**** BEGIN LOGGING AT Tue Jun 14 23:59:56 2005 Jun 15 00:01:15 morning Jun 15 00:27:19 morning all Jun 15 00:28:40 morning RP Jun 15 01:35:10 morning Jun 15 01:36:15 hi hrw|work Jun 15 01:36:35 morning hrw|work Jun 15 01:46:52 hi hrw|work, RP, do13 Jun 15 01:49:41 ~nslookup home.hrw.one.pl Jun 15 01:49:47 home.hrw.one.pl is 192.168.1.134 Jun 15 01:49:55 ibot: gracias Jun 15 01:49:57 i guess gracias is thank you Jun 15 01:49:59 hi schurig Jun 15 01:50:03 hrw|work: hi! Jun 15 01:50:32 mickey|sports: bbimage has a bug, it creates the log.image_cmd.* and run.image_cmd.* files directly in TMPDIR Jun 15 01:51:19 hi schurig Jun 15 01:51:31 bluelightning: hiho Jun 15 01:52:00 ~nslookup home.hrw.one.pl Jun 15 01:52:02 home.hrw.one.pl is 62.21.36.51 Jun 15 01:52:35 noip works Jun 15 01:53:29 openwrt use busybox init.. it so simple that looks strange for me Jun 15 01:55:52 hrw|work: hey, I just see that you are also in Bitbake at Berlios. When you delete the line data.setVar('T', '${TMPDIR}', localdata) near the end of bitbake/bin/bbimage, then the log&run file will be created at the usual place Jun 15 01:58:58 moin Jun 15 01:59:52 schurig: you are not in devel list of bitbake? Jun 15 01:59:55 hi CoreDump|home Jun 15 02:00:02 hi CoreDump|home Jun 15 02:03:01 hrw|work: unfortunately not Jun 15 02:03:16 hrw|work: I didn't ask kergoth`zzz so far, when I'm online he's sleeping Jun 15 02:03:51 hrw|work: for OE itself, we have several "project managers" that can add users, but for bitbake he's the only one that can add people Jun 15 02:05:54 hi koen Jun 15 02:06:04 good morning all Jun 15 02:06:14 hey bluelightning Jun 15 02:06:15 hi koen Jun 15 02:06:24 hey RP Jun 15 02:06:53 hi koen Jun 15 02:06:56 schurig: ok - will look Jun 15 02:06:57 hey CoreDump|home Jun 15 02:09:40 03CoreDump 07 * r1.3523.1.1 10openembedded/packages/altboot/ (3 files in 3 dirs): altboot: Add a timeout on boot and only show the menu when requested Jun 15 02:10:10 who maintains network-suspend-scripts (no MAINTAINER set)? Jun 15 02:10:31 noone? Jun 15 02:11:24 I was wondering what /etc/default/usbd was intended to do (it's referred to by /etc/apm/scripts.d/usbnet) Jun 15 02:13:09 bluelightning: it keep name of usbd module on Zaurus Jun 15 02:13:22 sa1100_bi on collie, pxa_bi on pxa zauruses Jun 15 02:14:02 is sa1100usb_core also used on collie? Jun 15 02:14:30 no Jun 15 02:14:44 hmm ok Jun 15 02:15:51 I wonder if that script needs to be reworked a bit... currently I'm looking at how to change it to handle g_ether on pxa iPAQs, and it would seem to me that some editable script like /etc/default/usbd would be the way to choose what module to use (g_ether or usb-eth) Jun 15 02:15:52 03hrw * r254 10bitbake/bin/bbimage: Keep all temporary files in proper place not in TMPDIR (patch by Holger Schurig) Jun 15 02:16:16 obviously I don't want any patch to screw it up for zaurus users either Jun 15 02:16:53 bluelightning: under 2.6 g_ether survive suspend/resume (atleast on Z) Jun 15 02:17:12 no network-suspend-scripts are not needed Jun 15 02:17:40 s/^no/so Jun 15 02:17:43 with 2.4 the ported g_ether is often non-functional on my h3970 Jun 15 02:17:54 non-functional after resume, I mean Jun 15 02:18:02 unloading it and loading it fixes that Jun 15 02:18:07 aha Jun 15 02:18:21 usb gadget backport to 2.4.. nice Jun 15 02:18:41 yeah, it works well, especially to be able to connect to WinXP for those who want that Jun 15 02:39:22 hi pb Jun 15 02:39:51 hi hrw|work Jun 15 