**** BEGIN LOGGING AT Mon Feb 12 02:59:57 2007 Feb 12 02:59:58 03Laibsch 07org.oe.documentation * r1f403efa... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: fixed and removed net-misc and net/misc sections Feb 12 03:31:08 03Laibsch 07org.oe.dev * r31e4af6f... 10/ (7 files in 4 dirs): kakasi: add kakasi, the "kanji-kana simple converter". It converts Kanji characters to Hiragana, Katakana or Romaji. Feb 12 03:31:14 03Laibsch 07org.oe.dev * rc5a0ce7d... 10/ (10 files in 6 dirs): divers: fix section "x11/apps" as per http://www.openembedded.org/user-manual&dpage=section_variable Feb 12 03:31:21 03Laibsch 07org.oe.dev * r14d71b93... 10/ (5 files in 5 dirs): divers: fix sections "net/misc" and "net-misc" as per http://www.openembedded.org/user-manual&dpage=section_variable Feb 12 03:40:11 03tmbinc 07org.oe.dreambox * r82b58911... 10/ (3 files in 3 dirs): tuxbox-plugins: fix buggy include, as upstream seems ignorant about this Feb 12 03:40:14 03tmbinc 07org.oe.dreambox * ra2978bdd... 10/ (1 packages/tuxbox/tuxbox-plugins.bb): tuxbox-plugins: fix buggy include, as upstream seems ignorant about this Feb 12 03:40:19 03tmbinc 07org.oe.dreambox * rb400f3b1... 10/ (1 packages/enigma/enigma_cvs.bb): enigma: update to 20070206 Feb 12 05:58:43 would anybody be willing to take a look at this question: (http://rafb.net/p/9OGZ0n81.txt), I make no claims about how well informed it is, but any attempts to answer/corrections/etc. would be appreciated Feb 12 06:42:14 Laibsch: someone delete the mainpage Feb 12 06:42:34 Laibsch: where 'someone' is host-196.218.19.207.tedata.net Feb 12 06:47:39 koen: just pushed the big slugos/ixp4xx rationalisation commit. Feb 12 06:47:59 I just read uour mail on gmane Feb 12 06:48:11 I suspect everyone is avoind uclibc till .29 gets released Feb 12 06:48:17 (at least I am) Feb 12 06:52:32 yeah Feb 12 06:58:04 had to draft a different email for oe-devel - just sent that now Feb 12 07:02:58 morning Feb 12 07:03:31 * hrw wants to mention that we have 20070212.. Feb 12 07:04:33 hey hrw Feb 12 07:04:43 hrw: you mean 401 on various sites? Feb 12 07:04:48 yes Feb 12 07:06:29 koen: https://wiki.openmoko.org/wiki/Disassembling_Neo1973 https://wiki.openmoko.org/wiki/Debug_Board https://wiki.openmoko.org/wiki/Debug_Board_v1 Feb 12 07:07:14 all 3 401 Feb 12 07:07:14 this wiki had to be open ;( Feb 12 07:07:38 koen: I know. Feb 12 07:07:49 morning all Feb 12 07:08:00 koen: I cant tell where I got those links but its annoying to see just links. Feb 12 07:08:25 hey gremlin[it] Feb 12 07:08:30 hi hrw ... can u give me the site where linux-rp is ?? so i chech the cvs and the sa1100 gadget ? Feb 12 07:08:38 hi koen Feb 12 07:09:36 gremlin[it]: moment Feb 12 07:10:17 http://www.do13.de/openzaurus/patches/collie/sa1100-udc-r1.patch Feb 12 07:13:32 mhh ok the r0 didn't work i'll try the r1 ... Feb 12 07:14:10 gremlin[it]: contact Dirk Opfer (do13) - this is his patches repo Feb 12 07:14:54 dok Feb 12 07:14:56 ok Feb 12 07:15:22 Installing xdm (X11R7.1-1.0.4-r0) to root... Feb 12 07:15:41 Package gnome-vfs-dbus wants to install file /usr/bin/gnomevfs-cat But that file is already provided by package gnome-vfs Feb 12 07:32:05 someone know WHY gpe-mini-browser SVN is NOT packaged? Feb 12 07:33:44 hm.. rebuild helped Feb 12 07:34:56 another question ... long time ago i remember i downloaded the data sheet of the ASIC (i don't remember if 1 2 or 3) ... but i'm unable to found again ;( Feb 12 07:46:45 fsck! Feb 12 07:46:54 do we have WORKING x11 login manager? Feb 12 07:47:28 gpe-login Feb 12 07:47:42 or rather gpe-dm + gpe-login Feb 12 07:47:42 gpe-login or gpe-dm? Feb 12 07:47:48 gpe-dm dies here Feb 12 07:48:06 in that case you're out of luck Feb 12 07:48:16 did I push that gdm .bb yet? Feb 12 07:48:25 gdm? on 112M device? Feb 12 07:48:50 yes Feb 12 07:49:06 I run it on my 128MB efika Feb 12 07:49:11 I tried it - hit swap under debian Feb 12 07:49:33 in shell I have 60M free now Feb 12 07:49:50 and progear has shitty vesafb which does not want to show getty Feb 12 07:53:22 * rwhitby syncs the ixp4xx/slugos changes Feb 12 07:54:06 03rwhitby 07org.oe.dev * rae2b5cf2... 10/ (1 packages/madwifi/madwifi-ng_r.inc): madwifi: made the ixp4xx TARGET override more generic Feb 12 07:54:09 03rwhitby 07org.oe.dev * rd9b0713e... 10/ (1 packages/busybox/busybox.inc): busybox: made the slugos override more generic Feb 12 07:54:13 03rwhitby 07org.oe.dev * re83f4141... 10/ (1 conf/distro/include/angstrom.inc): angstrom.inc: made the nslu2 distro features override more generic. Feb 12 07:54:17 03rwhitby 07org.oe.dev * rdb8b961c... 10/ (6 files in 3 dirs): arm-kernel-shim: Now does the endianness switch prepend code as well. Tested on slugosbe and slugosle. Feb 12 07:54:21 03rwhitby 07org.oe.dev * reb9139a6... 10/ (3 files in 3 dirs): ixp4xx-kernel: Enabled EABI compatibilty mode Feb 12 07:54:25 03rwhitby 07org.oe.dev * rd43c215c... 10/ (3 files in 3 dirs): ixp4xx-kernel: Modified to use arm-kernel-shim instead of hard-coded devio incantations Feb 12 07:54:29 03rwhitby 07org.oe.dev * rcf6150fa... 10/ (4 files in 2 dirs): ixp4xx-kernel: Modified to use arm-kernel-shim instead of hard-coded devio incantations Feb 12 07:54:34 03rwhitby 07org.oe.dev * r3e37a72b... 10/ (5 files in 3 dirs): ixp4xx.conf: Deprecated in favour of ixp4xxle.conf and ixp4xxbe.conf, and moved the OVERRIDES and MACHINE_ARCH stuff into nslu2.conf machine files. Feb 12 07:54:37 03rwhitby 07org.oe.dev * rc20ea006... 