**** BEGIN LOGGING AT Tue Oct 07 02:59:57 2008 Oct 07 04:07:47 03  07org.openembedded.dev * r2bb32b1919 10org.openembedded.dev.git/conf/distro/include/slugos.inc: Oct 07 04:07:47 slugos.inc: set PREFERRED_VERSION for pango and glib-2.0 Oct 07 04:07:47 (the default pango does not like the older glib) Oct 07 06:34:20 Hi there. Waht is the right strategy for selecting usb-bluetooth adapter for using with openembedded-powered device? Oct 07 06:34:49 it's "gumstix" really, relatively famouse ARM platform Oct 07 06:59:59 There are Chinese people Oct 07 07:00:10 There are Chinese people? Oct 07 07:01:04 There are Chinese people? Oct 07 07:01:39 There are Chinese people? Oct 07 07:32:22 libwww-perl ... Are errors about not matching architecture ok? http://jama.mk.cvut.cz/Others/libwww-perl.do_configure.log Oct 07 07:35:17 03  07org.openembedded.dev * r6a7fb4706e 10org.openembedded.dev.git/contrib/angstrom/sort.sh: angstrom feed sorted: add mpc831{3,5}e-rdb, try harder to create directories Oct 07 07:35:28 03  07org.openembedded.dev * r3058a9326e 10org.openembedded.dev.git/packages/xorg-driver/xf86-video-omapfb_git.bb: xf86-video-omapfb git: bump SRCREV for XV driver Oct 07 07:50:45 bonjour Oct 07 08:02:06 hi Oct 07 08:03:08 is there a different between the git and the monotone repository? Oct 07 08:10:16 julien: diff it? Oct 07 08:13:55 yo zecke Oct 07 08:17:24 broonie: Oct 07 08:17:29 sorry :) Oct 07 08:17:40 zecke: i am not going to fetch both repositories, then doing a diff on all files just to get to know which of both repos is more accurate. so i think it must have been defined which of both repository the primary is, right? Oct 07 08:18:17 julien: sure, and it has been on the ml a lot of times :) Oct 07 08:25:39 zecke : hi ! Oct 07 08:26:16 zecke : is it possible to access a variable foo defined in one .bb (in my case an image.bb) from another .bb ( a task.bb) ? Oct 07 08:26:36 steliosk: not in a sane way Oct 07 08:27:21 zecke : so i need to have something declared "globally" i.e distro or machine configs ? Oct 07 08:27:42 steliosk: yes, you do. Oct 07 08:29:02 zecke : ok thanks Oct 07 08:33:26 morning Oct 07 08:37:42 03  07org.openembedded.dev * rf8a4d30d84 10org.openembedded.dev.git/packages/xorg-driver/xf86-video-omapfb_git.bb: xf86-video-omapfb: bump PR Oct 07 08:44:54 03  07org.openembedded.stable * r28e3135df8 10org.openembedded.dev.git/contrib/angstrom/build-feeds.sh: Add tested Opie applications to stable feed Oct 07 08:44:55 03  07org.openembedded.stable * r18e1534d56 10org.openembedded.dev.git/packages/qpe-nmap/ (.mtn2git_empty qpe-nmap_0.02.bb): Oct 07 08:44:55 Add qpe-nmap package Oct 07 08:44:55 Previously added to .dev branch in rev 47c8cc979dde8a4e69611fa492570b57a97bcd1e Oct 07 08:44:57 03  07org.openembedded.stable * r7f7da03daf 10org.openembedded.dev.git/contrib/angstrom/build-feeds.sh: build-feeds.sh: add qpe-nmap and opie-rdesktop Oct 07 08:45:00 03  07org.openembedded.stable * r793e2bc7fb 10org.openembedded.dev.git/: Oct 07 08:45:02 merge of '4ac47db7fc2d351953133bf60c4f1af51000942b' Oct 07 08:45:04 and 'd81d8ce4431b8e050518b3a6ddf8c6b828365238' Oct 07 08:45:08 03  07org.openembedded.stable * rbc340d206b 10org.openembedded.dev.git/contrib/angstrom/build-feeds.sh: build-feeds.sh: add opie-jumpx, opie-kjumpx, opie-networkapplet, opie-notesapplet, opie-lockapplet Oct 07 08:45:12 03  07org.openembedded.stable * r96a8bb00db 10org.openembedded.dev.git/packages/ (16 files in 9 dirs): Oct 07 08:45:17 Fix qcop paths in postinst/postrm for multiple Opie/QPE packages Oct 07 08:45:19 Applied to .dev in: 231120beb086196600461c605e8af972d11ff350, Oct 07 08:45:21 f55b0eb745c295a90ea8e6e188f87c0493386d01 Oct 07 08:45:25 03  07org.openembedded.stable * rc5285a34f3 10org.openembedded.dev.git/packages/libopie/ (libopie2/c7x0_w100_disable.patch libopie2_1.2.3.bb): Oct 07 08:45:28 libopie2: disable W100 driver for c7x0 (addresses oebug #4191) Oct 07 08:45:30 * Add a patch to disable the w100 driver for c7x0 which is currently Oct 07 08:45:32 disabled in qte. Fixes Opie not starting on c7x0 until the w100 Oct 07 08:45:34 driver is fixed and reenabled. Oct 07 08:45:36 Applied to .dev in: b4fa1d211a61db498fb41875679b1cc49e399545 Oct 07 08:54:57 libwww-perl and liburi-perl build problem is maybe related to change of staging by base package arch.. because this change helped me http://jama.mk.cvut.cz/Others/libwww-perl_target_sys.change Oct 07 08:56:23 I don't know how to fix it properly, but perl/config.sh was looked for in staging/arm-angstrom-linux-gnueabi