**** BEGIN LOGGING AT Wed Oct 08 02:59:57 2008 Oct 08 07:56:25 bonjour Oct 08 08:29:54 morning Oct 08 08:34:02 morning Oct 08 08:40:50 <[X]Spot> good morning Oct 08 08:40:53 <[X]Spot> If i did all the steps here http://www.angstrom-distribution.org/building-%C3%A5ngstr%C3%B6m then how can I build a package (vlc for example) using my cutom configuration settings ? Oct 08 08:42:33 hi Oct 08 08:42:46 found a litle problem in the modell j7xx Oct 08 08:42:58 one link is wrong Oct 08 08:43:11 it refers to http://kernel.org//pub/linux/kernel/v2.6/testing/patch-2.6.19-rc6.bz2;patch=1 Oct 08 08:43:28 but should link to http://kernel.org/pub/linux/kernel/v2.6/testing/v2.6.19/patch-2.6.19-rc6.bz2 Oct 08 08:51:56 <[X]Spot> mifritscher Oct 08 08:51:58 <[X]Spot> If i did all the steps here http://www.angstrom-distribution.org/building-%C3%A5ngstr%C3%B6m then how can I build a package (vlc for example) using my cutom configuration settings ? Oct 08 08:53:29 normally bitbake vlc Oct 08 08:53:34 but I'm a beginner, too^^ Oct 08 08:56:37 <[X]Spot> :) Oct 08 08:56:59 <[X]Spot> thanks Oct 08 08:57:05 <[X]Spot> Do you have any doc about this ? Oct 08 09:12:05 morning all ! Oct 08 09:13:41 does anyone know if setting a var with bb.data.setVar is actually a local reference of the var or other modules get to see the modified var ? My tests up till now show that all modifications are local (i.e withing the .bb that changes the var) and other modules see the "original" values. Oct 08 09:13:54 03  07org.openembedded.dev * r73984c9666 10org.openembedded.dev.git/packages/python/python-native_2.5.1.bb: python-native: attempt to fix python-native build (it's more of a python-cross builds now) Oct 08 09:13:57 That said my python knowledge is at the 101 level Oct 08 09:14:04 03  07org.openembedded.dev * r9fd4cc281c 10org.openembedded.dev.git/packages/linux/ (linux-omap/dvb-fix-dma.diff linux-omap_git.bb): linux omap git: add patch that should make dvb behave better on arm, it doesn't fix the oops in the dma code Oct 08 09:20:49 03  07org.openembedded.dev * re4d8365ce5 10org.openembedded.dev.git/ (4 files in 2 dirs): gpe-sketchbook: add missing dependency (patch by raster), stylize, convert SVN build to SRCREV Oct 08 09:20:49 03  07org.openembedded.dev * r6ae564fa4a 10org.openembedded.dev.git/: Oct 08 09:20:49 merge of '1ad3e9affaa2235122936cd88f0c15b045a4ac3b' Oct 08 09:20:49 and '9ecdd345fb03610fa596e8cd5e8274e166e80087' Oct 08 09:27:27 03  07org.openembedded.dev * recbba2eac1 10org.openembedded.dev.git/packages/ttf-fonts/ttf-liberation_0.2.bb: ttf-liberation 0.2 apply enlightenment/TMP/oe/patches/0005-fix-src_uri-because-redhat-moved-original-files-away.patch (from raster) Oct 08 09:28:41 ~ping RP Oct 08 09:28:42 pong RP Oct 08 09:31:04 steliosk: yes, they are all local Oct 08 09:31:18 steliosk: only "config" time parsing is inherited by everything else Oct 08 09:31:40 zecke : Is there a way to make "global" changes from within a package ? Oct 08 09:32:03 steliosk: not in a sane way Oct 08 09:32:24 steliosk: the approach other use is to store stuff in the TMPDIR Oct 08 09:33:08 zecke : any example of a package that uses it ? Oct 08 09:33:26 steliosk: the shared lib magic in a bbclass is doing it :) Oct 08 09:33:35 steliosk: or kernel version is read from the tmpdir Oct 08 09:34:08 * steliosk looking Oct 08 09:35:50 don't hesitate to write such documentation on http://wiki.openembedded.net/index.php/Bitbake ;-) Oct 08 09:40:05 03  07org.openembedded.dev * r39d11c4b03 10org.openembedded.dev.