10/ (4 files in 2 dirs): nslu2.conf: Deprecated in favour of nslu2le.conf and nslu2be.conf. Feb 12 07:54:43 03rwhitby 07org.oe.dev * r879c2e35... 10/ (13 files in 5 dirs): Feb 12 07:54:43 slugos: Major build structure change - subsumed openslug and debianslug into a Feb 12 07:54:43 single slugos distribution which works with either nslu2be or nslu2le machines. Feb 12 07:54:43 Kept the openslug and debianslug conf files around to enable people to migrate Feb 12 07:54:44 smoothly, but eventually they will disappear. Feb 12 07:55:01 that's all. Feb 12 07:56:57 someone tried to run arm kernel without fpu emulation? Feb 12 08:04:43 hrv, me long time ago ... it don't boot ;( Feb 12 08:18:41 mornig all Feb 12 08:19:49 hey Ifaistos Feb 12 08:20:25 ho Feb 12 08:22:19 hey zecke Feb 12 08:25:59 fair opens, I will go to the beach soon (hopefully) Feb 12 08:26:21 or drop by at the Trolltech booth and see their Open Platform Feb 12 08:26:44 "open" Feb 12 08:29:28 there are a lot of Open Linux platforms here... Feb 12 08:29:44 "open" Feb 12 08:29:55 haven't seen a single 100% open one Feb 12 08:30:17 koen: your definition might just be plain wrong? Feb 12 08:30:28 could be Feb 12 08:31:04 koen: Open means, if you pay enough money you get a copy of the source Feb 12 08:31:05 can someone run bitbake-trunk interactive shell and build something? Feb 12 08:33:44 hrw: done Feb 12 08:34:00 Hi Feb 12 08:34:25 (saladino is trying to get the uclibc version of SlugOS to build) Feb 12 08:34:42 morning Feb 12 08:34:43 yep, me and a friend are working on that Feb 12 08:34:44 hrw: done Feb 12 08:34:47 (so I asked him to discuss here, unless it is nslu2-linux specific) Feb 12 08:34:58 koen, pH5: check packages - are they correct or not? Feb 12 08:35:01 good morning Feb 12 08:35:19 hrw: http://rafb.net/p/dYl5Nk41.html Feb 12 08:35:27 saladino: how far have you got in building uclibc? which version are you building? Feb 12 08:35:32 hrw: I get a lot of nano packages Feb 12 08:35:46 hrw: heh, I see Feb 12 08:35:48 koen: content of package Feb 12 08:35:52 hrw: only locale packages Feb 12 08:35:55 rwhitby we've been compiling the uclibc from svn Feb 12 08:35:58 right, same here Feb 12 08:35:59 using the svn bb file Feb 12 08:36:23 rwhitby: btw, I build angstrom/uclibc a while ago Feb 12 08:36:29 rwhitby: seemed to work Feb 12 08:36:30 as Ville Haapala told us Feb 12 08:37:05 in fact, "make ucslugc-image" compiled after some modifications Feb 12 08:37:38 saladino: modifications to nslu2-linux stuff, or generic uclibc stuff? Feb 12 08:37:50 koen: which machine/arch? Feb 12 08:38:04 rwhitby: armv5te Feb 12 08:38:25 rwhitby modifications in the uclibc .config file and some other autobuild files Feb 12 08:38:53 but we don't know the autobuild system yet Feb 12 08:38:55 saladino: have a look at conf/distro/include/angstrom-uclibc.inc to see what angstrom does. Feb 12 08:39:18 rwhitby: set ANGSTROM_MODE = "uclibc" and it will build uclibc Feb 12 08:39:30 koen: nod Feb 12 08:39:36 it will even allow you to build glibc packages after that without cleaning tmp/ Feb 12 08:40:24 koen: uclibc, or uclibcgnueabi? Feb 12 08:40:55 I guess eabi, since it was arm Feb 12 08:41:39 eabi with linux-threads (old) Feb 12 08:41:45 rwhitby: uclibcgnueabi for arm, uclibc for the rest Feb 12 08:43:05 that reminds me, I should release all the pinned versions in slugos.inc (e.g. glibc 2.3.5+cvs20050627) and see what happens Feb 12 08:43:08 ~lart bitbake shell Feb 12 08:43:08 * ibot cuts bitbake shell into thin stripes Feb 12 08:43:52 saladino: I would suggest building angstrom/uclibc, then seeing what versions of tools and libraries are different between angstrom/uclibc and slugos/uclibc Feb 12 08:44:09 aham Feb 12 08:44:20 angstrom/ucslugc is suposed to build correctly? Feb 12 08:44:31 yes, koen maintains that combo. Feb 12 08:45:06 ok, i'll do some tests Feb 12 08:45:28 and actually, if you want to only install packages in internal flash, then it might run on an nslu2 (since Angstrom boots fine on an nslu2 in le mode) Feb 12 08:45:53 (e.g. we could replace ucslugc with angstrom/uclibc. Feb 12 08:47:08 what is exactly angstrom? is it very similar to slugos/ucslugc or is intended for other devices? Feb 12 08:49:23 werner almesberger reported a bitbake bug Feb 12 08:49:50 wow Feb 12 08:56:42 zecke: he is working on OpenMoko Feb 12 08:56:59 where do i need to set MACHINE var? . ./setup-env; exec bitbake -k "ucslugc"-packages Feb 12 08:56:59 ERROR: You must set MACHINE to nslu2le or nslu2be. Feb 12 08:57:09 ups Feb 12 08:57:11 brb Feb 12 08:57:23 saladino: www.angstrom-distribution.org Feb 12 08:57:38 saladino: conf/auto.conf Feb 12 08:57:51 set it to nslu2be for ucslugc Feb 12 08:58:39 XorA: oh he is paid? great Feb 12 08:58:49 http://pastebin.ca/351674 - result of running gpe-mini-browser Feb 12 08:59:16 hrw: thats a new one Feb 12 09:03:53 ok thanks rwhitby Feb 12 09:12:18 ~fish lrg Feb 12 09:12:27 * ibot slaps lrg around with a