**** BEGIN LOGGING AT Wed Feb 25 02:59:57 2009 Feb 25 05:34:41 03Khem Raj  07org.openembedded.dev * rc08753e465 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:42 03Khem Raj  07org.openembedded.dev * rf6c034ed91 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:47 03Khem Raj  07org.openembedded.dev * r8f20816b25 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:47 03Khem Raj  07org.openembedded.dev * r8c2955ce84 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:48 03Khem Raj  07org.openembedded.dev * redf6af8da8 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions.inc: Use pidgin version 2.5.3 Feb 25 05:34:49 03Khem Raj  07org.openembedded.dev * r36d3e3dfc0 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:59 03Khem Raj  07org.openembedded.dev * r0773447006 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:34:59 03Khem Raj  07org.openembedded.dev * r5f635c671b 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 05:35:00 03Khem Raj  07org.openembedded.dev * r170188bce7 10openembedded.git/conf/checksums.ini: Feb 25 05:35:05 checksums: Remove spurious control characters in ++DFB-0.9.25.tar.gz url Feb 25 05:35:07 * This led to 'NOTE: Creating the CheckSum parser failed' but the build Feb 25 05:35:09 continues without checking for the checksums. Ideally it should have Feb 25 05:35:11 errored if OE_STRICT_CHECKSUMS was set. But I will leave that for Feb 25 05:35:13 other day. Feb 25 05:41:38 03Dmitry 'MAD' Artamonow  07org.openembedded.dev * r8ff9eac827 10openembedded.git/ (5 files in 3 dirs): Feb 25 05:41:38 xdiskusage: xdiskusage shows disk space usage information, it can produce a PostScript version of the display. Feb 25 05:41:38 Signed-off-by: Khem Raj Feb 25 05:44:06 * * OE Bug 4562 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Feb 25 05:44:08 * * New package - xdiskusage-1.48 Feb 25 05:44:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4562 Feb 25 05:46:06 * * OE Bug 5042 has been created by shawn.t.rutledge(AT)gmail.com Feb 25 05:46:08 * * vala doesn't cross-compile Feb 25 05:46:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5042 Feb 25 06:50:33 03Stefan Schmidt  07fso/milestone5.5 * r0d5a98248b 10openembedded.git/ (conf/checksums.ini packages/cellhunter/cellhunter_0.4.0.bb): cellhunter: Move to the 0.4.1 bugfix release Feb 25 08:14:39 good morning Feb 25 08:47:06 hi, hi, I've an error with the new fakeroot-native_1.12.1.bb: Feb 25 08:47:09 fakeroot-native-1.12.1: ftp://ftp.debian.org/debian/pool/main/f/fakeroot/fakeroot_1.12.1.tar.gz has no entry in conf/checksums.ini, not checking URI Feb 25 08:47:38 any idea how to fix that? Feb 25 08:49:00 either add checksum of fakeroot_1.12.1.tar.gz into conf/checksums.ini or disable checksum checking Feb 25 08:50:35 ynezz: thanks, giving a try Feb 25 08:53:41 ynezz: thanks, it works! how can I disable checksum checking? Feb 25 08:58:20 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * reb1b5e3f91 10openembedded.git/ (9 files in 3 dirs): linux-openmoko-devel: bump to latest andy-tracking; remove patches applied upstream Feb 25 08:58:31 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r8791e4ebac 10openembedded.git/conf/distro/minimal.conf: minimal.conf: use newer glibc and gcc Feb 25 09:01:38 03Koen Kooi  07org.openembedded.dev * r98238944c7 10openembedded.git/conf/checksums.ini: checksums: add fakeroot 1.12.1 Feb 25 09:09:54 mpoullet: look at conf/bitbake.conf Feb 25 09:09:57 http://www.marvell.com/products/embedded_processors/developer/kirkwood/sheevaplug.jsp Feb 25 09:11:32 for 99$ ? Feb 25 09:11:33 heh Feb 25 09:13:46 the price must be wrong Feb 25 09:14:13 512/512 NAND/RAM Feb 25 09:16:16 mwester: which configuration are you building with? i had no problems with fakeroot Feb 25 09:17:13 ynezz: why must be wrong? looks like NSLU2 Feb 25 09:18:23 nslu2 with 512MB RAM? Feb 25 09:19:05 no of course :-) Feb 25 09:37:05 * * OE Bug 5035 has been RESOLVED (FIXED) by Feb 25 09:37:08 * * source archive for fakeroot 1.9.6 no longer downloadable Feb 25 09:37:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5035 Feb 25 09:37:17 * * OE Bug 5039 has been RESOLVED (FIXED) by Feb 25 09:37:19 * * netbase 4.21 no longer downloadable Feb 25 09:37:22 * * http://bugs.openembedded.net/show_bug.cgi?id=5039 Feb 25 09:45:06 mickey|zzZZzz: do you plan to upgrade qemu svn rev? Feb 25 09:45:35 mickey|zzZZzz: I see you're doing a lot of commits :) Feb 25 09:45:52 didn't touch qemu since ages Feb 25 09:45:57 is there a known good srcrev? Feb 25 09:46:05 (as in... compiling and