**** BEGIN LOGGING AT Tue Jun 03 02:59:57 2008 Jun 03 03:44:07 * * OE Bug 4351 has been VERIFIED (REMIND) by xjqian(AT)gmail.com Jun 03 03:44:09 * * navit cvs migrated to svn; edit navit_cvs.bb Jun 03 03:44:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4351 Jun 03 04:20:06 * * OE Bug 4351 has been RESOLVED (INVALID) by udovdh(AT)xs4all.nl Jun 03 04:20:09 * * navit cvs migrated to svn; edit navit_cvs.bb Jun 03 04:20:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4351 Jun 03 04:20:30 ? Jun 03 04:23:06 * * OE Bug 4351 has been UNCONFIRMED by udovdh(AT)xs4all.nl Jun 03 04:23:08 * * navit cvs migrated to svn; edit navit_cvs.bb Jun 03 04:23:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4351 Jun 03 07:35:16 morning Jun 03 07:35:33 oe wiki is demolished in some pages... is it known? Jun 03 07:35:48 like OEandYourDistro Jun 03 07:51:46 google cache still contains last good page Jun 03 08:06:48 gints: damn wiki software :-( Jun 03 08:07:50 gints: looks like someone already repaired it Jun 03 08:10:56 03koen 07org.oe.dev * rfe8c6cc9... 10/ (6 files in 5 dirs): linux-neuros git: add linux recipe for the osd1 and osd2 devices Jun 03 08:11:01 03koen 07org.oe.dev * r0ae573df... 10/ (4 files in 3 dirs): fontconfig: fix configure Jun 03 08:15:16 <_acid-burn_> Hello Jun 03 08:15:28 <_acid-burn_> Someone from The Openembedded Team here ? Jun 03 08:17:12 morning Jun 03 08:17:20 03koen 07org.oe.dev * r792c430f... 10/ (3 files in 2 dirs): mplayer: enable v4l2 (not v4l2-tv), usefull for people with overlay hardware (e.g. omap, davinci, x86 video cards) Jun 03 08:17:25 03koen 07org.oe.dev * rd644ad9b... 10/ (1 packages/makedevs/makedevs_1.0.0.bb): makedevs: fix compile with gcc 4.3.x Jun 03 08:25:37 03koen 07org.oe.dev * r408b7cc4... 10/ (1 packages/linux/linux-neuros_git.bb): linux-neuros git: bump SRCREV and switch branch Jun 03 08:26:10 hi hrw Jun 03 08:29:05 _acid-burn_: whats up? Jun 03 08:30:31 <_acid-burn_> hello XorA just wanted to know if its possible to add a mail list for the dreambox branch commits, its somehow hard to find only the ones i am intrested in the general commits list Jun 03 08:30:59 03koen 07org.oe.dev * rdd5efe86... 10/ (1 conf/machine/neuros-osd.conf conf/machine/neuros-osd2.conf): neuros-osd*: build uImage as kernel image Jun 03 08:31:05 _acid-burn_: best to email the OE list with that request, the sysadmins arent around at the moment Jun 03 08:31:18 <_acid-burn_> ahhh great. thx Jun 03 08:35:33 03koen 07org.oe.dev * rf275c97e... 10/ (1 conf/machine/neuros-osd.conf conf/machine/neuros-osd2.conf): neuros-osd*: add uboot adresses (taken from Makefile.boot) Jun 03 08:56:36 :( shasum-native does not work at my box after build Jun 03 09:24:03 good morning Jun 03 09:29:02 morning Jun 03 09:53:33 hi Jun 03 09:54:27 after building an image with bitbake, there is a way to build the .ipkg for all programs in the image? Jun 03 09:55:21 The ipks should already be there Jun 03 09:56:00 image is built from packages Jun 03 09:57:12 ok, i'll double check... Jun 03 10:38:17 hmm, I need to disable the automounter on my Z. With swap on a CF card, it's become a liability Jun 03 10:38:52 it actually manages to mount the swap partition as vfat Jun 03 10:41:42 i tried to tar my tmp-dir. tar cvf tmp.tar tmp Jun 03 10:41:44 tar: Error exit delayed from previous errors Jun 03 10:42:03 is this not a good idea? Jun 03 10:48:41 markus: you need to scroll back through the tar output to find the actual error(s), that one doesn't really say anything Jun 03 10:49:11 sometimes the error is just a timestamp warning, and