02:39:58 morning all Jun 15 02:40:56 morning all Jun 15 02:41:07 hi ade|desk, pb_ Jun 15 02:56:29 mickey|sports: bitbake interactive lack "get" Jun 15 02:56:36 getvar even Jun 15 02:56:56 ops.. there is a print.. Jun 15 02:57:36 ya, there's print Jun 15 02:57:39 morning, btw. Jun 15 02:57:59 NOTE: [11:54] package gcc-cross-4.0.0: failed Jun 15 02:58:02 BB>> print PREFFERED_VERSION_gcc-cross Jun 15 02:58:02 3.4.4 Jun 15 02:58:28 hey mickeyl Jun 15 02:58:28 03koen 07 * r1.3535 10openembedded/packages/maemo/ (10 files in 6 dirs): Jun 15 02:58:29 Many meamo files: source -> ., courtesy Florian Boor Jun 15 02:58:29 Add gconf-osso, part of the maemo.org suite, courtesy Florian Boor Jun 15 02:58:31 03koen 07 * r1.3534 10openembedded/packages/base-files/ (base-files_3.0.14.bb base-files/nokia770/fstab): base-files_3.0.14.bb: support nokia770, courtesy Florian Boor Jun 15 02:58:59 mickeyl: good morning Jun 15 02:59:01 ~lart me Jun 15 02:59:02 * ibot whips out a sword and chops hrw|work in half Jun 15 03:03:13 mickeyl: can getvar be alias to print? Jun 15 03:03:36 hrw|work: sure. Jun 15 03:03:44 while we are talking about that... Jun 15 03:03:56 hi mickeyl Jun 15 03:04:00 setvar/getvar deals with outer env. variables Jun 15 03:04:06 how to deal with package specific data ? Jun 15 03:04:09 pkggetvar? Jun 15 03:04:30 doens't look good to me Jun 15 03:04:31 no idea - didtn had to tweak them Jun 15 03:04:59 in my tree, i have made print to work with package data Jun 15 03:05:09 print nano do_package Jun 15 03:05:37 i may do it with 'peek' and 'poke' Jun 15 03:05:41 as a reminiscent to the old C64 days Jun 15 03:05:42 heh Jun 15 03:05:47 heh Jun 15 03:06:03 mickeyl: to old PET2001 days Jun 15 03:06:12 old 8bit BASIC days Jun 15 03:06:14 that one was earlier with peek and poke Jun 15 03:06:15 right from PET over VIC20 to C64 Jun 15 03:06:36 I guess you should make "sys 49152" do something exciting, in that case. Jun 15 03:06:49 sys 6502,n was also nice Jun 15 03:06:53 not sure what, maybe display an ascii art picture of kergoth or something. Jun 15 03:07:03 heh Jun 15 03:07:13 wasn't it ? sys 64738 ? Jun 15 03:07:16 to reboot Jun 15 03:07:34 ah yes, right Jun 15 03:07:49 49152 was the base address for the gfx controller if I remember correctly Jun 15 03:08:00 long time ago Jun 15 03:08:15 mickeyl: sys 6502,n printed some text n-times out, I think it was Microsoft or Basic or something Jun 15 03:14:58 I have a package which wants to use makeinfo and latex in its Makefile, now some people will have these available, some won't (like me) Jun 15 03:15:15 I suppose I should add these as -native packages and make then DEPENDS? Jun 15 03:15:17 if its just for docs, disable it Jun 15 03:15:23 ah, ok Jun 15 03:15:27 phew ;) Jun 15 03:15:49 yeah, native latex might be too much. Jun 15 03:16:13 if you'd like to have a go at adding it, though, feel free. Jun 15 03:17:43 hello, anyone can explain what is the meaning of "INHERIT"? Jun 15 03:18:36 it contains a list of classes to be inherited. Jun 15 03:19:05 pb_: hmm, got a few things to sort out first ;) Jun 15 03:19:15 03mickeyl 07 * r1.3536 10openembedded/packages/mathomatic/ (mathomatic_unstable.bb mathomatic_12.4.2.bb mathomatic.inc): add mathomatic, a small, portable, general purpose CAS. Patch courtesy Simon Pickering Jun 15 03:19:24 thx mickeyl Jun 15 03:19:37 np. can you clos the bug? Jun 15 03:19:40 close even Jun 15 03:19:50 pb_, thx Jun 15 03:20:11 yep - give me a tick to work out how Jun 