large trout Feb 12 09:12:50 lrg: XorA : your Wolfson booth at 3GSM is small :) Feb 12 09:13:12 zecke: Wolfson dont have much clue about GSM :-) Feb 12 09:13:31 XorA: it is so small you don't even have girls around... Feb 12 09:13:41 XorA: you only provide stuff for gsm phones Feb 12 09:13:44 zecke: Im not lucky enough to go anyway Feb 12 09:14:07 barcelona is great Feb 12 09:14:12 apparently the booth babe quality as 3GSM is quite good Feb 12 09:14:28 morning all Feb 12 09:14:33 hey lrg Feb 12 09:14:51 XorA: actually I don't know, the fair just opened and I'm sitting in a cold room and guarding the server :) Feb 12 09:15:03 zecke: heh heh Feb 12 09:15:13 * lrg doesn't know much about GSM booth, although Barcelona is a nice place to visit :) Feb 12 09:15:14 zecke: find some of the geek babes and show them your hardware Feb 12 09:15:35 zecke: Aspects SOftware my old company are there as well Feb 12 09:15:35 yup Barcelona is a very nice place Feb 12 09:16:02 zecke: the ones who unwittingly paid me for OE dev work for 3 years Feb 12 09:16:13 psst Feb 12 09:19:43 *puh* i have spare time again Feb 12 09:21:30 * ade|desk wonders what the babe ratio is at focus on imaging, surely there must be models to test camera with ;) Feb 12 09:22:51 ade|desk: I don't like this sexism, some time I hate to be man... Feb 12 09:23:41 http://farm1.static.flickr.com/143/387797024_b02a0bb950_o.jpg the rack I guard, and a hint on what I'm doing here Feb 12 09:25:15 doh cant read it too small :( Feb 12 09:27:31 make ucslugc-packages fails -> ERROR: Nothing provides dependency ucslugc-packages Feb 12 09:39:32 03pfalcon 07org.oe.dev * r08c06c17... 10/ (4 files in 3 dirs): Feb 12 09:39:32 opie-bluetoothapplet cvs: Patch to get BT params from /etc/sysconfig/bluetooth Feb 12 09:39:32 * instead of hardcoding in C++ code. Feb 12 09:39:32 * Originally by Slavek Banko, http://handhelds.org/hypermail/opie-devel/29/2946.html Feb 12 09:39:32 * Submitted upstream as http://opie-bugs.oszine.de/view.php?id=1838 Feb 12 09:43:02 okay, later guys :) Feb 12 09:43:13 will expose me to the light now Feb 12 09:44:18 Hi! Feb 12 09:47:19 koen, hrw: I tested state of BT in OPIE, and it is havoc after forced upgrade to bluez3. It is indeed completely dbus bound, w/o even debug option to enter pin by other means except dbus calls. So, I'd like to re-add bluez2 recipes, under diffent PN, to preclude any clashes between 2 & 3. How that sounds? Feb 12 09:47:47 psokolovsky: sounds like opie sucks Feb 12 09:47:53 psokolovsky: if needed then let it be Feb 12 09:48:00 koen: you are repeating itself Feb 12 09:48:21 koen, I'd like it to be upgraded to dbus, but it's completely separate task. Feb 12 09:48:22 psokolovsky: I'd rather not want bluez2 back in OE Feb 12 09:48:59 koen, what issues it brings with itself, if properly separated by namespaces? Feb 12 09:49:53 koen, such dbus-less verion would be useful for other things, like routers, rescue disks, etc Feb 12 09:50:02 psokolovsky: exactly what you say, seperated namespaces Feb 12 09:50:20 psokolovsky: dbus isn't that big Feb 12 09:50:21 koen, so, ok to add under bluez2-* ? Feb 12 09:50:28 no,. not OK Feb 12 09:50:35 that sux Feb 12 09:51:00 "big" is not criterion here. "API bloat" is Feb 12 09:51:39 bluez-*-old_2.25.bb? Feb 12 09:51:42 if I can put pin in static file, why do need to write whole dbus app? Feb 12 09:52:13 hrw, I'd think use major release number is safe and cleaner Feb 12 09:52:34 hrw: shlib renaming makes that useless Feb 12 09:52:47 hrw: and bluez2 has broken shlibs, which is why it was removed Feb 12 09:53:00 koen: its libbluetooth1 contra 2 iirc Feb 12 09:53:09 hrw: sadly not Feb 12 09:53:14 ;( Feb 12 09:53:35 which is why I don't want bluez2 back into OE in *any* form Feb 12 09:54:15 psokolovsky: disable auth in bluez.conf and you don't need a pinhelper Feb 12 09:54:35 fwiw, pin auth is broken in gpe-bluetooth as well for bluez3 Feb 12 09:54:53 but since pin-auth can be disabled that isn't a real problem Feb 12 09:55:07 koen, latest 3.x doesn't seem to have any concept of pin_helper any longer, tehy just make dbus call ;-( Feb 12 09:56:30 psokolovsky: if there is no passkey agent registered, it can read the passkey for incoming connections from /etc/bluetooth/passkeys Feb 12 09:57:39 pH5, show silly me the code or config which does that Feb 12 10:01:53 hm, I must have been blind, I finally see codepath which skips dbus stuff. but all my tries to configure it were vain Feb 12 10:02:09 psokolovsky: security and passkey options in hcid.conf Feb 12 10:02:28 pH5, do you have working example? Feb 12 10:02:35 I alreday used them Feb 12 10:03:13 Feb 12 01:19:54 h4000 daemon.info hcid[3644]: link_key_request (sba=08:00:28:90:9E:FD, dba=00:0A:84:00:84:AA) Feb 12 10:03:13 Feb 12 01:19:54 h4000 daemon.info hcid[3644]: pin_code_request (sba=08:00:28:90:9E:FD, dba=00:0A:84:00:84:AA) Feb 12 10:03:13 Feb 12 01:19:54 h4000 daemon.debug hcid[3644]: call_passkey_agent(): no agent available Feb 12 10:04:04 doing a "make ucslugc-packages" the bitbake script searches in self.status.providers a package named "ucslugc-packages", as it is not a package, it fails, how can i fix that? Feb 12 10:09:07 hummm, there is not a