**** BEGIN LOGGING AT Fri Jul 07 02:59:56 2006 Jul 07 06:18:41 morning all Jul 07 06:32:41 morning Jul 07 06:35:12 hey tkp Jul 07 06:35:23 hey do13 Jul 07 06:36:13 hi Jul 07 06:36:31 do13_: which ver of wlan-ng crap is needed for 2.6/tosa? Jul 07 06:36:43 | /home/hrw/devel/build/3541/tmp/work/tosa-linux/wlan-ng-modules-0.2.3+svn20060707-r0/trunk/src/p80211/Modules.symvers: No such file or directory Jul 07 06:36:53 do13_: I got this with 0.2.3 and svn Jul 07 06:40:05 03lenehan 07org.oe.dev * rca2a1042... 10/packages/glibc/ (glibc-2.4/no-z-defs.patch glibc_2.4.bb): Jul 07 06:40:05 glibc 2.4: To build glibc for sh we need to pass no-z-defs=yes via Jul 07 06:40:05 configparams. Add a patch to do this so it builds for sh3 and sh4. Jul 07 06:41:34 If I want to enable frame buffer support, I will need to edit the kernel config for my machine (epia) Jul 07 06:41:41 yes Jul 07 06:42:00 epia_defconfig? Jul 07 06:42:13 but it says at the top: Automatically generated make config: don't edit Jul 07 06:42:25 are my changes just going to get overwritten? Jul 07 06:42:33 tkp: did you ever configured linux kernel? Jul 07 06:42:39 yes Jul 07 06:42:53 (usually though menuconfig) Jul 07 06:43:18 tkp: then let bitbake unpack/patch/configure epia kernel and then go into tmp/work/*epiakernel* and reconfigure it Jul 07 06:43:37 hi hrw Jul 07 06:43:41 tkp: then copy .config as epia_defconfig for your target Jul 07 06:43:45 hrw: which kernel version Jul 07 06:43:50 hrw: ah ok Jul 07 06:43:59 do13_: 2.6.17-r6 Jul 07 06:44:10 .dev? Jul 07 06:44:24 .oz354x with kernel from .dev and wlan-ng svn from .dev Jul 07 06:44:54 framebuffer support requires /dev/fb0 (I think) which is not on my target currently Jul 07 06:45:01 hrw: use 0.2.3 from .dev Jul 07 06:45:04 do13_: ok Jul 07 06:45:07 you know where I would need to go to add in the required dev nodes? Jul 07 06:45:44 tkp: device_table.txt Jul 07 06:45:59 hrw: .oz354x needs the module symvers patch (included in .dev) Jul 07 06:46:08 ok Jul 07 06:46:18 do13_: will push it then Jul 07 06:46:25 NOTE: package wlan-ng-modules-0.2.3-r3: task do_build: completed Jul 07 06:46:50 do13_: wlan-ng-utils also need update? Jul 07 06:47:23 hrw: probably not Jul 07 06:47:36 ok Jul 07 06:47:45 morning ppl. :) Jul 07 06:48:04 hi marcos Jul 07 06:48:11 hi hrw :) Jul 07 06:48:12 NOTE: package wlan-ng-utils-0.2.3-r2: task do_build: completed Jul 07 06:48:42 tosa/2.6 beware! I'm coming Jul 07 06:48:47 hehe Jul 07 06:49:09 opie image? Jul 07 06:49:13 yep Jul 07 06:49:32 hrw: FYI. my 1st payment got delayed. actually a lot of google payments got delayed... international money transfer issues... :( Jul 07 06:49:35 currently have gpe on tosa Jul 07 06:49:46 hrw: any ease way to use menuconfig in cross compilation mode (ie, with the correct flags for it to recognise that I'm cross compiling), or should I just edit by hand? Jul 07 06:49:46 thejapa: happens Jul 07 06:50:07 gpe with 2.4? Jul 07 06:50:21 do13_: yes Jul 07 06:50:38 tkp: ARCH=value_of_target_arch make menuconfig Jul 07 06:51:10 do13_: wlan-ng-modules build ok for 2.6.16 when module.symvers patch applied? Jul 07 07:06:00 cu Jul 07 07:06:28 hrw|gone: yep Jul 07 07:17:57 ok, I reconfigured the kernel, should I copy the new .config to /stv/trunk/oe/build/tmp/work/epia-linux/linux-epia-2.6.12-r2/defconfig, or into the package folder? Jul 07 07:22:26 tkp: i believe you should copy it into your .dev/packages/linux-epia-2.6.12-r2/defconfig, in other words, you package folder. Jul 07 07:22:45 hrw must be coming back in a few minutes anyway. :) Jul 07 07:22:47 right, and then bump the revision number to get it to rebuild Jul 07 07:30:01 tkp: if you bump PR, then the package folder is correct Jul 07 07:30:05 hey koen Jul 07 07:30:38 good morning all Jul 07 07:30:47 morning koen Jul 07 07:31:01 cool, is bumping the PR the only way to get it to rebuild from the package dir, or would -c clean linux-epia do the same? Jul 07 07:32:29 morning Jul 07 07:32:49 hey XorA Jul 07 07:36:11 tkp: -c clean does the same. But you have the same version number so ipkg can't upgrade these packages Jul 07 07:37:01 ah ok, so best to just use clean after a failure, and use PR when making modifications Jul 07 07:37:59 tkp: -c clean won't hurt after PR. :) Jul 07 07:38:47 s/PR./PR bump./ Jul 07 07:43:18 morning Jul 07 07:51:44 ~lart gtk+ Jul 07 07:51:44 * ibot turns gtk+ into a lifesized tux doll Jul 07 07:52:15 do13_: what would you say about tosa/2.6 for 3.5.4.1? Jul 07 07:53:43 hrw|work: showstoppers are ohci pm and a replacement for usbctrl Jul 07 07:54:04 ok Jul 07 07:54:48 do13_: I'm now finishing opie-image build for tosa - will build gpe-image later. Jul 07 07:55:21 do13_: both based on 2.6.17/.dev and both added into feed (directory got 711 perms so noone will see it ;) Jul 07 07:58:52 hrw|work: Image flashed? Jul 07 07:58:59 not yet Jul 07 08:09:41 NOTE: package opie-image-1.0-r20: task do_build: completed Jul 07 08:09:44 finally Jul 07 08:10:32 cool Jul 07 08:18:11 the device node I need (/dev/fb0) is in device_nodes, yet it isn't being created in my arget image Jul 07 08:18:26 what controls which nodes are actually created? Jul 07 08:19:04 tkp: iirc makedevs Jul 07 08:19:46 tkp: which image you are building? Jul 07 08:19:59 its for an epia Jul 07 08:20:18 I'm trying to add framebuffer support Jul 07 08:20:33 I enabled it in the kernel, but need to get the dev node created Jul 07 08:21:33 stupid question: what about udev? isn't it used with epia and kernel 2.6? Jul 07 08:21:55 tkp: org.oe.dev/files/ Jul 07 08:22:13 tkp: see http://www.openembedded.org/viewmtn/revision.psp?id=4059b35d988a80b56a7e6c4290a657321f196ac2 for some doc on the modified behaviour Jul 07 08:22:41 thejapa: sometimes you need the devnode before udev is launched (like /dev/console) Jul 07 08:23:09 koen: got it. Jul 07 08:24:47 do13_: NOTE: package gpe-image-1.0-r22: task do_build: completed Jul 07 08:25:39 flashing 2.6 opie... Jul 07 08:25:47 koen: just a question: why is the makedevs handling in image_ipk.bbclass the name sounds misleading Jul 07 08:27:06 hrw|work: how long does it take to on your machine to builld the gpe-image, this sounds very fast Jul 07 08:27:24 do13_: I used 3.5.4.1 build Jul 07 08:27:31 hrw|work: ok Jul 07 08:27:36 do13_: so I had to rebuild only tosa stuff Jul 07 08:28:21 yellow dragon took my tosa Jul 07 08:28:38 cool Jul 07 08:28:58 altboot need tosa love Jul 07 08:29:24 hrw|work: arh, you used 1.0.6? Jul 07 08:29:27 bootsplash too Jul 07 08:29:33 do13_: 1.0.7 Jul 07 08:29:54 do13: what's wrong with altboot 1.0.7? Jul 07 08:30:14 it starts but I did not pressed key for it Jul 07 08:30:28 will look after reboot (maybe some keypress left after altboot pass) Jul 07 08:30:58 I can't work out which file is actually being used to generate my device nodes. None of the files in org.oe.dev/files/ look like the layout I have, and nor does device_table.text in initscripts Jul 07 08:31:56 hrw|work: altboot waits for keypresses? Jul 07 08:32:23 do13_: 'press eny key to launch altboot' on tosa looks like key is pressed even if not Jul 07 08:33:21 hrw|work: ok haven't realized this :) Jul 07 08:34:37 now it died on calibration... touched all 5 points Jul 07 08:35:39 hrw|work: do a reset, I saw this 2 time Jul 07 08:40:21 do13_: booted Jul 07 08:40:35 do13_: that pcmcia 'empty socket' is jacket slot? Jul 07 08:41:01 hrw|work: yes. Jul 07 08:41:22 ~lart gaim Jul 07 08:41:23 * ibot stuffs gaim into a shiny new tin can and vacuum seals it Jul 07 08:42:20 tkp: What's in tmp/rootfs/dev ? (pastebin it) Jul 07 08:44:17 tkp: Look for makedev calls in the log: grep makedev tmp/work/bootstrap-image-1.0-r0/temp/log.do_rootfs.