**** BEGIN LOGGING AT Wed Nov 15 02:59:57 2006 Nov 15 03:06:32 I'm having a problem with linux-hotplug when trying to compile the opie-image. It fails on the install process, http://pastebin.ca/248887 is the error log, it seems like something really trivial, but I can't figure it out. Any help would be appreciated greatly. Nov 15 04:54:47 koen|awa1: the answer, btw, was ".mtn-ignore" Nov 15 05:06:14 njs: ping Nov 15 05:11:15 HopsNBarley|zzz: pong (sorta, cooking) Nov 15 06:47:19 morning Nov 15 06:55:10 morning hrw - thanks for finding and fixing the other nslu2 compatible machine instances. Nov 15 06:55:31 rwhitby: np. it stopped my fetch build Nov 15 06:55:41 rwhitby: I work on making source mirror Nov 15 06:56:22 hrw: if you're missing something, nslu2-linux has saved *every* single source file we've ever used for any slugos build in the last two years in sources.nslu2-linux.org/sources/... Nov 15 06:56:41 rwhitby: good to know Nov 15 06:56:43 (except the Intel licensed files) Nov 15 06:57:11 rwhitby: btw, how you support those intel files building? user have to fetch them and put in FILESPATH or DL_DIR? Nov 15 06:57:18 yep Nov 15 06:57:34 I have marvell cf card here - driver is no-license one Nov 15 06:57:41 with the new ixp4xx open source driver, that will reduce down to a single file. Nov 15 06:59:28 ok, time to make some todos for next days Nov 15 07:00:22 morning all Nov 15 07:00:27 hi valLongfield Nov 15 07:03:43 ~lart mwester for not pulling before push Nov 15 07:03:43 * ibot puts on some milking gloves. "All right, now, mwester, this won't hurt a bit...." for not pulling before push Nov 15 07:04:39 03rwhitby 07org.oe.dev * r4d4d5791... 10/ (4 files in 4 dirs): slugos: Bumped version to 4.1, updated slugimage and ixp4xx-kernel to latest versions. Nov 15 07:04:49 03rwhitby 07org.oe.dev * r0881ef75... 10/ (1 packages/linux/ixp4xx-kernel_2.6.19-rc5.bb): ixp4xx-kernel: Latest changeset now boots (openslug at least). Nov 15 07:05:26 hrw: mwester should be using the nslu2-linux MasterMakefile, which always pulls before pushing. Nov 15 07:05:44 hrw: yep, just noticed the multiple candidates ! Nov 15 07:05:50 (morning pull) Nov 15 07:06:03 If he didn't, he'll get some "education" from me too :-) Nov 15 07:06:24 I'm merging/merging now Nov 15 07:06:34 SlugOS is now using Apex as a second-stage bootloader :-) Nov 15 07:07:49 merged, pushed Nov 15 07:07:49 thx Nov 15 07:14:11 who maintain ossie? Nov 15 07:19:51 03hrw 07org.oe.dev * rfe0e9ab6... 10/ (11 files in 2 dirs): ossie/*: mark as BROKEN due to #1589 (HTTP AUTH on fetch) Nov 15 07:21:33 hi IvoC Nov 15 07:32:27 someone knows what this strange package location is ? http://ewi546.ewi.utwente.nl/mirror/www.openzaurus.org/official/unstable/3.5.4/sources/git_anongit.freedesktop.org.xorg.proto.calibrateproto.tar.gz Nov 15 07:32:53 vlo|work: its OpenZaurus 3.5.4.x source archive Nov 15 07:33:43 because the download keeps failing: the .tar.gz archive is not there Nov 15 07:40:50 my angstrom-gpe-image build keeps failing because it cannot download this archive which is needed by xcalibrateext-0.0+git20060814-r0 Nov 15 07:43:00 hi Oyvind Nov 15 07:43:29 03hrw 07org.oe.oz354x * r7175dc61... 