**** BEGIN LOGGING AT Tue Nov 20 02:59:57 2007 Nov 20 03:05:03 koen|away: the autobuilder from http://www.angstrom-distribution.org/unstable/autobuild/ isn't building spitz images for some reason. And the latest in "unstable" are kinda ancient. Nov 20 08:54:27 morning! Nov 20 09:55:51 hi lumag Nov 20 09:56:07 hi hrw! Nov 20 09:57:02 It seems you put great efforts in building that tiny kernels :) Nov 20 09:59:44 ;D Nov 20 10:00:08 lumag: clean pm on tosa Nov 20 10:00:28 That's my task for the weekend. Nov 20 10:00:43 hrw: I don't think, I'll have enough time before Nov 20 10:02:19 ;) Nov 20 10:02:41 * koen realizes it's tuesday Nov 20 10:02:57 its not n810 day yet though :-( Nov 20 10:03:27 today I realised one thing... Nov 20 10:03:55 with my c760 dying n810 can be next hack device too... Nov 20 10:04:14 as long as xterminal fullscreen will be an option Nov 20 10:05:38 ask ross to make a devils pie for maemo or something Nov 20 10:06:15 ? Nov 20 10:07:36 * XorA is mildy amused that his n810 will have to use Windows Mobile to get internet access Nov 20 10:08:16 my wireless devices already use closed-source to get to net Nov 20 10:09:12 XorA: since the n810 has usb host, a usb umts thingie should worl Nov 20 10:09:16 work* Nov 20 10:09:28 koen: is it powered host? Nov 20 10:09:33 ehm.... Nov 20 10:09:37 no idea Nov 20 10:10:01 * XorA supects not, and battery life of phone is probaby better than kludgy hub :-) Nov 20 10:10:12 anyway there would be less amusement Nov 20 10:10:13 XorA: you'll love the 6Mbyte/s flashing speed Nov 20 10:10:33 koen: you put the new flash image on your n800 yet? Nov 20 10:10:43 yes Nov 20 10:11:03 compared to it2007 it2008 is amazing Nov 20 10:11:24 woowoo Nov 20 10:11:42 the gui is something the openmoko gui dudes should study Nov 20 10:12:29 XorA: youtube from inside mozilla works pretty good Nov 20 10:13:05 koen: well at least 2 openmoko guys got n810 :-) and I assume OH has some lieing around Nov 20 10:13:36 the panel has true transparency Nov 20 10:19:18 XorA: http://www.openembedded.org/~koen/20071120-F2FA8157.jpg Nov 20 10:19:23 excuse the bad picture Nov 20 10:19:34 no screenshot tools available yet Nov 20 10:21:24 thats looking quite sweet Nov 20 10:23:10 the menu now has thumb sized tiles Nov 20 10:23:46 so they may have a lot of hot air, but at least someone is coding :-) Nov 20 10:24:05 yeah Nov 20 11:54:40 morning Nov 20 11:54:53 lardman: pong :) Nov 20 11:54:58 hey koen :) Nov 20 11:55:22 I'd like to create an n800 image using the old abi, do you know if this is workable (i.e. is the metadata still there)? Nov 20 11:56:15 it should be possible Nov 20 11:57:04 grep -rm for 'collie' in conf/distro and change the three statements with collie, simpad, etc in it and the n800 will suddenly build with OABI :) Nov 20 11:57:07 lardman: long time Nov 20 11:57:22 quote of the day: Nov 20 11:57:32 "Is there an X11 port to the N810?" - pdaX developer Nov 20 11:57:41 talk about being clueless Nov 20 11:58:25 ;DDDD Nov 20 11:58:29 oesf? Nov 20 11:58:38 what else :) Nov 20 11:58:46 hrw: yes, sorry been doing Nokia stuff mainly Nov 20 11:58:57 koen: cool, thanks Nov 20 11:59:15 I'm seeing if i can hack an existing powervr mbx module into an image to rev eng it Nov 20 11:59:26 zecke volunteered to combine those three statements into one, but I haven't seen a patch yet Nov 20 12:01:08 How's