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs: add SRCREVs for Enlightenment ST stuff (from raster) Oct 08 09:44:57 * steliosk thinks his python "kung fu" is weak Oct 08 09:46:35 03  07org.openembedded.dev * rc7ec367c3a 10org.openembedded.dev.git/packages/e17/e-wm_svn.bb: e-wm svn apply 0012-disable-no-cursor-patch-will-break-e-now-that-its.patch (from raster) Oct 08 10:12:57 zecke : How difficult will be to get bitbake to support "global" vars ? Oct 08 10:13:25 zecke : The other option i see is to get a bbclass to handle read/writting to tmpdir Oct 08 10:15:19 fscking monotone Oct 08 10:32:35 steliosk: hard, the question is how long should they live? Oct 08 10:33:51 zecke : globals should live for ever i.e reset during bitbakes start and let packages set them. kind of "shared memory" Oct 08 10:35:50 steliosk: you would want to have the same state between restarts and this is where it gets messy Oct 08 10:55:34 zecke, the public key I use on amethyst is suitable for git? Oct 08 11:01:44 Crofton|work: no :) Oct 08 11:02:00 ok Oct 08 11:02:21 make a special git key with ssh? Oct 08 11:02:56 Crofton|work: no, just send it to us, I think we need to put that into gitosis Oct 08 11:03:11 the amethyst one Oct 08 11:03:22 I'd prefer to not have too many keys :) Oct 08 11:04:29 Crofton|work: yes, but I'm too lazy to search it :) Oct 08 11:19:56 03  07org.openembedded.dev * r8b4f045ed8 10org.openembedded.dev.git/ (5 files in 2 dirs): update FSO stuff: frameworkd, mickeydbus, mickeyterm, zhone Oct 08 11:19:58 03  07org.openembedded.dev * ra1fc9cc99e 10org.openembedded.dev.git/packages/freesmartphone/ (frameworkd_git.bb mickeydbus_git.bb): frameworkd,mickeydbus: needing pprint now (no PR increase, same push) Oct 08 11:29:46 iirc I sent mine Oct 08 11:30:56 * Crofton|work goes down to the nw store and gets a key made Oct 08 11:31:03 er hardware store Oct 08 11:35:33 is git broken? Oct 08 11:36:23 dcordes: git the tool? Oct 08 11:39:20 re Oct 08 12:09:09 can OE unpack an .src.rpm? Oct 08 12:12:41 perharps base_srpm.bbclass Oct 08 12:12:54 ;) Oct 08 12:12:58 that's what my grep returned as well Oct 08 12:13:04 I was looking at rpm_xxx.bb Oct 08 12:13:09 and it did not inherit anything special Oct 08 12:13:17 but it gets an src.rpm from ftp Oct 08 12:13:32 it had a double do_unpack, to unpack the tar that was in the rpm, but I could not see any special stuff for unpacking the src rpm Oct 08 12:14:50 sure , anonymous undocumented bbclass doesn't help Oct 08 12:14:58 that's not really it, it does unpack the rpm Oct 08 12:37:16 hi Oct 08 12:45:46 I am new to OE, and I have a little question. I would like to build a target image for an embedded device , but I need to select packages. WHere can i find a tutorial or documentation that could help me to add a new machine and distro to OE ? Oct 08 12:46:07 My target is a Freescale i.MX27 on a custom board ( WDG-002 ). Oct 08 12:46:36 lucky boy Oct 08 12:47:02 my very old french doc if you want Oct 08 12:47:21 http://www.scourge.fr/mathdesc/documents/NotesPaperBook.html Oct 08 12:47:27 on i.mx21 Oct 08 12:47:38 drakou: http://pokylinux.org/handbook/ Oct 08 12:47:53 thanks :) Oct 08 12:48:49 drakou : start on a angstrom distro that you 'll derive for custom it Oct 08 12:49:46 hrw , your link is dead Oct 08 12:50:07 http://www.pokylinux.org/doc/poky-handbook.html Oct 08 12:50:11 documentation is on wiki ... Oct 08 12:50:12 the right one Oct 08 12:50:41 so my DISTO is : DISTRO = "angstrom-2008.1" Oct 08 12:50:59 and Machine ? I need to create a new machine file, I can't see the i.MX27 Oct 08 12:51:19 you need yes Oct 08 12:51:41 exept if you're running an already supported one Oct 08 13:27:04 RP, ping? Oct 08 13:30:10 humm , little problem : /proc/sys/vm/mmap_min_addr is