10/ (1 packages/qemu/qemu-gcc-check.inc): qemu-native: added gcc34 and gcc33 as available CC - close #1584 Nov 15 07:43:36 03hrw 07org.oe.oz354x * rde3cb8aa... 10/ (3 files in 2 dirs): qemu-native: added 0.8.2 Nov 15 07:43:43 03hrw 07org.oe.dev * rd17d39a6... 10/ (1 packages/qemu/qemu-native.inc): qemu-native: added gcc34 and gcc33 as available CC - close #1584 Nov 15 07:51:40 hi leoncamel Genesis Nov 15 07:51:42 hi Nov 15 07:51:57 hi, hrw Nov 15 07:53:10 hi Genesis Nov 15 07:53:25 bijour :) Nov 15 07:54:15 NOTE: package glibc-2.3.5+cvs20050627-r3: task do_populate_staging: started Nov 15 07:54:32 now it is time to check does qemu 0.8.2 on amd64 will generate arm locales Nov 15 07:55:41 ~dict entail Nov 15 07:55:43 03hrw 07org.oe.dev * r21027f08... 10/ (1 packages/linux/mnci-ramses_2.4.21-rmk2-pxa1.bb): mnci-ramses: set COMPATIBLE_MACHINE to 'mnci' Nov 15 07:55:43 Dictionary 'entail' (1 of 8): add a codicil, affect, allegorize, allude to, argue, assume, be indicative of, be significant of, be symptomatic of, bequeath, bequeathal, bequest, bespeak, betoken, birthright, borough-English, bring, bring to mind, call for, cause, characterize, coheirship, comprise, connote, contain, coparcenary, demand, denominate, denote, ... Nov 15 07:55:58 xorg 1.3 looks good Nov 15 07:56:17 Genesis: quite old it is... Nov 15 07:57:23 hu ? Nov 15 07:58:35 1.3 Nov 15 07:59:40 as far as i know , 1.3 is not released Nov 15 08:02:39 Genesis: you mean 7.3 ? Nov 15 08:02:59 http://lists.freedesktop.org/archives/xorg-announce/2006-November/000152.html Nov 15 08:03:02 xorg-server Nov 15 08:03:03 sorry Nov 15 08:03:08 that should be same Nov 15 08:04:41 ok. just like xora told: qemu-arm segfault on amd64 Nov 15 08:05:18 64bits issues are not dead :( Nov 15 08:06:45 hi likewise Nov 15 08:06:57 brb Nov 15 08:07:36 hi Nov 15 08:11:27 hi koen Nov 15 08:11:56 hey vlo|work Nov 15 08:13:07 hrw: hehe, it's a long time that I made a newer kernel for MNCI using OE :-( Nov 15 08:13:34 hrw: I have a 2.6 kernel in my pipeline, but mostly no time and no customer gives me a penny for finishing that work :-( Nov 15 08:13:54 koen: hi Nov 15 08:14:10 hey likewise Nov 15 08:15:36 schurig: :) Nov 15 08:16:32 hmm.. I have openzaurus-3.5.4.3.conf ;D Nov 15 08:17:20 * Genesis notice for the loto Nov 15 08:17:27 do we have any FedoraCore/RHEL/Mandriva developer here? Nov 15 08:20:32 _law_: you use FC6? Nov 15 08:20:42 <_law_> hrw, yes Nov 15 08:20:51 03hrw 07org.oe.oz354x * ra53884cd... 10/ (1 packages/qemu/qemu-gcc-check.inc): qemu-native: revert code from .dev, add gcc33/gcc34 in .oz354x way Nov 15 08:21:21 _law_: if there will be bitbake and openembedded-essential packages for it will you test them? Nov 15 08:21:40 <_law_> hrw, sure Nov 15 08:21:43 ok Nov 15 08:22:07 I need to catch those guys which started work on bitbake for debian Nov 15 08:22:52 BTW: http://www.hrw.one.pl/2006/11/15/compactflash-80211g-card/ Nov 15 08:25:10 hrw: I can test on RHEL4 if you like Nov 15 08:25:22 great Nov 15 08:27:05 hi [lala] & greentux Nov 15 08:27:25 cu all - time to go to work Nov 15 08:27:31 morning all Nov 15 08:27:43 NAiL: morning Nov 15 08:29:01 likewise: any success on the toolchain? Nov 15 08:30:10 NAiL: not the 0.9.27 uClibc yet Nov 15 08:30:23 NAiL: but today for me is IXP4xx 2.6.18 day Nov 15 08:31:02 NAiL: because it's nice to have