15 03:20:32 mickeyl: you should be able to close the bug yourself Jun 15 03:21:04 change it to 'fixed'? Jun 15 03:21:09 right Jun 15 03:21:18 pb_: sure i am... i wanted to delegate that though ;) Jun 15 03:21:24 ah, heh Jun 15 03:21:44 lardman|work: take a look at the changeset, i've added a .inc for the similarities Jun 15 03:22:28 ok, sorry about that - I've got a few different bbs in various states of finishedness and wanted to get that one out so I could forget about it Jun 15 03:22:46 np, just a recommendation for similar cases in the future Jun 15 03:22:52 I think I've closed it now, takes enough different stages though! Jun 15 03:23:03 pb_: i'm getting to like bugzilla Jun 15 03:23:39 i had my reservations, but once you get started it turns out nice Jun 15 03:24:57 lardman|work: just marking it as RESOLVED FIXED is sufficient. Jun 15 03:25:19 there is no need to change it from RESOLVED to CLOSED. Jun 15 03:25:32 pb_: ok Jun 15 03:25:37 mickeyl: excellent :-) Jun 15 03:26:50 pb_: wrt xlibs and dietx, is there a way of making the full-fat ones a DEPEND rather than whichever the user is building with (as diet won't work)? Jun 15 03:29:36 Or should I just add a comment to the .bb which someone will presumably read if it won't build? Jun 15 03:30:19 lardman|work: not at the moment. that's a bit of a hole in our xlibs packaging. Jun 15 03:31:05 pb_: I'll add the comment then Jun 15 03:32:11 righto Jun 15 03:39:33 heh... microsoft have router/AP powered by WinCE... Jun 15 03:47:50 mickey|lunch: mathomatic looks cool Jun 15 03:47:57 (for a console app ...) Jun 15 03:50:41 I thought it looked interesting Jun 15 03:51:06 How can one perform patching after configure has been run? Jun 15 03:51:13 (just out of interest) Jun 15 03:51:47 lardman|work: addtask extrapatching after configure and do_extrapatching() { .... } Jun 15 03:52:00 ok, cool Jun 15 03:52:00 "addtask extrapatching after configure" Jun 15 03:52:28 hmm.. gcc-cross 3.4.4 failed for wrt54 ;( Jun 15 03:54:07 I think you need "addtask extrapatching after do_configure before do_compile" or some such. Jun 15 03:55:33 03koen 07 * r1.3537 10openembedded/packages/ (7 files in 3 dirs): Jun 15 03:55:33 sapwood_1.20-1.bb: extra install_append, courtesy Florian Boor Jun 15 03:55:33 osso-gnome-vfs2_2.8.4.4-1.bb: adjust depends, courtesy Florian Boor Jun 15 03:55:33 Many files: added, courtesy Florian Boor Jun 15 03:55:33 hildon-libs_0.9.6-2.bb: adjust staging, courtesy Florian Boor Jun 15 03:58:52 mickey|lunch: http://pastebin.ca/14610 Jun 15 03:59:46 mickey|lunch: in other words: bug #88 strikes back Jun 15 04:14:33 mickey|lunch: http://pastebin.com/299783 - bitbake -i failed when had to post large log Jun 15 04:23:40 keturn: ping Jun 15 04:27:36 how to get gcc-cross build for mipsel? Jun 15 04:27:45 mipsel/uclibc Jun 15 04:38:14 someone built gcc-cross 3.4.4 with uclibc recently? Jun 15 04:38:56 grr fecking xerox, lo all btw Jun 15 04:39:06 morning ade Jun 15 04:39:47 http://www.hrw.one.pl/_pliki/oe/tmp/gcc-cross-mipsel-uclibc.error.txt is my error output which I got when trying to build Jun 15 04:44:16 hrw|work: i seem to remember something on the uclibc ml about 3.4.4 and mips ... but can't find it atm Jun 15 04:44:22 will keep looking Jun 15 04:44:54 I can test 3.4.3 Jun 15 04:46:36 i know that 3.4.2 is a working one Jun 15 04:47:57 uclibc 0.9.27 or svn btw ? Jun 15 04:48:04 0.9.27 Jun 15 04:48:10 righty ho Jun 15 04:48:22 I want to get builddir for my wrt54gs Jun 