not 0. This will cause problems with qemu so please fix the value (as root). Oct 08 13:30:15 I can't change that value .. Oct 08 13:32:13 anyone could help ? Oct 08 13:33:05 drakou: so? Oct 08 13:33:57 I 've just installed OE and try to make an image : but bitbake x11-image returns the following Oct 08 13:34:05 ERROR: Openembedded's config sanity checker detected a potential misconfiguration. Oct 08 13:34:05 Either fix the cause of this error or at your own risk disable the checker (see sanity.conf). Oct 08 13:34:05 Following is the list of potential problems / advisories: Oct 08 13:34:05 /proc/sys/vm/mmap_min_addr is not 0. This will cause problems with qemu so please fix the value (as root). Oct 08 13:34:26 drakou: echo 0 >/proc/sys/vm/mmap_min_addr as root Oct 08 13:35:42 it's not working with sudo, it's strange ... Oct 08 13:35:59 I did a sudo -s , and then the command and it's working Oct 08 13:36:00 yeah, becuase the redirect is working on the sudo, not the echo Oct 08 13:36:15 sudo doesnt run as root :-) Oct 08 13:38:43 3740 packages ... humm .. I thing I can take a break :p Oct 08 13:39:44 03  07org.openembedded.dev * r43a601d4e9 10org.openembedded.dev.git/packages/totem/totem-pl-parser_2.24.1.bb: totem-pl-parser: add 2.24.1 Oct 08 13:52:27 03  07org.openembedded.dev * re26a271428 10org.openembedded.dev.git/packages/totem/totem_2.24.2.bb: totem: add 2.24.2 Oct 08 13:59:11 hi Oct 08 14:02:05 got a compile during compiling: Oct 08 14:02:06 In function 'open', Oct 08 14:02:06 | inlined from 'collect_execute' at /mnt3/oe/build/tmp/work/arm-linux/gcc-cross-kernel-3.4.4-3.4.4-r11/gcc-3.4.4/gcc/collect2.c:1537: Oct 08 14:02:06 | /usr/include/bits/fcntl2.h:51: error: call to '__open_missing_mode' declared with attribute error: open with O_CREAT in second argument needs 3 arguments Oct 08 14:08:54 03  07org.openembedded.dev * r6ae73a87ee 10org.openembedded.dev.git/: Oct 08 14:08:54 merge of 'b2f17b7b5e7d3d3578983194a6e5733cc0d72d2b' Oct 08 14:08:54 and 'c03724e1451ef9439e526ac8b27ec4af5defecf9' Oct 08 14:08:55 03  07org.openembedded.dev * r7c7dd04703 10org.openembedded.dev.git/packages/freesmartphone/ (mickeydbus_git.bb mickeyterm_git.bb): mickey{dbus|term} git set PACKAGE_ARCH to all Oct 08 14:09:29 yes, it ises the HOSTs file Oct 08 14:10:42 ah Oct 08 14:10:44 see the problem Oct 08 14:10:46 NOTE: multiple providers are available for linux-libc-headers (linux-libc-headers, external-toolchain); Oct 08 14:10:46 NOTE: consider defining PREFERRED_PROVIDER_linux-libc-headers Oct 08 14:13:21 you'd have to wirte a multicast proxy for SSDP Oct 08 14:13:23 oops wrong window Oct 08 14:13:24 but why does it try to use external headers first? Oct 08 14:14:03 hi all.. I'm trying to get aplay etc from alsa-utils to end up in my rootfs Oct 08 14:14:46 I put it into my IMAGE_INSTALL, cleaned my rootfs build recipe and it builds alsa-utils and all of the packages, but I don't see them in my rootfs Oct 08 14:15:21 and if I list each of the alsa-utils packages individually (alsa-utils-amixer) etc, it complains that no recipe exists for them Oct 08 14:17:05 any ideas? much appreciated Oct 08 14:19:28 jconnolly: take a look at packages/alsa/alsa-utils_1.0.17.bb Oct 08 14:20:07 jconnolly: seems alsa-utils is an empty package. What you need is alsa-utils-alsamixer, for example. Just look at the .bb file. Oct 08 14:20:29 kalev: thanks I will. I have 1.0.11, 1.0.13 1.0.15 locally, I'll take a look in the repo Oct 08 14:20:56 hmm, that wans't the problem Oct 08 14:21:03 kalev: when I try to list these in a REQUIRES or IMAGE_INSTALL list, I get that they're unbuildable Oct 08 14:21:12 I'll have a look at 