git/packages/xorg-driver/xf86-input-joystick_1.3.2.bb: xf86-input-joystick_1.3.2.bb : new version for X11R7.4 Nov 04 17:34:27 03Graeme Gregory  07shared/xorg-7.4-update * r6065431961 10openembedded.git/packages/xorg-driver/xf86-video-glint_1.2.1.bb: xf86-video-glint_1.2.1.bb : new version for X11R7.4 Nov 04 17:34:34 03Graeme Gregory  07shared/xorg-7.4-update * r6aa1679946 10openembedded.git/packages/xorg-driver/xf86-video-geode_2.10.1.bb: xf86-video-geode_2.10.1.bb : new version for X11R7.4 Nov 04 17:34:36 03Graeme Gregory  07shared/xorg-7.4-update * r5e22be1f2b 10openembedded.git/packages/xorg-driver/xf86-input-synaptics_0.15.0.bb: xf86-input-synaptics_0.15.0.bb : new version for X11R7.4 Nov 04 17:34:47 03Graeme Gregory  07shared/xorg-7.4-update * r1d660ef558 10openembedded.git/conf/checksums.ini: checksums.ini : more new checksums for X11R7.4 Nov 04 17:34:47 03Graeme Gregory  07shared/xorg-7.4-update * r6f19fd3b15 10openembedded.git/packages/xorg-driver/xf86-video-i128_1.3.1.bb: xf86-video-i128_1.3.1.bb : new version for X11R7.4 Nov 04 18:02:57 anyone know whats up with bitdoc? Nov 04 18:16:56 is berlios acting up for anyone else? Nov 04 18:17:03 Hi, all. I have trouble when building gcc-cross_4.2.4. configure: error: GMP 4.1 and MPFR 2.2.1 or newer versions required by fortran. Can anybody help? Nov 04 18:34:10 re Nov 04 18:35:32 hey hrw, zecke Nov 04 18:38:03 ~seen rschuster Nov 04 18:38:07 rschuster was last seen on IRC in channel #oe, 19h 56m 39s ago, saying: 'kgilmer: this reminds me, that we (= OE) should ask for a developer room and not only a booth'. Nov 04 18:44:28 03Koen Kooi  07org.openembedded.dev * r44a09d14b3 10openembedded.git/packages/sysvinit/sysvinit_2.86.bb: sysvinit: avoid u-a race condition by raising 'init' to 60 Nov 04 18:50:38 * kergoth hugs git svn Nov 04 18:56:13 git-svn roxx Nov 04 18:56:36 it just takes too much time with some repos Nov 04 18:57:26 yeah, true Nov 04 18:58:07 that reminds me that for one repo I have to do mass 'svn propset svn:mime-type text/plain" ;( Nov 04 19:08:26 If i can not build gcc-cross_4.2.4, can i build any earlier gcc-cross instead? Nov 04 19:50:06 so, no more Bugzilla bot here? Nov 04 19:50:33 !test Nov 04 19:50:33 Failed Nov 04 19:50:44 hmm Nov 04 19:51:55 ~bug 4444 Nov 04 19:52:06 hm Nov 04 19:52:23 ~oebug 4444 Nov 04 19:54:45 !oebug 4444 Nov 04 19:55:12 the bot will re-join momentarily Nov 04 19:56:48 hihi Nov 04 19:59:17 * * OE Bug 4643 has been created by  Nov 04 19:59:19 * * guile-1.8.5-autobuild Nov 04 19:59:21 * * http://bugs.openembedded.net/show_bug.cgi?id=4643 Nov 04 19:59:29 * * OE Bug 4644 has been created by  Nov 04 19:59:31 * * opencv-1.0.0-autobuild Nov 04 19:59:33 * * http://bugs.openembedded.net/show_bug.cgi?id=4644 Nov 04 19:59:40 uh-oh... Nov 04 19:59:41 * * OE Bug 4645 has been created by  Nov 04 19:59:43 * * classpath-native-0.97.2-autobuild Nov 04 19:59:45 * * http://bugs.openembedded.net/show_bug.cgi?id=4645 Nov 04 19:59:53 * * OE Bug 4646 has been created by  Nov 04 19:59:55 * * qt4-x11-free-4.4.1-autobuild Nov 04 19:59:57 * * http://bugs.openembedded.net/show_bug.cgi?id=4646 Nov 04 20:00:05 * * OE Bug 4647 has been created by  Nov 04 20:00:07 * * enchant-1.3.0-autobuild Nov 04 20:00:08 oops Nov 04 20:00:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4647 Nov 04 20:00:13 * * Flood control triggered, clearing queue... Nov 04 20:00:16 :) Nov 04 20:00:17 * * [268] pending lines removed. Nov 04 20:00:25 ;) Nov 04 20:01:24 java hurts Nov 04 20:01:25 aha. cool! Nov 04 20:01:44 woglinde: hi, have you time/desire to help me with my gmp/mpfr problem? Nov 04 20:02:21 hrw speaks the truth. Nov 04 20:02:24 kalemas_ hm Nov 04 20:02:51 woglinde , please