openembedded/packages/meta/ucslugc-packages.bb Feb 12 10:09:15 shall i copy it from openembedded/packages/meta/slugos-packages.bb ?? Feb 12 10:09:50 why do you need *-packages? Feb 12 10:10:11 pH5, ok, so it seems that what I said is true: a new pin for outgoing connection can be entered only via dbus. passkey is only for incoming. but at least this crap keeps cache of pins, so I'll try to inject pin there. Feb 12 10:11:10 koen, rwhitby told me to try to compile "make ucslugc-packages" and that is what i am trying Feb 12 10:11:16 any hint? Feb 12 10:11:27 that obviously doesn't work Feb 12 10:11:35 try to compile some other packages Feb 12 10:13:37 koen, the fact is that at this moment ucslugc is suposed to be abandoned, but a friend and me are trying to make it work, rwhitby told me that if i want to mantain ucslugc i need to keep ucslugc-packages compilable Feb 12 10:13:43 so if it does not work Feb 12 10:13:50 i'll try to fix it Feb 12 10:13:50 :) Feb 12 10:14:09 im trying using the slugos-package.bb Feb 12 10:16:00 psokolovsky: as I said, static passkey for incoming connections. for outgoing connections this would be sort-of braindead anyway. Feb 12 10:16:43 pH5, damn. if I want to connect to my gps, what kind of connection that is? Feb 12 10:20:33 saladino: make slugos-packages Feb 12 10:21:16 it's just a list of packages that were known to work with uclibc Feb 12 10:21:56 thanks rwhitby , i'll try Feb 12 10:28:34 good morning Feb 12 10:29:25 good morning Feb 12 10:35:15 rwhitby im doing "make slugosbe-packages" after a "make setup-host-debian", is this correct to prove ucslugc compilability :? Feb 12 10:44:01 slugos-packages Feb 12 10:44:11 oh, makefile, Feb 12 10:44:42 discuss the MAkefile in #nslu2-linux, not here. The MasterMakefile is not an OE product. Feb 12 10:45:03 ok Feb 12 10:47:23 florian: good morning Feb 12 10:54:15 * koen fixes wiki mainpage Feb 12 10:54:19 damn wiki vandals Feb 12 11:09:11 03Laibsch 07org.oe.documentation * re4d3dacc... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: fixed and removed section 'kernel/module' Feb 12 11:09:13 03Laibsch 07org.oe.documentation * ref653e2d... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: fixed and removed section 'bootloader' Feb 12 11:09:35 hi jonbaine Feb 12 11:09:42 hi saladino !!! Feb 12 11:28:43 re Feb 12 11:42:06 morning all Feb 12 11:42:08 03Laibsch 07org.oe.dev * rc75a7b6d... 10/ (1 packages/fbvncserver/fbvncserver-kmodule_0.9.4.bb): fbvncserver-kmodule: fix SECTION Feb 12 11:42:16 03Laibsch 07org.oe.dev * rbfbdf129... 10/ (1 packages/uboot-utils/uboot-utils_1.1.2.bb): uboot-utils: add to SECTION "bootloaders" Feb 12 11:42:22 03Laibsch 07org.oe.dev * r350577dc... 10/ (7 files in 2 dirs): uboot: unify, fix SECTION Feb 12 11:43:45 hey Dirk Feb 12 11:43:59 hi koen Feb 12 12:24:11 i am compiling with a machine that identifies itself as "ppc-linux" Feb 12 12:24:32 as in openembedded/classes/siteinfo.bbclass there is no "ppc-linux" but "powerpc-linux" it does not work Feb 12 12:24:43 but as far as i know it is the same Feb 12 12:25:18 could somebody add a line for ppc-linux in target-info at openembedded/classes/siteinfo.bbclass ? Feb 12 12:25:27 just copy like powerpc-linux Feb 12 12:26:16 powerpc == ppc Feb 12 12:26:43 yep Feb 12 12:27:46 ppc is being migrated to powerpc Feb 12 12:27:47 hrw shouldn't it be added? as some machines are identified like that with those the script will fail unless someone adds that line Feb 12 12:28:36 saladino: ask koen Feb 12 12:29:13 koen? Feb 12 12:29:21 ppc = powerpc Feb 12 12:29:26 no need to add ppc Feb 12 12:30:00 but it fails when machines are identified as ppc, i know that it is the same, but some machines are identified like that and the script fails Feb 12 12:30:11 what is the solution for those machines? manual changes? Feb 12 12:30:26 i've changed myself, but for the future.. Feb 12 12:30:52 mmm the name of the arch deppends on the OS??? Feb 12 12:31:28 beacause saladino runs a Debian Etch, so everyone that runs that OS would have the same problem Feb 12 12:31:31 :S Feb 12 12:32:02 You don't get it Feb 12 12:32:14 :? Feb 12 12:32:26 you tell OE to build from powerpc and it will build binaries that run on powerpc cpus Feb 12 12:32:38 it doesn't matter that your machine claims to be 'ppc' Feb 12 12:32:46 since OE is *cross*compiling Feb 12 12:32:49 yep Feb 12 12:32:51 i know that Feb 12 12:33:02 the fact is that if i dont change that line all the script fails Feb 12 12:33:10 and it does not cross-compile nor nothing Feb 12 12:33:14 'all the script'? Feb 12 12:33:20 the autobuild Feb 12 12:33:26 I can compile for powerpc just fine Feb 12 12:33:36 I don't know of any autobuild Feb 12 12:33:55 I know that 'TARGET_ARCH=powerpc' works in OE Feb 12 12:34:21 ERROR: Information not available for target 'ppc-linux' Feb 12 12:34:21 NOTE: exceptions.TypeError:iterable argument required while evaluating: Feb 12 12:34:21 ${@siteinfo_get_bits(d)} Feb 12 12:34:21 ERROR: Information not available for target 'ppc-linux' Feb 12 12:34:30 maybe i am not understanding well Feb 12 12:34:49 but the last week i changed that line and i was able to compile for my nslu Feb 12 12:35:08 if i dont change that line it fails