* Jul 07 08:46:57 ~lart kde & cups combo Jul 07 08:46:57 * ibot forces kde & cups combo to use Outlook Express Jul 07 08:49:54 ~lart automake1.4 Jul 07 08:49:54 * ibot squeezes automake1.4 till automake1.4 turns blue like papa smurf Jul 07 08:50:48 ibot: botsnack Jul 07 08:50:49 koen: aw, gee Jul 07 08:56:52 hrw|work: how is the speed compared to your other devices? Jul 07 08:58:49 do13_: will check during day - no time now Jul 07 08:58:57 ok Jul 07 09:02:22 koen: I was told to suggest you try compiling binutils Jul 07 09:02:32 liam and graeme were suggesting it in the pub Jul 07 09:04:05 hrw|work: http://venge.net/monotone/wiki/DaggyFixes Jul 07 09:04:09 hey mikearthur_ Jul 07 09:04:23 good day XorA Jul 07 09:04:47 mikearthur_: which one? -cross -cross-sdk, the one for the target? Jul 07 09:04:52 hey XorA Jul 07 09:04:56 angstrom Jul 07 09:05:22 they'd had a few pints though Jul 07 09:05:27 so it could well be rubbish :) Jul 07 09:05:30 koen: Liam had a theory that the changes in -mblah=iwmmxt might have upset the target libtool, which I think mikearthur_ has a problem building Jul 07 09:05:47 oh sux, I'm mikearthur_ Jul 07 09:05:55 or at least that what I rember Jul 07 09:06:27 not libtool binutils Jul 07 09:06:28 there we go Jul 07 09:06:37 * XorA has libtool on the brain from fixing gtk+ Jul 07 09:06:57 so I guess that means the -cross one Jul 07 09:07:39 mikearthur: did Liam make his next train :-) Jul 07 09:07:55 well, he's here and doesn't smell horrific, so I guess so Jul 07 09:09:54 v8jlene: http://rafb.net/paste/results/71lKr762.html Jul 07 09:10:22 but, that does not reflect the nodes that are in the .ext2 file Jul 07 09:12:38 v8jlene: actually... saying that, it does. Jul 07 09:12:49 tkp: ?? Well, it should... that where it's made from... grep for genext2fs in the log file and see what it's doing.. Jul 07 09:13:20 yet when I mounted the file and copied it's content's to my flash drive, there seem to be some missing Jul 07 09:13:38 tkp: ? Jul 07 09:14:14 hmm... something odd. Must have copied it wrong Jul 07 09:14:20 I have some missing, and some extras! Jul 07 09:16:55 mikearthur: NOTE: package binutils-cross-2.16.91.0.7: completed Jul 07 09:16:58 tkp: So if you mount the ext2 file locally it's ok? Like mine shows: http://pastebin.ca/81283 Jul 07 09:17:27 mikearthur: NOTE: package binutils-2.16.91.0.7: completed Jul 07 09:18:39 yes, it is Jul 07 09:18:42 koen: what is state of Angstrom at moment BTW? Jul 07 09:19:03 v8jlene: it's when I boot into the machine that it changes Jul 07 09:19:24 XorA: for some reason init hangs right after boot for the past few weeks Jul 07 09:19:49 tkp: Umm, is it that /dev/.static/dev matches but /dev doesn't? Jul 07 09:20:02 XorA: but the kernels boot, so the toolchain is doing *something* right Jul 07 09:20:37 tkp: If you using udev, then the static dev is in /dev/.static/dev after boot and /dev is udev generated. Jul 07 09:21:01 just booting into it to have a look Jul 07 09:21:35 tkp: You issues was with fb0 right? Have you booted a kernel with the fb driver in it? Udev may well have just not created it because you had no fb... I don't use fb so have no dea... Jul 07 09:21:55 03koen 07org.oe.dev * rd7d2b282... 10/conf/distro/angstrom-2006.9.conf: Angstrom 2006.9: prefer a busybox with EABI support Jul 07 09:22:03 v8jlene: yes, I added frame buffer support to my kernel Jul 07 09:22:33 and your right, it's there in .static, but not in /dev Jul 07 09:23:42 tkp: Ok, so that's something to do with udev and how it deals with fb. I have no idea on that one... Jul 07 09:23:57 v8jlene: I'm doin somee reading... thanks Jul 07 09:27:05 03xora 07org.oe.dev * rf69048cb... 10/packages/ (4 files in 4 dirs): Jul 07 09:27:05 glib-2.0_2.12.0.bb, gtk+_2.10.0.bb : rediff and add the patches to fix Jul 07 09:27:05 the hardcoded libtool path in both these packages configure.in as they Jul 07 09:27:05 are essential. Jul 07 09:27:05 Both are running on my c7x0 now so they have at least been tested. Jul 07 09:28:23 koen: I am now running gtk+ 2.10 :-D Jul 07 09:28:36 XorA: cheers Jul 07 09:28:58 XorA: now send those patches upstream ;) Jul 07 09:30:45 koen: problem is those patches are not correct in the non cross compile situation where ./libtool is what is created Jul 07 09:35:02 XorA: florian mentioned that the .bb probably needs adjusting to package the cups stuff seperately Jul 07 09:35:57 XorA: when will glib-2.0.inc gtk+2.inc happen? Jul 07 09:36:26 hey lrg Jul 07 09:36:44 hey Koen Jul 07 09:36:49 hi liam Jul 07 09:36:51 morning all Jul 07 09:36:56 hey Marcin Jul 07 09:36:57 hrw|work: when someone cleans all the old version which are not used out Jul 07 09:37:12 * koen discovered redboot supports http Jul 07 09:37:22 hrw|work: we have many versions with many different packaging differences and I dont know which are used which are not Jul 07 09:37:33 hi Liam Jul 07 09:37:59 XorA: then do it for used ones? this will allow newer versions to pickup your fixes faster... Jul 07 09:39:07 hey Graeme Jul 07 09:43:32 I have one nice fix for rootfs_ipk Jul 07 09:44:09 hi Pendalar Jul 07 09:44:26 someone here has !GNU grep as default in system? Jul 07 09:46:43 hi lrg Jul 07 09:47:42 heh Jul 07 09:47:45 lrg: I have feeling you copied the wrong patch last night Jul 07 09:47:50 hrw|work: even darwin has gnu grep :( Jul 07 09:49:44 hey pb_ Jul 07 09:49:59 XorA: I was quite intoxicated ;) Jul 07 09:50:09 koen: netbsd/freebsd/opensd rather not Jul 07 09:50:36 lrg: is "apps.shared@wolfsonmicro.com" still the right place to send applications queries? I sent a note to that address on Monday but haven't heard anything back. Jul 07 09:51:11 * lrg checks mail... Jul 07 09:51:44 lrg: yes, it is pre3 patch you gave me last night, not the one with extra debug Jul 07 09:52:33 pb_: cant see it in the system. Can you send it to apps@ instead Jul 07 09:52:47 XorA: doh Jul 07 09:52:48 woooh Jul 07 09:52:54 progress on glibc Jul 07 09:52:57 lrg: will do Jul 07 09:52:59 thanks Jul 07 09:53:07 should I be worried when I notices like: NOTE: Couldn't find shared library provider for libgcc_s.so.1, when running bitbake udev? Jul 07 09:53:09 03hrw 07org.oe.oz354x * r86f5bec0... 10/classes/rootfs_ipk.bbclass: (log message trimmed) Jul 07 09:53:09 rootfs_ipk.bbclass: show which packages are lacking - may require GNU grep Jul 07 09:53:09 before: Jul 07 09:53:09 | Collected errors: Jul 07 09:53:09 | ERROR: Cannot satisfy the following dependencies for gpe-task-base: Jul 07 09:53:10 | + '[' -f /home/hrw/devel/build/3541/tmp/rootfs/usr/lib/ipkg/info/libgpg-error0.postinst ']' Jul 07 09:53:13 after: Jul 07 09:53:37 after: Jul 07 09:53:38 | Collected errors: Jul 07 09:53:38 | ERROR: Cannot satisfy the following dependencies for gpe-task-base: Jul 07 09:53:38 | gpe-session-scripts Jul 07 09:53:38 | -- Jul 07 09:53:54 cool Jul 07 09:54:31 koen: I do not know does !GNU grep has -C option so will not push it into .dev Jul 07 09:54:40 03hrw 07org.oe.oz354x * rd393178f... 10/packages/wlan-ng/ (3 files in 2 dirs): wlan-ng-modules: added patch from .dev to get it buildable with 2.6.17 kernel Jul 07 09:55:06 hrw|work: well right now OE only supports building on linux, so pushing it would be pretty safe Jul 07 09:55:31 pb_: got it Jul 07 09:55:34 can anyone explain why gettext depends on gettext-dev which depends on gettext Jul 07 09:55:39 lrg: rock Jul 07 09:56:34 pb_: btw, is this design using i.MX21 ? Jul 07 09:57:21 ade|desk: gettext-dev got lib.so* inside? Jul 07 09:57:41 koen: .dev does not have 'images from feeds'? Jul 07 09:57:50 hrw|work: iirc not Jul 07 09:57:54 feel free to apply that Jul 07 09:58:19 lrg: nope. s3c2410, same as our other designs. Jul 07 09:58:24 koen: can you check your .dev image does it contain /var/lib/lists or /usr/lib/ipkg/lists dirs? Jul 07 09:59:52 only /usr/lib/ipkg/lists (empty) Jul 07 10:00:07 koen: ok. Jul 07 10:00:31 hi Jul 07 10:00:38 koen: .oz354x force ipkg to use /tmp/lists and remove them before creating image Jul 07 10:00:52 if i want that the symbols in my binaries gets removed Jul 07 10:00:54 how can i do that Jul 07 10:01:12 is there a class that's doing it for me? Jul 07 10:01:55 Cobelius: strip is default - you want to not strip? Jul 07 10:02:13 I want so strip, but it don't work Jul 07 10:02:25 the libraries gets installied unstriped Jul 07 10:03:24 and the modules, too Jul 07 10:03:37 nothing in my packages gets striped Jul 07 10:04:22 03hrw 07org.oe.dev * r6434189c... 10/classes/rootfs_ipk.bbclass: (log message trimmed) Jul 07 10:04:22 rootfs_ipk.bbclass: show which packages are lacking - may require GNU grep Jul 07 10:04:22 before: Jul 07 10:04:22 | Collected errors: Jul 07 10:04:22 | ERROR: Cannot satisfy the following dependencies for gpe-task-base: Jul 07 10:04:23 | + '[' -f /home/hrw/devel/build/3541/tmp/rootfs/usr/lib/ipkg/info/libgpg-error0.postinst ']' Jul 07 10:04:25 after: Jul 07 10:04:36 but all of the files are not the the standard directorys Jul 07 10:04:51 there are in /opt/softwarename/lib Jul 07 10:05:01 or /opt/softwarename/modules Jul 07 10:05:16 could this be a reason that they don't get striped? Jul 07 10:09:13 does noone have an idea? Jul 07 10:10:43 command.c: undefined reference