doesn't really affect anything at all Jun 03 10:51:26 markus: leave away the 'v' Jun 03 10:51:36 in that way you'll only see errors and warnings Jun 03 11:01:24 hi. could someone point me in the right direction for the src_publish / distribute_sources thingie, i understand it's supposed to help in releasing relevant source packages easily but can't make it "just work" Jun 03 11:02:36 Tv: SRC_DIST_LOCAL = "symlink" and INHERIT += "src_distribute_local" into conf/local.conf Jun 03 11:02:55 ah, thanks, i'll read up on those Jun 03 11:03:30 hm.. OE has old version of those classes Jun 03 11:04:20 now i see http://oe.linuxtogo.org/user-manual&dpage=src_uri_variable#id2609292 Jun 03 11:04:48 putting the pieces together without that usage example was just not obvious Jun 03 11:05:21 moment - I am merging Poky version which is much simplier Jun 03 11:06:07 http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commit;h=ec171eff844bfb15568a82401bcd6efdfbd08156 Jun 03 11:07:17 RP: do we have git->mtn sync too? Jun 03 11:10:02 hrw: i'm trying to reproduce a bug that claims that dependencies of things installed aren't included in source distribution Jun 03 11:10:16 hrw: waiting for things to build, but if you know something about that off hand, i'd appreciate it Jun 03 11:10:50 Tv: change which I just pushed to git tree makes every source distributed Jun 03 11:11:02 for each built package Jun 03 11:11:15 hrw: umm, that's definitely not what some companies want Jun 03 11:11:54 or wait, i guess it still does the license check? i see the | splitting etc Jun 03 11:12:23 umm, but i don't see any conditionality inside the license loops Jun 03 11:12:49 oh you split the source tree by license? Jun 03 11:13:04 yes Jun 03 11:13:08 that'd work Jun 03 11:13:26 so the person who did build can then select which licenses sources needs to publish Jun 03 11:13:28 but the bug sounds more like that whole code wasn't triggered for the dependencies Jun 03 11:13:40 please report any wrong license btw Jun 03 11:14:00 Tv: that class is quite old and probably no one checked it during last half year Jun 03 11:14:44 well, as soon as i fight out these things in the way of a good build, i'll see if it's buggy or not Jun 03 11:14:49 just haven't built this project in a while Jun 03 11:14:51 bitrot Jun 03 11:15:04 hrw: no Jun 03 11:20:15 RP: ok - will have to push to mtn then too Jun 03 11:27:54 03hrw 07org.oe.dev * r6c9b3832... 10/ (1 classes/src_distribute.bbclass): src_distribute.bbclass: merge Poky version which handle all licenses Jun 03 13:13:18 Crofton: ping Jun 03 13:13:46 we kind of got sound working on 2.6.26-rc4 kernel git Jun 03 13:13:57 http://marc.info/?l=linux-omap&m=119634462131463&w=2 Jun 03 13:13:58 how so? Jun 03 13:14:59 can you email that to philip at balister dot org Jun 03 13:15:10 and remind me next week? Jun 03 13:15:19 aplay plays homer "doh", but there's annoying messages about ... Jun 03 13:15:19 I am swamped until then Jun 03 13:15:32 tps65010: power off button Jun 03 13:15:36 yeah Jun 03 13:15:38 i2c_omap i2c_omap.1: Transmit overflow Jun 03 13:15:41 I am seeing those also Jun 03 13:16:11 there may be some i2c regressions that need working out with the guys on linux-omap Jun 03 13:16:11 Crofton: ok ok, just wanted to inform you. Jun 03 13:16:23 thanks a lot Jun 03 13:16:39 np Jun 03 13:16:42 I really want to get this working, but the OSK is not high on my personal list :( Jun 03 13:17:08 once I get soundworking, I want to see if we can get the mistral screen going :) Jun 03 13:17:34 It is great people are trying to use this and helping debug :) Jun 03 13:19:15 =) Jun 03 13:24:36 Crofton: what did you want me to include in the email? Jun 03 13:26:29 just the link for now Jun 03 13:50:36 03mickeyl 07org.oe.dev * r3aaa97a4... 