instead of staging/armv4t-angstrom-linux-gnueabi where was prepared Oct 07 09:12:28 perl is broken Oct 07 09:12:32 !oebug perl Oct 07 09:12:33 * * Bug perl, Status: HTTP transfer failed :( Oct 07 09:13:28 JaMa: or we need to merge Poky changes to perl Oct 07 09:14:22 hm.. there are no changes Oct 07 09:14:32 !oebug 3403 Oct 07 09:14:32 * * Bug 3403, Status: NEW, Created: 2007-11-30 04:40 Oct 07 09:14:34 * * openembedded(AT)haerwu.biz: Perl-native depends on target architecture Oct 07 09:14:34 * * http://bugs.openembedded.net/show_bug.cgi?id=3403 Oct 07 09:14:42 JaMa: check this bug Oct 07 09:19:18 ~curse python manifest Oct 07 09:19:19 May the fleas of a thousand camels infest your most sensitive regions, python manifest ! Oct 07 09:21:46 03  07org.openembedded.dev * r73fd982b28 10org.openembedded.dev.git/classes/ (4 files): Oct 07 09:21:46 cpan* bbclass: fix for staging layout changes Oct 07 09:21:46 sdk bblcass: ditto, this time the fix is from poky Oct 07 09:21:48 03  07org.openembedded.dev * rbda2f9afcc 10org.openembedded.dev.git/packages/linux/ (linux-omap/musb-fix-endpoints.diff linux-omap_git.bb): linux-omap git: add another MUSB fix Oct 07 09:22:56 hrw: thanks Oct 07 09:31:28 thats the same problem? Oct 07 09:31:31 hi rschuster Oct 07 09:33:17 hrw: yes it seems like koen fixed it in last commit Oct 07 09:33:21 hrw: hi marcin Oct 07 09:37:48 mornin' Oct 07 09:38:53 ant_work: hi Oct 07 09:39:12 ant_work: I wrote you yesterday just before split, did you read it? Oct 07 09:39:29 ant_work: about menuconfig.. (worked here) Oct 07 09:40:47 I got it work setting xhost +local Oct 07 09:41:09 I had to change -x to -e in xterm atgument.. Oct 07 09:42:45 I finally compiled in kernel ext3..but it does not boot Oct 07 09:42:51 .-( Oct 07 09:43:06 ext2 is fine Oct 07 09:47:45 what for you need ext3 in kexecboot kernel? Oct 07 09:47:53 kernel can mount ext3 as ext2.... Oct 07 09:48:13 hi hrw Oct 07 09:48:32 some developer (cortez) asked for this in order to boot Android Oct 07 09:49:36 does kexecboot kernel do something other then 'boot, scan, let user choose, kexec' then? Oct 07 09:50:07 nope Oct 07 09:50:23 so what for ext3 if ext2 works? Oct 07 09:51:30 dunno the problem with Android...he wanted to boot it instead of chroot'ing Oct 07 09:54:09 btw in gumstix they even do CONFIG_EXT3_FS_XATTR=y Oct 07 10:00:47 you can't always boot ext3 as ext2. it will only work if the filesystem is clean (nothing in the journal) and you aren't using any of the ext3 "advanced features". Oct 07 10:01:18 depending on that for booting is a bad idea, because a single unclean shutdown could leave you unable to boot ever again. Oct 07 10:13:00 03  07org.openembedded.dev * r5ccd7695d5 10org.openembedded.dev.git/: Oct 07 10:13:00 merge of 'd94f0c673a273066ff54e6843a2b270742cbb1dd' Oct 07 10:13:00 and 'e2ebba6df8e583819b643aeb9f262e9f7803837d' Oct 07 10:13:07 03  07org.openembedded.dev * rb57e3557b0 10org.openembedded.dev.git/classes/java.bbclass: java.bbclass: Fixed oe_makeclasspath classpath generation. Oct 07 10:22:37 so it looks like kexecboot needs to use modules. Oct 07 10:22:43 ~curse sharp bootloader Oct 07 10:22:44 May the fleas of a thousand camels infest your most sensitive regions, sharp bootloader ! Oct 07 10:50:20 Anybody know how I compile for ARM ABI instead of SYSV? Oct 07 10:50:35 ??????? Oct 07 10:50:47 you can compile for arm oabi or eabi Oct 07 10:50:53 re Oct 07 10:51:07 sysv is for how the machine starts and other unix stuff Oct 07 10:51:33 woglinde: Not that well informed in these things =/ Oct 07 10:51:37 normaly the machine knows if it is eabi or abi Oct 07 10:51:56 or do you want to build obai for eabi machone? Oct 07 10:52:13 woglinde: But I have binaries that say ARM, version 1, from file command Oct 07 10:52:34 arm version 1? Oct 07 10:52:39 woglinde: and when I try to run a binary that says ARM, version 1(SYSV) it fails Oct 07 10:53:04 woglinde: It's eabi i guess Oct 07 10:53:39 hm Oct 07 10:54:10 hm seems this have nothing to do with eabi or oabi Oct 07 10:54:19 woglinde: As I said I'm a noob when it comes to these things =( Oct 07 10:55:25 hm I never saw arm version 1 or 2 before too Oct 07 10:56:21 woglinde: Whats the difference that "file" reports with (SYSV)? Oct 07 10:58:24 hm seems there soemthing really strange arm version 1 is stoneage Oct 07 10:58:54 two3four do you have a arm binary which was compiled with oe? Oct 07 10:59:04 what is this doing Oct 07 10:59:13 The version there refers to the ABI ratherthan the CPU variant IIRC. Oct 07 10:59:32 broonie hm Oct 07 10:59:49 No, I have a system on the computer already, trying to test if I can compile something that runs on it in OE Oct 07 11:04:43 what device it is? Oct 07 11:04:54 hi XorHL Oct 07 11:05:38 hi hrw Oct 07 11:06:02 hrw: A CC Systems CC Pilot XS Oct 07 11:09:47 two3four: what ld-linux.so.