things in the Zaurus world or have people moved on? Nov 20 12:01:37 every distro except angstrom is dead or dying Nov 20 12:01:50 well we knew that would be the case :) Nov 20 12:01:54 and a few people are trying debian, but keep ending up needing angstrom packages Nov 20 12:02:26 angstrom zaurus is starting to slow down though :-) Nov 20 12:02:34 I suppose you chaps are more involved with openmoko now than the Zaurus? Nov 20 12:02:56 * XorA does a little OpenMoko work Nov 20 12:03:59 lardman: openmoko is just a rebranded angstrom :) Nov 20 12:04:14 I know, but in terms of hardware I ment Nov 20 12:04:20 lardman: more and more zaurus devs abandon ship Nov 20 12:04:20 s/ment/meant Nov 20 12:04:29 lardman: nothing new in zaurus hardware Nov 20 12:04:38 sl-c3200 was last Nov 20 12:04:40 I suspect OM will become less Angstrom when devel stage finishes and release stage starts Nov 20 12:04:57 hrw: yes, I can understand that, new Nokia looks ok keyboard, etc. Nov 20 12:05:06 XorA: what will it go to then? Nov 20 12:05:18 lardman: right Nov 20 12:05:32 lardman: anyway european price is also not low Nov 20 12:05:57 lardman: well Angstrom will grow as a generic distro, OM will of course tighten belts to only supply what is actually needed for a phone Nov 20 12:06:03 hrw: no, not at the moment, the n800 price is pretty good though if you can do without a keyboard Nov 20 12:06:26 XorA: ah, okay. Nov 20 12:06:27 XorA: you're confusing distro and image :) Nov 20 12:06:45 ROM image....? ;D Nov 20 12:07:01 lardman: n800 cant replace c760 - lack of keyboard Nov 20 12:07:24 hrw: yep, I'm hoping the N810 will do the job for me, I hate not having a keyboard Nov 20 12:07:35 and my c760 hangs too often Nov 20 12:08:33 I do hate the debian build system Nov 20 12:08:42 I miss bitbake :) Nov 20 12:09:19 koen: no Im not Nov 20 12:09:28 koen: there will be policy differences Nov 20 12:09:40 koen: like what SSL implementation gets used Nov 20 12:10:08 gnutls, what else? Nov 20 12:10:57 well at moment it is looking like we will use OpenSSH which means OpenSSL which means probably switching a lot of packages to openssl if built for moko Nov 20 12:11:25 I object to the "we might differ in the future, so we'll fork preemptively" line of thinking Nov 20 12:11:26 Hello there! Nov 20 12:11:47 koen: we already are, angstrom didnt follow in MACHINE_CLASS Nov 20 12:11:56 XorA: you know you can't legally distribute various apps that link again ssl? Nov 20 12:12:11 koen: not my problem Nov 20 12:12:17 XorA: that's because no-one mentioned the machineclass on angstrom mailinglists Nov 20 12:12:45 XorA: I'd love to use machineclass, but noone mentioned it, proposed or discussed it Nov 20 12:13:33 I need to get MIPv6 working on a Qemuarm Nov 20 12:13:38 Can Anybody help me? Nov 20 12:13:50 koen: if you're referring to http://www.openssl.org/support/faq.html#LEGAL2, then won't OpenMoko get around that by being an OS that normally distributes openssl ? Nov 20 12:15:14 rwhitby: it also depends at what point "linking" occurs :-) I can prove it doesnt occur until after distribution :-) Nov 20 12:15:53 rwhitby: It's an area where you have to send in the lawyers to investigate Nov 20 12:16:07 As long as the OpenMoko distro (as delivered on the phone) includes openssl by default, then the GPL is very clear. Nov 20 12:16:29 libmokossl could be GPL :-) Nov 20 12:18:06 "However, as a special exception, the source code distributed need not include anything that is normally distributed (in either source or binary form) with the major components (compiler, kernel, and so on) of the operating system on which the executable runs, unless that component itself accompanies the executable." Nov 20 12:18:40 So as long as you don't prelink openssl into openssh, and you distribute openssl by default with the OS, then you're fine. Nov 20 12:18:51 rwhitby: let me guess, SlugOS already looked at that closely :-) Nov 20 12:19:13 XorA: yep. Nov 20 12:19:42 rwhitby: ah good to know thanks Nov 20 12:20:05 After dealing with the Intel NPE code license, the GPL is a piece of cake :-) Nov 20 12:20:24 of course we are not laywers, blah blah blah ... Nov 20 12:20:42 rwhitby: the lawyers decide what they are paid to decide anyway Nov 20 12:20:53 (have to say that in case anyone reading is in the USA) Nov 20 12:21:03 the only true opinion comes from trial Nov 20 12:21:33 and even then it's only in one jurisdiction and subject to appeal or overturn by a higher court. Nov 20 12:22:45 XorA: I haven't looked to see what GPLv3 says about this issue though. Nov 20 12:23:54 lets leave licenses to the lawyers :-) Nov 20 12:25:01 no, let's just ignore the lawyers and do what makes sense and is ethical. Nov 20 12:25:33 rwhitby: It's not at all clear when the system exception kicks in. Nov 20 12:26:09 broonie: seemed pretty clear to me reading the GPL, but I won't try and change your opinion. Nov 20 12:26:46 I mean they would have highlighted it as a special exception if they didn't want it to be an exception to be used. Nov 20 12:26:56 s/would/wouldn't/ Nov 20 12:26:58 rwhitby: I'm just relaying what Debian ended up deciding - we were worried that since we don't distinguish between packages in any meaningful way all packages in Debian are part of the "system" so we can't use it. Nov 20 12:27:30 broonie: yeah, but Debian is a big organisation with laywers, so they had to make that decision. Nov 20 12:27:36 ahahahahah Nov 20 12:27:52 (and they don't have a firmware rootfs like an embedded system) Nov 20 12:28:22 Big organisation, yes but the lawyers are generally amateur. Nov 20 12:28:43 anyway, getting back to on-topic, how should I tell Angstrom/OE to use GCC 3.x when doing the configure script for glibc 2.2.5 ? Nov 20 12:29:20 I've tried PREFERRED_VERSION for gcc-cross-initial, gcc-cross and gcc, but the configure script still finds gcc-4.x and barfs. Nov 20 12:29:55 broonie: amateur shysters :-D Nov 20 12:30:13 (glibc 2.2.5 is required for vendor library compatibility in the Unslung distribution, so I can't use a more recent glibc to solve the problem) Nov 20 12:32:45 hmm - actually it's glibc-initial 2.2.5 which has the problem, but I presume glibc 2.2.5 will have it too. Nov 20 12:36:57 * lardman keeps his ears open as my opengl libs are wanting glibc 2.2.x Nov 20 12:37:56 lardman: the problem arises when you move your host distro from gcc 3 to gcc 4 (i.e. from Debian Sarge to Ubuntu 7.10) Nov 20 12:38:26 rwhitby: ah, okay, I've no idea what mine is atm Nov 20 12:38:54 Can't you bitbake a native toolchain too? Nov 20 12:39:08 Looks like I need to change BUILD_CC in glibc-initial.inc to a native toolchain of gcc 3.x Nov 20 12:43:07 lardman: does gcc_3.4.4.bb stage by default? Nov 20 12:43:32 if so, I could just