10/ (1 conf/machine/include/motorola-ezx-base.inc): motorola-ezx-base.inc: set XSERVER Jun 03 13:50:40 03mickeyl 07org.oe.dev * r74c5551c... 10/ (1 packages/images/fso-image.bb): fso-image: only include openmoko-alsa-scenarios for machines that need it Jun 03 13:54:49 hello Jun 03 13:57:20 hi udovdh Jun 03 13:57:39 anyone busy with `my` other bugs? Jun 03 13:57:41 ;-) Jun 03 13:58:17 I am still busy with the QA errors I get Jun 03 13:58:32 QA_LOG=1 does not really log anything extra Jun 03 14:11:38 03koen 07org.oe.dev * r430ec523... 10/ (1 packages/dsplink/codec-engine_2.10.bb): codec-engine: add skeleton for 2.10, we only need cmem.ko from it Jun 03 14:23:18 umm, it seems that the Jun 03 14:23:19 addtask distribute_sources before do_build after do_fetch Jun 03 14:23:31 line in src_distribute.bbclass does nothing Jun 03 14:23:48 i get tmp/deploy/sources if i call bitbake -c distribute_sources foo Jun 03 14:23:54 but not if i just build a rootfs image Jun 03 14:24:21 do i need to do something extra to actually make it be used by default? Jun 03 14:36:56 what sets the default task if -c is not given? Jun 03 14:37:33 berlios.de seems to be down :( Jun 03 14:39:23 hmm BB_DEFAULT_TASK Jun 03 14:40:06 but if that addtask says "before do_build" doesn't that mean it should do distribute_sources always? Jun 03 15:03:51 hmm maybe i just wasn't cleaning things enough Jun 03 15:14:41 03mickeyl 07org.oe.dev * re112dc95... 10/ (1 packages/libusb/libusb packages/libusb/files): libusb-native: fix build Jun 03 15:20:06 * * OE Bug 3938 has been REOPENED by yan(AT)seiner.com Jun 03 15:20:08 * * Marvel 8385 libertas driver will not wake from suspend 2.6.24 Jun 03 15:20:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3938 Jun 03 15:45:14 who can help fix the QA bug in http://bugs.openembedded.net/show_bug.cgi?id=4348 ? Jun 03 15:52:28 bye all Jun 03 16:15:55 hrw|gone: your new nokia n810 usb record is heavy. I remember your zaurus usb hub blog entry :) Jun 03 16:17:06 * * OE Bug 4354 has been created by udovdh(AT)xs4all.nl Jun 03 16:17:08 * * mtd-utils don't work... Jun 03 16:17:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4354 Jun 03 16:17:46 I created a separate bug for the mtd issue, apart from the LAB problem Jun 03 16:35:54 udovdh: did you try using mtd-utils 1.1.0 or git? Jun 03 16:36:12 can I do that on 2007.12? Jun 03 16:36:46 http://www.angstrom-distribution.org/repo/?action=details&pnm=mtd-utils Jun 03 16:37:00 I can't build programs yet. I get QA errors Jun 03 16:37:12 so if you have a 1.1.0 for 2007.12: please! Jun 03 16:37:21 (thanks for the suggestion) Jun 03 16:38:13 stable doesn't know of 1.1.0 for mtd Jun 03 16:38:46 but I can try to build. Jun 03 16:47:06 hi guys. when i bitbake virtual/kernel .. am I getting the modules too? Jun 03 16:50:10 what would be right way to set up angstrom to build with uclibc? Jun 03 16:51:08 target_os is overided by distro conf Jun 03 16:53:37 gints: set ANGSTROM_MODE="uclibc" in a conf file Jun 03 16:53:52 cbrake many thx Jun 03 16:53:59 gints: sure :-) Jun 03 16:54:17 http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2008-February/001920.html ? Jun 03 16:55:06 * * OE Bug 3379 has been marked as DUPLICATE of bug 3378 by c.romain(AT)laposte.net Jun 03 16:55:08 * * gmp-native-4.2.1-r2-do_compile Jun 03 16:55:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3379 