to `XGetIMValues' Jul 07 10:10:56 means 'sorry bad x11'? Jul 07 10:11:00 hrw|work: diet-x11 needs more stubs Jul 07 10:11:10 aha Jul 07 10:11:11 hrw|work: iirc ade|desk had added some Jul 07 10:20:08 03xora 07org.oe.dev * r01588e7c... 10/packages/gnome/libgnomecups_0.2.2.bb: libgnomecups_0.2.2.bb : needs inttool to build Jul 07 10:22:47 ??? Jul 07 10:24:55 should I call strip in the installation prepend method? Jul 07 10:25:03 it would be a way to fix it Jul 07 10:25:27 or does exist an envirnment variable in which i have to add the path to files that gets striped? Jul 07 10:27:36 koen: i still need to add Jul 07 10:27:36 XUnsetICFocus Jul 07 10:27:36 XGetIMValues Jul 07 10:27:36 XSetICValues Jul 07 10:27:36 XDisplayOfIM Jul 07 10:27:37 Xutf8TextPropertyToTextList Jul 07 10:27:39 XwcLookupString Jul 07 10:27:41 XSetICFocus Jul 07 10:27:43 XVaCreateNestedList Jul 07 10:27:50 in which phase does the files normally gets striped? Jul 07 10:27:52 aah oops sorry Jul 07 10:28:05 do_compiler_append? Jul 07 10:28:10 do_install? Jul 07 10:29:26 I found a RUNSTRIP script in the work directory of busybox Jul 07 10:29:55 diet-x11 (new xorgyified) in .dev doesn't use the xim.patch i added to append to mallum's bits Jul 07 10:34:58 hmn Jul 07 10:35:10 okay i reading the source of package.bbclass Jul 07 10:35:44 what dies os.path.isabs() meaning? Jul 07 10:40:44 hi Jul 07 10:42:01 * Bernardo just made a big mistake... ctrl-c on a ipkg upgrade Jul 07 10:49:08 koen: gtk-engines needs some magic so it looks in correct dir for engines, at moment is hardcocded to 2.4.0 which fails with later gtk+ Jul 07 10:50:47 XorA: the interversion got bumped to 2.10 Jul 07 10:50:48 is there a simple guide to exporting your private key and importing for monotone ? Jul 07 10:50:58 XorA: I suspect gtkmm is broken as well with 2.10.x Jul 07 10:51:10 ade|desk: ls ~/.monotone Jul 07 10:51:44 ade|desk: note the "keys" subdir :) Jul 07 10:52:00 no such file or directory Jul 07 10:52:00 inside is your keypair Jul 07 10:52:11 koen: only if monotone >= 0.25 Jul 07 10:52:27 ade|desk: monotone 0.24? Jul 07 10:52:32 hrw|work: which we have written all over the place, so I'm assuming people are using that Jul 07 10:52:32 0.25.2 Jul 07 10:53:38 ade|desk: monotone --db yourdb.db privkey/readkey Jul 07 10:53:55 13:11 hrw@work:hrw$ monotone --db devel/oe/OE.db privkey hrw@openembedded.org Jul 07 10:56:17 doh, no keys set on this server, only on my other server Jul 07 10:56:27 ~lart me Jul 07 10:56:28 * ibot eats ade|desk and falls over dead Jul 07 10:59:28 koen: is busybox-1.2.0 working for you, it seems to try linking to host stuff not cross for me Jul 07 10:59:51 it built for me without troubles Jul 07 11:00:10 hmm ok Jul 07 11:08:05 koen: have you tried running it ? Jul 07 11:08:11 on a target Jul 07 11:08:13 not yet Jul 07 11:08:35 cos i think you maybe building it happy with your libc Jul 07 11:08:55 host libc? Jul 07 11:09:29 can you look at your busybox-1.2.0/Rules.mak for the CROSS define Jul 07 11:09:39 CROSS =$(strip $(subst ",, $(strip $(CROSS_COMPILER_PREFIX)))) Jul 07 11:10:22 and CROSS_COMPLIER_PREFIX is not set in .config file Jul 07 11:10:55 grrr Jul 07 11:11:16 busybox: ELF 32-bit LSB executable, Intel 80386 Jul 07 11:11:24 i thought so Jul 07 11:11:46 bad busybox, naughty busybox Jul 07 11:18:39 yup buildroot does a $(SED) "s,^CROSS.*,CROSS=$(TARGET_CROSS)\n\ Jul 07 11:18:39 PREFIX=$(TARGET_DIR),;" $(BUSYBOX_DIR)/Rules.mak Jul 07 11:19:06 OE need to do it too Jul 07 11:19:42 i'm surprised that previous busyboxs worked Jul 07 11:43:50 ade|desk: the next commit might interest you Jul 07 11:44:59 03koen 07org.oe.dev * ra7ccc19b... 10/packages/busybox/busybox_1.2.0.bb: busybox 1.2.0: fix horrendous bug where busybox wouldn't use the crosscompiler, spotted by the good mister Davey Jul 07 11:48:09 koen: nice one Jul 07 11:48:25 mister davey.. wow how formal Jul 07 11:48:49 koen: will test asap Jul 07 11:51:47 'file busybox' agrees with my change :) Jul 07 11:51:58 thats a good thing Jul 07 11:58:18 it built :) Jul 07 11:58:43 which it didn't before Jul 07 11:58:47 * koen hugs sed Jul 07 11:59:06 so it must be using either my host libc or more likely uclibc Jul 07 12:00:09 * ade|desk hugs koen, then stands way back belts far apart Jul 07 12:02:58 * Kristoffer screams "Its gonna blow!" and then runs for cover Jul 07 12:03:32 hrm Jul 07 12:03:44 compiling it for arm increased the size Jul 07 12:03:54 252k -> 372k Jul 07 12:04:05 ah well Jul 07 12:04:12 its a lot Jul 07 12:04:16 saves me from installing a 1mb coreutils Jul 07 12:04:21 is the defconfig the same ? Jul 07 12:04:35 no, I enabled some more "long options" to fix some bugs Jul 07 12:04:39 Whats this new busybox file? Jul 07 12:04:42 koen: add less, fsck, fuser Jul 07 12:04:57 I added fuser and fsck iirc Jul 07 12:05:12 and fuser is not needed, so thats psmisc out and gettext and so libstc++ too Jul 07 12:05:37 yep, fsck and fuser are enabled Jul 07 12:05:39 sounds a good deal to me Jul 07 12:05:39 03koen 07org.oe.dev * r94657225... 10/conf/distro/angstrom-2006.9.conf: angstrom-2006.9: use target prefix instead of hardcoding Jul 07 12:05:52 and less Jul 07 12:06:51 i still have the problems that the binaries in my packages don't gets stripped Jul 07 12:06:55 hmm not so good is that my image uncompressed is not 8.9 mb not 6.7 mb it was last year Jul 07 12:06:58 does someone have an idea? Jul 07 12:07:09 i also read the source of packages.bbclass Jul 07 12:07:14 but it seems to be okay Jul 07 12:07:48 Cobelius: is your package autofoo'ed ? Jul 07 12:08:11 'autofoo' says nothing to me Jul 07 12:08:11 and is it finding the strip tool to use ? Jul 07 12:08:15 what does it mean? Jul 07 12:08:20 autotools ? Jul 07 12:08:24 well I don't get an error message Jul 07 12:08:54 the RTAI package inherits from Autotools Jul 07 12:09:05 03koen 07org.oe.dev * rcbbb31a3... 10/packages/busybox/busybox-1.2.0/angstrom/defconfig: busybox 1.2.0: save 8k by disabling httpd in angstrom defconfig Jul 07 12:09:10 and nothing gets striped inside of it Jul 07 12:09:46 well a solution is to call strip in install_append by hand. Jul 07 12:10:04 But I think this will 'break' the the concept Jul 07 12:11:44 Cobelius: show us you local.conf Jul 07 12:11:47 ~pastebin Jul 07 12:11:53 have you built this package outside of OE?, it is likely that its defaulted to not strip or it is not defining which strip to use in the configure.in Jul 07 12:12:03 hmm... pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste, or for #oe use http://oe.pastebin.com, or http://bzflag.pastebin.ca/ Jul 07 12:13:04 Cobelius: or configure.ac even Jul 07 12:13:44 Cobelius: rdesktop-1.4.1/strip.patch is an example of another package which has this problem Jul 07 12:14:33 hmn Jul 07 12:15:04 how far i have seen it inside packages.bbclass the binaries gets stripped bevore he is doing the packages Jul 07 12:15:15 so it has nothing to do with the autotools Jul 07 12:15:34 @hrw: I have not uncommened INHIBIT_PACKAGE_STRIP Jul 07 12:15:54 most of the packages gets stripped Jul 07 12:15:54 Cobelius: you have it commented - right? Jul 07 12:16:02 but no one that I have written by my own Jul 07 12:16:07 my self Jul 07 12:16:43 Cobelius: what is the package and where is the source for it Jul 07 12:17:14 you mean the .bb file? Jul 07 12:17:50 koen: you could also strip out most of the minix crap, who ever uses minix nowadays? ext2 or jffs2 mostly Jul 07 12:18:17 I'm trying to write a new bb file for mplayer... mplayer-codecs, which will install the codecs into the correct location on the target Jul 07 12:18:36 tkp: mplayer-win32-codecs? Jul 07 12:18:44 yes Jul 07 12:18:59 I have this so far: http://rafb.net/paste/results/egjOJw99.html Jul 07 12:19:04 tkp: dont forget LICENSE field Jul 07 12:19:17 yeah! Jul 07 12:19:26 Here is the .bb file http://pastebin.de/8661 Jul 07 12:19:41 but the codecs don't end up on the target system Jul 07 12:20:01 (and I have mplayer-codecs in DEPENDS in the main mplayer.bb file) Jul 07 12:20:23 they end up in /tmp/work/i586-linux/mplayer-codecs-20060611-r0/image Jul 07 12:20:49 but I think they should be in /tmp/work/i586-linux/mplayer-codecs-20060611-r0/install instead? Jul 07 12:21:30 tkp: FILES_${PN} = "place where stuff is installed" Jul 07 12:21:56 hrw|work: right, which is set to /usr/locsl/lib/codecs Jul 07 12:22:10 locsl? Jul 07 12:22:12 but they still dont end up on the target system Jul 