running stable) Feb 25 09:46:22 hrw|gone should probably do that, he knows more about qemu than me Feb 25 09:46:24 or RP Feb 25 09:46:28 but he's always away ;) Feb 25 09:47:42 i see new one in poky Feb 25 09:50:10 and it works with bug 4814 Feb 25 09:50:25 (not really new, but compiles qemu) Feb 25 09:50:45 if you want i send a patch to de oe-dev ml Feb 25 09:51:17 yes, that'd help. thanks Feb 25 09:52:31 ok Feb 25 09:52:38 this afternoon Feb 25 09:52:40 :) Feb 25 09:58:02 Hi all Feb 25 09:59:00 I have few question. I start using openembedded distro and bitbake Feb 25 09:59:07 to create my own filesystem for imx31 Feb 25 09:59:34 bitbake gpe-image Feb 25 09:59:44 BB_VERSION = "1.8.13" Feb 25 09:59:44 METADATA_BRANCH = "org.openembedded.dev" Feb 25 09:59:44 METADATA_REVISION = "d4ed85c5499e8eb693b7564e60d7038598659e51" Feb 25 09:59:44 TARGET_ARCH = "arm" Feb 25 09:59:44 TARGET_OS = "linux-gnueabi" Feb 25 09:59:45 MACHINE = "mx31ads" Feb 25 09:59:47 DISTRO = "chinook-compat" Feb 25 09:59:49 DISTRO_VERSION = "test-20090225" Feb 25 09:59:51 TARGET_FPU = "" Feb 25 10:00:10 Can I use that disto? Feb 25 10:00:19 or is there a better distro? Feb 25 10:03:38 panicking, don't paste lots of stuff in the channel. Use pastebin instead Feb 25 10:04:13 as for the question: you can try angstrom, it's one of distros that have quite good support in OE Feb 25 10:04:41 can any1 fix mips port ? Feb 25 10:05:14 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rd6894a7ffd 10openembedded.git/packages/vala/vala.inc: vala: touch gobject.vala.stamp before compiling; apparantly there's a timeglitch in the tarball. Closes OE Feb 25 10:05:26 angstrom distribution, cp: cannot stat `/stuff/openembedded/packages/linux/linux-2.6.25.20/defconfig': No such file or directory - is missing there Feb 25 10:06:09 chinoook-compat is only useful for the n8x0 machines Feb 25 10:06:25 try minimal or angstrom as distros Feb 25 10:06:30 tahnks Feb 25 10:07:42 hi Feb 25 10:16:40 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r898589d824 10openembedded.git/packages/gnome/ (16 files in 7 dirs): Feb 25 10:16:40 gnome-vfs 2.22.0: fix *.a and *.la packaging Feb 25 10:16:40 remove old versions Feb 25 10:20:33 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r9b1380c48b 10openembedded.git/conf/distro/include/sane-srcdates.inc: sane-srcdates.inc: remove first bunch of SRCDATES for cvs packages that are no longer existing Feb 25 10:49:59 morning Feb 25 10:50:27 methril|work: newer qemu from Poky would be nice to have in OE Feb 25 10:51:31 hrw: i get a successful compiled Feb 25 10:51:40 version (ported) at home Feb 25 10:51:57 methril|work: did binary locale generation worked with it? Feb 25 10:52:11 i've to check Feb 25 12:41:52 hey hrw Feb 25 12:42:30 re Feb 25 12:44:12 * ant_work asks for everyone's opinion about oebug 1182 (it looks like modules are not stripped!) Feb 25 12:46:04 !oebug 1182 Feb 25 12:46:05 * * Bug 1182, Status: REOPENED, Created: 2006-07-16 22:53 Feb 25 12:46:06 * * janjans321(AT)gmail.com: Kernel modules are not stripped before they are packaged Feb 25 12:46:07 * * http://bugs.openembedded.net/show_bug.cgi?id=1182 Feb 25 12:46:33 why modules has to be stripped? Feb 25 12:46:45 to save size? Feb 25 12:46:53 *can* they be stripped? Feb 25 12:46:59 did you ever tried to load stripped module? Feb 25 12:47:10 * NAiL tried and got a panic Feb 25 12:47:15 ant_work: do an experiment - strip module and use it Feb 25 12:47:49 I think on Gentoo I'm using stripped modules, have to check Feb 25 12:48:44 * NAiL is pretty sure ant_work doesn't use stripped modules Feb 25 12:56:26 NAiL: http://lkml.indiana.edu/hypermail/linux/kernel/0812.2/00094.html Feb 25 12:56:36 patch was applied Feb 25 13:13:41 /lib/modules/2.6.27.7-9-xen/vdso/vdso32-syscall.so: ELF 32-bit LSB shared object, Intel 80386, version 1 (SYSV), dynamically linked, stripped Feb 25 13:14:04 x86 opensuse 11.1 Feb 25 13:17:20 27M against 2M Feb 25 13:17:35 thats with BUG modules Feb 25 13:42:49 03Cliff Brake  07org.openembedded.dev * rcd0a56bb41 10openembedded.git/conf/checksums.ini: checksums.ini: add checksum for pengutronix canutils Feb 25 13:42:50 03Cliff Brake  07org.openembedded.dev * rab066eb3b7 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 13:42:50 03Cliff Brake  07org.openembedded.dev * r68e1ea7aed 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 13:42:51 03Cliff Brake  07org.openembedded.dev * rd512927cbf 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 13:42:54 03Leon Woestenberg  07org.openembedded.dev * r29387a6668 10openembedded.git/packages/rt-tests/rt-tests_0.21.bb: rt-tests: Moved from 0.21 to 0.28. Feb 25 13:42:57 03Cliff Brake  07org.openembedded.dev * rcdbe198009 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 13:42:58 hello, I'm trying to cross compile emacs for the openmoko distro and the om-gta02 machine: I made my own recipe for qemu(heavily patched) and now it can't load libgcc_s.so.1 during the install because it's in /oetmp_openmoko/cross/armv4t/arm-angstrom-linux-gnueabi/lib/ and the path given to qemu is /home/embedded/oetmp_openmoko/staging/armv4t-angstrom-linux-gnueabi...what will it break if I make a simlink from the place wher Feb 25 13:42:58 e reside libgcc to the path given to qemu? Feb 25 13:43:00 03Cliff Brake  07org.openembedded.dev * r8a7ae36c69 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 25 13:43:03 03Theodore A. Roth  07org.openembedded.dev * r54f42088e2 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 25 13:43:08 getting_oe.xml: Update documentation. Feb 25 13:43:10 * New section about directory structure. Feb 25 13:43:12 * Update 'Getting BitBake' section. Feb 25 13:43:14 Resubmitted against .dev instead of .documentation. Feb 25 13:43:16 Signed-off-by: Theodore A. Roth Feb 25 13:43:18 03Theodore A. Roth  07org.openembedded.dev * ra8d2a6706e 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 25 13:43:21 getting_oe.xml: Update getting openembedded section. Feb 25 13:43:25 Resubmitted against .dev branch instead of .documentation. Feb 25 13:43:27 getting_oe.xml: Update getting openembedded section. Feb 25 13:43:29 Signed-off-by: Theodore A. Roth Feb 25 13:43:31 03Theodore A. Roth  07org.openembedded.dev * r28671fcc6a 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 25 13:43:34 getting_oe.xml: Add Configuring OpenEmbedded section. Feb 25 13:43:36 This patch fills in the stub for the "Configuring OpenEmbedded" section. Feb 25 13:43:42 Signed-off-by: Theodore A. Roth Feb 25 13:56:15 or maybe I should create a new directory with all the simlinks? Feb 25 14:40:44 03Koen Kooi  07org.openembedded.dev * r25b0f22d36 10openembedded.git/packages/tasks/task-gstreamer-ti.bb: task-gstreamer-ti: add gst-plugins-bad Feb 25 14:51:19 03Koen Kooi  07org.openembedded.dev * rd6a0ddf225 10openembedded.git/packages/automake/automake-1.10/path_prog_fixes.patch: Feb 25 14:51:19 automake: rediff path_prog_fixes for 1.10 Feb 25 14:51:19 * people should really test things before making it a global default. 'sane default' implies that at least do_patch succeeds Feb 25 15:07:02 eFfeM, did you resolve your cdrtools problem? Feb 25 15:07:59 eFfeM, by the way why do you package cdrtools and not cdrkit? Feb 25 15:11:47 Gnutoo: because no one added cdrkit Feb 25 15:12:36 hrw, ok lol...but cdrtools has licences issues no? Feb 25 15:14:09 hrw, by the way I've an issue with qemu: qemu-arm only support one path in -L and so if I give him /home/embedded/oetmp_openmoko/staging/armv4t-angstrom-linux-gnueabi it doesn't find libgcc that is in /oetmp_openmoko/cross/armv4t/arm-angstrom-linux-gnueabi/lib/...what should I do? Feb 25 15:14:26 I'm trying to cross-compile emacs Feb 25 15:18:47 how does binary locales work? Feb 25 15:26:43 mmm...QEMU_OPTIONS is set to "" Feb 25 15:27:00 (i'm looking how binary locales are generated with qemu) Feb 25 15:27:02 thanks a lot Feb 25 15:30:53 QEMU_OPTIONS variable is used only to provide proper CPU Feb 25 15:31:13 so armv6 binaries are used with arm1163 cpu emulation etc Feb 25 15:36:18 yes I just found out that Feb 25 15:36:27 i'm now looking to cmd Feb 25 16:03:41 03Philip Balister  07org.openembedded.dev * r1b720c087d 10openembedded.git/conf/checksums.ini: Feb 25 16:03:41 checksums.ini : Add checksum for libnetfitler_queue-0.0.16.tar.bz2. Feb 25 16:03:41 * I know I am slow adding this. Feb 25 16:11:37 ah ok it uses local-tree...i'll look at how the local-tree was generated Feb 25 16:13:46 Gnutoo: man issues of cdrtools are fixed, but can't get it to compile Feb 25 16:13:58 Gnutoo: btw was afk hence the late reaction Feb 25 16:14:04 eFfeM, did you try changing the ?