Jun 03 16:55:25 morning all Jun 03 16:56:21 hi? Jun 03 17:13:21 mickey_away: Hi. Does the bread and butter image build from OE already? Jun 03 17:21:54 likewise, 1.1.0 builds but has QA errors (which look for unrelated packages) Jun 03 17:21:54 when I use 1.1.0 for erasing flash it oopses Jun 03 17:22:05 udovdh: argh Jun 03 17:22:09 udovdh: not good Jun 03 17:22:27 wnat do I do next? Jun 03 17:22:57 could you build 1.1.0 for h2200 for 2007.12? Jun 03 17:22:57 if that also oopses the oops issue is real Jun 03 17:23:18 udovdh: uhm, I don't have that hardware Jun 03 17:23:35 don't need it Jun 03 17:23:39 just a machoen that runs OE Jun 03 17:23:44 machine... Jun 03 17:23:52 set machine to h2200 Jun 03 17:23:59 and distro to angstrom-2007.1 Jun 03 17:24:03 and bitbake mtd-utils Jun 03 17:24:06 udovdh: what oopses? the host machine oopses??? Jun 03 17:24:10 no Jun 03 17:24:12 the ipaq Jun 03 17:24:34 host is pretty stable Jun 03 17:24:34 so what Jun 03 17:24:38 ssorry Jun 03 17:24:40 so what Jun 03 17:24:44 ARGH Jun 03 17:24:59 is anyone of you experienced with adding custom recipes? Jun 03 17:25:05 so what's the use of me building mtd-utils 1.1.0 (sorry for the typos, hit enter when I meant ') Jun 03 17:25:11 i've got some troubles with integrating x264 into mplayer Jun 03 17:34:03 ? Jun 03 17:34:03 brb Jun 03 17:41:04 likewise, it could avoid local building issues on my setup Jun 03 17:41:16 I get QA errors where others dont have them Jun 03 17:42:01 hi, all! Jun 03 17:42:04 it could also avoid operator stupidity Jun 03 17:42:05 ;-) Jun 03 17:42:08 udovdh: then you should fix those first Jun 03 17:42:23 udovdh: can you compare with "others" ? Jun 03 17:42:43 yes, if only I knew how. Jun 03 17:42:48 I can't get the details on the QA errors. just know that there are errors but not why. Jun 03 17:42:53 is there some package in OE which reads/writes /dev/rtc* alarm? Jun 03 17:43:03 date? Jun 03 17:43:10 * slapin needs to test a driver Jun 03 17:43:23 if I knew the QA error details I could try to fix them or perhaps see a common cause Jun 03 17:43:24 udovdh, it soesn't read alarm. Jun 03 17:44:24 something which uses RTC_ALM_SET, RTC_ALM_READ ioctls Jun 03 17:44:24 hmmm. dunno then.. was guessing. Jun 03 17:51:14 shutdown? Jun 03 17:51:23 suspend kinda tools? Jun 03 17:51:29 wake up at xx time? Jun 03 17:55:11 Crofton: ping! Jun 03 17:55:47 I'm getting a failure on the tiff package with a fresh pull/update Jun 03 17:56:06 seem to recall you having a similar issue Jun 03 17:56:36 please don't say "clean build"! Jun 03 18:35:58 sakoman, not me Jun 03 18:36:42 hi crofton Jun 03 18:36:48 gm Jun 03 18:37:03 hi there Jun 03 18:37:18 could someone point me in a direction on how to use bitbake with distcc? Jun 03 18:37:42 tomboy??? I said you can use icecream Jun 03 18:37:47 sakoman, I'm going to do a pull rebuild now though Jun 03 18:38:12 woglinde: i also found hints that distcc can also be used. and since i don't have to use a centralized server i'd prefer that solution Jun 03 18:40:45 tomboy64: icecream is like distcc, it's purpose is to distribute compilation job Jun 03 18:41:17 cyrilRomain: but for icecream i'd have a centralized topology Jun 03 18:41:20 tomboy64: icecream was even created from distcc code iirc Jun 03 18:41:33 yes, it was. Jun 03 18:41:47 and i'd much prefer a decentralized one. Jun 03 18:43:42 tomboy64: hm ok. Shame that icecream doesn't seem to be able to run decentralized :-/ Jun 03 18:44:28 i never said it can't. it just says on it's homepage it uses a centralized server. Jun 03 18:45:03 and