dozens of lines like that i've pasted Feb 12 12:35:12 and does not compile Feb 12 12:35:20 am i doing something wrong? Feb 12 12:36:11 your nslu isn't powerpc, but arm Feb 12 12:36:20 i know that Feb 12 12:36:51 there are no machines in OE that set TARGET_ARCH=ppc Feb 12 12:37:10 that target is for the Arch of the system that supports the cross copiling not?? Feb 12 12:37:45 not the arch the compiling is -- for -- i think ¿? Feb 12 12:38:04 am i right?? Feb 12 12:38:23 * koen thinks TARGET_ARCH is pretty self explanatory Feb 12 12:38:31 koen yes Feb 12 12:38:36 wait a minute Feb 12 12:38:43 target = bb.data.getVar('HOST_ARCH', d, 1) + "-" + bb.data.getVar('HOST_OS', d, 1) Feb 12 12:38:43 targetinfo = {\ Feb 12 12:38:56 as HOST_ARCH is suposed to be the cross-compiling machine Feb 12 12:38:59 not the target ones Feb 12 12:39:10 or no? Feb 12 12:39:15 right Feb 12 12:39:39 so in that HOST_ARCH my x86 identifies as x86 and builds an arm image Feb 12 12:40:04 I see the problem Feb 12 12:40:05 but my ppc-linux identifies as ppc-linux and it is not able to build the arm image unless i put that line Feb 12 12:40:18 am i doing something wrong? Feb 12 12:40:18 powerpc host identify as ppc instead of powerpc Feb 12 12:40:22 ah, your machine is running an old kernel Feb 12 12:40:29 2.6.20 Feb 12 12:40:46 so ppc should fallthrough to powerpc Feb 12 12:40:46 * jonbaine O_o" Feb 12 12:41:33 should i do something aditional? or just change the siteinfo.bbclass? Feb 12 12:44:22 03koen 07org.oe.dev * rdede9079... 10/ (1 classes/siteinfo.bbclass): siteinfo.bbclass: add support for oldstyle powerpc hosts Feb 12 12:45:01 thanks koen Feb 12 12:45:03 thanks a lot koen Feb 12 12:46:46 I'm going to change task-base Feb 12 12:47:15 instead of one big task-base there will be task-base-FEATURE packages on which task-base will depend Feb 12 12:47:26 right Feb 12 12:47:36 I always wondered why it worked this way Feb 12 12:48:00 task-base-wifi, task-base-usbhost BUT... task-base-usbhost-wifi, task-base-pcmcia-wifi? Feb 12 12:48:29 koen: no one bothered to change it and I only added stuff Feb 12 12:52:03 task-base-pcmcia-wifi-kernel26 Feb 12 12:54:30 but where to put kernel-module-irnet which is irda (machine) and also ppp (distro) Feb 12 12:54:46 task-distro-ppp-irda or task-base-irda-ppp? Feb 12 12:56:20 task-{distro/combined/machine}-foo? Feb 12 12:58:36 will consider Feb 12 13:12:06 hi all Feb 12 13:14:30 koen: task-base-acpid task-base-alsa task-base-apm task-base-bluetooth task-base-core-default task-base-ext2 task-base-ipsec task-base-irda task-base-kernel26 task-base-keyboard task-base-minimal task-base-nfs task-base-oh-minimal task-base-pci task-base-pcmcia task-base-phone task-base-ppp task-base-screen task-base-smbfs task-base-touchscreen task-base-usbgadget task-base-usbhost task-base-wifi task-base Feb 12 13:14:40 Which file do I have to change to get a /dev/console entry created in the image? Feb 12 13:53:21 morning Feb 12 14:55:08 Depends: task-base-core-default, kernel, task-base-kernel26, task-base-acpi, task-base-alsa, task-base-ext2, task-base-irda, task-base-pci, task-base-pcmcia, task-base-usbhost, task-distro-nfs, task-distro-ppp, task-distro-wifi, task-machine-extras, task-distro-extras Feb 12 15:07:41 Depends: task-base-core-default, kernel, task-base-kernel26, task-base-acpi, task-base-alsa, task-base-ext2, task-base-irda, task-base-pci, task-base-pcmcia, task-base-usbhost, task-distro-nfs, task-distro-ppp, task-distro-wifi, task-base-machine, task-distro-base Feb 12 15:07:41 Recommends: task-base-kernel26-extras, task-base-alsa, task-base-pcmcia, task-base-irda, task-base-usbhost, task-distro-ppp, task-distro-nfs, task-distro-wifi, task-base-machine-extras, task-distro-extras Feb 12 15:10:31 03Laibsch 07org.oe.documentation * r4b34e229... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: fixed and removed sections console/net, console/networking, console/util, gnome/libs, gnome/multimedia, gnome/office and kde/dev Feb 12 15:35:19 is gnome/office a banned section? Feb 12 15:41:03 XorA: According to http://www.openembedded.org/user-manual&dpage=section_variable it was. Feb 12 15:43:12 damn I think gnome/* needs added, as abiword is quite defineately gnome application seeing as it uses libgnome* Feb 12 15:43:45 hrw, RP: so, let's talk about circular deps in opie-mediaplayer. do you say that even if opie-mediaplayer1 just RRECOMENDS its plugins, it's bad and leads to cycles? Feb 12 15:43:47 best one to talk to is probably hrw. Feb 12 15:45:28 Laibsch: I shall remeber to hopefully do something about that when I have spare time Feb 12 15:48:43 lrg: ah, just noticed the SOC audios stuff hit linus - cheers.. :) Feb 12 15:49:49 I've some errors compiling ucllibc Feb 12 15:49:54 CosmicPenguin: np, thanks. :) Feb 12 15:50:06 a warning armeb-linux-uclibc-ld: WARNING: libc/libc_so.a(clnt_generic.os): 100cab60(__GI_strcmp): PLT refcount was -1 (set to 1) Feb 12 15:50:27 and some of these errors-> closedir.c:(.text+0x3a): undefined reference to `_call_via_r3' Feb 12 15:54:37 lrg: 2.6.21 will have asoc? Feb 12 15:54:56 koen: yep Feb 12 15:55:01 cool Feb 12 15:56:36 koen: fwiw, neo phone audio should follow for 2.6.22 Feb 12 