* files do you have in /lib? Oct 07 11:10:17 if you have ld-linux.so.2, that's oabi. if you have ld-linux.so.3, that's eabi. Oct 07 11:10:25 if you have both, that's the jackpot. Oct 07 11:10:59 pb__: So oabi then =/ Oct 07 11:12:44 pb__: hehe Oct 07 11:13:08 pb__: So how do I use OE to compile binaries that run on it? Oct 07 11:28:44 hrw: why would kexecboot need modules? The kernel size seems ok Oct 07 11:28:59 compiling all in kernel Oct 07 11:29:13 ant mwester want to ask you about kexece Oct 07 11:29:45 ah, ok Oct 07 11:30:17 he needs a menu to choose from varous kernel and rootfs Oct 07 11:30:29 *g* isnt it what you are doing? Oct 07 11:30:47 well .... yes Oct 07 11:31:23 if there are no size limits one could just ude initramfs-bootmenu-image (uclibc, using ncurses and busybox) Oct 07 11:31:37 otherwise, kexecboot (klibc) Oct 07 11:32:09 two3four: that's what an arm-linux configuration will get you. Oct 07 11:32:33 yo zecke Oct 07 11:32:44 03  07org.openembedded.dev * r5f496b0365 10org.openembedded.dev.git/packages/tasks/task-openmoko-feed.bb: task-openmoko-feed: Remove gpsdrive and diversity due to failure to build Oct 07 11:32:46 morning Oct 07 11:33:37 hi pb btw. Oct 07 11:34:09 hi woglinde, pb__ btw Oct 07 11:34:23 ant *g* Oct 07 11:34:43 pb__: Thanks, will try it out Oct 07 11:34:50 methril: +inherit 'hi' Oct 07 11:35:06 bitbake hi Oct 07 11:35:12 *_* Oct 07 11:40:30 hi woglinde, ant_work Oct 07 11:50:14 hi diego Oct 07 11:58:18 hm Oct 07 11:58:24 readahead Oct 07 11:58:44 thats what we probably get into for faster booting Oct 07 12:15:34 03  07org.openembedded.dev * ra11ef3e654 10org.openembedded.dev.git/packages/u-boot/ (4 files in 2 dirs): u-boot: update omap3 stuff Oct 07 12:21:59 multimachine.bbclass is gone from .dev? Oct 07 12:22:11 does anything replace it or was it completely removed? Oct 07 12:22:24 ???? Oct 07 12:22:29 just build Oct 07 12:22:39 angstroem will tookmcare of it Oct 07 12:23:12 I had INHERIT += "multimachine" Oct 07 12:23:15 in my site.conf Oct 07 12:33:17 jo raster Oct 07 12:37:27 jow Oct 07 13:38:21 !oebug 4598 Oct 07 13:38:22 * * Bug 4598, Status: NEW, Created: 2008-10-02 21:56 Oct 07 13:38:23 * * kalev(AT)smartlink.ee: New recipe: splashutils Oct 07 13:38:24 * * http://bugs.openembedded.net/show_bug.cgi?id=4598 Oct 07 13:38:26 !oebug 4599 Oct 07 13:38:27 * * Bug 4599, Status: NEW, Created: 2008-10-02 22:23 Oct 07 13:38:27 * * kalev(AT)smartlink.ee: New recipe: pmount 0.9.17 Oct 07 13:38:30 * * http://bugs.openembedded.net/show_bug.cgi?id=4599 Oct 07 13:38:30 !oebug 4600 Oct 07 13:38:32 * * Bug 4600, Status: NEW, Created: 2008-10-02 22:41 Oct 07 13:38:34 * * kalev(AT)smartlink.ee: New recipe: ivman Oct 07 13:38:36 * * http://bugs.openembedded.net/show_bug.cgi?id=4600 Oct 07 13:38:39 can someone please commit those recipes? Oct 07 13:39:23 kalev when I am at home this evening Oct 07 13:39:49 kalev okay? Oct 07 13:40:04 hmm thats odd.. uclibc-initial does not select the config correctly anymore Oct 07 13:40:17 jineld hm Oct 07 13:40:25 but i have to got Oct 07 13:40:29 till later Oct 07 14:05:52 ok guys.. how to handle this DEFAULT_PREFERENCE stuff? I want to use uclibc-svn, so I set PREFERRED_VERSION_uclibc = "0.9.29-svn" (same for _initial) in my local.conf, I also tried using := but it does not help Oct 07 14:06:04 so I guess I need to adjust the DEFAULT_PREFERENCE but that would mean touching the actual .bb recipe Oct 07 14:06:21 what's the clean way to select uclibc-svn from the local configuration? Oct 07 14:10:56 is anyone building wxwidgets Oct 07 14:24:55 03  07org.openembedded.dev * rc900036a25 10org.openembedded.dev.git/ (3 files in 3 dirs): (log message trimmed) Oct 07 14:24:55 mpc8313e-rdb: Merge Freescale's patches for u-boot 1.2.0 and linux 2.6.24.3. Oct 07 14:24:55 Also make the u-boot version set conditionally rather than unconditionally, Oct 07 14:24:55 in the machine configuration file. Oct 07 14:24:55 Although u-boot 1.2.0 and 2.6.24.3 are reasonably outdated, the PCI Express Oct 07 14:24:59 support has never been merged upstream, neither to u-boot nor to linux, so Oct 07 14:25:01 use the above versions if you want to toy with PCI Express. Oct 07 14:25:06 03  07org.openembedded.dev * r34292021a1 