DEPEND on that, and then set BUILD_CC in the glibc-initial_2.2.5.b to point to that staged host gcc Nov 20 12:44:18 XorA: Well, it is pretty much the intent of that clause in the GPL and the FSF have made noises about it for their own stuff. Nov 20 12:45:10 rwhitby: no idea, been out of the loop for a while Nov 20 12:45:37 nice to see you all, got to go to a funeral, should be back later on, bye Nov 20 12:45:40 lardman: looks like I need a gcc-native_3.4.4.bb instead of a gcc_3.4.4.bb Nov 20 13:32:02 spitz autobuild images available at http://www.angstrom-distribution.org/unstable/autobuild/spitz/ Nov 20 13:34:35 hmm, it seems GPL3 allows advertising clauses, so as long as programs say that you can use GPL2 or later version all is fine Nov 20 13:37:36 koen: where you run autobuilder? ewi? Nov 20 13:41:35 hrw: yes Nov 20 13:42:08 hrw: and as soon as I unbreak fuse on my workstation we'll have a 2008 autobuilder as well Nov 20 13:42:29 hrw: powerpc machines are probably moving to digital opsis next week Nov 20 13:42:40 when stelios and I have some time to work out the details Nov 20 14:10:20 bye all! Nov 20 15:31:33 yay Nov 20 15:31:35 hello Nov 20 15:36:52 ho cofeineSunshine Nov 20 15:52:20 hi chaps Nov 20 15:52:46 Right, so for oabi, gcc <4 should naturally work? Nov 20 15:53:20 yes Nov 20 15:53:28 and gcc >=4 as well Nov 20 15:54:17 http://www.angstrom-distribution.org/unstable/autobuild/collie/ is oabi with gcc 4.1.2 Nov 20 15:54:51 so how is it selected? I had a quick look through the conf files and couldn't spot anything obvious? Nov 20 15:54:58 gcc or oabi? Nov 20 15:55:04 oabi :) Nov 20 15:55:13 why is collie still oabi anyway? Nov 20 15:55:19 kernel? Nov 20 15:55:27 strongarm cpu doesn't support oabi Nov 20 15:55:31 ah, ok Nov 20 15:55:34 you need the 'bx' instruction Nov 20 15:55:39 which is thumb Nov 20 15:56:00 so is it a configure switch for gcc then? Nov 20 15:56:12 strongarm supports oabi Nov 20 15:56:19 it does not support eabi Nov 20 15:56:21 ah right, I typo'ed Nov 20 15:56:35 ah, I read through it and understood Nov 20 15:56:41 Powerbook-2:distro koen$ grep simpa . -rn | grep collie Nov 20 15:56:41 ./angstrom-2007.1.conf:202:ARM_ABI ?= "${@['','oabi'][bb.data.getVar('MACHINE',d) in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Nov 20 15:56:41 ./angstrom-2008.1.conf:164:ARM_ABI ?= "${@['','oabi'][bb.data.getVar('MACHINE',d) in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Nov 20 15:56:41 ./include/angstrom-eglibc.inc:17:TARGET_OS = "linux${@['','-gnueabi'][bb.data.getVar('TARGET_ARCH',d,1) in ['arm', 'armeb'] and bb.data.getVar('MACHINE',d) not in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Nov 20 15:56:43 ./include/angstrom-glibc.inc:7:TARGET_OS = "linux${@['','-gnueabi'][bb.data.getVar('TARGET_ARCH',d,1) in ['arm', 'armeb'] and bb.data.getVar('MACHINE',d) not in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Nov 20 15:56:46 ./include/angstrom-uclibc.inc:10:TARGET_OS_UC = "linux${@['-uclibc','-uclibcgnueabi'][bb.data.getVar('TARGET_ARCH',d,1) in ['arm', 'armeb'] and bb.data.getVar('MACHINE',d) not in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Nov 20 15:57:00 the machines in those statement get oabi Nov 20 15:57:20 ah, good good Nov 20 15:57:24 selecting eabi is just a matter of changing TARGET_OS to foo-bar-gnueabi Nov 20 15:57:43 can the opposite be done so I