15:57:31 cool Feb 12 16:00:51 lrg: wm9705 is supported? Feb 12 16:01:55 hrw: only through generic ac97 driver, too busy Feb 12 16:14:29 * koen gives up on matlab for today and goes to watch the A-team reruns on tv Feb 12 16:31:14 What does the SECTIONS (not SECTION!) line in packages/opie-taskbar/opie-taskbar.inc do? Is it necessary? Feb 12 16:41:46 03pfalcon 07org.oe.dev * rc97365fe... 10/ (7 files in 2 dirs): (log message trimmed) Feb 12 16:41:46 opie-mediaplayer1: Start measures to get rid of circular package dependencies. Feb 12 16:41:46 * The idea is simple and standard: introduce a node in graph "above" current ones, Feb 12 16:41:46 and make it depend on existing nodes, instead of them depend on each other. Feb 12 16:41:46 Thus, we'll have opie-mediaplayer1-core & opie-mediaplayer1-plugins*, and Feb 12 16:41:48 opie-mediaplayer1 virtual package above them. Feb 12 16:41:56 * Step 1: rename existing opie-mediaplayer1 recipe to opie-mediaplayer1-core Feb 12 16:41:56 03pfalcon 07org.oe.dev * rfbc33c15... 10/ (3 files in 2 dirs): Feb 12 16:41:56 opie-mediaplayer1: Continue cycle-breaking. Feb 12 16:41:58 * opie-mediaplayer1-ogg-mp3 is what becomes opie-mediaplayer1. For now we don't Feb 12 16:42:00 want adhoc packaging variants. Feb 12 16:42:02 03pfalcon 07org.oe.dev * reb6addb7... 10/ (3 files in 2 dirs): opie-mediaplayer1: Finish refactor to get rid of circular dependencies. Feb 12 16:42:05 03pfalcon 07org.oe.dev * r75595250... 10/ (3 files in 2 dirs): Feb 12 16:42:09 preferred-opie-versions.inc, preferred-opie-cvs-versions.inc: Add opie-mediaplayer1-core. Feb 12 16:42:11 * + allow to override few package versions. Feb 12 16:42:16 Laibsch, yep, it is required, look at file carefully Feb 12 16:42:24 it's "normal" var Feb 12 16:42:44 feel free to rename to "SECTION_LIST" or sth Feb 12 17:01:54 I see that there was some talk over the weekend about Bluez and PINs...has anyone tried the example passkey-agent.c from Bluez-utils? It doesn't build by default so maybe we could make a separate recipe/package for it? Feb 12 17:03:39 apaulsen, there was talk today. my personal grief that it's completely stupid for core kernel bluetooth stack to cage users unconditionally into dbus. Feb 12 17:03:56 OMG DBUS BAD!!!! Feb 12 17:03:59 LOL!!! Feb 12 17:04:01 OMG!!! Feb 12 17:04:47 actually it's only a build time dependency or is it? Feb 12 17:04:57 koen, dbus may be superflous for many usecases. let's do pin entering via web with embedded apache. Feb 12 17:05:06 mickeyl, runtime Feb 12 17:05:25 hmm so it really _needs_ dbus to run at all now? Feb 12 17:05:30 no Feb 12 17:05:39 only for outgoing pin requests Feb 12 17:05:45 user agnet which actually sniffs pin from user must talk via dbus to hcid Feb 12 17:06:04 yeah, i.e. for any real use ;-) Feb 12 17:06:15 ? Feb 12 17:06:27 my gps, mouse and keyboard work just fine with the hx4700 Feb 12 17:06:28 you cannot even test if it works on no w/o dbus Feb 12 17:06:51 koen, and my gps & phone happen to require pins ;-) Feb 12 17:08:40 my headsets require PINs and the only recipe I was able to find that used the new DBUS passkey system is bluez-gnome and that requires a huge gnome icon pack and a custom .desktop file to get working Feb 12 17:08:58 hrw: http://projects.linuxtogo.org/plugins/scmsvn/viewcvs.php?rev=8968&root=gpe&view=rev Feb 12 17:12:09 koen: hrm... that was important for the release :-/ Feb 12 17:24:32 I guess my question was, has anyone tried the passkey-agent.c? If it works at all, someone should make a recipe for it. I'll give it a try if no one else is interested, but it may not be pretty Feb 12 17:28:53 What is this? I have had several failures lately same as this one. http://pastebin.ca/352220 Feb 12 17:29:52 apaulsen, good plan. AFAI get it, it's pin agent dispatcher, kinda what you could do previously using pin_helper Feb 12 17:30:05 goxboxlive: # Feb 12 17:30:05 | arm-angstrom-linux-gnueabi-g++: Internal error: Segmentation fault (program as) Feb 12 17:30:10 goxboxlive: faulty RAM? Feb 12 17:30:37 Is it possible to define a python function in a bb file, and call it from within do_install for example? Feb 12 17:30:45 goxboxlive, or too little of it ;-) Feb 12 17:30:58 gcc loves to segfault on OOM Feb 12 17:31:42 tkp, sure Feb 12 17:32:03 psokolovsky: cool... know where there is an example I can look at? Feb 12 17:32:15 tkp, OE is full of examples Feb 12 17:32:30 hmm strange i have 1GB Ram, and OE is the only thing that runs on it. Maybe it is because i havent rebooted the machine for over 100 days. Feb 12 17:32:34 ${@func()} is the easiest way to call func and use its value Feb 12 17:32:38 koen: finally! Feb 12 17:32:50 well, I defined a function python myfunction( d ) {...