10org.openembedded.dev.git/packages/initscripts/initscripts-1.0/volatiles: volatiles: create/verify /var/log/lastlog (silences dropbear's complaining) Oct 07 14:25:06 03  07org.openembedded.dev * r9dd822d248 10org.openembedded.dev.git/: Oct 07 14:25:10 merge of '7d34cad8c1cc77f4e891350dd5960331ff47128c' Oct 07 14:25:12 and 'f8b5444079bbbbe46a1effb15957cc79e8753dca' Oct 07 14:25:14 03  07org.openembedded.dev * rec252ef5e4 10org.openembedded.dev.git/: Oct 07 14:25:16 merge of '06439bd12faeb366d7d703a017d5c609d41cb96e' Oct 07 14:25:18 and '8334fa112d76dd1bff6086d9cbd2007755c42e05' Oct 07 14:42:47 curious ~ does anyone here use Eclispe for development? Oct 07 14:49:18 sometime Oct 07 14:49:52 but i've not yet integrate eclipse in OE in a good workflow Oct 07 14:50:04 03  07org.openembedded.dev * r8fc535a2f6 10org.openembedded.dev.git/packages/initscripts/initscripts_1.0.bb: initscripts: package contents changed with the volatiles commit, so bump PR Oct 07 14:53:20 kgilmer_, is this based on eclipse? Oct 07 14:53:22 http://labs.o-hand.com/anjuta-poky-sdk-plugin/ Oct 07 14:53:43 hi Crofton|work, no that's anjuta :) Oct 07 14:54:07 ah Oct 07 14:54:14 I get confused Oct 07 14:54:30 yeah I see 'sdk' and i get confused too :) Oct 07 14:55:08 so many good ideas, so little time Oct 07 14:58:53 Crofton|work ~ http://bugblogger.com/in-the-oven-openembeddedpoky-eclipse-225/ Oct 07 15:00:30 kgilmer_, does it work with eclipse on F8/9? Oct 07 15:01:16 hmmm...not sure Crofton|work. but it (sometimes) works with AMD64 Gentoo, which means it will probably work on any linux environment :). Oct 07 15:01:39 the free java stuff, not suns Oct 07 15:01:49 ? Oct 07 15:03:05 well, eclipse runs on your host jvm. Oct 07 15:03:18 so, i guess it's as free as whatever you have installed Oct 07 15:03:31 but when we release this stuff it will be EPL or BSD Oct 07 15:07:30 kgilmer_: the screenshot looks nice :) Oct 07 15:07:57 kalev: try anjuta sdk plugin one day Oct 07 15:08:55 kalev: thx Oct 07 15:09:09 hrw: yeah anjuta is good, i prefer eclipse tho Oct 07 15:09:43 I've been thinking that I should probably switch to something more modern than vim Oct 07 15:09:53 all GUI editors seem so ... clumsy :) Oct 07 15:10:14 * mickey|tw likes kdevelop Oct 07 15:10:29 kalev: gvim then :D Oct 07 15:10:59 hehe :) Oct 07 15:11:24 hrw is certainly no peach Oct 07 15:11:39 haha Oct 07 15:11:43 i mean eclipse is no peach Oct 07 15:11:46 not hrw :) Oct 07 15:12:30 for me eclipse has one thing which makes me not touching it Oct 07 15:12:38 this reminds me... I'm new around here and I was thinking of filing a bug report... task-base in org.openembedded.dev always builds task-base-wifi and task-base-bluetooth... they aren't conditional on keywords in MACHINE_FEATURES or something of the sort Oct 07 15:13:09 took me a while to figure out what was pulling in madwifi-ng when I didn't want it ;) Oct 07 15:15:03 RP, do you have the bitkeeper exports somewhere? Oct 07 15:16:13 hrw, what's that? Oct 07 15:16:27 kgilmer_: java Oct 07 15:17:30 shoragan: http://ewi546.ewi.utwente.nl/tmp/oe-bitkeeper-cvs-export.tar.bz2 Oct 07 15:17:54 hrw, thanks Oct 07 15:18:20 404 :/ Oct 07 15:18:27 moment Oct 07 15:20:19 hrw ~ haha Oct 07 15:20:43 shoragan: try http://ewi546.ewi.utwente.nl/~hrw/oe-bitkeeper-cvs-export.tar.bz2 Oct 07 15:21:00 fair enough Oct 07 15:21:03 shoragan: in few minutes as I need to copy Oct 07 15:21:06 now Oct 07 15:21:43 kgilmer_: I used JEdit for some time few years ago and it was sluggish like hell. kdevelop/kate/kvim/gvim was running much much nicer Oct 07 15:22:11 kgilmer_: I also tried to use azureus (also java) and failed when it took 60% of ram (had 1GB total) Oct 07 15:23:23 hrw: yeah imho java on linux is a problem. it's a sad situation. i have to shuffle slotted jvms in gentoo because each release has some issues. Oct 07 15:23:57 i'm hoping things can stabilize with openjdk Oct 07 15:24:11 in addition eclipse supports many platforms, and each team has it's preferences Oct 07 15:24:23 so depending on what you are working on, linux support may be great or non-existant Oct 07 15:27:14 the main problem i'm trying to solve with recipe work in eclipse is eleminating all the directory movement. i spend too much time cd some/super/long/path/ Oct 07 15:36:27 how's openjdk support in OE? Oct 07 15:38:11 good question kalev :) Oct 07 15:38:31 afaik there are some that having it building against arm targets. Oct 07 15:39:10 but there is nothing really in OE yet as far as I know. rschuster1 may know more... Oct 