07 12:22:27 oops! Jul 07 12:22:39 tkp: are they added into image? Jul 07 12:23:20 yes Jul 07 12:23:31 but they are noy in rootfs Jul 07 12:23:50 are they inside of the ipk package? Jul 07 12:24:13 so is there a .ipk file in the deploy directory? Jul 07 12:25:05 no Jul 07 12:26:03 okay Jul 07 12:26:12 * koen hints at ${S} Jul 07 12:26:25 koen: was just looking at that Jul 07 12:26:28 so the package isn't be createt Jul 07 12:26:30 what is $S? Jul 07 12:26:39 the 'source' dir Jul 07 12:26:40 koen: and TFTP ? surely nc is enough Jul 07 12:26:42 when bitbake calls do_package Jul 07 12:26:55 is there an error like : "Don't creating empty archive"? Jul 07 12:27:11 I think so yes Jul 07 12:27:21 ade? Jul 07 12:27:26 Cobelius: yes empty packages are not build Jul 07 12:27:33 do you have an idea about the strip probblem? Jul 07 12:27:41 unless you allow empty packages Jul 07 12:28:01 @tkp: This was a question fot tkp, not somthing about my problem Jul 07 12:28:45 http://rafb.net/paste/results/qg5T0V61.html Jul 07 12:28:56 Cobelius: do you have the .bb, source etc available ? Jul 07 12:29:12 http://pastebin.de/8661 Jul 07 12:29:43 don't forget to set COMPATIBLE_HOST Jul 07 12:30:04 hang on. it creates an ipk when I run it standalone Jul 07 12:30:08 @tkp: unplack /data/stv/trunk/oe/build/tmp/deploy/ipk/mplayer-codecs_20060611-r0_i586.ipk at look whats inside of it Jul 07 12:30:15 but not when I run it via mplayer Jul 07 12:30:35 Have you written the Package in DEPENDS and RDEPENDS? Jul 07 12:30:47 not both... I't in DEPENDS Jul 07 12:30:56 when I add it to RDEPENDS I get errors Jul 07 12:31:06 What errors are you get then? Jul 07 12:31:16 hold on, i'll rerun Jul 07 12:32:07 koen, good way to express "if arch=sh* then arch=sh" Jul 07 12:33:08 Cobelius: you have the SRC_URI as local .. is that intended ? Jul 07 12:33:39 yes thats okay. The version of RTAI that I need is not longer available on the servers Jul 07 12:36:17 ok. 'bitbake mplayer-codecs' works, 'bitbake mplayer' works, but 'bitbake tkp-image' on do_rootfs Jul 07 12:36:33 http://rafb.net/paste/results/a3IRd445.html Jul 07 12:38:20 thats with it in both DEPENDS and RDEPENDS Jul 07 12:38:42 XorA: what's your opinion about gtk+ 2.10? Jul 07 12:39:14 koen: well it is working here, I think its worth making default and letting people actually find the bugs Jul 07 12:39:21 and tkp-image.bb has mplayer in REDEPENDS Jul 07 12:39:26 RDEPENDS* Jul 07 12:39:26 koen: it does just need that printing stuff packages Jul 07 12:39:43 tkp: it need it in IPKG_INSTALL too Jul 07 12:39:54 hrw|work: IPKG_INSTALL = "${RDEPENDS}" Jul 07 12:40:15 koen: it does have slight ABI problems coming from 2.8.X though, I had to rebuild sylpheed-claws and gpe-todo, but I had the same sort of problems from 2.6 -> 2.8 Jul 07 12:40:18 but the error: ERROR: Cannot satisfy the following dependencies for mplayer: Jul 07 12:40:20 tkp: it has mplayer - but no mplayer-codecs? Jul 07 12:40:38 Hmn you could try to make a: bitbake package-index Jul 07 12:40:39 hello Jul 07 12:40:41 no, but mplayer.bb has mplayer-codecs in DEPENDS & RDEPENDS Jul 07 12:40:51 do I need it in tkp-image.bb too? Jul 07 12:41:06 I thought adding mplayer to tkp-image would be enough Jul 07 12:41:47 Cobelius: that is one crappy build in rtai Jul 07 12:42:25 it looks like your $(STRIP) may not be exported correctly for the build Jul 07 12:42:40 GNUmakefile.in is a mess Jul 07 12:42:51 and the error says it 'Cannot satisfy the following dependencies for mplayer:', but doesn't go on to say which dependency Jul 07 12:43:24 @tkp: can you post both current versions of the .bb files? Jul 07 12:44:02 @ade: Okay but this is just one example. I created another packages at nowhere does the strip work. Jul 07 12:44:53 s/at/and/ Jul 07 12:45:46 autotools works well if the package uses autotools correctly ;) Jul 07 12:47:16 okay. so it will maybe help if i would set the STRIP variable by hand? Jul 07 12:47:23 Cobelius: http://rafb.net/paste/results/hIY7oO46.html Jul 07 12:48:28 Cobelius: as the build is using both kconfig and autotools i can't see how install-strip target is working Jul 07 12:48:48 gah Jul 07 12:48:57 Active Directory ate my account Jul 07 12:49:17 ~lart active directory Jul 07 12:49:17 * ibot puts active directory into a headlock and administers a mighty noogie, rubbing half of active directory's hair of Jul 07 12:49:28 ~lart utwente for using AD Jul 07 12:49:29 * ibot executes killall -KILL utwente for using AD Jul 07 12:49:40 ewwwww Jul 07 12:49:42 good thing I have my own mail server... Jul 07 12:49:44 Active Directory Jul 07 12:50:00 morning jkp_ ! Jul 07 12:50:00 AD ain't so bad Jul 07 12:50:08 afternoon Jul 07 12:50:12 AD sucks Jul 07 12:50:19 mind you, LDAP in general is a pita Jul 07 12:50:36 well, they had ldap working quite well Jul 07 12:50:52 but for some reason they decided to switch to exchange and AD Jul 07 12:51:00 * koen suspects some well placed bribes Jul 07 12:51:15 heh Jul 07 12:51:22 AD == LDAP with fluff anway Jul 07 12:51:28 koen: that does seem a bit strange to go from ldap to AD. Jul 07 12:51:37 virtually all the network admins are debian zealots Jul 07 12:51:53 someone took you Chancellor out for dinner? Jul 07 12:52:00 I guess so Jul 07 12:52:05 hmn Jul 07 12:52:10 I have no Idea Jul 07 12:52:11 either have one or the other. it doesn't matter which, but changing is just plain waste of time Jul 07 12:52:27 hmm. :( Jul 07 12:52:33 * tkp cries Jul 07 12:52:46 wimp Jul 07 12:52:54 :p Jul 07 12:52:55 maybe I beed to bitbake -c clean stv-image Jul 07 12:53:12 where are the deps? which package has the deps? Jul 07 12:53:48 jkp_: http://rafb.net/paste/results/hIY7oO46.html Jul 07 12:54:14 maybe I'll try with ${S} Jul 07 12:55:56 tkp: FEED_URIS_append_familiar OUT Jul 07 12:56:20 hrw|work: oh yes, thats left in from the one I copied it from Jul 07 12:56:34 export IPKG_INSTALL = "task-bootstrap" Jul 07 12:56:37 IPKG_INSTALL = "${RDEPENDS}" Jul 07 12:56:44 I see differency here... Jul 07 12:56:57 RDEPENDS = "mplayer-codecs mplayer-common" is not nice Jul 07 12:57:15 mplayer-common was already in there Jul 07 12:57:25 thats where I copied mplayer-codecs from Jul 07 12:57:26 tkp: 1. remove mplayer-codecs from (R)DEPENDS. 2. add them into RRECOMMENDS_yourdistro Jul 07 12:58:19 hrw|work: ok, makes sence Jul 07 12:58:46 tkp: OE is not for one arch/machine/distro but for unlimited amount of them Jul 07 12:59:08 tkp: your change was strictly ARCH/distro/machine related which is bad Jul 07 12:59:21 hrw|work: yes, I understand.. I'm just trying to get it working first Jul 07 12:59:43 most of these variables seem to be quite undoccumented Jul 07 12:59:45 and mplayer-codecs should have LICENSE = "totally illegal warez" Jul 07 12:59:48 koen: binutils still broken for me :( Jul 07 13:00:13 XorA: iirc they are legal if you have mswindows license Jul 07 13:00:28 yuck, who wants that ;) Jul 07 13:00:38 * hrw|work has MSIE4 cd at home Jul 07 13:00:38 hrw|work: only with a wishful thinking I would guess Jul 07 13:00:45 hrw|work: is there a special code for that licence? Jul 07 13:00:51 XorA: anyway who needs them ;) Jul 07 13:01:05 tkp: LICENSE = "totally illegal warez" sounds nice Jul 07 13:01:22 hrw|work: Im pretty sure the MS license doesnt allow you to chop bits of windows out and ship them seperately Jul 07 13:01:24 hehe... it's not illeagal if you have a licence tho Jul 07 13:02:10 if you own a ipaq you have a license right? Jul 07 13:02:11 hehe Jul 07 13:02:47 so what's the difference between 'export IPKG_INSTALL' and 'IPKG_INSTALL'? Jul 07 13:03:21 hi, does one of you know what the memory footprint of kdrive is? Jul 07 13:04:30 ArjanS: fairly small Jul 07 13:05:12 ArjanS: mainly depends on the driver for kdrive and the arch it is for Jul 07 13:05:32 kdrive Xvesa is rather good Jul 07 13:05:33 ade|desk: thanks Jul 07 13:06:35 I don't know about the kernel framebuffer version Jul 07 13:06:46 i would think it is similar Jul 07 13:11:49 ArjanS: its hard to say but my Xvesa has vmSize of 5396 Jul 07 13:21:16 what does ${D} represent? Jul 07 13:21:47 dir_which_will_Be_packaged Jul 07 13:21:57 ah ok Jul 07 13:22:37 glasses, mustache, open mouse and a goaty beard , no ? Jul 07 13:23:13 perhaps i'm not up on the young folks smilies these days Jul 07 13:23:29 hehe Jul 07 13:25:04 03hrw 07org.oe.oz354x * r03023d97... 