="gcc" Feb 25 16:14:10 Gnutoo: and i am unaware of cdrkit, never used it Feb 25 16:14:10 ok Feb 25 16:15:02 Gnutoo: i tried all kind of things, the bad thing is that the author apparently ran autoconf and configure etc then processed the results and threw the srces away Feb 25 16:15:44 eFfeM, lol...maybe you should try cdrkit then Feb 25 16:15:58 actually now looking at it Feb 25 16:16:15 eFfeM, if the program was GLP it would have been againt the license but I bet the program is not GPL Feb 25 16:16:35 cdrtools ? Feb 25 16:16:57 eFfeM, yes it's said in the GPL that you need to include the build scripts Feb 25 16:17:25 well he hacked the output so the original scripts are not really needed to rebuild everything Feb 25 16:18:09 cdrtools license is OSS Feb 25 16:26:37 is there a machine/distro setup I can use to build packages for a nokia n810 running the stock maemo filesystem? Feb 25 16:27:19 nokia800 + chinook-compat Feb 25 16:27:28 no warranty about 100% compatibility Feb 25 16:27:55 I'm happy to experiment. Thanks hrw :) Feb 25 16:35:56 morning Feb 25 16:39:50 g'day kergoth Feb 25 16:40:16 hey pb_ Feb 25 16:40:19 pb_: was it you that did the -sdk recipes? i can't recall Feb 25 16:46:03 um, dunno. I think I might have done an early version of them but I don't think there is any of my code left. Feb 25 16:46:18 I'm pretty sure the sdk stuff has been rewritten at least once since I last touched it. Feb 25 16:46:36 ah, okay, thanks Feb 25 16:46:54 many of *-sdk came from Poky Feb 25 16:47:57 eFfeM, yes but look at http://en.wikipedia.org/wiki/Cdrkit ... there i a license incompatibility in it Feb 25 16:48:02 man, bitbake should really hide these exception tracebacks by default, shove them in a log or something. it's disgusting to most Feb 25 16:48:21 kergoth: try bitbake-dev? Feb 25 16:49:23 that would be a nice GSoC project, improve bitbake error reporting .... Feb 25 16:50:18 the 'package foo started' 'package foo ended' messages have got to go, also. they're completely useless now that the scheduler operates on a task level Feb 25 16:51:38 even confusing :) Feb 25 16:51:57 that too Feb 25 16:52:11 seeing a package build start/stop 12 times in 20 seconds .. Feb 25 16:54:33 NOTE: exceptions.ValueError:substring not found while evaluating: Feb 25 16:54:34 ${@base_detect_branch(d)} Feb 25 16:54:34 Traceback (most recent call last): Feb 25 16:54:37 anyone seen this one? Feb 25 17:04:00 03Frans Meulenbroeks  07org.openembedded.dev * ra987ef9819 10openembedded.git/ (conf/checksums.ini packages/cdrkit/cdrkit_1.1.9.bb): cdrkit: created package; cmd line tools to create CDs/DVDs Feb 25 17:04:07 03Frans Meulenbroeks  07org.openembedded.dev * r209062cb1c 10openembedded.git/packages/cdrtools/cdrtools-native_2.01.bb: cdrtools-native: fixed LICENSE Feb 25 17:04:28 haven't seen that, but it sounds like it's blowing up in the crazy code that tries to figure out which git branch you're on. Feb 25 17:04:38 are you running this in a git checkout of the dev tree? Feb 25 17:05:05 i had multiple directories in my $BBFILES, and one of them was just /path/to/dir rather than /path/to/dir/*/*.bb Feb 25 17:05:12 when i changed it from the former to the latter, the problem went away Feb 25 17:05:16 ah, heh Feb 25 17:05:16 guess its making assumptions about BBFILES content Feb 25 17:05:21 yeah. it seems Feb 25 17:05:51 * kergoth adds to his list of OpenEmbedded issues he should report someday Feb 25 17:06:00 gnutoo read the wikipedia page, cdrkit builds without any problem Feb 25 17:06:14 eFfeM, ok wow Feb 25 17:11:36 hi Feb 25 17:13:40 heh, "tmp" should probably be either renamed or split up at some point. packages and images are build output, not "temporary", really. they're associated with this build, yes, but not tmp. maybe just call it "output". one of those relics.. Feb 25 17:14:05 anybody with iPAQ ? Feb 25 17:14:19 yes... some Feb 25 17:14:32 who ? Feb 25 17:14:48 also .. anybody with kernel 2.6 on iPAQ 36** ? Feb 25 17:14:50 kergoth: now, thats an easy change to make - just make it happen Feb 25 17:14:50 I have soem old ones e.g. Feb 25 17:15:22 flameman: 36xx iirc did not had 2.6 finished Feb 25 17:15:32 i know Feb 25 17:15:33 From what I know the 2.6 port to the older ones is still not ready for use Feb 25 17:15:33 CosmicPenguin: aye.. just one of those things those of us that have been using it for so long rarely think of.. you just get used to it.. nevermind that it doesn't always make sense.. :) Feb 25 17:15:40 as i opened a specif ewiki page Feb 25 17:16:08 hrw i am reporting exactly what is working and what is not Feb 25 17:16:26 but i need to find out people that want to collaborate Feb 25 17:16:27 flameman: and you plan on working on getting 2.6 workin on it? Feb 25 17:16:32 yes Feb 25 17:16:46 at this moment my h3660 is running 2.6.26 Feb 25 17:16:51 kergoth: I've been renaming mine for a while now - not sure why I started Feb 25 17:16:54 and it is emerging x11 Feb 25 17:17:07 flameman: contact thesing (Thomas