1.0.17.bb Oct 08 14:27:02 jconnolly: ah, I didn't notice you said it doesn't work if you list the packages individually, sorry :) Oct 08 14:28:18 kalev: nm, if I list the individual packages in IMAGE_INSTALL it ends up in my rootfs Oct 08 14:28:23 not in REQUIRES though Oct 08 14:28:26 for other packages Oct 08 14:28:31 this is good enough for now Oct 08 14:28:35 thanks for your help. ;D Oct 08 14:30:08 good! Oct 08 15:49:09 bye Oct 08 15:49:50 zecke: will the "final" git tree be made available before next Wed? I mean the one that has the final metadata but is not yet fully synced with today's mtn? Oct 08 15:53:22 zecke: I'd like to start to rebase the patches I against this one to merge them as soon as possible, after the full switch has happened. Oct 08 15:53:49 s/patches I/patches I have/ Oct 08 15:59:57 morning Oct 08 16:01:02 kergoth: hey! Oct 08 16:01:19 otavio: unlikely, the rebase should be easy Oct 08 16:01:55 zecke: right. Oct 08 16:24:07 anyone worked on getting parrot (http://www.parrot.org/) into OE? Oct 08 17:25:29 <[X]Spot> Hi all.. If i build vlc.. bitbake vlc.. and have error: No package 'fribidi' found Oct 08 17:25:36 <[X]Spot> how can I add this package ? Oct 08 17:26:18 <[X]Spot> ERROR: Nothing PROVIDES 'fribidi' (but '[]' DEPENDS on or otherwise requires it) Oct 08 17:26:25 <[X]Spot> bitvake fribidi Oct 08 17:31:09 find openembedded/packages -name '*.bb' | xargs grep fribidi Oct 08 17:31:40 (assuming that find openembedded/packages -name '*fribidi*' comes up empty first) Oct 08 17:38:53 [X]Spot: are you using latest oe.dev? Oct 08 17:49:22 <[X]Spot> yes Oct 08 17:49:29 <[X]Spot> I just installed it using git Oct 08 17:50:46 <[X]Spot> mwester Oct 08 17:50:46 <[X]Spot> OE:beagleboard [bitbake@localhost oe]$ find org.openembedded.dev/packages -name '*.bb' | xargs grep fribidi Oct 08 17:50:46 <[X]Spot> org.openembedded.dev/packages/vlc/vlc-gpe_0.8.4.bb: --disable-fribidi \ Oct 08 17:50:46 <[X]Spot> org.openembedded.dev/packages/vlc/vlc-gpe_0.7.2.bb: --disable-fribidi \ Oct 08 17:50:46 <[X]Spot> org.openembedded.dev/packages/vlc/vlc-gpe_0.8.1.bb: --disable-fribidi \ Oct 08 17:50:48 <[X]Spot> .. Oct 08 17:54:44 find openembedded/packages/ -name '*fribidi*' Oct 08 17:54:47 openembedded/packages/libfribidi/libfribidi_0.10.4.bb Oct 08 17:57:07 which vlc version is it that you're compiling? Oct 08 18:13:13 zecke, the phrase book look sgood Oct 08 18:15:07 it is rather good. :) Oct 08 18:17:25 I think I learned more about git in the few minutes it took me to look it over Oct 08 18:17:45 Indeed Oct 08 18:17:51 than in months of hacking about Oct 08 18:41:17 hi!!! Oct 08 18:41:32 is there any useful documentation for begginers on OE? Oct 08 18:41:50 for very very newbies :D Oct 08 18:44:47 odesus did you see the wiki? Oct 08 18:45:11 yes but I got some problems with bitbake.. Oct 08 18:45:15 kind of difficult to install.. Oct 08 18:45:26 on linux? Oct 08 18:45:32 http://bec-systems.com/oe/html/index.html Oct 08 18:45:36 is it true the learning curve for OE is kind of complex=? Oct 08 18:45:38 OpenEmbedded User Manual Oct 08 18:45:49 odesus: bitbake is in the repos of most distros already, might be as easy as as apt-get install bitbake or yum intsall bitbake Oct 08 18:45:55 http://wiki.openembedded.net/index.php/Getting_Started Oct 08 18:46:36 jin and Jay7: Is it true the learning curve of OE is kind of complex? Oct 08 18:46:44 odesus: yes :> Oct 08 18:47:12 but probably depends on what you want to do Oct 08 18:47:28 as I understand OE offers more advantages than other tools like freshmeat, busybox and buildroot Oct 08 18:47:47 odesus: can't say.. for me it was enough easy