hardware to test on Nov 15 08:31:50 any reason you're not building 2.6.19-rc5? :) Nov 15 08:32:36 NAiL: not really actually, well the rc part scares me a bit. Nov 15 08:33:49 NAiL: Moving follow-ups to #nslu2-linux Nov 15 08:33:58 ok :) Nov 15 08:35:42 hi hrw Nov 15 08:40:38 morning all Nov 15 08:43:10 hi rp Nov 15 08:47:08 likewise: rc5 works ok on my x86 machines Nov 15 08:47:45 hrw|work: thanks, this is for ixp4xx. Good to know though. Did you any work with the real-time patches (maybe for celinux)? Nov 15 08:48:15 likewise: not yet Nov 15 08:48:40 likewise: you got a loft? Nov 15 08:58:49 morning all Nov 15 08:58:53 hey RP Nov 15 08:59:26 I said that before, didn't I... Nov 15 08:59:34 RP: fetchall built libopie ;D Nov 15 08:59:42 15 10:04 < RP> morning all Nov 15 08:59:44 15 10:22 < RP> morning all Nov 15 08:59:47 RP: you did Nov 15 08:59:57 * RP needs to wake up Nov 15 09:02:07 I will create ugly tslib_1.0.bb for .oz354x Nov 15 09:02:22 hrw|work: Why did it build libopie though? :-/ Nov 15 09:02:56 RP: did not traced it yet Nov 15 09:03:53 ~lart FUSE people Nov 15 09:03:54 * ibot takes out FUSE people with the trash Nov 15 09:04:18 "if you cross compile , you 'd better modify the Makefile by own ." Nov 15 09:04:37 koen: what's FUSE? Nov 15 09:04:54 likewise: filesystem in userspace Nov 15 09:05:05 likewise: sshfs, gmailfs, etc Nov 15 09:05:17 koen: ah of course. I was thinking of JUCE, which is something else... Nov 15 09:05:25 koen: They won't take patches? Nov 15 09:05:59 Some user got offended because I called the behaviour 'stupid' Nov 15 09:06:29 the patch: http://www.openembedded.org/repo/org.openembedded.dev/packages/fuse/files/not-run-updaterc.d-on-host.patch Nov 15 09:07:13 I bet kergoth would've screamed in agony when noticing that behaviour Nov 15 09:08:19 koen: True and it is a rather bad idea but people seem to get upset about words like stupid :) Nov 15 09:08:19 does anyone know if Xcalibrate has been fixed in kdrive 1.3? Nov 15 09:15:04 hrw|work: I just dumped the runqueue for some world builds and there are only fetch and fetchall tasks listed :-/ Nov 15 09:15:18 hrw|work: The continue on error behaviour is also working ok here :-/ Nov 15 09:20:14 RP: with clean tmp? Nov 15 09:21:30 hrw|work: For the actual queue of tasks, it doesn't matter, it doesn't look at tmp before that point... Nov 15 09:21:40 aha Nov 15 09:24:13 hrw|work: It could be the actual task execution is broken somehow :-/ Nov 15 09:27:32 good morning Nov 15 09:27:55 moin! Nov 15 09:29:07 hi florian_kc Nov 15 09:32:11 03koen 07org.oe.dev * ra92b3f38... 10/ (1 packages/tinymail/tinymail_svn.bb): tinymail: add maxdate for configure patch Nov 15 09:40:20 I need someone with oz 3.5.4.x on machine Nov 15 09:43:58 NOTE: Pruned 36598 inactive tasks, 3385 left Nov 15 09:44:03 nice list Nov 15 09:50:48 Who worked on the dependency stuff in bitbake lately? Nov 15 09:51:23 About from some wrong warnings about missing providers i get errors like this: Nov 15 09:51:26 ERROR: Nothing provides runtime dependency libmimedir-0.4-dev Nov 15 09:55:27 florian_kc: this dependency is wrong written Nov 15 09:56:31 RP: [hrw@misc1 celinux-test]$ time bitbake -cfetchall world -vk|tee fetchall2.log Nov 15 09:56:35 RP: started Nov 15 09:57:13 