Kunze) for some strongarm stuff - he works on collie (zaurus sl-5500) 2.6 kernel Feb 25 17:17:25 CosmicPenguin: i think itd be nice to keep things like workdirs and staging in tmp, and move build output, packaged staging files, etc into an 'output' directory or so Feb 25 17:17:46 hrw do you know his email ? Feb 25 17:17:57 flameman: check in OE ML archives Feb 25 17:18:22 could you check it out, please ? Feb 25 17:18:48 if you aren't willing to help yourself, you're not likely to get much help, flameman Feb 25 17:20:03 kergoth i am working, and i am new here, it should be not so hard to help Feb 25 17:20:08 anyway thank you Feb 25 17:20:15 searching a mailing list isn't difficult Feb 25 17:21:52 kergoth OK Feb 25 17:26:51 flameman: what's the address of your wiki page? Feb 25 17:27:09 flameman: it seems there is progress http://linuxtogo.org/gowiki/Angstrom Feb 25 17:27:28 lot of ipaqs seem supported Feb 25 17:27:42 http://www.elinux.org/Ipaq-Flameman Feb 25 17:27:54 angstrom is ... do not like it Feb 25 17:28:25 flameman: if you intend to do native compilation, OE is not for you (waste of yime) Feb 25 17:28:41 ant_work ... yes i know Feb 25 17:28:51 but spend some days and read..perhaps some bits are interesting Feb 25 17:28:55 but for the boot strap steps we need cross compile Feb 25 17:29:04 (btw I use Gentoo on my boxes) Feb 25 17:29:17 flameman: kexecboot is working at least on 2 ipaqs Feb 25 17:29:32 good :) Feb 25 17:29:34 native compilation? on ipaq 36xx? Feb 25 17:29:42 we don't have any other testers :-) Feb 25 17:29:44 yes hrw Feb 25 17:29:55 it is emerging x11 just now Feb 25 17:30:03 flameman: you are insane ;d Feb 25 17:30:10 :-] Feb 25 17:30:13 aahahaha Feb 25 17:30:19 :) Feb 25 17:30:46 03Frans Meulenbroeks  07org.openembedded.dev * r2255b27c27 10openembedded.git/packages/cdrkit/cdrkit_1.1.9.bb: cdrkit: fix install issue Feb 25 17:31:47 I can understand building on BUG (i.mx31 128MB ram, microsd as rootfs) but 36xx with 32/64MB ram... Feb 25 17:32:15 flameman: btw you can ask koen about hx4700 and h2200 (and see Angstrom front-page) Feb 25 17:32:55 yes i could Feb 25 17:33:08 i am emailing him :P Feb 25 17:33:35 i was asking to check out the email cause i was far from an x11 machine Feb 25 17:33:54 so it was a bit hard to use links2 for browsing :P Feb 25 17:34:13 * hrw -> off Feb 25 17:34:20 i am now on my notebook so i can see web pages :P Feb 25 17:35:24 flameman: this page is not updated (well, Zaurus are updated ;-) http://linuxtogo.org/gowiki/LinuxDevices Feb 25 17:42:17 ant_work thank U a lot :P Feb 25 17:44:43 is there a command or setting I can use to have bitbake remove old packages when it build a newer version? My tmp dir is over 50G and I need to free up some space... Feb 25 17:44:45 emacs compiled fine!!! thanks a lot all, let's see if it runs... Feb 25 17:45:30 INHERIT += "rm_work" Feb 25 17:45:38 from oe/openembedded/classes/rm_work.bbclass Feb 25 17:45:48 I am not exactly certain what it does though Feb 25 17:47:23 what file would that go in? local.conf? Feb 25 17:47:48 Crofton|work: it removes the work directories after do_build (everything but the temp dirs). Feb 25 17:48:00 bbl, bye Feb 25 17:48:51 what about when a package version changes, does it clean up after that? Feb 25 17:50:20 Yes and no -- it does no specific work when a version changes, but it accomplishes the general goal. Feb 25 17:51:32 rm_work deletes everything in the work/arm-angstrom-gueeabi-whatever/mypackage_1.3.2 directory except the logs in the temp directory, after the build of that package is complete. Feb 25 17:51:45 it leaves the contents behind if the build for the package fails. Feb 25 17:52:40 so that is a setting in local.conf then? Feb 25 17:53:04 So as long as you never need to refer to the contents, rm_work is fine ----- but I don't use it because I'm a dummy, and I often need to diff or examine the files generated by the configure step for a good build in order to resolve a problem with a newer broken version of the same package. Feb 25 17:53:11 scruggs: yes, local.conf Feb 25 17:54:01 anyone from #angstrom channel ? if so could please PM me ? Feb 25 17:54:07 thanks :) Feb 25 17:54:07 My tmp never gets to 50GB because I delete it regularly -- starting from scratch is often the only way to uncover long-latent problems in OE dependencies and metadata. Feb 25 17:54:43 are there any other strategies that are similar? I do diffs and such as well after a build so rm_work will work but its not optimal Feb 25 17:55:02 mabye I should just get on a