} Feb 12 17:32:52 aggg.. Feb 12 17:32:57 ah, k Feb 12 17:32:58 cool Feb 12 17:33:04 hrw: I'd rather have an eds backend, but this will do in the meantime :) Feb 12 17:33:30 koen: yep Feb 12 17:34:07 hmm... just resulting in a bb parsing error Feb 12 17:34:15 ${@myfunction( "${D}${libdir}" )} Feb 12 17:34:51 oh... your not allowed to pass args? Feb 12 17:35:57 tkp, of course allowed Feb 12 17:36:14 cu Feb 12 17:36:39 odd... works when defined as python myfuntion()... but not as python myfunction( d ) Feb 12 17:37:51 hmm... still doing something wrong: Feb 12 17:37:54 NOTE: exceptions.NameError:name 'bytecompile_python_code' is not defined while evaluating: Feb 12 17:38:21 the function is defined before do_install_append in my bb file Feb 12 17:39:31 do I have to define it in a bbclass and inherit in order to be able to use it in a task like do_install? Feb 12 17:54:47 the ${@} notation seems to let me run the code in the main bit of the bb file, and assign the result to a variable Feb 12 17:55:24 thats not quite what I'm after... I need to run my function from within the do_install routine Feb 12 18:01:47 03pfalcon 07org.oe.dev * r5ff0c6a7... 10/ (4 files in 3 dirs): (log message trimmed) Feb 12 18:01:47 bluez-utils 3.9: Add patch to handle EINTR while being attached to uart properly. Feb 12 18:01:47 * EINTR is not really an error, but a Unix (mis)feature, and requires special Feb 12 18:01:47 handling (i.e. retrying). Bluez authors/maintainers really should do a bit of trivial Feb 12 18:01:47 Unix programming reading before wring BT stacks. Feb 12 18:01:49 * Specifically EINTR happens when resuming from suspend. So, with this patch Feb 12 18:01:51 (and assuming there're no weird APM scripts), configured BT finally survives Feb 12 18:32:59 has anyone been able to successfully use cpan to install modules on an oe device? Feb 12 18:34:15 I'm having a hell of a time Feb 12 18:49:58 lrg: Hello; when/if you will have time, could I ask you few questions about ASoC (mostly ak4535 driver related)? Feb 12 18:52:32 mmp: it's ok now Feb 12 18:54:33 lrg: I'm not sure whether I understand correctly whole way how ASoC works, but in ak4535 driver I noticed that many DAPM widgets are routed through one big 'Switch'. Feb 12 18:54:48 How can one modify its state? Feb 12 18:55:23 It is not exported into device's mixer Feb 12 19:32:32 bitbake 1.6.2 is old ??? Feb 12 19:32:42 s/old/stable/ Feb 12 19:32:42 pretty old Feb 12 19:32:51 * koen uses 1.7.4 without much problems Feb 12 19:32:54 too much old ? Feb 12 19:33:23 i have a dependency problem, with angstrom-gpe-image Feb 12 19:33:26 | ERROR: Cannot satisfy the following dependencies for angstrom-gpe-task-base: Feb 12 19:33:27 | libopenobex1 (>= 1.2) Feb 12 19:34:09 good dinner :) Feb 12 19:34:12 thanks Feb 12 19:34:24 libopenobex built and in tmp/deploy/pkg ? Feb 12 19:34:25 ~bon appetit Feb 12 19:34:27 arch matching? Feb 12 19:34:31 i guess bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Smaklig måltid!. Hyvää ruokahalua. Bo Proveito Feb 12 19:34:31 thanks Feb 12 19:34:54 h3600 so arm-oabi Feb 12 19:47:50 Hi Feb 12 19:53:10 greentux, oh, just in case - do you build angstrom-gpe-image? you should build it, not gpe-image Feb 12 19:53:46 yes yes ... i did angstrom-gpe-image Feb 12 19:53:47 psokolovsky: tnx for info. will talk to a dev tomorrow Feb 12 19:54:24 greentux, sorry, meant that for gremlin[it] , that's all completion in irc client ;-) Feb 12 19:54:28 gremlin[it], ^^^^ Feb 12 19:54:44 yes yes i answer ... Feb 12 19:55:11 koen, can I build angstrom & angstrom-oabi in the same build dir? Feb 12 19:55:26 psokolovsky: yes Feb 12 19:55:36 cool Feb 12 19:55:45 psokolovsky: one uses the gnueabi suffix, the other oabi Feb 12 19:56:06 ok Feb 12 19:56:31 psokolovsky: I'm pretty proud of being able to mix and match different machines, archs and ABIs Feb 12 19:56:58 I'm trying to get cpan working on my board, everything seems to configure ok, but it can't install anything, it always complains "Warning: Cannot install , don't know what it is." Feb 12 19:57:49 koen, yes, that's very nice and cool. thanks for working on that! Feb 12 19:58:16 psokolovsky: it follows from multimachine, which was made by kergoth and improved by RP Feb 12 19:58:50 mhhh doing grep on org.openembedded.dev looking for libopenobex1 but nothing match ... how is possible ??? Feb 12 19:59:10 hi folks Feb 12 19:59:45 * philippe wishes there were more integrators like koen Feb 12 19:59:50 gremlin[it]: openobex Feb 12 20:00:20 openobex is build ... Feb 12 20:00:44 gremlin[it], openobex-1.2-r3 is what built for eabi for me Feb 12 20:00:58 1.2-r4 for me ... Feb 12 20:02:02 gremlin[it], have a look that you have .ipk for it. you known, ipkg-utils are still (?) pretty buggy, ahev bugs in version matching, put stuff in morgue randomly, etc Feb 12 20:02:19 ok found ... openobex create libopenobex-dbg_1.2-r4_arm-oabi.ipk, libopenobex-dev_1.2-r4_arm-oabi.ipk, but no libopenobex ... and '1' is still missing ;( Feb 12 20:02:45 gremlin[it], have a look in morgue Feb 12 20:04:28 ?? Feb 12 20:05:43 /deploy/glibc/ipk/morgue Feb 12 20:06:51 nothing about obex there Feb 12 20:07:42 all i have is in /deploy/glibc/ipk/ Feb 12 20:09:10 gremlin[it], ok, so that's likely better ;-). maybe just some random glitch, try to -c rebuild it (make sure to kill Packages.* files first) Feb 12 20:14:20 rebuild all angstrom-gpe-image or only openobex ? Feb 12 20:15:07 gremlin[it], only openobex Feb 12 20:15:15 bitbake openobex -c rebuild Feb 12 20:15:30 just make suer you kill Packages, or will get busted image Feb 12 20:15:56 done :) Feb 12 20:16:55 libopenobex1.