hrw|work: oh, right... that changed somehow. obviously this one was never changed... what do i need to change? Nov 15 09:58:17 florian_kc: *-dev should get automatically rdeps on other -dev Nov 15 10:01:10 hrw|work: right... then the dependency settings in the sdk target might be wrong Nov 15 10:02:54 florian_kc: task-sdk* and meta-sdk* need to get rewritten Nov 15 10:03:36 RP: thx for git-native update. now it works with git fetcher Nov 15 10:03:54 NOTE: SITE files /home/hrw/devel/oe/org.openembedded/site/endian-little /home/hrw/devel/oe/org.openembedded/site/x86_64-linux Nov 15 10:03:57 florian_kc: RDEPENDS = 'libmimedir-dev' will pull in libmimedir-0.4-dev and all the -dev packages it depends on Nov 15 10:09:31 koen: ok... well, i'm quite sure i didn't catch all latest information about changes in oe. but i have some time to spend on it... so what do we need to update in these? Nov 15 10:11:02 replace all the libfoo-5.6.4-dev with foo-dev, or remove them is gpe-bar already depends on them Nov 15 10:12:09 koen: sounds like an easy job :-) Nov 15 10:12:20 NOTE: Task failed: Fetch failed: http://www.minisip.org/snapshots/libmutil-0.2.2.tar.gz Nov 15 10:17:32 NOTE: Pruned 36681 inactive tasks, 3302 left Nov 15 10:17:45 progress Nov 15 10:18:58 <[lala]> morning guys Nov 15 10:19:07 hi lala Nov 15 10:19:24 <[lala]> maybe a stupid question: how can i prevent that bitbake deletes the run.* scripts? Nov 15 10:19:29 <[lala]> hi hrw Nov 15 10:20:32 <[lala]> tmp/work//temp/ only contains the log files Nov 15 10:20:38 [lala]: no idea Nov 15 10:21:31 koen: looks like only a few of these are in task-sdk Nov 15 10:21:35 hi mallum Nov 15 10:21:40 hey mallum Nov 15 10:22:02 hey florian_kc, koen Nov 15 10:22:09 florian_kc: In theory we don't need much of task-sdk anymore. Hang on one second as I've fixed up some of the really broken references locally... Nov 15 10:24:24 one useful thing for any frontend: "XXXX of YYYY tasks left" info somewhere Nov 15 10:24:57 03koen 07org.oe.dev * r38b07284... 10/ (1 packages/libmutil/libmutil0_0.2.2.bb): libmutils: minisip.org removed all tarballs, use SVN Nov 15 10:25:25 koen: you forgot " - close #xyz" Nov 15 10:25:30 RP: ok, that's nice Nov 15 10:26:03 koen: thx for fixing Nov 15 10:26:20 | /usr/bin/ld: fad-getad.o: relocation R_X86_64_32 against `a local symbol' can not be used when making a shared object; recompile with -fPIC Nov 15 10:26:23 ;( Nov 15 10:26:33 libpcap is broken Nov 15 10:27:31 hi XorA|gprs Nov 15 10:27:33 hrw|work: I'll look at the complete mini{sip,splat} stack this week Nov 15 10:27:54 I'm thinking of taking few free days from work to work on OE Nov 15 10:28:02 howdee Nov 15 10:28:09 hey XorA|gprs Nov 15 10:28:15 hrw|work: cool Nov 15 10:28:20 hi XorA|gprs Nov 15 10:28:42 good find by mreimer last night Nov 15 10:29:00 florian_kc: I have some paid OE work to do Nov 15 10:29:27 florian_kc: and some OZ/Angstrom one too - read my last blog post Nov 15 10:29:37 hrw|work: even better :-) Nov 15 10:29:42 * florian_kc does so Nov 15 10:30:45 03rpurdie 07org.oe.dev * r7c5dba6a... 10/ (6 files in 2 dirs): xproto: Drop pre X11R7 versions Nov 15 10:31:00 03rpurdie 07org.oe.dev * rf513f8c4... 