07 15:39:47 at some point though kalev, we'll be looking at possibly using openjdk in our product. Oct 07 15:40:20 * rschuster1 is putting jakarta commons stuff into Jalimo repo like hell currently :) Oct 07 15:40:36 jakarta-commons stuff -> ant -> openjdk Oct 07 15:41:05 kgilmer_: I had the idea of using openjdk as the native java runtime first Oct 07 15:41:25 this means it would not have to be cross-compiled. I hoped that this would be easier Oct 07 15:41:35 that would be great rschuster1 :) Oct 07 15:41:56 unfortunately the openjdk build does not allow disabling the AWT stuff right now Oct 07 15:42:24 if that would be the case it would have much less dependencies and would be easier to build Oct 07 15:44:04 ic. anything to make the java/oe stuff simpler would be great. Oct 07 15:44:22 i did some work with jakarta-commons as well, but didn't put it in the jalimo repo Oct 07 15:44:26 the biggest pain was BSF Oct 07 15:44:49 lemme know if you want me to send you my work rschuster1 Oct 07 15:45:00 kgilmer_: yeah please send me your bsf work Oct 07 15:45:44 kgilmer_: I have commons-configuration and commons-jxpath in the repo now Oct 07 15:45:49 ok. i opted to remove the netrexx support :) i'll email it to you. Oct 07 15:46:35 once upon a time a made list of packages needed for ant and those two where needed and where the most complicated Oct 07 15:47:03 kgilmer_: you will also find an example of using -sourcepath to work around circular dependencies Oct 07 15:47:19 ant is rife with those! Oct 07 15:47:53 :) Oct 07 15:48:49 kgilmer_: I have no problem with disabling particular library functionality Oct 07 15:49:16 if that is causing trouble at some point we can look at providing it later Oct 07 15:49:39 make sense Oct 07 15:49:41 at least one time I saw this being done in a debian package as well Oct 07 16:08:11 rschuster1 I've emailed you the recipes Oct 07 16:08:22 bsf depends on some other stuff that i included like rhino. Oct 07 16:08:50 kgilmer_: ok. thanks Oct 07 16:08:53 again it would be best for me to check into jalimo repo and that's my eventual goal but still not comfortable that a recipe checkin will be stable. Oct 07 16:29:26 i think i'm done with the mtn2git fixes now... Oct 07 16:29:42 the only thing left now is the user id mapping Oct 07 16:35:51 shoragan: the one in wiki I would consider complete Oct 07 16:36:14 ok Oct 07 16:38:33 just few entries are missing and I was not able to map them Oct 07 16:39:22 wiki page? Oct 07 16:41:01 http://wiki.openembedded.net/index.php/Talk:Monotone_to_Git_ID_mapping Oct 07 16:41:08 it is in channel topic too Oct 07 16:42:00 oops. Sorry, yes, it's up there. :( Oct 07 16:42:11 hi woglinde, ant_work Oct 07 16:42:17 oops Oct 07 16:49:08 03  07org.openembedded.dev * r93f18bbca0 10org.openembedded.dev.git/: Oct 07 16:49:08 merge of '8abbc6b27bdcde45d495d401b09e67def2993302' Oct 07 16:49:08 and 'f309dfd6b287a0b52a295b00b2a304b1badf5c57' Oct 07 16:49:16 03  07org.openembedded.dev * r58d51d2084 10org.openembedded.dev.git/packages/linux/ (3 files in 2 dirs): Oct 07 16:49:16 mpc8315e-rdb: Provide initial defconfig for Linux 2.6.24.3 with PCIe enabled. Oct 07 16:49:16 PCI Express is tested to work, with an Intel PRO/1000 GbE x1 PCIe card. Oct 07 16:58:19 bye Oct 07 16:59:56 gn Oct 07 18:52:25 03  07org.openembedded.dev * ra65ca965c9 10org.openembedded.dev.git/packages/xorg-driver/xf86-video-omapfb_git.bb: xf86-video-omapfb: bump SRCREV to get windowed xv working Oct 07 19:16:08 woglinde: hi Oct 07 19:17:57 he kalev Oct 07 19:18:04 show me the 2 bugs Oct 07 19:18:15 3! :) Oct 07 19:18:26 !oebug 4598 Oct 07 19:18:27 * * Bug 4598, Status: NEW, Created: 2008-10-02 21:56 Oct 07 19:18:29 * * kalev(AT)smartlink.ee: New recipe: splashutils Oct 07 19:18:29 * * http://bugs.openembedded.net/show_bug.cgi?id=4598 Oct 07 19:18:32 !oebug 4599 Oct 07 19:18:33 * * Bug 4599, Status: NEW, Created: 2008-10-02 22:23 Oct 07 19:18:34 * * kalev(AT)smartlink.ee: New recipe: pmount 0.9.17 Oct 07 19:18:36 * * http://bugs.openembedded.net/show_bug.cgi?id=4599 Oct 07 19:18:39 !oebug 4600 Oct 07 19:18:40 * * Bug 4600, Status: NEW, Created: 2008-10-02 22:41 Oct 07 19:18:41 * * kalev(AT)smartlink.ee: New recipe: ivman Oct 07 19:18:44 * * http://bugs.openembedded.net/show_bug.cgi?id=4600 Oct 07 19:20:09 kalev hm why did you rewrite do_configure for splashutils? Oct 07 19:20:54 hm and ln -s fbcondecor_helper ${D}${base_sbindir}/splash_helper Oct 07 19:20:57 will not work Oct 07 19:23:38 the stock do_configure failed to regenerate something, I think