don;t need to mess with the defaults? Nov 20 15:57:59 e.g. nokian800-gnuoabi Nov 20 15:58:09 yes, add you machine to the ['collie\, simpad ] statements Nov 20 15:58:15 blech Nov 20 15:58:18 I can't type today Nov 20 15:58:22 :) Nov 20 15:58:48 then by default nokia800 with be oabi? But what if I want the default to be eabi, can I use the same construct as above as the machine type? Nov 20 15:59:04 to force oabi Nov 20 16:00:31 or should that be glibc-gnueabi? Nov 20 16:01:09 any arm machine in the collie, simpad, etc list will get oabi, otherwise eabi Nov 20 16:01:21 ok, will modify Nov 20 16:01:49 Next questions: any ideas of matching binutils for gcc 3.4.4, does it matter (I assume so)? Nov 20 16:02:03 I'm thinking 2.16 Nov 20 16:02:19 I have no idea, sorry Nov 20 16:02:27 I also think I need to use, probably glibc 2.1.x Nov 20 16:03:27 Nokia-N800-:/home/user/pvr# /lib/ld-linux.so.3 --list ./power_test Nov 20 16:03:27 ./power_test: /lib/libc.so.6: version `GLIBC_2.0' not found (required by ./power_test) Nov 20 16:03:28 ./power_test: /lib/libc.so.6: version `GLIBC_2.1' not found (required by /home/user/pvr/libsrv_um_1.1.33.561.so) Nov 20 16:03:28 ./power_test: /lib/libc.so.6: version `GLIBC_2.0' not found (required by /home/user/pvr/libsrv_um_1.1.33.561.so) Nov 20 16:03:29 libsrv_um_1.1.33.561.so => /home/user/pvr/libsrv_um_1.1.33.561.so (0x40004000) Nov 20 16:03:30 libdl.so.2 => /lib/libdl.so.2 (0x41220000) Nov 20 16:03:32 libc.so.6 => /lib/libc.so.6 (0x41028000) Nov 20 16:03:34 /lib/ld-linux.so.2 => /lib/ld-linux.so.3 (0x2a000000) Nov 20 16:03:45 that's my problem, I'd like to build an image on which I can run these things Nov 20 16:05:03 how backward compatible is glibc? Might those errors be caused by my trying to run it on an eabi system? Nov 20 16:05:13 with glibc 2.5.x Nov 20 16:10:47 hmm, strings \usr\lib\libc.so.6 | grep GLIBC on the n800 produces GLIBC_2.4, GLIBC_2.5, GLIBC_PRIVATE Nov 20 16:23:40 I'll try 2.4 and see if that can do the backwards compatibility Nov 20 16:50:09 NOTE: preferred version 2.4 of glibc not available (for item virtual/libc) Nov 20 16:50:12 hmm? Nov 20 16:51:47 there seems to be a glibc_2.4.bb file. Any ideas? Nov 20 16:56:02 hmm, typo on my part probably, nokian800 rather than nokia800 Nov 20 17:13:50 aaargh, why is it grabbing glibc-2.7? Nov 20 17:17:28 anyone out there with a current angstrom build running on something oabi? Nov 20 17:18:32 if so, or actually probably easier on the build machine, could you tell me what this outputs?: strings \usr\lib\libc.so.6 | grep GLIBC (alter path to libc.so.6) Nov 20 17:24:12 right, from the feed, looks like it contains the strings we need Nov 20 17:24:14 good good Nov 20 17:39:27 <_charly_> hi :) Nov 20 17:45:18 koen: is somewhere described, how autobuilder chooses packages for the images? It bugged me a bit, that after flashing today's image, there were a lot of things to 'ipkg upgrade' Nov 20 17:45:45 there's a lockdown file that sets all the versions Nov 20 17:45:49 but it needs updating Nov 20 17:46:09 see http://bugs.openembedded.org/show_bug.cgi?id=3338 Nov 20 17:49:36 Hmm. Why is it necessary if there is already preferred-versions file? Nov 20 17:59:24 package glibc-intermediate-2.5-r8: task do_configure: failed -> configure: error: cannot compute suffix of object files: cannot compile Nov 20 17:59:43 evening! Nov 20 17:59:52 hi