15 04:49:09 openwrt use 3.4.3 or 3.4.4 Jun 15 04:49:11 ah the openwrt OE'ed Jun 15 04:49:40 hmm ok will search uclibc ml harder then ;) Jun 15 04:51:24 not quite OE'ed.. they use own buildsystem Jun 15 04:51:43 hi guys Jun 15 04:54:20 i'm having a problem with bitbake (gcc-cross and opie-image for now) - the flex-native-2.5.31 package is missing (404) - what mirror should i specify for it? Jun 15 04:55:51 heanet? Jun 15 04:56:26 it is default one, it has flex-2.5.31.tar.bz2 archive (just downloaded to test) Jun 15 05:00:30 hrw|work: so instead of GNU_MIRROR = "ftp://ftp.gnu.org/gnu" in local.conf it will be "ftp.heanet.ie/mirrors/ftp.gnu.org/"? (that dir looks empty to me) Jun 15 05:01:54 flex will be fetched from sourceforge Jun 15 05:04:10 Archer: flex is from SOURCEFORGE_MIRROR Jun 15 05:05:24 hrw|work: do you have gcc-cross-initial ? Jun 15 05:05:56 ade|desk: yes Jun 15 05:06:58 what do I do after bitbake opie-image finishes to make an initrd.bin and a zImage? Jun 15 05:07:03 hrw|work: thanks - trying with ftp.heanet.ie/mirrors/download.sourceforge.net/ Jun 15 05:07:43 I've got two scripts in /stuff/build/tmp: run.image_cmd.30422 run.image_cmd.30425 Jun 15 05:07:57 jumpkick_: if i'm not mistaken the zImage is the kernel image - you need to transfer them to the device/board somehow (i've just started so I don't know any more) Jun 15 05:08:40 Archer: remove SOURCEFORGE_MIRROR setting from your local.conf and it will work Jun 15 05:08:46 hrw|work: have you tried without make -j 4 ? Jun 15 05:08:50 jumpkick_: ignore Jun 15 05:08:53 ade|desk: not yet Jun 15 05:09:36 ade|desk: rebuilding... Jun 15 05:10:00 ignore? Jun 15 05:10:14 sec Jun 15 05:10:22 think I found it in deploy.... Jun 15 05:10:57 jumpkick_: recent bitbake put that files in correct place now Jun 15 05:10:58 looks too small though: -rw-r--r-- 1 bitbake bitbake 9.7M 2005-06-14 23:26 opie-image-collie-20050613055205.rootfs.tar.bz2 Jun 15 05:11:23 nevermind Jun 15 05:11:28 the one above it is 14M Jun 15 05:11:31 good stuff Jun 15 05:11:42 jumpkick_: jffs2 are padded Jun 15 05:12:07 gotya, so the padding compresses really well I'd guess then Jun 15 05:13:57 its 0 Jun 15 05:14:06 initrds need to be ext2.gz or cramfs Jun 15 05:16:38 ade|desk: same ;( Jun 15 05:18:22 ok, am getting an OE tree atm, will then take a look at making a mipsel toolchain Jun 15 05:19:45 thx ade Jun 15 05:22:26 this may take some time... am at the wrong office today for this sort of thing :( Jun 15 05:22:36 happens Jun 15 05:22:41 I can wait Jun 15 05:22:52 have router to configure ;) Jun 15 05:22:56 days at this rate Jun 15 05:23:13 * jumpkick_ nerviously waits for his opie-image to finish flashing.... Jun 15 05:24:11 oh dear packages/opie could take a while Jun 15 05:27:30 * jumpkick_ boots it Jun 15 05:28:21 wooohooo! Jun 15 05:40:09 hmm... doesn't mount SD... :( Jun 15 05:40:39 hangs in /bin/sh /etc/sdcontrol insert Jun 15 05:41:15 gotta go to work... Jun 15 06:04:31 hrw|work: removed the sourceforge declaration, but it still tries to get flex from oesources.org Jun 15 06:04:50 Archer: a... Jun 15 06:05:13 Archer: why you didnt said that on start... Jun 15 06:05:18 should I'll try with ftp.heanet.ie/mirrors/download.sourceforge.net/sourceforge Jun 15 06:05:18 ? Jun 15 06:05:42 you can comment CVS_TARBALL_STASH so it will be fetched from upstream location not from our source mirror Jun 15 06:05:46 keturn: hi Jun 15 06:05:48 kergoth: hi Jun 15 06:06:04 kergoth: can you add schurig to