Oct 08 18:48:35 odesus: freshmeat and oe are completely different, and busybox is a tool that oe can build. you sound very confused. Oct 08 18:49:09 yes very very confused.. Oct 08 18:49:26 but I have already had experience in building software and making freebsd ports and gentoo ebuilds :) Oct 08 18:50:14 i have worked with buildroot Oct 08 18:50:19 but It's kind of difficult. Oct 08 18:50:40 then OE sounds the natural way for bring quickly applications to market.. Oct 08 18:50:50 odesus: just read Getting_Started for beginning Oct 08 18:51:40 ok many thanks for your suggestions to all Oct 08 19:02:33 yo kergoth Oct 08 19:09:51 how could a task cause a segmentation fault with ipkg ? Oct 08 19:11:12 http://rafb.net/p/7jOUZ555.html Oct 08 19:11:53 a breath of wind could cause a segfault in ipkg... Oct 08 19:11:57 (ducks) Oct 08 19:12:01 ;) Oct 08 19:12:29 hey pb_ Oct 08 19:15:34 I'm building OE for beagle and designing a VMWare image for others and can't get the build to not use /tmp. TMPDIR helps but I still see .c,.s,.o files in /tmp. Does anyone know a way around this or do I need to link /tmp to another partition? Oct 08 19:18:41 03  07org.openembedded.dev * r9ff620dfdf 10org.openembedded.dev.git/: Oct 08 19:18:41 merge of '88676b03209e4a8147b6f9ab4d6f5c66d0132128' Oct 08 19:18:41 and 'bcae381ccf0226faafe756ae8699c9b1623bf951' Oct 08 19:18:49 03  07org.openembedded.dev * r30760b6c0a 10org.openembedded.dev.git/packages/ (6 files in 3 dirs): Oct 08 19:18:49 beecrypt: merge duplicate recipes Oct 08 19:18:49 * Merge do_stage from portabase/beecrypt into beecrypt Oct 08 19:18:49 * Delete portabase/beecrypt (patches no longer seem to be necessary) Oct 08 19:41:40 bluelightning : changing the nesting order of the tasks solved the issue. Probably the wind changed ;) Oct 08 19:41:55 steliosk: :) Oct 08 19:48:22 <[X]Spot> mwester Oct 08 19:49:04 hi everybody : How could I do to build an image (like x11-image ) without any kernel ? I don't need a kernel. Oct 08 19:49:31 <[X]Spot> [bitbake@localhost oe]$ find org.openembedded.dev/packages/ -name '*fribidi*' Oct 08 19:49:31 <[X]Spot> org.openembedded.dev/packages/libfribidi Oct 08 19:49:31 <[X]Spot> org.openembedded.dev/packages/libfribidi/libfribidi_0.10.4.bb Oct 08 19:49:57 <[X]Spot> just did bitbake vlc Oct 08 19:51:23 <[X]Spot> Crofton|work did you do anything else (instead of "bitbake vlc") to build vlc ? Oct 08 19:52:06 drakou, you can't really Oct 08 19:52:12 You'd have to fake it Oct 08 19:52:27 because, the kernel build is failing Oct 08 19:52:54 cp: cannot stat `/home/drakou/OE/org.openembedded.dev/packages/linux/linux-2.6.21/defconfig': No such file or directory Oct 08 19:52:54 NOTE: Task failed: Oct 08 19:55:31 I think I know why Oct 08 19:55:37 03  07org.openembedded.dev * r14cca7b461 10org.openembedded.dev.git/packages/portabase/metakit_2.4.9.3.bb: metakit: fix library installation/packaging Oct 08 19:57:50 Where are located the bb files for the kernels ? Oct 08 19:58:09 I created a new machine, and in my machine file, there is : PREFERRED_PROVIDER_virtual/kernel = "${MACHINE}-kernel" Oct 08 19:58:41 but, since this is a new machine, i guess the ${MACHINE}-kernel doesn't exists ... Oct 08 20:03:04 hi Oct 08 20:03:50 hi woglinde Oct 08 20:04:18 hi jineld Oct 08 20:07:04 enough for me today, bye everyone Oct 08 20:07:12 bye jineld Oct 08 20:38:06 hi all. Oct 08 20:38:37 i want to set up an oe distribution in an "enterprise" setting. Oct 08 20:38:45 ???? Oct 08 20:38:50 this menast what? Oct 08 20:38:53 means Oct 08 20:39:00 which means that a bunch of developers (some of which are quite daft) will be using it. Oct 08 20:39:31 