10/ (1 packages/tasks/task-sdk.bb): task-sdk: Fix broken package references to remove parsing errors. Ideally needs totally rewriting Nov 15 10:31:06 03rpurdie 07org.oe.dev * rc018d6b3... 10/ (1 packages/tasks/task-pivotboot.bb): task-pivotboot: Give variables some defaults - avoid parsing errors Nov 15 10:31:11 03rpurdie 07org.oe.dev * ra0e25807... 10/ (1 packages/meta/meta-maemo.bb): meta-maemo: Fix broken package references to remove parsing errors. Ideally needs totally rewriting Nov 15 10:31:17 03rpurdie 07org.oe.dev * r3c5d934a... 10/ (1 packages/e17/entice_20060416.bb): entice: Fix broken package references to remove parsing errors Nov 15 10:34:18 RP: what do you think about adding empty {MACHINE,DISTRO}_FEATURES into bitbake.conf? Nov 15 10:34:50 hrw|work: No need as task-base gives them defaults Nov 15 10:35:16 RP: but not task-opie - so I need to copy some code Nov 15 10:35:57 hrw|work: If we have other users then the code should move to bitbake.conf Nov 15 10:36:29 RP: same with combined_features Nov 15 10:37:55 RP, koen: What should be the basic ideas and new concepts for a new sdk target? Nov 15 10:39:39 florian_kc: The idea is any packages -dev package has the right RDEPENDS so if your image is made from task-gpe-base, task-gpe-base-dev will give you the -dev packages for the sdk Nov 15 10:41:31 RP: ok... that should make the sdk targets pretty small :-) Nov 15 10:42:03 florian_kc: yes, and much less error prone Nov 15 10:42:36 indeed Nov 15 10:43:06 an and you removed this patch from the compiler Nov 15 10:43:12 s/an/ah Nov 15 10:49:10 3115 tasks to go Nov 15 10:51:36 hrw|work: sounds like my personal task list :-) Nov 15 10:51:41 hi zecke Nov 15 10:51:46 hey zecke Nov 15 10:51:48 florian_kc: its bitbake tasklist Nov 15 10:52:08 :-) Nov 15 10:56:10 03koen 07org.oe.dev * r212c3e97... 10/ (1 packages/tsclient/tsclient_0.140.bb): tsclient: fix DEPENDS Nov 15 10:58:02 koen: where libpanelapplet was? Nov 15 10:58:12 hrw|work: is that going to be a real source mirror, or only a celinux/omap5192osk mirror? Nov 15 10:58:17 hrw|work: inside gnome-panel Nov 15 11:00:05 koen: I'm doing full fetch now. will it be full or limited will be decided later Nov 15 11:00:47 if it's going to be a full source mirror, bitbake_qa would be a better tool to fetch stuff Nov 15 11:01:50 koen: I know, but I wanted to test fetchall task Nov 15 11:06:42 hi stefan_schmidt Shoragan Nov 15 11:07:14 hi :) Nov 15 11:08:48 Shoragan: can you remind me url to bitbake debian packages? Nov 15 11:10:12 sure Nov 15 11:10:22 http://sicherheitsschwankung.de/~jluebbe/debian/unstable/ Nov 15 11:10:34 i posted to the wiki, too Nov 15 11:12:42 Shoragan: I have a customer which want to fund effort of making usage of OE easier for Debian and Fedora based systems Nov 15 11:12:58 Shoragan: this also means bitbake packaging. Nov 15 11:13:37 i'll probably upload the package after the etch release, because of the freeze Nov 15 11:14:18 Shoragan: are you interested in? Nov 15 11:14:37 yes :) Nov 15 11:15:02 good Nov 15 11:15:13 but i'll have to go in 20 min and still have some stuff to do, i'll be available later today Nov 15 11:15:25 ok Nov 15 11:15:35 Shoragan: can you mail me at haerwu@hrw.one.pl? Nov 15 11:16:24 k Nov 15 11:17:59 Shoragan: got mail Nov 15 11:20:23 morning all Nov 15 11:21:07 hi dirk Nov 15 11:21:40 hey Marcin