since i have a running distcc environment and am quite happy with it it's much easier to get bb to run on that - along with the topology feature Jun 03 18:46:53 tomboy64: I think the choice for a central server was to make the setup more easy Jun 03 18:47:09 uhm Jun 03 18:47:12 probably Jun 03 18:47:26 but once you know how distcc works it's easy to setup, too Jun 03 18:49:48 * XorA|gone wonders how the hell you do a compile without at least one node being the boss Jun 03 18:50:00 in that way icecream and distcc are the same Jun 03 18:50:19 well, you run a server on every node that acts as a client. Jun 03 18:50:30 ok, that sounds idiotic. Jun 03 18:51:04 distcc: every computer that runs distccd offers its cpu for compilation. with restrictions imposed by the config. Jun 03 18:51:24 every other computer can connect to that computer and tell what it wants to have compiled. Jun 03 18:52:16 distcc every computer has to be identical though :-( Jun 03 18:52:23 nope. Jun 03 18:52:39 yes, as it uses local gcc /usr/lib/* etc etc Jun 03 18:53:06 unless they fixed that since I gave up on it Jun 03 18:53:33 well, you'd have to properly setup cross-compiling, that's true. Jun 03 18:53:52 AFAIk icecream pushes the right env to the machines Jun 03 18:54:09 but using amd64-machines to produce 32-bit code works quite well - w/o the cross-compiling stuff Jun 03 18:54:21 well, that would be a feature i'm not in need of. Jun 03 18:54:50 maybe its just because I lived through the gcc abi breakage Im less trusting :-) Jun 03 18:54:58 :-D Jun 03 18:55:06 2->3? Jun 03 18:55:26 erm something like 4.0.0 -> 4.0.1 broke C++ Jun 03 18:55:36 never noticed. Jun 03 18:55:48 only started using the 4.1 branch Jun 03 18:55:51 it was a bug in the original release that just had to be fixed, but it was a pain Jun 03 18:56:43 boy, glibc sucks! Jun 03 18:57:27 Ill agree with that :-) Jun 03 18:57:30 mhm Jun 03 18:57:48 do you happen to have an idea on how to change the $PATH bitbake sees? Jun 03 18:58:00 nope Jun 03 19:16:21 morning Jun 03 19:16:33 morning Jun 03 19:17:02 hi hrw Jun 03 19:18:22 hi fique Jun 03 19:18:35 hello woglinde Jun 03 19:37:18 has anyone tried to build with llvm instead of gcc ? (I remember koen wanting to give it a try, but I'm wondering about the results) Jun 03 19:44:16 Anyone else running into a tiff build issue? Error is: Jun 03 19:44:18 | ../libtool: line 784: X--tag=CC: command not found Jun 03 19:44:47 which libtool ver? Jun 03 19:45:34 hmm . . . seems to have built 1.5.10 Jun 03 19:45:45 should work Jun 03 19:47:11 hmm . . . "koen org.oe.dev * r74076120... / (1 packages/libtiff/tiff_3.8.2.bb): tiff: update to 4.0.0beta2 for libtool fixes" Jun 03 19:47:35 have a nice evening guys Jun 03 19:48:05 he had me check that some ncurses fixes did not break libtool 1.510 yesterday Jun 03 19:49:58 well, tiff is definitely breaking now :-( Jun 03 19:50:37 can you change the preferred version Jun 03 19:50:56 RP: ayt? Jun 03 19:52:02 anyone, RP: Does anybody know, if there are some known problems with "__attribute__ ((unused))" in with linux-kernel and arm-gcc compiler ? Jun 03 20:00:38 helgeD: are you sure that the __ at the end is required? Jun 03 20:22:38 mwester: ayt? Jun 03 20:23:57 mwester-laptop: ayt? Jun 03 20:32:11 hi florian Jun 03 20:32:40 re Jun 03 20:46:48 03koen 07org.oe.dev * r91e97fa5... 