Nov 04 20:03:32 kalemas could you please pastebin the config.log again? Nov 04 20:03:44 kalemas and what is your hostsystem? Nov 04 20:03:45 http://en.pastebin.ca/1244316 Nov 04 20:03:53 http://en.pastebin.ca/1244255 Nov 04 20:03:59 mandriva 2009 Nov 04 20:04:11 64bit? Nov 04 20:04:16 no, 32 Nov 04 20:04:17 mwester: good that someone pays to get it done Nov 04 20:04:48 kalemas_ what is your hostcompilet gcc -version saying? Nov 04 20:04:55 hrm Nov 04 20:05:31 anyone look into implementing versioning for our classes? Nov 04 20:05:36 hi kergoth Nov 04 20:05:40 hey Nov 04 20:06:15 kergoth like .m4 serial number? Nov 04 20:06:47 hrw: true. I managed to lessen the pain somewhat by convincing my boss that if I had to work with Java, then Android would be a good way to test the build system -- so at least I get to work with some interesting code. Nov 04 20:07:01 gcc (GCC) 4.3.2 Nov 04 20:07:12 well, packages are occasionally very tightly coupled to the class, itd be nice to at least have an api version, if not a minor version, for the class Nov 04 20:07:37 course, i guess you could just shove that into the filename, then could keep old ones around Nov 04 20:07:40 * kergoth scratches head Nov 04 20:08:11 kalemas_ hm Nov 04 20:08:14 why did google have to use xml for the repo manifest. hrmph Nov 04 20:09:31 kalemas could you please check if mpfr is build? Nov 04 20:09:58 bitbake -b ...mpfr.bb? Nov 04 20:10:13 no on the i686-linux dir Nov 04 20:10:30 find stamps -name mpfr-native* Nov 04 20:10:31 maybe the gcc is trying to use the hostlibs Nov 04 20:10:44 ok Nov 04 20:13:09 i686-linux contains only mpfr-native-2.3.1.r1 dir Nov 04 20:13:28 kalemas no gmp? Nov 04 20:14:00 gmp-native-4.2.4 Nov 04 20:14:10 hm okay Nov 04 20:14:39 could you try an older host compiler? Nov 04 20:14:57 thats the last clue I have Nov 04 20:18:41 bye Nov 04 20:22:52 hoi zecke Nov 04 20:24:38 ERROR: log data follows (/home/wweng/tmp/work/i486-linux/apache2-2.2.3-r5/temp/log.do_install.3315) Nov 04 20:24:40 | NOTE: make DESTDIR=/home/wweng/tmp/work/i486-linux/apache2-2.2.3-r5/image install Nov 04 20:24:42 | make: *** No rule to make target `install'. Stop. Nov 04 20:24:43 | FATAL: oe_runmake failed Nov 04 20:25:02 hubar: trapped in a warp hole? Nov 04 20:25:18 zecke: I wish Nov 04 20:25:37 zecke: This thing is more troublesome! Nov 04 20:25:55 hubar: engineering principle: 1.) understand the issue 2.) look for possible solutions 3.) try them until it works Nov 04 20:26:20 I guess I don't understand the issues. Nov 04 20:26:25 hubar: I see you think pasting four lines is a possible solution, it didn't work the day before yesterday, yesterday, today Nov 04 20:26:41 hubar: and I will not touch apache before the hell freezes (I'm a cherokee fan) Nov 04 20:27:11 hehe Nov 04 20:27:12 ye Nov 04 20:27:13 s Nov 04 20:27:17 try lighthttp Nov 04 20:27:22 its smaller its faster Nov 04 20:27:37 hubar: and to your error. so make install fails Nov 04 20:27:41 the config is as hard as apaches Nov 04 20:27:44 there are altogether too many little webservers around, i have no idea what the current contenders are anymore Nov 04 20:27:59 hubar: this can be a) there is no Makefile in the directory Nov 04 20:28:03 hubar: b) the Makefile has no install... Nov 04 20:28:08 I tend to think it is a) Nov 04 20:28:39 kergoth: cherokee is beating them all, and I met alonso (main developer of cherokee) this year in spain... :) Nov 04 20:28:56 nice Nov 04 20:29:05 * kergoth adds to his list of projects to start paying attention to Nov 04 20:29:09 zecke: which directory is it looking for makefile in? Nov 04 20:29:24 * kergoth is trying to get back in the open source groove, and that means