bitbake at berlios? Jun 15 06:06:16 kergoth: hi! Jun 15 06:06:23 hey Jun 15 06:06:40 been a long time since i've woken up at 8am. Jun 15 06:06:42 * kergoth yawns Jun 15 06:07:24 hrw|work: trying Jun 15 06:11:08 schurig: whats your berlios username? Jun 15 06:11:17 kergoth: holgerschurig Jun 15 06:11:21 k Jun 15 06:12:27 good morning kergoth Jun 15 06:14:41 keturn: ping Jun 15 06:44:08 hrw|work: sorry for the delay .. still getting gcc Jun 15 06:44:32 ade|desk: no problem Jun 15 06:44:50 heh.. my network cards does not support WOL :( Jun 15 06:45:11 will have to enable via-rhine one (onboard) and check does it support wol Jun 15 06:45:17 hi kergoth Jun 15 06:45:27 hrw|work: why do you need WOL ? Jun 15 06:47:01 ade|desk: I can remote login to my router so if I will need to have my home machine working then ssh router, wake home machine Jun 15 06:47:40 ah groovy Jun 15 06:49:28 hrw|work: flex is still still missing (even with CVS_TARBALL_STASH commented) - actual dl location is mesh.dl.sf.net Jun 15 06:56:36 Archer: how you set mesh? Jun 15 06:56:50 mesh? Jun 15 06:57:20 Archer: why not just download it from debian or some such and put it in your sources/ Jun 15 06:57:54 mesh seems to be the default here.. Jun 15 06:58:00 a sec.. Jun 15 07:01:40 http://rafb.net/paste/results/41zkVf40.html Jun 15 07:03:38 MACHINE = "tune-xscale" Jun 15 07:03:43 nice one ;) Jun 15 07:04:47 Archer: I see. Jun 15 07:05:25 03zecke123 * r255 10bitbake/ (MANIFEST TODO lib/bb/data_dict.py): Jun 15 07:05:25 BitBake (minor update): Jun 15 07:05:25 -Remove data_dict.py as it was not needed for sooo long Jun 15 07:05:25 -Remove TODOs related to data_dict.py Jun 15 07:05:25 -Remove data_dict from the MANIFEST Jun 15 07:05:49 rFF Jun 15 07:07:02 hrw|work: it's a pretty bare arm board - at first i'd just be curious if th fbdev is recognized and it it loads anything Jun 15 07:07:46 should i change it to something else? Jun 15 07:08:19 grab http://heanet.dl.sourceforge.net/sourceforge/lex/flex-2.5.31.tar.bz2 and place in DL_DIR Jun 15 07:08:33 I will push change in few minutes Jun 15 07:11:42 Is there a list somewhere of the allowable SECTION sections? Jun 15 07:13:27 lardman|work: ipkgfind.handhelds.org Jun 15 07:13:42 the sections are on your left Jun 15 07:15:38 03koen 07 * r1.3538 10openembedded/ (4 files in 3 dirs): osso-af-startup_0.28-1.bb: remove -oknodo, courtesy Florian Boor Jun 15 07:17:02 03hrw 07 * r1.3537.1.1 10openembedded/packages/flex/flex.inc: flex: changed SRC_URI back to SOURCEFORGE_MIRROR as mesh.dl.sf.net does not have file Jun 15 07:17:42 Archer: with that change flex is fetchable Jun 15 07:18:32 nice - should i run update bk://openembedded.bkbits.net/openembedded to get it? Jun 15 07:20:32 change the first openembedded into oe-devel Jun 15 07:20:44 or wait an hour till the repo's get synced :) Jun 15 07:21:22 i'll go with the devel ver ;) Jun 15 07:23:59 done - well, fingers crossed... Jun 15 07:34:58 What's the flag to pass to make which means variables outside of the Makefile take priority? -s? Jun 15 07:35:06 -e Jun 15 07:35:11 thx Jun 15 07:35:20 np Jun 15 07:39:57 hrw|work: grr went away for a bit .. bloody well got gcc-cross-4.0.0-r1 being built .. oops Jun 15 07:49:58 people are never satisified Jun 15 07:51:05 so true Jun 15 07:52:20 hrw|work: ok 3.4.4 started now Jun 15 07:52:45 cool Jun 15 07:52:59 may take a while on this laptop Jun 15 07:53:11 ( in a vmware session) Jun 15 07:53:38 pb_: When I get some breathing