Nov 15 11:21:46 hey Dirk Nov 15 11:21:53 hi Koen Nov 15 11:22:26 hrw|work: does qemu-native compile for you on oz354x? Nov 15 11:22:37 do13_: 0.8.2 yes Nov 15 11:22:47 ok Nov 15 11:22:52 do13_: for amd64. but does not work on amd64 Nov 15 11:23:13 but according to xora there is no working version Nov 15 11:24:00 can you update qemu in oz354x? Nov 15 11:24:13 I updated it today to latest upstream Nov 15 11:24:24 0.8.2 ? Nov 15 11:24:56 yes Nov 15 11:26:18 hmm, strange 0.8.1 here Nov 15 11:27:43 will sync Nov 15 11:28:02 15 09:07 < CIA-4> hrw org.oe.oz354x * rde3cb8aa... / (3 files in 2 dirs): qemu-native: added 0.8.2 Nov 15 11:30:09 03hrw 07org.oe.oz354x * reddaf353... 10/ (14 files in 5 dirs): qpealarmclock(applet): updated to 1.0.9 - close #1576 Nov 15 11:33:26 03koen 07org.oe.dev * r424d220b... 10/ (1 packages/libmutil/libmutil0_svn.bb): libmutil: first pass at updating to current SVN Nov 15 11:38:48 hrw|work: thx. It works. Nov 15 11:39:50 hrw|work: new buildmachine, forgot to add pull branch. So it used the snapshot date :) Nov 15 11:39:55 ;D Nov 15 11:49:36 hey pH5 Nov 15 11:50:47 hey koen Nov 15 11:54:40 mallum: is the kdrive-iwmmxt code online somewhere? Nov 15 11:57:49 koen: nope not yet, its pretty broken Nov 15 11:57:59 koen: and it dont give you much in performance Nov 15 11:58:08 mallum: the immediate value stuff? Nov 15 11:58:11 koen: just a load of server crashes Nov 15 11:58:32 koen: it wants all pixmaps 8 byte aligned Nov 15 11:59:03 mallum: so we need to wait for intel to make an arm11 core with iwmmxt Nov 15 12:00:10 mallum: did you see http://www.logicpd.com/eps/devkits/freescale/devkit/imx_devkit/ ? Nov 15 12:02:44 * mallum checks Nov 15 12:03:34 koen: nice, I think RP mentioned this the other day Nov 15 12:03:50 mallum: I bet :) Nov 15 12:04:04 the other mx31 devboard is like $5k Nov 15 12:04:39 mx31L does not have the gpu though right ? Nov 15 12:08:25 mallum: yes, the L lacks the GPU Nov 15 12:09:33 :( Nov 15 12:11:25 anybody here Nov 15 12:11:26 ? Nov 15 12:11:29 it does have the mpeg4 stuff by the looks of it Nov 15 12:11:45 Added 15864 packages in 0 days, 0 hours, 19 minutes and 19 seconds Nov 15 12:11:58 15864 potential QA nightmares... Nov 15 12:13:58 hello siomai Nov 15 12:19:59 hi florian_kc :) Nov 15 12:20:22 im looking for a howto or any kind of documentation of flashing back to the SHARP rom, from pdaxrom + uboot Nov 15 12:21:28 siomai: oh... that's the wrong place to ask here, but maybe there is a pdaxrom channel. Nov 15 12:22:25 :) i'll try that, thank you very much. Nov 15 12:22:40 siomai: yw Nov 15 12:47:05 ~lart opencvs Nov 15 12:47:05 * ibot turns opencvs into a lifesized tux doll Nov 15 12:48:23 greetings ! Nov 15 12:49:55 hey all Nov 15 12:50:17 hey Liam Nov 15 12:52:39 hi Ifaistos, hi lrg Nov 15 12:53:17 :) Nov 15 12:56:26 ~seen lenehan Nov 15 12:56:28 lenehan was last seen on IRC in channel #debian, 885d 18h 43m 29s ago, saying: 'do u know an website with how to in order to make it work ?'