knowing a lot of project status off the top of his head, among other things.. Nov 04 20:29:46 hubar: what do you think it could be? Nov 04 20:30:28 woglinde 42 - will be enought? Nov 04 20:31:11 woglinde gcc-4.2 or needs earlier? Nov 04 20:31:23 probably /home/wweng/tmp/work/i486-linux/apache2-2.2.3-r5/httpd-2.2.3. Nov 04 20:31:40 Because there is no files at all in that directory. Nov 04 20:31:59 kalemas yes try gcc-4.2 Nov 04 20:32:17 strange. It should have been stuffed with http 2.2.3 files. Nov 04 20:33:29 woglinde is it able to try help me throught remote terminal? Nov 04 20:34:35 kergoth: http://www.alobbs.com/ Nov 04 20:34:53 kalemas where do you live? Nov 04 20:34:55 hubar: good finding, is there any else in the acpahe2-2.2.3-r5 directory (there is temp for sure) Nov 04 20:35:30 kalemas I do not want to be suid for breaking into computers Nov 04 20:35:40 zecke: sure, there is also image/ directory. Nov 04 20:36:02 hubar: okay, temp, image, install are create by OE, anything that contains source files? Nov 04 20:36:26 nope Nov 04 20:37:10 hubar: does this look right? Nov 04 20:37:27 hubar: it looks like you manually tried to cover up and error and now get bit anyway? Nov 04 20:38:06 zecke: hmm all I did was whacking dftables' patch. Nov 04 20:38:28 hubar: so where did the source go? Nov 04 20:39:16 zecke: no idea... Nov 04 20:39:29 XorA: ping Nov 04 20:39:42 hubar: bitbake -cclean apache2 and try again? Nov 04 20:39:53 zecke: pong Nov 04 20:41:00 XorA: I wonder if you successfully launched Xorg on a neo/freerunner yet (just curious) Nov 04 20:41:30 zecke: I launched Xorg 1.4.2 using fbdev, but the framebuffer was totally messed up Nov 04 20:41:54 zecke: I should finish the upreving tomorrow, then Ill try 1.5.1 and see what happens Nov 04 20:42:43 XorA: when is this new Xorg release due? are all individial parts released yet? Nov 04 20:43:07 zecke: 7.4 is out. 7.5 is in progress Nov 04 20:44:03 XorA: what do you feel about attempting to ditch kdrive? Nov 04 20:44:36 zecke: I think its the way to go, kdrive was never maintained really anyway Nov 04 20:44:49 zecke: it just seems to get randomely broken releases Nov 04 20:47:00 hmm different error msg this time. | gen_test_char > test_char.h Nov 04 20:47:01 | /bin/bash: gen_test_char: command not found Nov 04 20:47:41 hubar: how do you debug this one? Nov 04 20:48:17 export PATH=.:$PATH Nov 04 20:50:27 hubar: did that work? Nov 04 20:51:00 sort of. Nov 04 20:54:32 Hello, I'm trying to compile gnome-utils and I get this error in the configure step: configure: error: cannot run test program while cross compiling. Any idea or tip how I could fix this? Nov 04 21:03:49 gregoiregentil fix the m4 macro in configure.in/ac Nov 04 21:04:18 gregoiregentil easiest test is do delete this macro at all Nov 04 21:07:49 woglinde: thanks for the advice. I'm trying to do it... Nov 04 21:09:54 woglinde: eh, you should only ever do that if its a sanity check. if it does anything that results in setting a variable that's used later, it should be reworked to be cached, and supplied that way. if it is in fact a sanity check, then you should probably surround it with if "xyes" != "x$cross_compiling"; then. more likely to get it accepted upstream that way, since you aren't interfering with the normal case functionality Nov 04 21:12:37 zecke: thinking about trying a proof of concept set of git repositories, one per package, with our changes as commits or topic braches, the .bb in the source tree, and android's repo tool or something to pull them all down Nov 04 21:12:38 heh Nov 04 21:14:06 kergoth: hehe, I see that all upstream projects will allow distribution branches Nov 04 21:14:31 kergoth: so you