schedule of nuking tmp when it gets larger than I like Feb 25 17:56:42 I encourage everyone to nuke their tmp weekly or more often. Many of the broken recipes and changes that get checked into OE would have been uncovered had the submitter built from scratch, IMO. Feb 25 17:58:06 And there are few people who cannot spare the cpu cycles to do a complete build weekly; surely most users do not have their dual-core or quad-core systems busy nonstop during the night all week long. (unless they reboot into Windows so they can play online Quake or Doom deathmatches!) Feb 25 17:58:46 I suspect you're also seeing people working against not quite up to date mainline. Feb 25 17:59:54 thanks. That will probably be the approach I take. Feb 25 18:01:08 I have a really nice build machine so builds are pretty fast, I'm just mostly trying to avoid the disk IO... too many people I know having hard drive failures lately :) Feb 25 18:03:35 * mwester noted that he can buy a brand-new 500GB Seagate hard drive for USD 65 -- making more storage than he'll ever need for OE tmp, and even the possibility of mirroring it! Feb 25 18:04:30 03Steffen Sledz  07org.openembedded.dev * r39fbd9c5af 10openembedded.git/conf/checksums.ini: checksums.ini: checksums for another e2fsprogs-1.41.4 url added Feb 25 18:07:02 heh, that reminds me Feb 25 18:07:12 i need to see about getting 2 new SATA drives soon i think Feb 25 18:07:21 getting prefail messages about one of mine from smartd/logwatch Feb 25 18:07:25 (RAID1 ftw) Feb 25 19:13:34 03Koen Kooi  07org.openembedded.dev * ra7e404e506 10openembedded.git/conf/checksums.ini: checksums.ini: add more sums Feb 25 19:35:10 okay, who's for moving nonworking/obselete/whatever out of packages/? Feb 25 19:37:46 packages/nonworking -> nonworking-packages Feb 25 19:37:48 ? Feb 25 19:37:48 <--votes "yes" Feb 25 19:38:06 then we could stop using a silly glob in BBFILES, and not have to mask them out Feb 25 19:39:24 who gets to define non-working? Feb 25 19:40:13 ? Feb 25 19:40:21 we aren't changing policy, i'm just talking about changing its location Feb 25 19:40:43 ah Feb 25 19:40:48 I see :) Feb 25 19:41:11 I completely agree Feb 25 19:59:06 Yeah, sounds right Feb 25 19:59:09 ga Feb 25 20:40:41 03Stanislav Brabec  07org.openembedded.dev * rdb29dc5fbe 10openembedded.git/packages/mc/ (5 files in 2 dirs): (log message trimmed) Feb 25 20:40:41 mc-4.6.2: Miscelaneous fixes: Feb 25 20:40:41 * Correctly build help whiles cross compiling. Feb 25 20:40:41 * Detect codeset from langinfo tables, not locale name. Feb 25 20:40:41 * Ignore modifiers with Esc + Numeral F-keys emulation. Feb 25 20:40:45 * Package locale dependent files to locale sub-packages. Feb 25 20:40:47 * Install profile. Feb 25 21:18:48 mwester: can i set the default version to pkgconfig 0.23 as well or will that make even more problems for your config? Feb 25 21:21:26 mickeyl, the new branch i notice today mickey/org.openembedded.dev substitute mickey/sane-defaults-and-updates or have different focus ? Feb 25 21:23:19 i was a bit annoyed today Feb 25 21:23:31 everything of that should land in oe.dev Feb 25 21:23:46 it's just a "leave me alone and let me work" branch Feb 25 21:23:47 ;) Feb 25 21:23:53 ok :) Feb 25 21:24:23 lets see what comes out of the .stable|testing|unstable discussion Feb 25 21:24:31 perhaps we can afford to maintain three branches these days Feb 25 21:24:39 i'm testing "minimal" distro ... so want to know ... Feb 25 21:24:39 since we have more manpower than previously Feb 25 21:24:56 yeah, better use mickey/org.oe.dev for minimal Feb 25 21:25:09 mickey/sane-defaults-and-updates can be removed Feb 25 21:25:38 ok Feb 25 21:25:53 go ahead and change pkgconfig defaults; I'll adjust if I can't get libtool and all to work correctly. Feb 25 21:26:28 mwester: ok thanks, and sorry again if i caused trouble for you Feb 25 21:26:31 I was just trying to get things stable long enough to find a good version to branch from the dev branch for the release branch for SlugOS 5. Feb 25 21:26:45 mickeyl: stable should be maintained continuously.. I'm not sure that here is enough manpower :) Feb 25 21:26:55 Hmmm Feb 25 21:27:04 well, i even think we can maintain 3 branches now Feb 25 21:27:07 stable should be really stable Feb 25 21:27:10 which is only bugfixes Feb 25 21:27:13 no updated packages IMO Feb 25 21:27:16 The icecc blacklist, is that maintained actively? Feb 25 21:27:19 but that's due to the one being in charge Feb 25 21:27:20 No trouble caused to me that shouldn't be normal on a dev branch... Feb 