*.ipk is still missing :( Feb 12 20:18:57 ok, I started build, will see what I'll get Feb 12 20:32:21 mhhh libopenobex.so.1.2.0 wasn't build only .a , looking in /work/arm-oabi-angstrom-linux/openobex-1.2-r4/install Feb 12 20:46:45 seem as ${@base_contains("COMBINED_FEATURES", "bluetooth", "gpe-bluetooth", "",d)} isn't correctly handled ... or for soem reason bluetooth is setted :( ... Feb 12 20:47:40 * koen reads http://linuxdevices.com/news/NS4018299974.html Feb 12 20:49:45 * gremlin[it] dreams about OMAP3430 :) Feb 12 20:50:04 no linux support yet Feb 12 20:50:52 give me a board and a fee and i'll do, i'll do :) :) Feb 12 20:51:20 * chouimat dreams about the new girl next door :) Feb 12 20:51:45 * philippe thinks chouimat is not a reall geek :-) Feb 12 20:51:59 philippe: :) Feb 12 20:52:10 hahahah :) ... Feb 12 20:52:38 ARM ltd has patches for the cortex-A8 core Feb 12 20:52:55 philippe: if you can get the imap3430 to hug you when you feel sad ... I want the source code ;) Feb 12 20:53:40 * philippe finally got around the battery empty-need-serial and reset button unreachable because hidden by hte cradle problem Feb 12 20:53:42 * hvontres|poodle thinks chouimat should just knock on the door an say hello...:) Feb 12 20:53:57 hvontres|poodle: done Feb 12 20:54:21 philippe: paperclip Feb 12 20:54:28 hvontres|poodle: I needed sugar earlier today ;) Feb 12 20:54:43 koen: that was sort of the solution indeed.... Feb 12 20:55:20 chouimat: yeah, riiight. :) Feb 12 20:55:48 Ok, I have tested the bluez-utils passkey-agent if anyone is interested Feb 12 20:55:49 * philippe thinks chouimat will be out of a lot of stuff the next few days :-) Feb 12 20:56:12 philippe: hehe Feb 12 20:56:24 apaulsen: and? Feb 12 20:56:28 turns out it was already being built so I only needed to install it Feb 12 20:56:32 it does work Feb 12 20:56:37 ah, cool Feb 12 20:56:53 you basically start it on the command line with the passkey and MAC address Feb 12 20:56:54 mhhh coming back about new omap 3430 ... wifi is connected throug sdio ... linux have some kind of sdio working /?? Feb 12 20:57:25 gremlin[it]: there is a patch for some sdio wifi cards Feb 12 20:57:40 ahh ok :) Feb 12 20:57:41 --- packages/bluez/bluez-utils.inc.orig 2007-02-12 14:37:46.000000000 -0600 Feb 12 20:57:41 +++ packages/bluez/bluez-utils.inc 2007-02-12 15:08:47.000000000 -0600 Feb 12 20:57:41 @@ -41,6 +41,7 @@ Feb 12 20:57:41 chmod u+s ${D}${base_sbindir}/hciattach ${D}${base_sbindir}/hciconfig Feb 12 20:57:41 install -m 0644 ${WORKDIR}/hcid.conf ${D}${sysconfdir}/bluetooth/ Feb 12 20:57:41 install -m 0755 ${WORKDIR}/02dtl1_cs.sh ${D}${sysconfdir}/apm/event.d/ Feb 12 20:57:43 + install -m 0755 ${S}/daemon/.libs/passkey-agent ${D}${base_bindir}/ Feb 12 20:57:45 } Feb 12 20:58:13 gremlin[it]: http://source.mvista.com/~dsingleton/sdio/ Feb 12 20:58:22 Windows Vista: XT emulator for dual Opteron Feb 12 20:58:44 koen: ewwww not mvista... Feb 12 20:58:55 mhhh mmvista ... Feb 12 20:59:03 hvontres|poodle: you read my mind :) Feb 12 20:59:36 wait... isn't *source*.mvista.com kind of an oxymoron anyways ? Feb 12 20:59:43 ok we have but need to be rewrite :) Feb 12 21:01:31 should inherit update-rc.d be used for everything that has an init.d script Feb 12 21:01:31 ? Feb 12 21:01:41 Jin^eLD: Yes Feb 12 21:02:40 koen: I am setting INITSCRIPT_NAME = "myname" in the .bb file inherits update-rc.d but doesn't set INITSCRIPT_NAME while parsing Feb 12 21:03:12 aem.. misconstructed sentence Feb 12 21:03:13 :) Feb 12 21:03:30 I mean - I do se the, var but bitbake complains that it is not set Feb 12 21:03:34 when parsing the .bb file Feb 12 21:03:39 why could that be? Feb 12 21:04:49 ok thats because I am an idiot :) Feb 12 21:04:50 03andrew.paulsen 07org.oe.dev * r55734c44... 10/ (3 files in 2 dirs): bluez-utils: ship pass-agent Feb 12 21:04:50 forget it Feb 12 21:04:51 Linux-SDIO was moved on sf.net :) Feb 12 21:13:13 g'night all Feb 12 21:20:38 03xora 07org.oe.dev * r26eef982... 10/ (14 files in 5 dirs): Feb 12 21:20:38 firefox_2.0.0.1.bb : add new version of firefox with a couple of patches Feb 12 21:20:38 that make it actually run further than before. If now segfaults after Feb 12 21:20:38 running at 100% cpu for a while. But I cant get arm gdb to give me sensible debugging at the segfault. Feb 12 21:58:08 anyone in here use bootchart? Feb 12 23:17:31 nite all, sweet dreams Feb 12 23:26:31 Has anyone seen weird errors from configure like this: Feb 12 23:26:38 configure: error: `CFLAGS' has changed since the previous run: Feb 12 23:31:14 and the only change in CFLAGS is in whitespace ... Feb 13 00:20:29 in order to work on the 2.6 SD driver for the zuarus 5500, I'm going to need openembedded right? Feb 13 02:22:01 03rwhitby 07org.oe.dev * raad5df38... 10/ (1 conf/distro/include/slugos.inc): slugos.inc: Put FULL_OPTIMIZATION flags in a single line, otherwise configure barfs when building cyrus-sasl and gdb. Go figure. Feb 13 02:27:41 has anyone gotten stuck building linux-libc-headers-2.6.18-r1? looks like the patch isn't applied properly (Applying patch arm-syscall-define.patch) Feb 13 02:28:44 dougt: built fine here on Debian Etch distro=slugos machine=ixp4xxbe Feb 13 02:30:17 using machine=ep93xx distro=generic. **** ENDING LOGGING AT Tue Feb 13 03:00:00 2007