10/packages/opie-taskbar/ (opie-taskbar.inc opie-taskbar_1.2.1.bb): opie-taskbar: select proper background for tosa Jul 07 13:27:21 cool... cleaning mplayer/-codecs and removing their ipk files seems to have sorted it out Jul 07 13:42:25 yay broke the under 8mb ram disk milestone, still 3mb pxe to the device :( Jul 07 13:44:58 heh Jul 07 13:45:11 One question, wouldnt it be a good idea to start the database from scratch again?.. Jul 07 13:45:35 ad lose history? Jul 07 13:47:19 which mpeg1/2 player is the smallest non-gui player which will just display in an X window Jul 07 13:48:50 Kristoffer: no, why should you? it'll take days and the generated snapshots contain the same content. Jul 07 13:49:21 mickey|thesis, just thought that It would clear out all the change history and thus increase speed? Jul 07 13:49:44 Kristoffer: ah globally? Of course it would increase speed, but keeping history is the point of an SCM. Jul 07 13:49:59 we are sad having lost history once when we moved from bk to monotone Jul 07 13:50:06 we don't want that again Jul 07 13:50:45 Kristoffer: but it wouldn't increase the speed of your working copy of OE which is where the real speed needs to be Jul 07 13:51:07 aha Jul 07 13:51:17 got it, thx for explaining Jul 07 13:51:25 mickey|thesis: only 6 days left! Jul 07 13:51:33 koen: til ? Jul 07 13:51:35 really? Jul 07 13:51:37 doh Jul 07 13:51:38 ade|desk: should be able to do a cutdown mplayer to fit that Jul 07 13:51:48 ade|desk: http://www.openembedded.org/wiki/OpenEmbeddedMigration Jul 07 13:52:08 ade|desk: smpeg? mpeg2? Jul 07 13:52:10 ade|desk: tried smpeg? Jul 07 13:52:28 oooh i'd forgotten smpeg Jul 07 13:52:40 koen: I will get back from vacations to get 0.27 ;) Jul 07 13:52:55 * hrw|work moved vacations - start on 20060714 instead of tomorrow Jul 07 13:53:22 hrw|work: going anywhere nice ? Jul 07 13:53:23 sounds good. where does it go? Jul 07 13:54:11 the dutch have 2 favourite destinations: 'balconia' and 'hintergarten' Jul 07 13:54:41 heh Jul 07 13:54:45 ade|desk: parents and then sea Jul 07 13:55:25 which sea? Jul 07 13:55:25 03kristoffer 07org.oe.dev * re1dfbc0a... 10/ (3 files in 2 dirs): Jul 07 13:55:25 conf/distro/preferred....opie-cvs-versions.inc : Typo fix Jul 07 13:55:25 * Fixed spelling of opie package (adress-book -> address-book) Jul 07 13:55:25 linux/linux-jlime-sh3-2.6.17/LinuxSH-2.6.17.patch : Fix localversion Jul 07 13:55:25 * Removed localversion so that it doesnt create /lib/modules/2.6.17-sh Jul 07 13:55:26 linux/linux-jlime-sh3-2.6.17/defconfig_jlime : updated defconfig Jul 07 13:55:30 * updated .config to better one Jul 07 13:55:31 ade|desk: Baltic Jul 07 13:55:34 03kristoffer 07org.oe.dev * r71d4e16e... 10/ (4 files in 2 dirs): (log message trimmed) Jul 07 13:55:34 conf/distro/jlime-donkey.conf: Updated SRCDATE for OPIE-version Jul 07 13:55:36 * Updated due to cvs errors (submitted bugreport) Jul 07 13:55:38 conf/distro/preferred-opie-cvs-versions: Fix of types and lost packages Jul 07 13:55:40 * Addition of lost packages (Opie-stumbler, Opie-examples) Jul 07 13:55:42 * Fixes of typos (address-book) Jul 07 13:55:44 Kristoffer: split your commits! Jul 07 13:55:46 linux/linux-jlime-sh3-2.6.17/defconfig_jlime: minor change Jul 07 13:55:50 03kristoffer 07org.oe.dev * rbacc445a... 10/packages/moc/moc_2.4.0.bb: Jul 07 13:55:52 packages/moc/moc_2.4.0.bb: New .bb file Jul 07 13:55:54 * MOC = Music On Consoler, is a music player using ncurses. Jul 07 13:55:59 Kristoffer: kernel change = 1st, opie change = 2nd etc.. Jul 07 13:55:59 hrw|work, sorry Jul 07 13:56:39 Kristoffer: and try to make summary in first line Jul 07 13:56:53 mickey|thesis: busybox we pushed had a big bug :) Jul 07 13:56:57 ade|desk: http://maps.google.com/maps?f=q&hl=en&q=olecko,+poland&ie=UTF8&om=1 first, then Baltic see Jul 07 13:57:44 koen: heh, read the log. funny. we _need_ that testing infrastructure of zecke Jul 07 13:57:57 it would've spotted that it doesn't contain an arm binary Jul 07 13:58:04 mickey|thesis: we need to find zecke first :) Jul 07 13:58:10 zecke is on a funeral Jul 07 13:58:16 and will be back tomorrow or sunday Jul 07 13:58:17 ah, right Jul 07 13:58:25 his grandpa Jul 07 13:58:30 mickey|thesis: ;( Jul 07 13:58:36 hrw|work, noted Jul 07 14:05:32 mickey|thesis, do13: few keys on tosa/2.6 are broken in opie: home, rotate, record does not work. left OK is cancel Jul 07 14:06:02 odd. last time i touched the keymap (1/2 year ago) they worked :/ Jul 07 14:06:11 oh 2.6 Jul 07 14:06:22 forget that then :) Jul 07 14:06:52 question is whether do13 wants to remap the keys in the driver before we start touching the keymap Jul 07 14:07:54 mickey|thesis: any plans to improve MachineStatus page in OZ wiki? Jul 07 14:08:09 D'oh Jul 07 14:08:22 * koen put his bag on the 1/0 button Jul 07 14:08:33 koen: ? Jul 07 14:08:54 hrw|work: let's say it went very quiet around here Jul 07 14:08:58 this is a stupid one....just copied bitbake elsewhere coz now we finally have thigns shaping up i want to setup our environ / repositorybetter locally Jul 07 14:09:07 ah Jul 07 14:09:18 i run bitbake and it tells me Unable to open conf/bitbake.conf Jul 07 14:09:25 hrw|work: i thought it was pretty complete. what does it need? Jul 07 14:09:27 is there an environ var to tell it where to look for that? Jul 07 14:09:44 jkp_: BBPATH Jul 07 14:10:24 Kristoffer: opie bug: bump SRCDATE by one day Jul 07 14:10:45 hrw|work, thx Jul 07 14:10:47 mickey|thesis: I was thinking about collecting info what works, what not there Jul 07 14:11:12 Kristoffer: 07 12:39 < CIA-13> mickeyl * opie/noncore/net/opietooth/applet/bluezapplet.cpp: s/Resource/OResource/ Jul 07 14:11:45 Kristoffer: basically: if opie cvs does not compile then try HEAD Jul 07 14:11:52 mickey|thesis: ta :) Jul 07 14:12:12 hrw|work, oki Jul 07 14:12:32 is there a way to make it so you dont always have to be in the build dir to run bitbake? Jul 07 14:12:54 jkp_: set TMPDIR in local.conf? Jul 07 14:13:12 hrw|work: ok :) i was going on what the docs on the wiki say Jul 07 14:15:11 jkp_: I do not remember those docs... Jul 07 14:15:28 jkp_: months passed since I read them... ;D Jul 07 14:26:21 :) Jul 07 14:28:57 * CosmicPenguin snaps Jul 07 14:29:43 CosmicPenguin: why so? Jul 07 14:31:23 initscripts has irritated me for the last time Jul 07 14:37:44 hrm Jul 07 14:38:10 bitbake is reading its bitbake.conf, but not my local.conf for some reason Jul 07 14:39:10 hi Jul 07 14:40:11 what tells bitbake where to read local.conf? Jul 07 14:41:27 morning all Jul 07 14:41:41 hey RP Jul 07 14:42:01 hey Richard Jul 07 14:42:17 lrg: How did you get on last night? Jul 07 14:42:53 RP: I think we've narrowed it down to 1 function. XorA is putting some more debug Jul 07 14:43:16 anyone? :S - ive got local.conf in the right place but it isnt being read Jul 07 14:43:38 the bitbake docs say it has no knowledge of this file and that it has to be pointed to it with an include or some such Jul 07 14:43:52 lrg: probaly next week before I get that done now, I destroyed me install on the Z and am now replacing it all Jul 07 14:44:15 * XorA is suprised the Z worked at all last night Jul 07 14:44:31 XorA: np, I'm in support hell for the forseable future :( Jul 07 14:44:44 lrg: what was that function name again? Jul 07 14:45:27 jkp_: its in a conf dir under where you're running bitbake from and that directory is in BBPATH? Jul 07 14:46:03 XorA: dapm_power_widgets. we specifically want to check each widget (by name and event type/handler) is being processed corectly Jul 07 14:46:04 yeah Jul 07 14:46:26 lrg: ok, I shall remember that now Ive had less beer Jul 07 14:46:44 * lrg had far too much Stella :( Jul 07 14:46:45 its telling em i have no files to build which cant be true if its reading the config Jul 07 14:47:02 using -D it shows its reading bitbake.conf, but not my own conf Jul 07 14:47:05 (local) Jul 07 14:47:30 jkp_: If it didn't see your local.conf, BBFILES is probably unset so that's expected Jul 07 14:47:52 lrg: another couple after I left :-) Jul 07 14:48:04 heh Jul 07 14:48:08 now i understant Jul 07 14:48:10 *understand Jul 07 14:48:19 XorA: unfortunately yes Jul 07 14:48:25 i pointed it to it's default conf, not thje OE one Jul 07 14:48:31 lrg: heh heh, I hope you stopped hacking before then Jul 07 14:52:42 RP: BBPATH should point to local.conf - BBFILES are defined in local.conf Jul 07 15:02:06 hrw|work: I didn't say otherwise ;-) Jul 07 15:02:27 okok Jul 07 15:16:58 do13, mickey|thesis: fixed keymap Jul 07 15:17:30 will push