can put your garbage there, upstream will then be able to quickly take a look and cherry-pick Nov 04 21:14:54 that'd be very nice. or at the very least, have a single repo somewhere where all the distros can play together and interact, patch wise Nov 04 21:14:59 which pulls from upstream Nov 04 21:15:09 either way, git makes this sort of setup a lot more viable Nov 04 21:15:19 * kergoth isnt yet sure whether it's optimal Nov 04 21:15:37 03  07holger/openmoko-merges * rff0d06489d 10openembedded.git/packages/wvdial/wvdial_1.60.bb: wvdial: Install the peers file into the right place Nov 04 21:15:48 03Julian_chu  07holger/openmoko-merges * r9e16706439 10openembedded.git/packages/python/python-xlib_0.14.bb: Nov 04 21:15:48 python-xlib: Add recipe python-xlib Nov 04 21:15:48 Erin_Yueh provided this recipe. Nov 04 21:15:48 It is a python binding for X library. Nov 04 21:15:49 03Holger Hans Peter Freyther  07org.openembedded.dev * rc1401f7b18 10openembedded.git/ (3 files in 3 dirs): (log message trimmed) Nov 04 21:15:52 python-pygame: Blacklist smpeg and build pygame without smpeg Nov 04 21:15:54 We can not build smpeg when enterprise is on. Put it into the Nov 04 21:15:56 BBMASK to make sure it is not buildable. Nov 04 21:15:58 Use base_conditional to not add the dependency to the DEPENDS Nov 04 21:16:00 when building for the enterprise branch. Remove the movie support Nov 04 21:16:02 (the one actually requiring smpeg) and only echo it into the Setup Nov 04 21:16:19 eeek Nov 04 21:18:52 wvdial... that bitch... I spend hours on wvstreams just to find out it is using makecontext and that lacks an arm implementation Nov 04 21:19:20 :( Nov 04 21:19:33 pb_: do you think implementing createcontext, swapcontext, getcontext for ARM in glibc would get me a @gnu.org address? Nov 04 21:22:00 I have to fix a harfbuzz bug first and then I found a nice kernel issue with sd cards Nov 04 21:22:23 mount -orw fails initially and WP is checked, if you mount ro first and then remount,rw it allows you to do so Nov 04 21:22:53 * kergoth feels lame, only has two patches in the upstream kernel. need to get more active Nov 04 21:25:12 I'm about to go down the road of trying to get an angstrom build running on the asus wl-520gu, there is a config for the wl-500g, they appear to be similar hardware, so I hope it will be relatively painless Nov 04 21:27:16 kergoth: I think it is really hard to contribute to the kernel nowadays, specially if you don't have a kernel position... Nov 04 21:27:28 aye, these two were due to my work at mv Nov 04 21:29:30 Unless you want to be a kernel janitor or something Nov 04 21:29:33 and thats no fun Nov 04 21:30:16 maybe I should work on GNU/hurd :( Nov 04 21:37:53 re kalemas Nov 04 22:01:16 hm? Nov 04 22:01:22 why is it hard to contribute to the kernel? Nov 04 22:02:33 getting things into the kernel isn't hard, but lkml can be kind of harsh if it's something big and isn't perfect. i expect most of the barrier is just getting up to speed with the codebase, so you have some idea of what you're going to be working on Nov 04 22:02:51 as cp said, kernel-janitors has lots of little things that need fixing for people looking to work on something Nov 04 22:13:36 built the angstrom FS - X11 image and tried it on an OMAPzoom board. It fails right when it starts gpe-display manager. /usr/bin/gpe-dm . i built my FS using beagleboard conf. The kernel is a TI 2.6.26 kernel. Nov 04 22:14:14 any tips? Nov 04 22:18:10 without an error message we cant say anything Nov 04 22:18:25 but it seems kernel starts init starts Nov 04 22:18:32 and then X or gpe is broken Nov 04 22:25:42 woglinde:where i can get a log/ i couldnt find /var/log/messages? Nov 