room, I'll merge in some of my x-compile hacks for xserver-xorg Jun 15 07:55:29 I was in the midst of xserver-xorg working (I had been using my own homebrew xorg for a while now) when I was distracted by other things Jun 15 07:58:02 CosmicPenguin: ah, cool. Jun 15 07:58:22 I patched up 6.8.99.10 well enough to make it work in my i686 -> i586 "cross", but I suspect there are still some bits that aren't right. Jun 15 07:59:02 hrw|work: right i now have the same error, good start :) Jun 15 07:59:08 I know that some guy was having trouble compiling it on ARM the other day. I think that was because his compiler was choking on x86 assembler code in one of the drivers. Jun 15 07:59:43 heh Jun 15 08:00:04 Yeah, I'm not making that big a leap either - just x86_64 -> 486 Jun 15 08:00:32 ade|desk: ;) Jun 15 08:11:35 hrw|work: seen http://buildroot.uclibc.org/cgi-bin/viewcvs.cgi/trunk/buildroot/toolchain/gcc/3.4.4/602-sdk-libstdc%2B%2B-includes.patch?rev=10397&view=markup Jun 15 08:13:33 ade|desk: it should help? Jun 15 08:14:04 well i cant see the patch for this in oe ... maybe i am not looking correctly Jun 15 08:16:17 ade|desk: sdk-libstdc++-includes.patch in packages/gcc/gcc-3.4.4/ Jun 15 08:16:50 ade|desk: but it is used only in 3.3.4 Jun 15 08:17:57 sorry .. its just that it seems not to find the headers properly Jun 15 08:18:55 maybe it should be added into gcc 3.4.4 also - I dont know Jun 15 08:20:12 hrw|work: i need to travel home now... will take a better look tomorrow Jun 15 08:20:25 ok Jun 15 08:20:31 I will do test at home Jun 15 08:25:03 bye Jun 15 09:44:41 pb_: where can I find valid -mtune params for gcc? Jun 15 09:44:55 I think I screwed up the arm926ejs one Jun 15 09:47:42 I think there's documentation in the code somehwere Jun 15 09:48:24 I was hoping for online list, but source will do Jun 15 09:48:27 * koen|tv fires up vi Jun 15 09:48:33 koen|tv: in the manual Jun 15 09:49:08 http://gcc.gnu.org/onlinedocs/gcc-4.0.0/gcc/ARM-Options.html#ARM-Options Jun 15 09:49:19 thanks Jun 15 09:51:13 aha Jun 15 09:51:27 arm926ej-s instead of arm926ejs Jun 15 09:51:28 mmmm, iwmmxt Jun 15 09:51:38 I wonder if that does anything yet Jun 15 09:52:04 iirc that doesn't work on hardfloat Jun 15 09:52:15 I think it just enables the wMMX intrinsics. Jun 15 09:53:34 afaik, gcc doesn't have any ability to vectorize code yet (though there is some work going on in that area for other platforms) Jun 15 09:58:24 ah Jun 15 10:00:23 03koen 07 * r1.3540 10openembedded/conf/machine/tune-arm926ejs.conf: tune-arm926ejs.conf: fix typo in mtune parameter Jun 15 10:01:36 this reminds me of a crazy idea I had Jun 15 10:02:01 what would be involved in booting an ipaq to linux in big-endian mode? Jun 15 10:03:40 could it be as simple as a modified version of haret ? Jun 15 10:07:31 well, the trouble is that all the I/O would be screwed up. Jun 15 10:08:12 other than that, yeah, you could just tweak haret to flip into big-endian mode before booting the kernel. Jun 15 10:10:02 pb_, which I/O ? I figured I was missing something, but I don't quite understand Jun 15 10:10:39 jacmet has successfully booted the nslu2 in LE mode Jun 15 10:12:06 jacques: iirc, any peripherals that are less than 32 bits wide will be wired up wrongly. Jun 15 10:12:25 ah ok, that sounds serious Jun 15 10:12:31 it's quite likely that the nslu2 just doesn't have any of those. Jun 15 10:12:49 you might get away with it on some ipaqs as well, but on the h5400 for example I think you