later Jul 07 15:17:57 guys - whats the deal with virtual ? im creating a stripped down oe version for our distro that we can commit to SCM .... im copying over bit by bit the packages we need Jul 07 15:18:16 now its telling me i need virtual/i586-oe-linux-gcc Jul 07 15:18:55 how does the virtual thing work? Jul 07 15:19:40 cu Jul 07 15:19:52 cya hrw|gone Jul 07 15:20:20 hrw|gone, bluetooth applet works fine now btw Jul 07 15:21:33 jkp_: The distro usually maps that to a non-virutal target Jul 07 15:22:24 yeah, i just got it@! Jul 07 15:22:27 heh :) Jul 07 15:22:34 so in this case i needed gcc Jul 07 15:28:34 RP - know a way to work out what version of the monotone repo you are at? Jul 07 15:31:34 jkp_: cat MT/revision :) Jul 07 15:34:13 You can tell why I like git ;-) Jul 07 15:40:56 im really starting to like bitbake man Jul 07 15:41:09 03coredump 07org.oe.dev * recb89dce... 10/packages/altboot/files/init.altboot: altboot: Fix Tosa/2.6 kbd detection Jul 07 15:41:16 OE seems to offer a log over buildroot in the end Jul 07 15:41:52 the *only* trouble ive got with it is it needs more documentation...but im a developer and im as guilty as anyone for not doing my part on the projects i work on ;) Jul 07 15:42:22 jkp_: OZ started in buildroot and OE was built as a replacement to do everything better/properly :) Jul 07 15:42:34 jkp_: documentation is a know issue and we're all guilty... Jul 07 15:42:41 heh :) Jul 07 15:42:55 we've done the whole host of options for our embedded system Jul 07 15:43:19 started taking some existing distros and trying to strip them back...then went to LFS, then BR, now finally OE Jul 07 15:44:00 i think we're happy here now we know how it works anyway :) Jul 07 15:50:36 v8jlene: there? Jul 07 15:52:15 v8jlene: just building a toolchain from scratch and it never build make-native first Jul 07 15:52:20 so it failed on glibc Jul 07 15:52:42 i forced it to build the later version of make and all is well...is it missing from the deps somewhere? Jul 07 16:26:52 jkp_: I suspect we assume make-native is provided... Jul 07 16:27:22 if we want to change that...where would you recommend? Jul 07 16:30:14 jkp_: You can't really use OE without make being present. You could add a sanity check to the sanity scripts to ensure the build system had the right version of make? Jul 07 16:30:45 right- its to work around build issues with the x86 toolchain on AMD64 Jul 07 16:31:31 Can you put some details about the problem to oe@ - it'd be nice to have it documented... Jul 07 16:32:41 (and the sanity checked improvement would be accepted into oe) Jul 07 16:35:57 hi all Jul 07 16:36:02 hey mr_nice_slacker Jul 07 16:36:11 hi koen Jul 07 16:37:01 RP: sure, v8jlene is the one who discovered the fixes Jul 07 16:38:09 koen: grr i have the kernel going through mounts the initrd frees kernel ram .. but no go for init :( Jul 07 16:38:16 just like you had Jul 07 16:38:23 on x86? Jul 07 16:38:28 yup Jul 07 16:38:30 ah Jul 07 16:38:33 koen: so, this is what I was thinking to overhaul the initscripts mess Jul 07 16:38:42 that's actually good news Jul 07 16:38:49 * koen stops cursing eabi Jul 07 16:39:04 koen: thats ulibc with 4.1.1 Jul 07 16:39:04 koen: for each architecture, specify a configuration file that lists the scripts that should be used, and what runlevels they should be put in Jul 07 16:39:30 koen: then, a python script could parse that file, fetch the scripts, and install them and create the symlinks Jul 07 16:39:35 all this replacing the existing do_install() stage Jul 07 16:39:56 i know this could be the false chanel, but i think that here are people who have many different pda. the most pda have a joypad - do you know if it is pressable in the middle on the most moddels? and if it is is it a 5th key or just all 4 keys pressed? Jul 07 16:39:58 CosmicPenguin: sounds ok to me Jul 07 16:40:12 mr_nice_slacker: yes and Jul 07 16:40:15 time to write some python Jul 07 16:40:21 ahhh - python, my old enemy Jul 07 16:40:39 anyway to get -c clean to remove related ipk files too? Jul 07 16:40:46 ade|desk: http://uclibc.org/lists/uclibc/2006-July/015804.html ? Jul 07 16:40:53 koen:thx Jul 07 16:42:09 oh i forgot to ask the jopads are most impemented as keyboards or? Jul 07 16:43:06 so kdb is the handler Jul 07 16:44:55 * koen hands rmk a jack-jack plug Jul 07 17:09:39 koen: segfaults :( Jul 07 17:12:09 ade|desk: right Jul 07 17:12:21 ade|desk: like the ones in the link I pasted? Jul 07 17:12:34 still reading Jul 07 17:12:46 something with binutils Jul 07 17:13:29 at least the static binary from debian works to i know the kernel that is produced is fine Jul 07 17:35:19 OE++fcin+ Jul 07 17:37:25 koen: doing an exact buildroot version and will see if it is a binutils thing Jul 07 17:44:27 right home time , cya folks Jul 07 18:36:52 hi Jul 07 18:37:20 ~seen do13 Jul 07 18:37:31 do13 is currently on #oe, last said: 'RP: ok'. Jul 07 18:39:38 dirk: i updated keymap for tosa/26 - see http://ewi546.ewi.utwente.nl/tmp/htw/t/keymap-2.6.map Jul 07 18:39:40 hey hrw|gone Jul 07 18:40:35 irc from tosa isnot so good as from c760.. Jul 07 18:45:41 do13: touchscreen is weird.. Jul 07 18:49:36 those yellow lines does not looks goood too Jul 07 18:50:14 http://handhelds.org/scap/port.21164.png Jul 07 18:50:58 thats TS madness Jul 07 18:55:17 hi koen Jul 07 18:55:50 hey Hrw|tosa Jul 07 18:57:39 bt usb worls under 2.6 Jul 07 18:58:13 I know Jul 07 18:58:26 I used it on my c700 and h2200 a while back Jul 07 18:58:32 and on my slug Jul 07 18:59:04 its one of my usb toys Jul 07 18:59:42 usbaudi has some problems on arm and with usb hubs Jul 07 18:59:46 audio* Jul 07 19:00:21 audio is something which i will test in august Jul 07 19:00:52 you can buy a usb sounds card for $0.99 in hongkong Jul 07 19:01:00 home delivered for <$10 Jul 07 19:01:08 eFfeM-away demod it on his slug Jul 07 19:01:17 o.. addy? Jul 07 19:01:34 some ebay shop Jul 07 19:03:19 and last test.. 4slot card reader Jul 07 19:04:50 you have multiple lun support in your kernel? Jul 07 19:05:05 looks like not Jul 07 19:05:15 will have to add it Jul 07 19:06:19 anyway reader say that want 90mA but no power led on it.. recognised anyway Jul 07 19:09:59 my card readers refuses to recognize my 4gb cf card Jul 07 19:10:07 yay for a camera with usb2 Jul 07 19:11:11 * koen takes a look at http://ewi546.ewi.utwente.nl/rrd/graph/traffic_eth0_604800.png Jul 07 19:13:50 Hrw|tosa: how's the automirror thing working out? Jul 07 19:14:43 feeds.hentges.net redirect betweeen but i didnot enabled that in images...... Jul 07 19:15:26 do13: backlight jumps when battery on 50% Jul 07 19:15:42 tosa/2.6 is not ready for users yet Jul 07 19:16:08 not so long ago c7x0/2.6 wasn't ready for users :) Jul 07 19:16:22 yep Jul 07 19:17:53 I bet getting out images to test would help a bit Jul 07 19:18:21 did someone add those apm scripts yet? Jul 07 19:18:49 i run 3541 now on tosa - will discusssss with dirk about releasing them as 3542-alpha1 Jul 07 19:18:59 for usb? not iirc Jul 07 19:19:37 45 pln for usb audio Jul 07 19:20:17 ~change 45 pln to eur Jul 07 19:20:22 45.00 Zloty (PLN) makes 11.0385 Euro (EUR) (from http://www.xe.com/) Jul 07 19:22:33 will buy one maybe when/if will have cxx00 Jul 07 19:24:19 jezz.. that usb cards are smaal Jul 07 19:24:41 about 3 sd cards glued together Jul 07 19:25:01 http://allegro.pl/item112972001_super_karta_muzyczna_usb_2_0_z_mic_fv_gw_od_ss.html Jul 07 19:25:43 koen: chip and minimal glue Jul 07 19:27:01 : Jul 07 19:27:04 :) Jul 07 19:28:07 heh Jul 07 19:28:35 Hrw|tosa: the post to oz-users 10 minutes ago is funny Jul 07 19:29:01 it was sent when we were discussing tosa/2.6 :) Jul 07 19:29:28 time for opiemail then Jul 07 19:30:59 Hrw|tosa: the usb audio dongle I used is http://cgi.ebay.nl/USB-KEY-SIZE-5-1-CHANNEL-AUDIO-SOUND-CARD-ADAPTOR-A07_W0QQitemZ280004500003QQihZ018QQcategoryZ74957QQcmdZViewItem Jul 07 19:31:10 koen actually it was < eur 10 Jul 07 19:31:40 ah yes Jul 07 19:31:50 stupid elcheapo dollar Jul 07 19:32:39 koen: will reply Jul 07 19:34:17 Hrw|tosa: touchscreen needs to be synced with the lcd controller. Jul 07 19:35:31 do13: need ttto be written it is? Jul 07 19:38:32 Hrw|tosa: lineo had a ugly loop to sync the touch, I wanted to use the hardware feature of the controller but I have some probs with the sync input. Jul 07 19:40:31 hey lrg Jul 07 19:40:43 hey koen Jul 07 19:41:55 interesting device at