10/ (1 packages/libtiff/tiff_3.8.2.bb): tiff 4.0.0beta2: def pref -1 Jun 03 20:54:20 g'morning Jun 03 20:55:40 hi ant Jun 03 20:56:10 woglinde: hallo Jun 03 21:00:28 sakoman, tiff dies for me also Jun 03 21:00:39 cd org Jun 03 21:00:53 Crofton: pull/update -- koen checked in a fix Jun 03 21:01:03 heh Jun 03 21:01:07 I was about to do that Jun 03 21:03:07 koen: you can skip that step and it will still work Jun 03 21:09:56 nobody volunteers for fixing a bad udev regression? Jun 03 21:09:59 Bug 4118 - udev-118 failed to create device nodes Jun 03 21:10:27 I've downgraded udev to 092 (same version from Angstrom 2007.12) Jun 03 21:18:55 xjqian: hi Jun 03 21:26:48 mwester, mwester-laptop: Are you there? Jun 03 21:32:53 libtiff is failing for me from a known bug in libtool. Is there an update to any bitbake files for this, or how can it be fixed? Jun 03 21:33:13 mtn pull, mtn up Jun 03 21:33:55 Crofton: if you're talking to me, i did that before asking ;) but thanks... I'll try it again to see if a fix has been checked in within the last hour or so Jun 03 21:34:08 I think it has :) Jun 03 21:34:47 oh wow, that would be great! I'll let you know shortly. thanks! Jun 03 21:34:56 I'm checking also Jun 03 21:36:18 mtn: modifying packages/libtiff/tiff_3.8.2.bb Jun 03 21:37:26 seems it was changed, now I'm trying to build Jun 03 21:38:43 looks good here Jun 03 21:39:16 is there a mailing list for monotone commits or an archive somewhere? Jun 03 21:39:53 helgeD_: Any luck with 2.6.24 updates? Jun 03 21:40:05 I'm new to OE and am just starting to use this. A lot of the documentation on the website is very un-helpful to me. There's a lot of "type this to build" but I don't know what it builds or does. Jun 03 21:40:53 it looks good here too! thanks Crofton and koen@openembedded.org (the person who checked in the fix!) Jun 03 21:42:33 hvontres|work: yes, I think I found the problem with WPA and kernel modules. Jun 03 21:42:48 hvontres|work: else, 2.6.24 works GREAT! Jun 03 21:42:57 hvontres|work: did you looked at the wiki page? Jun 03 21:46:41 helgeD_: I saw you updated it. So what is the problem with the crypto modules? Jun 03 21:50:13 it seems they are not in the default image, if you didn't compiled them into the kernel statically. Jun 03 21:50:21 pretty simple, but hard to find. Jun 03 21:53:24 helgeD_: ahh, so they get compiled bu not installed... tricky Jun 03 21:59:34 so, i got INHERIT += "src_distribute_local" in my distro .conf file, and i can call "bitbake -c distribute_sources foo" to get some sources into tmp/deploy/sources -- is that it, or should my rootfs image building automatically trigger the source publishing? i sort of understood it would happen without explicit request Jun 03 21:59:48 i don't see how to make it happen for *all* things i biuld Jun 03 22:00:41 hmmph Jun 03 22:00:58 except now that i cleaned and rebuilt one package, its source was there Jun 03 22:01:25 it seems there's an element i'm not seeing, as my last full clean build definitely didn't even create tmp/deploy/sources Jun 03 22:05:44 hvontres|work: I just sent a mail with details. Will try tomorrow again. Jun 03 22:05:49 Late now. Night! Jun 03 22:05:58 helgeD_: Sounds cool Jun 03 22:18:02 hey materdaddy Jun 03 22:18:10 yo yo yo senor loudawg Jun 03 22:19:31 can we please slow down the chattiness a bit in here? I'm having trouble keeping up with it all :-P Jun 03 22:20:05 if I were an OP, I'd kick/ban you Jun 03 22:20:24 haha Jun 03 22:21:25 can someone explain 'recrdeptask' to me? or point be to TFM Jun 03 22:23:26 so here's my theory: src_distribute only includes sources of things explicitly listed to bitbake, but none of the dependencies built automatically Jun 03 22:36:09 if i add this at the end of src_distribute.bbclass, it starts to work (i think) Jun 03 22:36:11 do_build[recrdeptask] = do_distribute_sources Jun 03 22:36:19 now anyone want to explain to me what i just did?