25 21:27:20 Or are people just making assumptions? Feb 25 21:27:43 mickeyl: then we should make new stable's frequently Feb 25 21:27:48 Jay7: agreed Feb 25 21:27:55 ~ per 3-6 month Feb 25 21:28:18 mwester: ok. i should probably have waited longer with that, but it was an ugly situation with all these packages that vanished from upstream and the respective updates requireing newer tools :/ Feb 25 21:28:19 not sure about unstable/testing.. Feb 25 21:28:41 at the end of the day, i want one branch where i can just toss stuff into Feb 25 21:28:47 if people don't think .dev is that branch Feb 25 21:28:53 then we need to dedicate another branch for that Feb 25 21:29:15 no problems with that, just a clear guideline needed Feb 25 21:29:44 this need coordinators team imho Feb 25 21:29:58 release team/core team :) Feb 25 21:30:37 i'm not sure i can stand more coordination :D Feb 25 21:30:43 too much friction in the core team Feb 25 21:30:46 *shrug* Feb 25 21:31:59 mickeyl: please don't think that the fakeroot was selected to highlight your change; it was just the most recent one where libtool did funny stuff, and it happened to be the one I finally decided to ask on the mailing list how to fix it, rather than just nailing it down with a PREFERRED_VERSION! Feb 25 21:32:19 I'm not trying to make any trouble for anyone Feb 25 21:32:36 (must run to a meeting, back later) Feb 25 21:32:39 no worries, Feb 25 21:32:40 cu Feb 25 21:32:55 BTW, is FSO MS5.1 on a branch, or a tag on the dev brnach? Feb 25 21:32:56 * kergoth kind of thinks someone should be elected the benevolent dictator to make the final call when the core team is divided Feb 25 21:32:59 * kergoth isn't volunteering, however Feb 25 21:33:03 mwester: it's on a branch Feb 25 21:33:17 hi mickeyl Feb 25 21:33:18 Ok, I'll update my env and pull that then. :) Feb 25 21:33:19 later! Feb 25 21:33:30 * mickeyl agrees with kergoth, though due to the diversity of OE i can't think of anyone who would be competent Feb 25 21:33:32 hey pb__ Feb 25 21:33:59 koen :) Feb 25 21:34:06 *cough* Feb 25 21:34:10 ;) Feb 25 21:34:17 Jay7: hah Feb 25 21:34:36 'benevolent dictator' :) Feb 25 21:36:11 mickeyl: well, i don't think the person has to know everything, as long as they only mediate. the core team would still make most decisions where viable, and this hypothetical person would just give the nod, or even do nothing unless there's a disagreement Feb 25 21:36:12 * kergoth shrugs Feb 25 21:37:23 kergoth: right, and of course that person is free to ask for advice or information from others if they don't feel competent to make a particular decision on their own. Feb 25 21:37:28 indeed Feb 25 21:37:48 it's in the nature of most big projects that no one person holds everything in their head all the time. Feb 25 21:38:20 Except for Andrew Morton Feb 25 21:38:55 well, look at linus. he delegates almost everything now Feb 25 21:39:13 heh Feb 25 21:39:16 but it's a full day job Feb 25 21:39:21 delegating Feb 25 21:39:37 right. much as I hate to use the kernel as a model of good development practice, it is indeed an example of this sort of thing working pretty well. Feb 25 21:39:43 hehe Feb 25 21:41:09 anyone know if i have access on the patchwork? i'd like to commit the patch i sent for opinions, but i don't want it to sit in the patchwork after its committed Feb 25 21:41:27 at other side is FreeBSD with core team, commiters and community Feb 25 21:41:30 make an ccount and I will give it maintainer status Feb 25 21:41:54 the current problem is we have everyone jsut commiting rrandom stuff to .dev Feb 25 21:42:01 so it is broken all the time Feb 25 21:42:19 aye Feb 25 21:42:44 i used to spend many hours every week just doing world builds repeatedly and fixing every error i saw, i couldn't imagine trying to do that now Feb 25 21:42:59 Crofton|work: the account's created, thanks Feb 25 21:43:31 I give anyone with an account maintainer status, there aren't many choices :) Feb 25 21:43:41 hehe Feb 25 21:44:03 * kergoth 's starting to get a backlog of local OpenEmbedded modifications he needs to sort through Feb 25 21:44:14 ok Feb 25 21:44:44 mickeyl: for base_get_scmbasepath: Feb 25 21:44:48 - return path_to_bbfiles[0][:path_to_bbfiles[0].rindex( "packages" )] Feb 25 21:44:48 + return [path.rfind("packages") for path in path_to_bbfiles][0] Feb 25 21:44:55 is there a better way to do that, or will that do? Feb 25 21:45:03 right now it freaks if the OpenEmbedded dir isn't the first item in BBFILES Feb 25 21:45:07 heh Feb 25 21:45:40 I recompiled