let me explain: Oct 08 20:39:43 h then sdk stuff might better for you Oct 08 20:39:55 instead of whole oe Oct 08 20:40:08 when i want to build (angstrom|poky|whatever) i just download the metadata and start working as you would expect. Oct 08 20:41:06 but at my workplace i have embedded linux developers that work from windows machines using putty, and don't have local storage. they are used to clearcase and want to ssh into a build machine and work off of some centralized tree. Oct 08 20:41:08 and indeed, it is older then 66 Oct 08 20:41:13 sorry Oct 08 20:41:26 sometimes those arrow keys move of their own accord Oct 08 20:41:44 turqy as I said build the sdk stuff transfer it to a servermachine Oct 08 20:41:53 and then they can work as they want Oct 08 20:42:05 even with clearcase Oct 08 20:42:40 woglinde: i'll look into it. most of my oe experience is based on poky. is the sdk you describe similar to Poky's idea of an sdk? basically just a toolchain and maybe some useful IDE plugins? Oct 08 20:42:41 and if you find errors in the sdk stuff improve it *g* Oct 08 20:43:17 toolchain yes, ide plugins hm I do not know one Oct 08 20:43:24 in terms of oe Oct 08 20:43:44 poky has a nice one for anjuta. Oct 08 20:43:58 http://labs.o-hand.com/anjuta-poky-sdk-plugin/ Oct 08 20:44:45 I never usedn anjuta Oct 08 20:45:04 I am somehow familar with eclipse Oct 08 20:45:26 ah. Oct 08 20:46:12 most of the folks around here just use vim/emacs/nano + ssh, or they use some windows editor combined with a window's clearcase tool and hack over samba shares. Oct 08 20:46:26 hm oh Oct 08 20:46:38 I thought they only know windows tools Oct 08 20:46:45 woglinde we have some oe integration stuff in eclipse and hope to release an initial version in the next week or so. Oct 08 20:46:57 kgilmer huh what stuff? Oct 08 20:47:24 a plugin that sits on top of CDT to make hosting the OE toolchain easier Oct 08 20:47:37 integration with QEMU similiar to how it works in anjuta Oct 08 20:47:40 and a recipe editor Oct 08 20:47:47 hm Oct 08 20:47:52 sounds nice Oct 08 20:48:13 screenshot: http://bugblogger.com/in-the-oven-openembeddedpoky-eclipse-225/ Oct 08 20:48:57 eclipse needs better git and mercurial supprt Oct 08 20:49:17 eclipse needs to stop crashing :) Oct 08 20:49:24 aeh hehe Oct 08 20:49:40 kgilmer_: work well with 10GB of ram :) Oct 08 20:49:41 yes eclipse is sometimes a shitty blob Oct 08 20:49:52 hahah chouimat|work Oct 08 20:49:55 dont look at the code of the profiling project Oct 08 20:50:11 i miss eclipse 2.11 Oct 08 20:50:30 I hacked one night to get the profiling stuff compiled with newer libxerces-c Oct 08 20:50:49 but i think the fundamental issue is that the jvm is unstable in linux. 1.5, 1.6 both. Oct 08 20:50:57 somehow they forget linux because under windows alls working fine Oct 08 20:51:08 kgilmer_ hm uh unstable? Oct 08 20:51:12 in which sense? Oct 08 20:51:22 agreed woglinde Oct 08 20:52:17 well i get a lot of 'out of memory' errors and jvm crashes Oct 08 20:52:41 googling around, regardless of distro, shows a large number of open issues with sun jvm in linux Oct 08 20:52:59 we use linux and eclipse at bug labs and pretty much everyone has issues Oct 08 20:52:59 hm Oct 08 20:53:16 seems they only happens with eclipse Oct 08 20:53:47 yeah, that's true. i think eclipse exersizes the jvm the most. Oct 08 20:54:10 i'm kind of frustrated about it at the moment woglinde :) Oct 08 20:54:22 kgilmer yes I can understand you Oct 08 20:54:47 ganymade made a good step Oct 08 20:54:53 but not enough Oct 08 20:56:10 hm I have to kill the inventer of AM_GLIB_GETTEXT Oct 08 20:58:52 heading out woglinde. ttyl :) Oct 08 21:00:36 woglinde: thanks for cleaning up splashutils do_configure mess yesterday Oct 08 21:00:54 kalev no prob Oct 08 21:00:59 pmount worries me Oct 08 21:01:17 the idiots didnt understand how to program for users which do not want NLS Oct 08 21:01:50 evening Oct 08 21:01:59 ah, hmm Oct 08 21:02:20 and I found mega old rotten macro in e2fs-libs Oct 08 21:02:44 I have a very annoying problem with fetching a kernel I dont know, where I have done the misconfiguration Oct 08 21:02:56 what do they use e2fs-libs for anyway? Oct 08 21:03:04 can someone please have a look at this bitbake output: http://pastebin.com/m4dc6ee5f Oct 08 21:03:10 thanks in advance! Oct 08 21:03:34 woglinde: Hello again :-) Oct 08 21:03:51 hi szsoftware Oct 08 21:04:02 kalev aeh youself put the depends in Oct 08 21:04:20 another stalker is after me Oct 08 21:04:26 http://www.linkedin.com/in/erindufresne Oct 08 21:04:45 szsoftware try to update your bitbake Oct 08 21:04:59 szsoftware otherwise I have no clue Oct 08 21:05:01 yes, but I have no idea what they use from e2fs-libs, I just noticed their configure script needed it Oct 08 21:05:46 you said that you found some old rotten macro, so I thought maybe you investigated something :) Oct 08 21:06:02 ah, you're just trying to get it to compile on your target? Oct 08 21:06:44 kalev hehe I tried to compile it with newer gettext Oct 08 21:06:57 kalev it is fixed Oct 08 21:07:03 but I have to fix pmount too Oct 08 21:07:06 woglinde: hmm, ok. The funny thing is, I dont need a kernel built with bitbake console-image, because I build that outside of oe. But I dont know, how to disable kernel building. Any hint? Oct 08 21:07:21 woglinde: awesome :) Oct 08 21:07:35 szsoftware, I've tried and failed to do that Oct 08 21:07:37 szsoftware go to the manual and read about PROVIDED stuff Oct 08 21:07:58 crofton hm cant you do it like the toolchain? Oct 08 21:08:17 woglinde, it was easier to build a kernell :) Oct 08 21:09:04 woglinde: thanks so far.CU Oct 08 21:09:39 kalev hm yes seems ted tso likes that his code bit rots Oct 08 21:10:07 or he did not update his gettext version for 5 years Oct 08 21:10:10 dont know Oct 08 21:11:21 at least that's something what OE is good at: we've got many pairs of eyes looking at every recipe :) Oct 08 21:42:52 hahah Oct 08 21:43:00 now seems I have fight it Oct 08 21:43:03 and won Oct 08 21:43:16 ~lart pmount authors for not thinking Oct 08 21:43:16 * ibot takes a rusty axe and swings it violently, taking pmount authors's head off for not thinking Oct 08 21:43:32 *g* Oct 08 21:43:54 what was the problem? Oct 08 21:44:54 ~clap woglinde Oct 08 21:45:06 silly ibot :) Oct 08 21:51:43 kalev they used it, only to think about that someone will not use NLS Oct 08 21:55:43 ok, bed time for me Oct 08 21:55:46 see you woglinde Oct 08 21:55:55 kalev pleas try pmount tomorrow Oct 08 21:56:01 and if you have Oct 08 21:56:01 sure Oct 08 21:56:07 test it with a mmc too Oct 08 21:56:12 best internal Oct 08 21:56:25 want to know if the remove mmc hack is needed Oct 08 21:57:17 ok, I will Oct 08 21:57:18 thanks Oct 08 23:21:09 he khem Oct 08 23:21:15 khem are you awake? Oct 08 23:24:19 he should be :) Oct 08 23:24:30 1630 on the left coast Oct 08 23:25:33 hm I wonder why we install uclibc utils in stage Oct 08 23:25:38 this is rather useless Oct 08 23:27:06 that is what I want to discuss with him Oct 08 23:27:19 because we are not using qemu Oct 08 23:27:25 this can go away Oct 08 23:27:29 jo raster Oct 08 23:27:47 boo Oct 08 23:28:02 hms I need a faster computer Oct 08 23:31:29 so nice Oct 08 23:31:33 pmount is fixed Oct 08 23:32:19 ~lart the morons