. Nov 15 12:56:43 ~seen twibble Nov 15 12:56:45 hrw|work: i haven't seen 'twibble' Nov 15 12:57:17 what was nick of our local Perl guru? Nov 15 13:00:51 perl-native for x86_64 need work Nov 15 13:02:31 ~seen v8jlene Nov 15 13:02:44 ah. Nov 15 13:02:54 #1581 submitted, he is in Cc: Nov 15 13:03:01 v8jlene is currently on #oe, last said: 'Re ldd (bugs #1484 & #878). Ldd calls "set -o pipefail" which makes the busybox sh exit. It can be fixed by either removing that from ldd (only needed to get output under SELinux), or by patching busybox sh not to die when it gets an unknown set option. Anyone have a preference on which way to fix this? I'm undecided..'. Nov 15 13:03:04 3.5" bug Nov 15 13:03:15 ~idle v8jlene Nov 15 13:03:26 15 14:27 [FreeNode] -!- v8jlene is idle for 226 hours, 35 minutes and 9 seconds. Nov 15 13:11:02 * koen waits for mickey|sports to start ncursing Nov 15 13:11:53 huhu Nov 15 13:29:59 ~lart hrw|work for jumping to conclusions Nov 15 13:29:59 * ibot nabs the moon and broadsides hrw|work with the sea of tranquility for jumping to conclusions Nov 15 13:48:50 morning guys Nov 15 13:49:04 koen: ncurses demo is done and sent to RP. now the ball is in his half ;) Nov 15 13:49:12 now back to paid work... Nov 15 13:51:21 ~lart me Nov 15 13:51:22 * ibot changes hrw|work's permissions to 0777 and tells the world Nov 15 13:51:25 mickey|devel: thanks :) Nov 15 13:51:37 np. hope you can get anything out of it :) Nov 15 13:51:45 at least it looks cool ;) Nov 15 13:51:52 dunno whether it's useful Nov 15 13:51:53 heh Nov 15 13:53:13 mickey|devel: It looks very useful and gives me the right kind of hints on how to use python which is good :) Nov 15 13:53:27 cool :) Nov 15 13:53:34 I'll have a fiddle with it later :) Nov 15 13:53:51 hey mickey|devel Nov 15 13:54:00 mickey|devel: hey Nov 15 14:05:54 has anyone noticed, that depending on how the Makefile.am is written some files may go or may not go to the ipk packge? In both cases make install will install them correctly, but in one case they are taken into the ipk, and in the other case you will get "NOTE: the following files were installed but not shipped in any package", is this a known problem? Nov 15 14:08:32 morning Nov 15 14:18:56 Jin^eLD: It depends whether they match FILES_${PN}. The variable has some defaults for common installs but specialist ones need extra definitons to be added Nov 15 14:33:45 oh ok.. I see Nov 15 14:33:49 I will look at that Nov 15 14:33:53 thanks Nov 15 14:35:21 RP: mc gcc-cross-sdk last night built (4.1.1) after the update this morning it suffers from the "two or more data types in declaration specifiers" issue too. Nov 15 14:35:27 s/mc/my Nov 15 14:36:57 RP: bitbake/trunk starts threads, wait for them to finish, start another set, wait for finish etc? Nov 15 14:37:31 hrw|work: starts X threads, when one finishes, starts another etc. Nov 15 14:38:14 florian_kc: I don't think I touched anything to do with that? Nov 15 14:39:03 RP: http://pastebin.ca/249184 looks like it start X, wait for end, start next X Nov 15 14:39:39 RP: right, that's quite unlikely... on my other machine and nils one we have this for a longer time now. Nov 15 14:39:54 hrw|work, im back :) Nov 15 14:40:06 hrw|work: When you interrupt, it waits for all tasks to finish before exitting Nov 15 14:40:49 florian_kc: I've not looked into that problem I'm afraid Nov 15 14:43:49 hi Nov 15 14:44:08 RP: ok Nov 15 14:44:20 Shoragan: and I'm bit busy now - 1h? Nov 15 14:44:27 k Nov 15 15:00:44 * koen wonders about http://bugs.openembedded.org/show_bug.cgi?id=1265 