it was the libtool script Oct 07 19:23:39 kalve better write a posinst for that Oct 07 19:23:53 postinst Oct 07 19:24:05 sure, I can do that Oct 07 19:24:05 okay I will look at it Oct 07 19:25:28 pkgconfig is not need either Oct 07 19:25:33 autotools took care of it Oct 07 19:25:53 aha, ok Oct 07 19:26:13 what's wrong with creating the symlink in do_install_append? Oct 07 19:27:05 hm where is ${D}${base_sbindir} pointing when it is in the ipk? Oct 07 19:29:05 it's the same as if I had written: cd ${D}${base_sbindir}; ln -s fbcondecor_helper splash_helper Oct 07 19:29:29 nope Oct 07 19:29:42 ln is taking the whole pfad hm Oct 07 19:29:46 let me try Oct 07 19:29:53 test it, I am sure Oct 07 19:30:26 ah okay Oct 07 19:30:36 ln -s does it only the other way Oct 07 19:30:42 when dirs differ Oct 07 19:30:43 okay Oct 07 19:30:46 my fault sorry Oct 07 19:31:08 :) Oct 07 19:34:20 splashutils only works with automake 1.10 though Oct 07 19:34:35 which has DEFAULT_PREFERENCE=-1 Oct 07 19:35:08 but it should work for Angstrom, for example Oct 07 19:35:38 yes Oct 07 19:35:48 dont know why we update Oct 07 19:35:52 1.10 is stable enough Oct 07 19:36:06 maybee its about openmoko Oct 07 19:36:12 I will mickeyl ask the next time Oct 07 19:36:42 I mean, it won't work with automake 1.9.3, but 1.10 is good Oct 07 19:36:52 I always test use the latest version available in oe of autotools Oct 07 19:37:17 kalev not working because of specify the version or the used macros? Oct 07 19:37:45 because there's this line in configure.ac: AM_INIT_AUTOMAKE([1.10 -Wno-portability]) Oct 07 19:38:11 I have no idea why it is there, but I suspect there's some reason for that Oct 07 19:42:24 hm splashutils pulls in perl-native Oct 07 19:42:33 seems I have to wait more time Oct 07 19:47:26 <[X]Spot> Crofton|work did you managed to compile VLC with http support ? Oct 07 19:58:53 evening Oct 07 20:10:21 <[X]Spot> Hi guys .. I am building Angstrom for beagleboard with OE following http://www.angstrom-distribution.org/building-%C3%A5ngstr%C3%B6m Oct 07 20:10:22 <[X]Spot> but when I start buinding bitbake base-image Oct 07 20:10:30 <[X]Spot> I receive the following error Oct 07 20:10:36 <[X]Spot> ERROR: Build of /OE/org.openembedded.dev/packages/git/git-native_1.5.3.bb do_compile failed Oct 07 20:10:42 <[X]Spot> ./OE/angstrom_stable//work/i686-linux/git-native-1.5.3-r2/temp/log.do_compile.30877: Oct 07 20:10:49 <[X]Spot> SUBDIR git-gui Oct 07 20:10:58 <[X]Spot> BUILTIN git-citool Oct 07 20:11:06 <[X]Spot> ln: accessing `git-gui': No such file or directory Oct 07 20:11:12 <[X]Spot> make[1]: *** [git-citool] Error 1 Oct 07 20:11:17 <[X]Spot> do you have any idea ? Oct 07 20:11:18 ~pastebin Oct 07 20:11:18 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Oct 07 20:11:22 nope Oct 07 20:11:27 never saw this git error Oct 07 20:11:38 and please use pastebin the next time Oct 07 20:12:25 <[X]Spot> ok Oct 07 20:12:27 <[X]Spot> soory Oct 07 20:12:38 <[X]Spot> but this directory git-gui exists ?! Oct 07 20:13:58 how do I override the DEFAULT_PREFERENCE? PREFERRED_VERSION_xxx is simply ignored Oct 07 20:14:33 meaning: how do I do that from the local conf? Oct 07 20:14:48 I would not want to touch anything inside oe.dev (i.e. I would not want to alter the .bb files) Oct 07 20:14:53 any hints on how to do it correctly? Oct 07 20:15:38 Jin^eLD can be the distribution conf has set it to = insted of ?= Oct 07 20:15:53 not sure what angstrom is doing.. Oct 07 20:16:30 I thought angstrom only sets PREFERRED_VERSION stuff, and thats for sure done with ?= Oct 07 20:16:48 check it to be sure Oct 07 20:18:33 PREFERRED_VERSION_uclibc ?= "0.9.29" Oct 07 20:18:55 if I want to use the -svn version the preferred setting is ignored Oct 07 20:19:07 unless I manually modify the DEFAULT_PREFERENCE value in the .bb recipe Oct 07 20:19:21 Jin^eLD yes Oct 07 20:19:31 this how I told you it works Oct 07 20:19:35 there is now workaroung Oct 07 20:20:13 well, yes, and there is really no "clean" way around this? Oct 07 20:20:17 doh... this is not nice Oct 07 20:20:32 sorry I do not know a other way Oct 07 20:20:55 if I want to ask for a review for this - post as [RFC] or something? Oct 07 20:25:10 [X]Spot, I think it builds with http now Oct 07 20:31:43 hello everybody. Can someone please look at this bitbake output: http://pastebin.com/m73480515 ? Is this a oe-bug or bb-file misconfiguration? Oct 07 20:33:01 szsoftware: you did build the whole toolchain right? Oct 07 20:33:21 szsoftware hm extermal tollchain? Oct 07 20:33:58 