-) Jun 03 23:05:03 no, doesn't work properly yet Jun 03 23:05:22 sometimes i get sources, sometimes i don't, depending on what package i tell bitbake to build Jun 03 23:06:02 i don't even see the do_distribute_sources task mentioned now Jun 03 23:31:53 anyone trying out gcc 4.3 Jun 03 23:36:15 later, after my appointment to have needles stuck in my eye Jun 03 23:37:55 >:o Jun 04 00:51:46 khem, I think we have tried 4.3 on the beagle Jun 04 00:53:14 http://groups.google.com/group/beagleboard/browse_thread/thread/893374ac8ff737b4/72650275e52a940c?hl=en&lnk=gst&q=koen+compiler#72650275e52a940c Jun 04 01:09:45 i'm lazy. i want to talk over the d-bus, but i don't already know glib and i don't care to write my nearly trivial d-bus program in c. Jun 04 01:10:23 what (if anything) do people use for a high-level interpreted language in the embedded environment? Jun 04 01:10:54 (i mean if this were a 16-bit microcontroller i would write C. i also wouldn't be running linux.) Jun 04 01:11:38 python works well enough for me, and has nice dbus bindings iirc ;) Jun 04 01:11:58 yes but it's big Jun 04 01:12:23 odds are i'll need to do the microSD root thing to step up from 16MB to 2GB - but i haven't yet Jun 04 01:12:33 i'm looking at an .ext2.gz that's 8MB Jun 04 01:12:39 *eyebrow* Jun 04 01:13:35 so you just took out most of /usr/lib/python2.x? Jun 04 01:13:49 pretty much went with the defaults Jun 04 01:13:59 no X though Jun 04 01:14:09 yeh - hmm Jun 04 01:15:11 i'm at 7.6M w/o python already Jun 04 01:15:20 heh, python-core is half a meg Jun 04 01:15:30 note that was .gz size Jun 04 01:15:33 yes. Jun 04 01:15:49 it's a 20+ MB image because there's still flash left over ;) Jun 04 01:32:04 garh! i try to bitbake python-dbus and it builds X :( Jun 04 01:40:28 Crofton|work: hey Jun 04 01:40:43 hi Jun 04 01:41:08 Crofton|work: is it the plain gcc 4.3 you tried ? Jun 04 01:41:15 or the csl something Jun 04 01:41:21 plain gcc Jun 04 01:41:24 cool Jun 04 01:41:25 koen did the test Jun 04 01:41:31 did the link work? Jun 04 01:41:33 what image did you try Jun 04 01:41:55 console-image? Jun 04 01:41:58 not sure Jun 04 01:42:21 ok Jun 04 01:42:36 hmmm there is a problem Jun 04 01:42:49 ? Jun 04 01:42:51 Were you able to compile libusb ? Jun 04 01:43:14 not sure Jun 04 01:43:22 I haven't reached to bottom of it Jun 04 01:43:31 personally, I have only tried "working" combinations Jun 04 01:43:34 but it seems the includes confuse the gcc Jun 04 01:44:02 4.3 and libusb? Jun 04 01:44:08 Crofton|work: yes Jun 04 01:44:20 the symptom shows up Jun 04 01:44:28 lemme point you to my build log Jun 04 01:45:15 http://tinderbox.openembedded.net/public/logs/167241.txt Jun 04 01:45:28 it can not find stdlib.h Jun 04 01:45:39 same commandline works well with gcc 4.2.2 Jun 04 01:45:50 Two things can happen Jun 04 01:46:16 Either the includes are hosed up or compiler include search is hosed up Jun 04 01:46:43 I also do glibc builds Jun 04 01:49:06 glibc build should work Jun 04 01:50:05 Crofton|work: do you have 4.3 installed somewhere and handy ? Jun 04 01:50:14 not at the moment Jun 04 01:50:25 we backed of to 4.2.x for beagle Jun 04 01:50:35 we are still debuging basic stuff Jun 04 01:51:26 ok Jun 04 01:51:49 is beagle board port still new ? Jun 04 01:52:13 http://beagleboard.org/ Jun 04 01:53:03 wow omap3530 Jun 04 01:53:48 Crofton: The boards must be expensive Jun 04 01:54:33 $150 hmm **** ENDING LOGGING AT Wed Jun 04 02:59:57 2008