http://linuxdevices.com/articles/AT6071673793.html Jul 07 19:43:17 lrg: compulabs is busy pushing their stuff upstream Jul 07 19:43:29 which yo probably know since you read l-a-k as well :) Jul 07 19:44:02 "However, Echo's wm8731-based audio subsystem is supported through the Alsa project's ASoC layer, Stewart said. " Jul 07 19:44:18 * koen reads "liamware" in that Jul 07 19:44:26 koen: heh Jul 07 19:46:58 lrg: too bad the compulab boards are so expensive Jul 07 19:47:03 hey Liam Jul 07 19:47:16 re Jul 07 19:47:17 hey dirk Jul 07 19:47:22 hey Marcin Jul 07 19:47:36 do13: does 0% of battery is normal? Jul 07 19:47:55 Hrw|tosa2: not really :) Jul 07 19:48:03 Off-line, battery status high: 0% Jul 07 19:48:29 Hrw|tosa2: can you do another suspend/resume cyle Jul 07 19:48:50 can. will have to disconnnect Jul 07 19:49:16 ok Jul 07 19:50:53 lrg: nice device Jul 07 19:51:21 do13: I hope I get one for testing... Jul 07 19:52:45 re Jul 07 19:53:00 dirk: after suspend it does not resume Jul 07 19:53:10 now tosa is on ac and booting Jul 07 19:53:50 ping? Jul 07 19:54:00 lrg: do you know any afforable arm boards with 128MB (or more) ram? Jul 07 19:54:16 koen: gumstix ? Jul 07 19:55:14 koen: there is a compulab board mentioned in the article - not sure about the ram though Jul 07 19:55:42 lrg: it's >$300 in single quantities Jul 07 19:55:45 hrw|husky: join #openzaurus for a second please Jul 07 19:55:53 :( Jul 07 19:56:26 lrg: and the gumstix only have 64MB Jul 07 19:57:35 koen: i think I've got a spare netbook board, iirc it's got 256MB Jul 07 19:57:44 oooh Jul 07 19:57:54 koen: I could send it to you. Jul 07 19:57:57 that would make a nice native compilation environment Jul 07 19:58:24 it's also got CF, MMC, PCMCIA Jul 07 19:58:48 yeah, psion promised to send me a compleet nb pro for OE support Jul 07 19:58:52 do13: loaded usbhost, got wifi. rmmod host, suspend == no resume.. known bug? Jul 07 19:59:01 about 2 years ago... Jul 07 19:59:11 koen: yeah,it was cancelled :( Jul 07 19:59:36 can it run a recent kernel? Jul 07 19:59:40 hrw|gone: Hmm. What's the state of the display? Jul 07 19:59:58 koen: yep Jul 07 20:00:02 ah cool Jul 07 20:00:13 do13: black. on power press green(alrm) led blink twice Jul 07 20:00:18 lrg: could you send me a mail so we could work out the details? Jul 07 20:00:39 koen: sure, I'll dig it out tomorrow night Jul 07 20:00:41 koen@dominion.kabel.utwente.nl Jul 07 20:00:46 cool, thanks Jul 07 20:00:48 hrw|husky: strange Jul 07 20:00:51 do13: cannot connect serial cable now Jul 07 20:01:31 * hrw|husky need nullmodem with tcx3106 on both ends:) Jul 07 20:01:50 lrg: The Trinity DAW has only inputs. Do you know such a device with inputs and outputs Jul 07 20:02:38 do13: iirc it's also got a headphone jack Jul 07 20:03:16 do13: maybe iit has something common with that 0% battery? I connected it for~2minutes to ac and after disconnect it shown 50% Jul 07 20:03:27 hi schurig Jul 07 20:03:44 hrw|husky: hi! Jul 07 20:03:49 lrg: no headphone jack. I'am looking for xlr connectors Jul 07 20:03:53 schurig: look at recent driver_hermes change in OE Jul 07 20:04:10 tvtime Jul 07 20:04:11 schurig: we removed need for hostap headers Jul 07 20:04:12 hrw|husky: did you adapt this to kernel 2.6.17 ? :-) Jul 07 20:04:15 later all Jul 07 20:04:44 schurig: i run 17 only on tosa. and it has shit called wlan-ng Jul 07 20:04:57 hrw|husky: does is resume after connecting the power Jul 07 20:05:11 do13: moment... Jul 07 20:06:22 do13:nope. Jul 07 20:06:36 guys Jul 07 20:06:45 does anyone know about the syslinux packages? Jul 07 20:06:47 do13: on power press alarmled blink. on release it blink again Jul 07 20:07:02 jkp_: what do you mean? Jul 07 20:07:13 i found a doc online from amd that detailed using syslinux and OE together, and lo, they are the ones who created the packages Jul 07 20:07:31 CosmicPenguin: endgame for us is to produce a .img file of the entire system Jul 07 20:07:52 stupid amd Jul 07 20:07:55 there's an AMD util called imgloader that can be used to image up machines from .img files Jul 07 20:07:55 always getting in our way Jul 07 20:08:00 CosmicPenguin: ? Jul 07 20:08:01 curse those handsome fellows Jul 07 20:08:08 heheh Jul 07 20:08:14 so, what do you know then? Jul 07 20:08:17 hrw|husky: This is for debugging purpose. The led blinks on suspend and resume. I suspect the battery is drained. Can you charge the device for a few mins? Jul 07 20:08:54 do13:it is charging Jul 07 20:09:03 CosmicPenguin: http://tree.celinuxforum.org/CelfPubWiki/ELC2006Presentations?action=AttachFile&do=get&target=jordan_crouse_celf_2006.pdf Jul 07 20:09:45 cosmic: heh those amd guys... oe lamers :) Jul 07 20:09:51 CosmicPenguin: do you know how the syslinux package works? Jul 07 20:11:24 actually all we want is a .bb package that can generate a filesystem with a bootloader on it Jul 07 20:11:38 whether that be grub, syslinux, or lilo i dont mind Jul 07 20:11:50 jkp_: actually, I can help - thats my presentation Jul 07 20:11:57 hahahahaha Jul 07 20:12:00 excellent! Jul 07 20:12:10 CosmicPenguin: nice one! Jul 07 20:12:10 grub and lilo won't work for you - the problem is that you need the actual media on hand to install them Jul 07 20:12:21 CosmicPenguin: right Jul 07 20:12:23 syslinux is the only bootloader that you can install without knowledge of the actual media Jul 07 20:12:29 and it was a good presentation Jul 07 20:12:35 can you talk me through how to set up what we want? Jul 07 20:12:46 we want to produce a finished image just like you documented Jul 07 20:12:55 and be able to load that on with imgloader Jul 07 20:13:30 i wasnt too clear on whether you can reimage running systems either, im hoping so coz that takes care of "firmare" updates Jul 07 20:13:50 The first thing you'll want ito examine is bootimg.bbclass Jul 07 20:13:59 nice :) Jul 07 20:14:05 i knew i was missing a link :p Jul 07 20:14:18 thats the code that takes the kernel and syslinux and combines them together into a vfat filesystem Jul 07 20:15:25 then, you can create a task of your own that will use the mkimg utility to combine the vfat partition and your rootfs partition together Jul 07 20:15:34 right....so thats our boot partition Jul 07 20:15:39 ok - what about MBR? Jul 07 20:15:57 imgloader uses the syslinux MBR by default Jul 07 20:16:10 excellent - did you have a hand in that code? Jul 07 20:16:16 I wrote it Jul 07 20:16:21 you da man! Jul 07 20:16:22 :p Jul 07 20:16:25 well, not the syslinux MBR Jul 07 20:16:26 but the rest Jul 07 20:17:16 is there an example of bootimage.bbclass in use? Jul 07 20:17:40 hmmm... I don't think so Jul 07 20:18:07 let me pastebin one Jul 07 20:18:10 i ont see any deps on the kernel in there Jul 07 20:18:22 cool :) man - you've saved us messing around for hours dude Jul 07 20:19:39 CosmicPenguin: maybe put it on the wiki instead of pastebin? this conversation/information has broader value, right? Jul 07 20:20:09 fair enough Jul 07 20:20:21 I just lost web access, so hold on while the proxy hamster takes a break Jul 07 20:20:26 heh Jul 07 20:21:10 sounds like you need more rodentia in your infrastructure. (i never thought i'd see myself type that) Jul 07 20:21:12 zwelch: good plan btw Jul 07 20:22:19 jkp_: yeah, well, call it directed self-interest ;) i was watching this conversation with interest, noting the time and date so i could go back to my logs and find the details Jul 07 20:22:40 heh - its good to get this stuff in the wiki for sure though Jul 07 20:23:02 hey zwelch, how ya beeN? Jul 07 20:23:17 kergoth: heyya. working hard and .... yeah, that's about it Jul 07 20:23:24 hehe Jul 07 20:23:28 how 'bout yerself? Jul 07 20:23:59 moved to arizona for a government/dod type job, then quit for an opportunity with openedhand before i even got out of the extended stay hotel :P Jul 07 20:24:20 timing is everything Jul 07 20:26:12 are you getting out of AZ at least? Jul 07 20:26:16 i hope that the opportunity has proven worthwhile Jul 07 20:26:19 re Jul 07 20:29:22 going to stay in az for another 6 months to a year.. its sunny here, i dont mind it Jul 07 20:29:25 will see after that Jul 07 20:29:50 !!tosa_tc6393_enable!! tosa_ac97_init sharpsl-pm sharpsl-pm: Fatal condition. Suspend. !!tosa_tc6393_disable!! tosa_ac97_exit Jul 07 20:29:51 Hrw|tosa2: I have no frelling idea! Jul 07 20:30:09 do13: it was discharged too much Jul 07 20:30:30 Hrw|tosa2: yep -> Fatal condition Jul 07 20:30:54 do13: is this ac97 related ? Jul 