sorry I can not find fbida in the oe repos Oct 07 20:34:03 until now everything built fine Oct 07 20:34:49 so you might provide us your recipe, which I guess is written by yourself Oct 07 20:35:03 the arm-angstrom-linux-gnueabi-strip Oct 07 20:35:08 is present Oct 07 20:35:52 but maybe not in the path Oct 07 20:36:13 woglinde: yes, its an edited from nonworking/fbi_1.31. It going better with fbida 2.07 except this install error. Will pastebin the recipe, wait Oct 07 20:36:17 when install arm-angstrom-linux-gnueabi-strip exiftran /home/sven/oe/stuff/tmp/work/armv6-angstrom-linux-gnueabi/fbida-2.07-r0/image/usr/bin is run Oct 07 20:38:00 this is the fbida_2.07.bb: http://pastebin.com/m39d9c901 Oct 07 20:39:38 hm then something inside the source is wrong Oct 07 20:39:44 package looks okay Oct 07 20:41:01 woglinde: I want to know what is doing install? The syntax looks weird in the error log, so if install wants to install "arm-angstrom-linux-gnueabi-strip" Oct 07 20:41:36 woglinde: on pc-host the fbida-source builds fine Oct 07 20:43:09 yes Oct 07 20:43:16 because you have strip always in the path Oct 07 20:44:00 woglinde: what to recheck, for that -strip command? maybe a bypass through bb-file options? Oct 07 20:44:55 woglinde: does not every package built need stripping? only certain, regarding the source-code? Oct 07 20:47:41 szsoftware nope I think you have to change the install line in the source package Oct 07 20:48:51 <[X]Spot> Crofton|work do you mean this from the repository ? Oct 07 20:48:54 woglinde: ahh, so patching the Makefile? Oct 07 20:49:37 or Makefile.am as you using autotools Oct 07 20:51:14 woglinde: will a unstripped binary work too? Oct 07 20:54:20 szsoftware sure Oct 07 20:54:25 it is only bigger Oct 07 20:54:55 woglinde: estimated how much bigger Oct 07 20:54:56 ? Oct 07 20:57:29 dont know Oct 07 20:57:33 but not about 50% Oct 07 20:57:33 [X]Spot, if you build your own image, it should be enabled Oct 07 20:57:38 hm args wait Oct 07 20:57:43 depends on the libc Oct 07 20:57:49 can be as twice as big too Oct 07 20:58:41 woglinde: ok, dirty workaround but will try avoiding stripping Oct 07 20:59:07 hms klibc is not building for sh4 Oct 07 20:59:15 <[X]Spot> Crofton|work do you have it built ? I still have that awfull issue with bitbake base-image Oct 07 20:59:16 <[X]Spot> :( Oct 07 20:59:27 szsoftware yes Oct 07 20:59:43 I really dont know why the strip command isnot found Oct 07 20:59:47 [X]Spot, I don't have it built Oct 07 20:59:56 the problem is the git thing? Oct 07 21:00:42 [X]Spot, do you have git installed on your machine Oct 07 21:00:57 do you have git-gui on your host? Oct 07 21:01:41 <[X]Spot> Crofton Oct 07 21:01:44 <[X]Spot> yes, I have Oct 07 21:01:48 sometimes you can work around native package issues by making sure the utility is installed Oct 07 21:01:53 <[X]Spot> [root@localhost ~]# which git-gui Oct 07 21:01:53 <[X]Spot> /usr/bin/git-gui Oct 07 21:01:58 <[X]Spot> I checkout it using git Oct 07 21:01:59 bother Oct 07 21:02:29 pastebin the enotire log file Oct 07 21:02:38 I gotta run bbl Oct 07 21:03:03 <[X]Spot> can I download this git clone git://git.openembedded.net/org.openembedded.dev.git Oct 07 21:03:05 <[X]Spot> using mtn Oct 07 21:03:13 ????? Oct 07 21:03:19 <[X]Spot> ok Crofton , bye for now ? Oct 07 21:04:35 woglinde: so, how's splashutils and friends? :) Oct 07 21:06:00 * Gnutoo has just sent the mplayer patch in their apropriate mailing-list...I hope it will be accepted at the end of the process... Oct 07 21:06:50 kalev klibc doesnot build for sh4 Oct 07 21:07:03 cannot zesz thaz for now Oct 07 21:07:06 splasutils Oct 07 21:07:11 willtry the others Oct 07 21:07:15 thanks Oct 07 21:13:11 I tested all of them on my own i586 toolchain and on Angstrom armv5te, compiled fine for both Oct 07 21:13:35 if you changed something in those recipes then we can't be so sure though Oct 07 21:14:21 kalev haha building klibc has nothing to do with the recpie Oct 07 21:14:43 yes of course :) Oct 07 21:15:11 I'm just saying that if you can't get klibc to compile for your target, then you could trust me that I tested them Oct 07 21:15:28 woglinde: ok, I've workarounded, have copied the binarys as data, so they are now in *.ipk without x-bit set. I must have a closer look at this, but thank you for your hint Oct 07 21:17:56 szsoftware hm try to use install -m 755 Oct 07 21:18:10 whithout -s Oct 07 21:18:16 which is for strip Oct 07 21:21:04 woglinde: will try to fiddle this into Makefile Oct 07 21:23:45 kalev splashutils is in Oct 