04 22:26:30 sundar can you ssh into it? Nov 04 22:26:38 sundar do you have serial console? Nov 04 22:28:10 i have serial support Nov 04 22:28:59 is there a standard way of developing within openembedded? At the moment I work on and compile my kernel elsewhere and then later get bitbake to fetch it from the source dir and compile it from scratch. Nov 04 22:29:44 but serial port is locked at 'starting GPE manager" & enventually i get a prompt on the display.i can use it as a terminal with a keyboard Nov 04 22:29:52 Since I'm doing the work under git i figured I can get it to fetch from the dev dir with git and then it'd just do a git-pull each time i run bitbake fetch Nov 04 22:31:02 that doesnt seem to work though, it just ends up saying "fatal: Not a valid object name 1" and I've no idea where the 1 comes from Nov 04 22:42:58 hi, i try to compile gnome-python-desktop but it complains about a missing libgatomic.la - where is that coming from? Nov 04 22:56:47 emdete ah try newer glib Nov 04 22:56:53 there was a fix for this Nov 04 22:57:01 woglinde: how? Nov 04 22:57:11 git pull Nov 04 22:57:18 bitbake -c clean glib Nov 04 22:57:23 bitbake glib Nov 04 22:57:39 i just pulled some hours ago... may that be too old? Nov 04 22:57:46 nope Nov 04 22:57:52 the fix was earlier Nov 04 22:58:03 i cleaned the whole work dir Nov 04 22:58:09 maybe om didn't get it? Nov 04 22:58:21 yes that could be Nov 04 22:58:29 crap Nov 04 22:58:37 which glib-2.0 version has om yet? Nov 04 22:59:04 2.0-2.16.4-r0 Nov 04 23:00:11 so what do you think - too old? Nov 04 23:01:19 hm I just investigate Nov 04 23:01:27 woglinde: thnx! :) Nov 04 23:02:32 hm Nov 04 23:02:40 I swear I checked something in Nov 04 23:03:07 :D Nov 04 23:04:05 * * OE Bug 4801 has been created by  Nov 04 23:04:07 * * aspell-0.60.6-autobuild Nov 04 23:04:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4801 Nov 04 23:06:47 emdete ah its pb thumb patch Nov 04 23:07:35 hm I wonder why it clashes with gnome-python-desktop Nov 04 23:09:15 hm how is it linking? Nov 04 23:13:20 woglinde: where can i see that? Nov 04 23:15:22 in the temp dir Nov 04 23:15:26 do_compile Nov 04 23:17:25 ah, okay - sry i also play with git to apply my changes on top of om's HEAD... Nov 04 23:25:29 woglinde: http://rafb.net/p/imqaz585.html Nov 04 23:26:03 ...it's in glib - doesnt glib contain that atomic lib? Nov 04 23:26:18 its a patch from pb Nov 04 23:26:28 so you can compile glib with thumb mode Nov 04 23:27:16 where can i find the patch? Nov 04 23:27:20 emdete could you look into the glib-2.0 dir Nov 04 23:27:33 shure - for what> Nov 04 23:27:35 ? Nov 04 23:27:35 and look if libatomic.la is there? Nov 04 23:27:40 shure, mom Nov 04 23:28:23 patch is /devel/arm/git/openembedded/packages/glib-2.0/files/atomic-thumb.patch Nov 04 23:28:27 ups Nov 04 23:31:23 no, no libatomic.la at all Nov 04 23:41:41 woglinde: any idea how i can get further? Nov 04 23:51:27 woglinde : is there a way i can generate logs during boot up, to understand why my gpe is broke?i have both serial port support and ssh Nov 05 00:14:05 Sundar login via ssh Nov 05 00:14:10 and start gpe manually Nov 05 00:14:18 time for bed now Nov 05 00:14:21 good nite Nov 05 02:00:00 'lo. i'm trying to build something that requires python-devel. during a 'bitbake -c rebuild pkgname' it died and gave me 'configure: error: cannot find python-devel support' Nov 05 02:00:32 so... do i find python devel and add it to oe/org.openembedded.org/packages?? Nov 05 02:00:37 or in my custom/packages? Nov 05 02:09:33 dbtid: try changing your package to depend on python-dev (OE uses the "-dev" suffix AFAIK) Nov 05 02:10:54 um, i'm a complete n00b w/ OE. Nov 05 02:11:02 