Nov 20 18:00:12 how do i properly restart my ipaq? Nov 20 18:00:28 * CoreDump wouldn't know Nov 20 18:00:42 but "reboot" should work for you as well` Nov 20 18:01:30 ok Nov 20 18:01:39 today i bought 2gb sd card Nov 20 18:02:33 installing new angstrom Nov 20 18:31:28 GCC 3.4.4 doesn't seem to like -mfpu -mfloat-abi and -mtune Nov 20 19:19:02 anyone know what version of GCC added nptl support? Nov 20 19:19:13 * CoreDump doesn't Nov 20 19:20:57 does __thread support live in GCC? That's where I ought to be looking? Nov 20 19:27:58 Presumably glibc 2.5 wasn't designed (or at least the .bb) to be compiled with gcc 3.4.4 Nov 20 19:28:20 glibc 2.5 has GLIBC_ADDONS ?= "ports,nptl,libidn" should I alter nptl to linuxthreads? Nov 20 19:29:54 or even better, how does one tell gcc to use nptl Nov 20 19:35:23 * lardman wonders if it's possible to tell the threading model of a binary somehow? Nov 20 19:36:17 __thread support in gcc is enable if you build it against a thread enable libc Nov 20 19:36:41 toolchain voodoo Nov 20 19:37:05 so my glibc_inital (or whatever it;s called) is lacking __thread? Nov 20 19:37:28 or rather thread support Nov 20 19:42:54 hi Nov 20 19:43:12 Can GCC 4.x.x binaries link against GCC 3.x.x libraries or are there name-mangling issues? Nov 20 19:43:25 hi mimecar Nov 20 19:43:33 evening lardman Nov 20 19:44:35 I don't think it can. hmm. Nov 20 19:45:16 me neither Nov 20 19:45:55 lardman: chrooting into a collie rootfs doesn't do the trick? Nov 20 19:48:41 koen: it's an option, but I would prefer to build an image as I know very little about chroot Nov 20 19:48:58 though this is painful, so I'm more and more tempted all the time :) Nov 20 19:49:04 lardmanwhy don't you compile with gcc 3? Nov 20 19:49:20 I want to compile with GCC 3.4.4 Nov 20 19:49:32 in fact I am compiling with that Nov 20 19:49:35 and ggc4...? Nov 20 19:49:59 the default images use gcc 4.x.x, I was just thinking it would be far less effort Nov 20 19:50:46 but they won't work, I need gcc 3.4.4. Not sure about libc version, if the compat flags are in there (they were for a collie libc 2.5 binary) then that's fine Nov 20 19:53:02 first, build with 3.4, after, if you want play with gcc 4 Nov 20 19:53:46 mimecar: I have some binaries I want to run and reverse engineer, I need to use gcc 3 anyway Nov 20 19:54:09 but I can't buld an image using gcc 3.4.4 Nov 20 19:54:24 if you need I compile some packages for collie... Nov 20 19:54:34 no, this is for a Nokia N800 Nov 20 19:54:44 ok Nov 20 19:54:46 koen said collie as I also need OABI Nov 20 19:55:39 OAB? Nov 20 19:56:17 http://wiki.debian.org/ArmEabiPort Nov 20 19:57:04 ok Nov 20 20:07:23 would be a change on build structure? Nov 20 20:08:10 yeah probably, I don't know if version 2.5 of glibc likes compiling with GCC 3.4.4 Nov 20 20:08:44 I should probably go back to the days before eabi and grab a working set of metadata from there and build an image with that Nov 20 20:08:59 I'll have to see what I have backed up on my computer Nov 20 20:09:11 then, should i stop on working with collie ? Nov 20 20:09:16 would be nice to use the latest versions though Nov 20 20:09:20 mimecar: no idea Nov 20 20:09:36 mimecar: my problems don't affect you Nov 20 20:10:01 if the're some plans to do the same with collie... Nov 20 20:10:06 althought you work on other device Nov 20 20:10:17 this work of mine is a