for use install with -o root and -g root Oct 08 23:32:19 * ibot teaches the morons the basics, including how to RTM for use install with -o root and -g root Oct 08 23:36:44 okay Oct 08 23:36:45 nite Oct 08 23:36:59 and hope that I didnt break oe again *g* Oct 08 23:37:56 hmmm Oct 08 23:38:32 setting owner and group may make sense if they they set setuid bits... Oct 08 23:40:02 03  07org.openembedded.dev * r374f41b58e 10org.openembedded.dev.git/packages/e2fsprogs-libs/ (3 files in 2 dirs): Oct 08 23:40:02 e2fsprogs-libs: update to versio 1.41.2 Oct 08 23:40:02 * add mkinstalldirs.patch which is necessary because Oct 08 23:40:02 newer gettext do not use MKINSTALLDIRS anymore but MKDIR_P Oct 08 23:40:03 03  07org.openembedded.dev * r4290c4b34a 10org.openembedded.dev.git/packages/gettext/gettext_0.17.bb: Oct 08 23:40:06 gettext: make version 0.17 the default Oct 08 23:40:08 * 0.14.1 is using some old macros never used in modern autotools scripts Oct 08 23:40:09 raster but not when you do it in oe Oct 08 23:40:10 * only e2fslibs so far clashes, while it needs some of the old macros Oct 08 23:40:12 (MKINSTALLDIRS) but this will be fixed really fast Oct 08 23:40:14 03  07org.openembedded.dev * r728b45761b 10org.openembedded.dev.git/: Oct 08 23:40:16 merge of 'ea367d406c83c35493f3e8c93c52fad15e9eac1e' Oct 08 23:40:18 and 'f2ebdb1826372581987a96a662d30d97b544c438' Oct 08 23:40:20 03  07org.openembedded.dev * r80bdb16c9d 10org.openembedded.dev.git/classes/gettext.bbclass: Oct 08 23:40:24 gettext.bbclass: make it simpler to understand Oct 08 23:40:25 * set the deps for package which can use NLS stuff Oct 08 23:40:27 * filter the deps out if NLS is not requested Oct 08 23:40:29 03  07org.openembedded.dev * rea59717145 10org.openembedded.dev.git/packages/pmount/ (4 files in 2 dirs): (log message trimmed) Oct 08 23:40:57 pmount: update to version 0.9.17 Oct 08 23:40:57 * recipe original provided by Kalev Lember Oct 08 23:40:57 * added gettext.patch because we want build without NLS stuff Oct 08 23:40:57 * added install.patch bescaue we want to install withou -o root Oct 08 23:40:57 and -g root Oct 08 23:40:59 * leave mmc problem open, if there was one Oct 08 23:41:22 hm Oct 08 23:41:25 okay Oct 08 23:41:37 forget the line closes bug 4599 Oct 08 23:43:13 woglinde: hey how are u Oct 08 23:45:15 he khem Oct 08 23:45:21 khem hm I am tired now Oct 08 23:45:23 a little Oct 08 23:45:24 bit Oct 08 23:45:36 khem did you read about my uclibc suggesion? Oct 08 23:45:52 we dont need ldconfig and ldd in staging Oct 08 23:50:35 hm okay Oct 08 23:50:38 nite now really Oct 08 23:52:39 03  07org.openembedded.dev * r448b0e5e3f 10org.openembedded.dev.git/packages/gettext/gettext-native_0.17.bb: Oct 08 23:52:39 gettext-nativ: added config.rpath back to automake dir Oct 08 23:52:39 * bump PR Oct 09 01:55:32 Hi!!! Oct 09 01:56:19 Is there any other place for downloaded OE.mtn.bz2.1 continuously openembedded.net cuts the transference.. Oct 09 02:01:43 any suggestion? Oct 09 02:05:49 mmmm Oct 09 02:43:56 03  07org.openembedded.dev * r9497fc9fd5 10org.openembedded.dev.git/: Oct 09 02:43:56 merge of '83df338c70f7f4819197794f3daebb64a89fcf0a' Oct 09 02:43:56 and '9f79c6b2ba1ff673c6a3fa4d02579b720046d1cf' Oct 09 02:44:04 03  07org.openembedded.dev * r6d4a8240a7 10org.openembedded.dev.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs.inc: bump EFL Oct 09 02:56:25 03  07org.openembedded.dev * rd86e7ea56f 10org.openembedded.dev.git/ (2 files in 2 dirs): ane-srcrevs.inc: bump openmoko kernels and move linux-openmoko-devel to stable-tracking branch (warning, may be broken) **** ENDING LOGGING AT Thu Oct 09 02:59:57 2008