would be in trouble with the graphics controller. Jun 15 10:13:06 which ones can you think of off-hand for the ipaq ? (ok you already said graphics) Jun 15 10:13:34 sleeve, possibly. maybe the asics on some of them. Jun 15 10:14:02 I don't remember exactly what the impact is of endian switching, though, so I could well be wrong. I'd have to look it up. Jun 15 10:14:07 hmm, sleeve would be bad - I was thinking about using the h5550 for this Jun 15 10:14:30 essentially turn it into an nslu2 on steriods Jun 15 10:14:35 steroids Jun 15 10:14:43 heh Jun 15 10:14:50 but I'd need to be able to use my pcmcia ide adaptor Jun 15 10:15:05 I should break out my nslu2 again. I'd like to use it as a print server. Jun 15 10:15:20 I guess it shouldn't be too hard to make CUPS run on there. Jun 15 10:15:25 really, all I'd need to work is serial, usb (usbnet), and pcmcia Jun 15 10:15:41 right Jun 15 10:15:50 well, you could give it a go. it'd be easy enough to make a hacked haret. Jun 15 10:16:01 well, wifi would be nice too, but certainly not required Jun 15 10:16:49 pb_, I think I shall try it :-) it may take me a day or two to get setup to build haret, but I really wanna try this Jun 15 10:17:40 cool Jun 15 10:17:43 let me know how you get on Jun 15 10:17:51 is BE faster on arm? Jun 15 10:18:04 pb_, count on it :-) Jun 15 10:18:09 for applications that are naturally big endian, yes Jun 15 10:18:21 the arm itself is endian neutral Jun 15 10:25:48 ere Jun 15 10:28:12 hi hrw Jun 15 10:28:19 I'm using gcc 3.4.3 here Jun 15 10:28:40 keturn: I had it failed Jun 15 10:43:18 hi zecke Jun 15 10:52:47 bye Jun 15 11:48:07 what argument do you use with oe_libinstall? Jun 15 11:48:19 good question Jun 15 11:48:28 heh Jun 15 11:48:45 grep oe_libinstall */*bb Jun 15 11:49:02 Crofton: that is what I do Jun 15 12:01:29 http://pastebin.ca/14667 Jun 15 12:01:46 Any thoughts on how to make oe_libinstall not barf like this? Jun 15 12:02:31 03koen 07 * r1.3541 10openembedded/conf/distro/familiar-0.8.3.conf: familiar-0.8.3.conf: 0.9.0 -> 0.8.3, as keenly spotted by Erik Hovland Jun 15 12:03:04 Crofton: what oe_install command line did you use? Jun 15 12:04:09 e_libinstall -so host/lib/.libs/libusrp ${STAGING_LIBDIR} Jun 15 12:04:24 I really only want the shared lib .... Jun 15 12:04:36 you forgot -C before the path Jun 15 12:04:50 oe_libinstall -so -C host/lib/.libs libusrp ${STAGING_LIBDIR} Jun 15 12:04:55 err, -C host/lib that is Jun 15 12:05:06 what does -C do? Jun 15 12:05:14 specifies a path. Jun 15 12:05:21 without that, it looks in the current directory Jun 15 12:05:23 heh Jun 15 12:05:47 grep oe_libinstall */*.bb is not the best way to figure things out :) Jun 15 12:05:49 Thanks Jun 15 12:05:56 hm? Jun 15 12:06:04 all the examples use correct syntax. yours doesnt Jun 15 12:06:11 heh Jun 15 12:06:24 regardless, i'm pretty sure the syntax is covered in comments at hte beginning of hte script, in base.bbclass Jun 15 12:06:25 Well, I didn't understand the -C, so I left it out :) Jun 15 12:06:49 hi Jun 15 12:07:36 qpe-gaim looks like working with irc. Jun 15 12:13:11 Thanks guys, that did the trick Jun 15 12:13:25 Do we have anywhere on the wiki we can collect stuff like that? Jun 15 12:15:17 theres something of an oe users manual in the wiki Jun 15 12:15:19 belongs in that Jun 15 12:30:39 Evening. Jun 15 12:31:31 evening SirFred Jun 15 12:32:09 RP: Hi. Jun 15 12:34:11 RP: Did you had time to work in the quest for the holy w100 decoding? Jun 15 12:34:45