everything (fresh angstrom, opie-image) and the image build but opie fails with qcop crashes, on arm920t.... I updated the git and now it fails on lack of checksum.ini for base-passwd and netbase Feb 25 21:48:50 kergoth: what's the contents of path_to_bbfiles at that point? Feb 25 21:48:55 is it already a list of strings? Feb 25 21:49:09 yeah, its a split of BBFILES Feb 25 21:49:38 its just trying to find the first entry in bbfiles with a packages/ dir. crappy way to detect the official upstream repo, but what the hell Feb 25 21:49:52 * kergoth just wants it to not die with a ValueError :P Feb 25 21:49:56 hehe, right Feb 25 21:50:00 let me think for a second Feb 25 21:50:56 (path.rfind("packages") for path in path_to_bbfiles).next() is better than making the unnecessary temporary list, at least Feb 25 21:51:07 heh Feb 25 21:51:11 yeah Feb 25 21:51:17 make it a generator expression Feb 25 21:52:14 * kergoth 's been trying to get a handle on python best practices lately Feb 25 21:52:23 good! :) Feb 25 21:52:43 lots of things have changed Feb 25 21:52:50 now if only pylint would stop yelling at me about stupid things, id' be happier Feb 25 21:52:50 many things can be expressed much more elegant nowadays Feb 25 21:52:59 "not enough methods in class foo".... bit eme Feb 25 21:53:02 with coroutines, generators, comprehensions, and the like Feb 25 21:54:59 hmmm - time to make X server 1.6 build, methinks Feb 25 22:06:53 yay - no need for drmfix.patch any more Feb 25 22:07:21 that patch has been around for years Feb 25 22:18:52 hmm Feb 25 22:19:25 hi Feb 25 22:26:11 So who is Michael "Mickey" Lauer ? he checked in base-passwd in git on the 23th without checksum.ini, and that breaks angstrom Feb 25 22:27:06 mickeyl over there :) Feb 25 22:27:33 OE_STRICT_CHECKSUMS = "" Feb 25 22:27:38 theres 'netbase' too. Feb 25 22:28:10 * BusError uses the 'gitk ' finger pointing of death :D Feb 25 22:32:24 mickeyl, tries that in my local.conf, doesn't make any difference Feb 25 22:32:55 BusError, try pulling again, someone else may have checked in the checksums by now Feb 25 22:33:14 then OE is broken Feb 25 22:33:19 it should work Feb 25 22:33:27 check for typos Feb 25 22:34:13 * BusError deletes the line and retype it Feb 25 22:34:44 you can cut'n'paste from bitbake.conf Feb 25 22:34:50 it's there with ?= "1" Feb 25 22:34:55 these files were checked in on the 23th, we're the 25th. Seemd that everyone is using the 'skip' flag Feb 25 22:35:24 there was a subtle error in checksums.ini which led bitbake to ignore checksums Feb 25 22:35:29 I fixed it last night Feb 25 22:35:43 so anything that got committed in between might have missed checksums Feb 25 22:40:04 doh - we were installing the .la files into the xserver-xorg package - thats an extra 6Mb we didn't need... :( Feb 25 22:40:22 eh Feb 25 22:47:07 03Khem Raj  07org.openembedded.dev * r2b729c7f1a 10openembedded.git/conf/checksums.ini: checksums: Add base-passwd checksum Feb 25 23:03:24 Hmmm Feb 25 23:03:30 Anyone around that uses icecc? Feb 25 23:06:49 Tartarus: I tried once Feb 25 23:07:18 I did not succeed in 1st attempt Feb 25 23:09:22 yeah, roman did some fixes that I think are all in now Feb 25 23:10:06 * * OE Bug 5043 has been created by ospite(AT)studenti.unina.it Feb 25 23:10:08 * * checksum dependent on URL? Feb 25 23:10:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5043 Feb 25 23:12:25 hehe.. bug of the day Feb 25 23:17:14 03Stanislav Brabec  07org.openembedded.dev * r50e689eec4 10openembedded.git/ (conf/checksums.ini packages/mc/mc.inc): mc: Added checksum, removed obsolete line in mc.inc. Feb 25 23:22:03 hmm, wonder what the best way to get a list of unique elements in a list, is.. probably best off using a set for that Feb 25 23:26:06 kergoth: x = [5,6,5] Feb 25 23:26:24 ? Feb 25 23:26:32 x_unq = dict(map(lambda i: (i,1),x)).keys() Feb 25 23:26:50 print x_unq Feb 25 23:26:51 ah, thats one way Feb 25 23:46:55 gpe-image task do_rootfs fails in always the same way, in installing locales : http://oomz.net/log.txt Feb 25 23:47:43 thats annoying because I'm 'this' close to get an image of that for the first time -- thst woild be a bonus as opie (the only one that used to work), now is borken Feb 25 23:49:13 "sh: rm: command not found" is not terribly helpful Feb 26 02:16:08 * kergoth ♥ devshell Feb 26 02:35:09 03Mike Westerhof  07org.openembedded.dev * ra64be907e3 10openembedded.git/ (5 files in 3 dirs): SlugOS: reinstate fakeroot 1.9.6 and set as the PREFERRED_VERSION **** ENDING LOGGING AT Thu Feb 26 02:59:57 2009