one-off Nov 20 20:10:23 ok Nov 20 20:23:03 Mu gcc-cross-initial has --disable-threads passed to its configure Nov 20 20:23:16 is this usual, or could it be part of my problem? Nov 20 20:25:12 nope, looks like that's normal, it's in gcc-cross-initial.inc Nov 20 20:36:21 yes, since you need a c library for thread support in gcc Nov 20 20:36:32 and since you need gcc to build a c library... Nov 20 20:37:34 but it's when I try building glibc-initial (I think - the one that's build by gcc-initial) that I have the problem Nov 20 20:38:16 glibc-intermediate in fact Nov 20 20:39:12 So I'm not sure how to sort this out or indeed where? Nov 20 20:39:46 My only thought is that my version of gcc (3.4.4) doesn't like nptl, which is what glibc (2.5) expects Nov 20 20:42:33 hmmm Nov 20 20:42:56 you could try building for the openzaurus branch Nov 20 20:43:04 s/for/from/ Nov 20 20:43:15 that gives you gcc, old glibc, oabi and no nptl Nov 20 20:45:01 could it be, that my ipaq cant boot using 2gb sd card instead 256mb Nov 20 20:48:35 koen: genius, sounds like a plan Nov 20 20:48:45 cofeineSunshine: could be Nov 20 20:49:18 koen: who's the n800/770 maintainer these days? Nov 20 20:49:48 lardman: we life off patches from poky and mamona Nov 20 20:50:02 I started needing it to get me where I need for work :) Nov 20 20:50:18 koen: any native dsp support, or all from the Nokia images? Nov 20 20:50:27 and I just sold my n800 to finance the n810, so that leaves the 770 to play with Nov 20 20:50:36 no idea on dsp support Nov 20 20:50:48 I have a blackfin board for DSP stuff :) Nov 20 20:51:07 koen: cool, you working doing dsp stuff then? Nov 20 20:51:16 going to Nov 20 20:51:29 try to find the right courses at the uni Nov 20 20:51:41 right now I'm a dsp n00b with lots of devboards :) Nov 20 20:52:11 it's quite interesting working on the N800/770 DSPs Nov 20 20:52:48 yeah, I followed your work on it Nov 20 20:52:55 the few emails I read Nov 20 20:52:59 I've got too many balls in the air though - ogg task, mp3 task, scaling/colourmap conversion tasks to all be debugged Nov 20 20:53:32 powervr mbx caught my fancy though so I'll do a bit of work on that then go back to the dsp stuff Nov 20 20:54:56 hmm, monotone tells me that branch 'org.openembedded.openzaurus' is empty Nov 20 20:55:16 .oz354x Nov 20 20:55:16 ah, name wrong Nov 20 20:55:26 see http://monotone.openembedded.org/ Nov 20 20:55:35 http://www.openembedded.org/wiki/DevelopmentBranches ? Nov 20 20:57:58 no :) Nov 20 20:58:16 http://monotone.openembedded.org/ lists all branches Nov 20 20:58:31 so you know how to (mis)spell org.oe.oz354x Nov 20 20:58:41 lol Nov 20 20:58:49 you probably need bitbake 1.4 or 1.6 as well Nov 20 20:59:02 I'm wondering if the oz354fam083 has 770 conf files already Nov 20 20:59:18 yes, I should have an older version somewhere Nov 20 20:59:37 .oz354fam083 is the predecessor of .oz354x Nov 20 20:59:49 it's young enough Nov 20 21:00:12 what I need to build has obviously been build using a very old system, glibc 2.1 Nov 20 21:00:16 gcc 3.4.4 Nov 20 21:01:18 anyway, I'll have to take a look at it tomorrow evening, being summoned to watch CSI & drink wine, not too bad a thing to have to do :) Nov 20 21:01:32 koen: thanks for your help Nov 20 21:01:51 enjoy the wine and CSI Nov 20 22:12:42 bye all Nov 21 00:19:50 night **** ENDING LOGGING AT Wed Nov 21 02:59:56 2007