07 21:25:08 thanks woglinde Oct 07 22:09:09 woglinde: thanks again, I'm off to bed now Oct 07 22:09:23 hm Oct 07 22:09:31 e2fs libs made some problem Oct 07 22:09:39 while trying to compile pmount Oct 07 22:10:11 gah :) Oct 07 22:12:04 hm yes MKINSTALLDIRS is not defined somehow Oct 07 22:59:45 jo raster Oct 07 23:03:33 woglinde: moo Oct 07 23:03:46 hms Oct 07 23:03:56 fighting with people using old gettext stuff Oct 07 23:04:16 and are to eager to fix ther autotools files Oct 07 23:06:28 haha now it works Oct 07 23:19:53 03  07org.openembedded.dev * rf07755cf80 10org.openembedded.dev.git/packages/splashutils/ (.mtn2git_empty splashutils_1.5.4.2.bb): Oct 07 23:19:53 splashutils: firs import to openembedded Oct 07 23:19:53 * recipe provided by Kalev Lember Oct 07 23:19:53 * closes bug 4598 Oct 07 23:20:04 03  07org.openembedded.dev * r1e4f2f12a9 10org.openembedded.dev.git/: Oct 07 23:20:04 merge of '434c4a5a7f3ee7f462db798e27cc0f7c2fd7458f' Oct 07 23:20:04 and 'aa5ca25b387398015df6ff4983a31df9f74a3dff' Oct 07 23:59:25 03  07org.openembedded.dev * r8305c3aa92 10org.openembedded.dev.git/classes/base.bbclass: base.bbclass: Use 'unzip' for *.jar files as well. Oct 07 23:59:30 03  07org.openembedded.dev * rf4f3f9c2e3 10org.openembedded.dev.git/packages/swt/ (7 files in 2 dirs): Oct 07 23:59:30 swt-gtk: Clean up and removal of obsolete versions Oct 07 23:59:30 * use fastjar instead of gjar Oct 07 23:59:30 * delete pre-releases 3.4M3 and 3.4M5 Oct 07 23:59:32 * added 3.4 Oct 07 23:59:34 * updated refreshed hildon patch Oct 07 23:59:36 03  07org.openembedded.dev * rf9da9c3bee 10org.openembedded.dev.git/: Oct 07 23:59:38 merge of '9706988717e3c2c70303850cb12a58843a12fa20' Oct 07 23:59:40 and 'd6e042ee5309203ac2054f2cbb7ead6cb61a4e38' Oct 07 23:59:42 03  07org.openembedded.dev * r5e974a3303 10org.openembedded.dev.git/conf/checksums.ini: checksums.ini: Updated swt entry. Oct 07 23:59:45 03  07org.openembedded.dev * rd340b4e964 10org.openembedded.dev.git/: Oct 07 23:59:47 merge of '414c6691b908eaab89fe4400b49ebf0e8d47aaad' Oct 07 23:59:49 and 'e17edbd4fcc354c30505a8c14cdc5944260864f7' Oct 07 23:59:51 03  07org.openembedded.dev * ra9f5dbb4fe 10org.openembedded.dev.git/conf/distro/chinook-compat.conf: chinook-compat: Updated preferred swt version. Oct 07 23:59:58 03  07org.openembedded.dev * r2562b16e8c 10org.openembedded.dev.git/: Oct 08 00:00:00 merge of '6347af950ec1a0ff98214a570984c55df668fad7' Oct 08 00:00:02 and 'aa5ca25b387398015df6ff4983a31df9f74a3dff' Oct 08 00:00:04 03  07org.openembedded.dev * r02267f61ad 10org.openembedded.dev.git/: Oct 08 00:00:06 merge of '7d22e72b69f8f66fb6c9a1517dd341b70559cfc7' Oct 08 00:00:08 and 'c82c3d791987b7feb095d7b6b0c12a52eef72fb0' Oct 08 00:00:10 03  07org.openembedded.dev * r7edfb13ff3 10org.openembedded.dev.git/: Oct 08 00:00:12 merge of '2a9c44cdb5873b8f43feb32a28672dc98ff82666' Oct 08 00:00:14 and 'dc0102fc89237b4da908e11a15fd36bef0dd2677' Oct 08 00:00:16 03  07org.openembedded.dev * r2525231ebb 10org.openembedded.dev.git/: Oct 08 00:00:18 merge of 'd3c8a06d1d4f7abaf8f97964d82c9c0811415c68' Oct 08 00:00:20 and 'e4568544a8390f2371814f5e9359920853e5b0b1' Oct 08 00:00:28 03  07org.openembedded.dev * r55d94b8d69 10org.openembedded.dev.git/classes/java.bbclass: Oct 08 00:00:31 java.bbclass: Make oe_makeclasspath behavior more sane. Oct 08 00:00:33 * return oe_makeclasspath result in variable Oct 08 00:00:35 * check for file existence only when -s is given Oct 08 00:00:37 03  07org.openembedded.dev * rf54a9127c2 10org.openembedded.dev.git/packages/miglayout/miglayout_3.0.3.bb: miglayout 3.0.3: New recipe. Oct 08 01:49:46 03zecke123 * r1093 10bitbake/ (ChangeLog lib/bb/utils.py): (log message trimmed) Oct 08 01:49:46 [bb.utils] Make prunedir cope with symlinks to directories Oct 08 01:49:46 2008-10-07 Julian Chu Oct 08 01:49:46 Reviewed by Holger. Oct 08 01:49:46 The packagekit git repository contains a symlink and on fetching the Oct 08 01:49:47 pruning failed as "packagekit" is not a directory but a symlink to a Oct 08 01:49:51 directory. Oct 08 01:58:37 03zecke123 07bitbake-1.8 * r1094 10/ (ChangeLog lib/bb/utils.py): (log message trimmed) Oct 08 01:58:37 [bb.utils] Make prunedir cope with symlinks to directories Oct 08 01:58:37 2008-10-07 Julian Chu Oct 08 01:58:37 Reviewed by Holger. Oct 08 01:58:37 The packagekit git repository contains a symlink and on fetching the Oct 08 01:58:37 pruning failed as "packagekit" is not a directory but a symlink to a Oct 08 01:58:41 directory. **** ENDING LOGGING AT Wed Oct 08 02:59:57 2008