do i do that with PACKAGES += ... Nov 05 02:11:06 or Nov 05 02:11:10 i'll look in all my .bb files Nov 05 02:11:11 1 sec Nov 05 02:11:43 DEPENDS Nov 05 02:12:22 ok. i will add DEPENDS += "python-dev" Nov 05 02:12:26 what is RDEPENDS? Nov 05 02:12:30 i see that in some of these Nov 05 02:12:49 run-time dependency (as opposed to a build-time dependency) Nov 05 02:13:13 thanks Nov 05 02:13:31 if DEPENDS is an empty string, will DEPENDS += "string" Nov 05 02:13:35 work properly? Nov 05 02:13:47 Yes. Nov 05 02:13:49 thanks Nov 05 02:13:50 ok Nov 05 02:14:00 it failed Nov 05 02:14:12 ERROR: Required build target 'name' has no buildable providers. Nov 05 02:14:30 Missin or unbuildable dependency chain was: ['name', 'python-dev'] Nov 05 02:14:52 Ok. So you'll have to search about to find what you need. Nov 05 02:15:12 by search you mean... google? or in my existing .bb files? Nov 05 02:15:15 search for the correct package to include in your BBPATH Nov 05 02:15:25 existing bb files Nov 05 02:15:54 note that one bb file may create many packages; commonly -dev packages are created as a side-effect of the main package Nov 05 02:16:01 hmmm Nov 05 02:16:28 so i would look in oe/org.openembedded.dev/packages? Nov 05 02:19:42 Did you try python-devel as a dependency at all? Nov 05 02:20:13 not yet Nov 05 02:20:27 Maybe python is an exception, give that a try. Nov 05 02:20:32 ok Nov 05 02:20:46 trying Nov 05 02:20:57 (and yes, you would look in the packages dir, probably one of the recipes in "python" Nov 05 02:20:59 ) Nov 05 02:21:19 oy Nov 05 02:21:20 one sec Nov 05 02:21:44 ok Nov 05 02:22:06 python/python-pycairo_1.4.0.bb: SECTION = "python-devel" Nov 05 02:22:09 no DEPENDS though Nov 05 02:22:41 Well, no it needs to be listed as a DEPENDS in your package. Nov 05 02:22:46 there's a task that has an RDEPENDS Nov 05 02:22:58 ok nope nothing in org.openembedded.dev/packages Nov 05 02:23:03 It would be listed differently in one of the recipes, or perhaps even created dynamically. Nov 05 02:23:09 find . -name \*.bb | xargs grep 'python.*dev' Nov 05 02:23:21 didn't show anything useful Nov 05 02:29:51 openmoko has somethihng Nov 05 02:30:59 dbtid: What I'm trying to say is that you may not find that exact text anywhere. Nov 05 02:31:07 ok Nov 05 02:31:34 dbtid: Just put "DEPENDS += "python-devel" in your recipe to see if that satisfies the missing dependency that you originally had. Nov 05 02:32:10 i did; it failed. sorry, i didn't report back. Nov 05 02:32:17 Ah, ok then. Nov 05 02:33:03 Is "python-devel" the common package name as used in Debian, or does Debian use some other name for that? Nov 05 02:33:21 i'm pretty ignorant about such things Nov 05 02:33:34 let me look at synaptic Nov 05 02:34:04 I've always found that OE has a pretty close correlation to Debian when it comes to package naming. Nov 05 02:34:11 ok Nov 05 02:35:11 there's a package in synaptic called python-dev Nov 05 02:41:06 mwester: there is a task called task-openmoko-python-devel that is in the images/openmoko-python-devel-image.bb Nov 05 02:41:10 might it be something like that? Nov 05 02:41:42 there's one in tasks/ too Nov 05 02:48:16 No, that wouldn't be it. Nov 05 02:48:40 i was wondering if it was some kind of product, along the lines of what you said how it might not be obvious Nov 05 02:49:05 I'd take a wild guess then that it's built as a side-effect of python, and therefore "DEPENDS += "python"" would do the job --- even if it sounds a bit overkill. Nov 05 02:49:16 ok i will try it Nov 05 02:51:06 heh, now it's going after tcl_8.4.11. Nov 05 02:51:08 not sure why Nov 05 02:51:14 but it wasn't doing that before :) **** ENDING LOGGING AT Wed Nov 05 02:59:57 2008