07 20:30:58 do13: from 50 to 0 without steps Jul 07 20:31:11 lrg: yep Jul 07 20:31:17 * lrg hides Jul 07 20:31:52 Hrw|tosa2: this sounds strange. I'll check the battery reading code Jul 07 20:32:04 do13: is the wm9712 used for battery monitoring via aux adc ? Jul 07 20:32:16 lrg: yes Jul 07 20:32:22 ok Jul 07 20:33:04 do13: so we are reading back vbat and the basing the remaining charge on this ? Jul 07 20:33:27 lrg: correct Jul 07 20:34:45 lrg: lineo implemented a table with adc values versus charging level in percent Jul 07 20:35:05 do13: I'll speak with someone that knows the details of this on monday. We may have to monitor the speed of vbat change versus time. Jul 07 20:35:48 looks like pastebin is acting up Jul 07 20:36:22 do13: I'll try and speak with the chip designer for his input as well. Can you email me the table. Jul 07 20:36:37 lrg: thx. Afaik this is needed for the charging control. But the charging is handled via an external chip. Jul 07 20:36:56 jkp_: oh, well - bootimg.bbclass is pretty easy to use - just make sure there is a file called bzImage in ${STAGING_KERNEL_DIR) and inherit the class in your meta file Jul 07 20:37:06 lrg: np. Jul 07 20:37:33 jkp_: you can have it automatically create a syslinux config file with these lines: Jul 07 20:37:38 AUTO_SYSLINUXCFG="1" Jul 07 20:37:41 lrg: its good to have you :) Jul 07 20:37:49 LABELS="default debug" Jul 07 20:37:53 APPEND_default=" " Jul 07 20:37:53 APPEND_debug="kgdbwait " Jul 07 20:37:59 or whatever Jul 07 20:38:09 right...ill give it a go in a mo :) Jul 07 20:43:22 do13: did you saw my modified keymap? Jul 07 20:44:13 Hrw|tosa2: thx. I found your comment, but haven't looked into it. Jul 07 20:45:25 do13: mapped leftOk to control and mapped home, rotate,record to Fkeys which opie want Jul 07 20:45:41 Hrw|tosa2: ok Jul 07 20:46:22 will push it tomorrow if time allow Jul 07 20:47:11 lrg: mail send Jul 07 20:47:21 cu Jul 07 20:48:05 do13: thx Jul 07 21:03:05 is anyone still using python 2.3 in OE? python-pyro has a hardcoded path for 2.3 :-\ Jul 07 21:13:23 03eFfeM 07org.oe.dev * r9e05b81f... 10/packages/meta/slugos-native.bb: Jul 07 21:13:23 slugos-native: added libc6-dev to DEPENDS list, Jul 07 21:13:23 otherwise crt1.o and friends are missing Jul 07 21:13:38 :q Jul 07 21:13:42 oops :) Jul 07 21:17:17 03cbrake 07org.oe.dev * rc6be6cd0... 10/packages/python/python-pyro_3.4.bb: fix sourceforge cvs URL Jul 07 21:17:21 03cbrake 07org.oe.dev * r0f2321c6... 10/packages/python/python-pyro/pyro-unattended-install.patch: Jul 07 21:17:21 python-pyro 3.4: update hardcoded path to python 2.4 Jul 07 21:17:21 this really should be fixed, but I think time is better Jul 07 21:17:21 spent getting pyro 3.5 integrated Jul 07 21:17:33 CosmicPenguin: am i right in thinking that if we installed imgloader on our end system we should be able to use tha to reimage the system from a remote system and reboot? Jul 07 21:18:27 jkp_: possibly Jul 07 21:18:39 it really depends on what you change Jul 07 21:19:13 if you just change the bits without resizing the partition, then imgloader won't gain you anything over just dd'ing the image onto the partition directoly Jul 07 21:19:14 want to reimage the entre system, ie - firwmare update Jul 07 21:19:15 directly even Jul 07 21:19:25 imgloader really only comes into play when you have the multiple partitions to deal with Jul 07 21:19:46 it would be interesting to have an imgloader like app that only wrote particular partitions Jul 07 21:19:48 yeah,but its handy to have one mechanism that handles everything in some ways Jul 07 21:19:50 a command line app maybe Jul 07 21:20:02 aye Jul 07 21:20:35 CosmicPenguin: i think you might have missed something in your explanation of the class Jul 07 21:20:53 it ran through but didnt create any files it seems Jul 07 21:21:38 no? Jul 07 21:21:48 it will end up in DEPLOY_DIR Jul 07 21:21:51 i have a kernel in the staging dir Jul 07 21:22:11 ive only got ipk there Jul 07 21:23:14 http://paste.lisp.org/display/22231 Jul 07 21:24:40 dont i need to actually call the routine do_bootimg from somewhere? Jul 07 21:25:02 no Jul 07 21:25:14 it will be inherited and run when you build the package Jul 07 21:25:24 hrm Jul 07 21:25:32 welll....ill paste the log of a run Jul 07 21:26:33 http://paste.lisp.org/display/22231#1 Jul 07 21:28:55 you're not bulding a rootfs Jul 07 21:29:38 i thought that was a seperate job Jul 07 21:29:48 after i get the bootimage bit done Jul 07 21:29:56 I typically combine them together Jul 07 21:30:00 since all the variables are right there Jul 07 21:30:10 but this should be doing something anyway right? Jul 07 21:30:31 but you could do them seperately, I guess Jul 07 21:30:39 You'll need to set PACKAGES="" Jul 07 21:30:44 to avoid that stage Jul 07 21:30:54 * jkp_ tries Jul 07 21:31:36 nada ainda Jul 07 21:31:41 still nowt :/ Jul 07 21:33:53 wierd Jul 07 21:34:02 im reading the code and it looks like it ought to just run Jul 07 21:35:13 hmm... Jul 07 21:35:25 Mabye its because the task list is confused Jul 07 21:35:32 I've only ever run it after rootfs Jul 07 21:36:19 You can try to change the last line to something like addtask bootimg before do_build after do_install Jul 07 21:36:35 ok Jul 07 21:36:42 IMAGE_ROOTFS seems to depend on rootfs Jul 07 21:37:53 Oh, that might be Jul 07 21:38:08 feel free to hack away Jul 07 21:38:18 yeah, that didnt work ... Jul 07 21:38:58 hey Jul 07 21:39:14 yo Jul 07 21:39:29 * RP returns Jul 07 21:39:33 koen better hurry up with the staging from packages stuff... :) Jul 07 21:39:34 hi zecke Jul 07 21:39:48 I'm tired of having various X proto packages writing all over each other Jul 07 21:39:52 * RP is awaiting that with interest too :) Jul 07 21:40:27 sorry to be dumb Jul 07 21:40:31 how do i do rootfs? Jul 07 21:40:39 its not a package...is it a class? Jul 07 21:40:56 aye Jul 07 21:40:56 inherit rootfs_ipk in your meta file Jul 07 21:40:58 rootfs_ipk Jul 07 21:41:11 RP: having some problems with the backlight on my spitz. When I wake it up sometimes I can't get the bl to come back. I don't know if it's zaurusd or bl or the kernel.... Jul 07 21:41:38 JustinP: Can you get in via other means when its in that state to debug? Jul 07 21:41:57 jkp_: Read a *-image.bb target like bootstrap-image to see how its done Jul 07 21:42:33 RP, cool thanks Jul 07 21:44:06 *home* Jul 07 21:44:10 ljp: ping Jul 07 21:44:44 NOTE: package cairo-1.0.4: started Jul 07 21:44:45 yay Jul 07 22:03:55 hrmmm Jul 07 22:04:06 nothing is being built into ipkgs on our system :/ Jul 07 22:04:22 distro.conf has INHERIT += "debian multimachine" Jul 07 22:04:28 and INHERIT += "package_ipk" Jul 07 22:06:45 hmmm - ok, now its building ipks coz i got ipk installed Jul 07 22:06:53 i wonder why it never build it on the way before Jul 07 22:14:44 ~seen RP Jul 07 22:14:46 rp is currently on #gpe #oe #openzaurus #handhelds.org. Has said a total of 179 messages. Is idling for 32m 48s, last said: 'jkp_: Read a *-image.bb target like bootstrap-image to see how its done'. Jul 07 22:16:17 RP: Am I correct to asume that the platform specific stuff for alsa-soc on corgi is handled in corgi.c? Jul 07 22:19:20 RP: yeah, it's still working. My music will still play and I'm sure I can ssh in if my wireless is on. I haven't yet had it happen at home so I could ssh in, though. Jul 07 22:34:12 JustinP: If it happens again and you can ssh in, have a look in sysfs and write down the status of the backlight properties Jul 07 22:35:36 hvontres|work: The machine specific stuff, yes Jul 07 22:37:39 jkp_: It should build ipkg itself Jul 07 22:37:51 wierd Jul 07 22:37:54 yeah, it didnt Jul 07 23:21:24 nite guys Jul 07 23:21:47 heh, koen's news item on oe.org Jul 07 23:21:53 we found out about that issue years ago Jul 07 23:22:06 yeah, others didn't :) Jul 07 23:22:40 "Read the license(s), bastard" would be the right topic :} Jul 07 23:22:46 hehe Jul 07 23:23:06 adding licenses to the .bb's... ah, memories Jul 07 23:23:13 What would slashdot be if they couldn't rile up the masses occassionally Jul 07 23:24:09 now he left :( Jul 07 23:24:13 tomorrow pals Jul 07 23:26:34 annoying Jul 07 23:26:36 I hit ctrl alt backspace on the wrong server... :) Jul 07 23:26:39 hah Jul 07 23:26:43 hate it when that happens Jul 07 23:32:49 kergoth :( Jul 07 23:32:52 could be worse. Jul 07 23:32:59 i hit ctrl alt del on the wrong server, last week **** ENDING LOGGING AT Sat Jul 08 02:59:59 2006