Nov 15 15:03:22 is there an easy way to set up my kernel bb file to generate an empty ipk kernel file or do I just need to manually edit the ipk file everytime and delete the kernel? Nov 15 15:03:45 FILES_kernel-image = "" Nov 15 15:03:55 ALLOW_EMPTY_kernel-image = "1" Nov 15 15:04:14 hi Gerrath Nov 15 15:04:21 Jin^eLD, hi. Nov 15 15:04:37 hi koen Nov 15 15:04:48 hey chouimat Nov 15 15:05:44 Jin^eLD, any luck with directfb? I'm using directfb 1.0 RC1, I can send you the bb file, although I hope to submit a patch to OE (bug tracker) soon Nov 15 15:05:59 koen, thank. Nov 15 15:07:24 Gerrath: no luck yet, my strace did hang somehow; however I figured I may have a problem with RGB / BGR Nov 15 15:07:24 Gerrath: what exactly did you have to patch for your version? Nov 15 15:07:56 Jin^eLD: DirectFB doesn't support BGR IIRC Nov 15 15:08:23 zecke: I know, my hope was - I read some other people had it - that the framebuffer simply returned the wrong information Nov 15 15:08:48 anyway, I would already be happy to see swapped colors or whatever; the main problem is that directfb just exits with various weird messages Nov 15 15:09:10 right now I thought I go back one step and get some fbdev apps Nov 15 15:09:16 and see how well or how bad they work Nov 15 15:11:21 Jin^eLD, actually what I had to fix was an error in the actual directfb code.. a missing header file, other than that the patch that already exists for directfb works with the directfb 1.0, there might have been a couple other things I had to add to the bb file. I can email you the files if you want. Also I'll try to get it in the bug tracker sometime this week. Nov 15 15:12:12 Jin^eLD, can you cat things to your fb? Nov 15 15:12:54 Gerrath: would be nice if I could look at your files... the only problem I had was related to asm/page.h, it was not finding the PAGE_SIZE, but apart from that it compiled well.. Nov 15 15:13:06 you mean like ls >/dev/fb0 ? Nov 15 15:13:23 cat /dev/urandom > /dev/fb0 Nov 15 15:13:24 no, cat /dev/urandom > /dev/fbo Nov 15 15:13:37 yes, what koen said :-) Nov 15 15:13:40 :) Nov 15 15:13:45 I get no error when doing that Nov 15 15:13:50 but I do not see anything on the screen... Nov 15 15:13:51 hmm Nov 15 15:13:59 you should get pretty colours Nov 15 15:14:04 Jin^eLD, you should see all kinds of noise :-) Nov 15 15:14:28 what do you get when you do a hexdump /dev/fb0? Nov 15 15:14:31 I thought that should be the case, yes :) I guess I will do some poking around to see why nothing is there... Nov 15 15:14:44 uhm.. a lot :) Nov 15 15:14:46 'bitbake fbgrab' Nov 15 15:14:58 fbgrab will grab a png image of /dev/fb0 Nov 15 15:15:07 ah, cool - compiling Nov 15 15:15:07 if you dump dev/urandom to the fb0 then do a cat you should see all the random data :-) Nov 15 15:15:48 Gerrath: well I can't tell how random the hexdump is :) but it does not seem to have any visible patterns in the hex values :) Nov 15 15:15:52 Jin^eLD, opps not cat but hexdump.. cat would not look very good. Nov 15 15:16:05 moment, I compiled fbgrab, lets see.. Nov 15 15:16:28 Jin^eLD, I wrote a little c program to dump specific values to my fb when working on it.. I'll post it for you, hold on.. Nov 15 15:16:31 cat /dev/urandom > /dev/fb0 Nov 15 15:17:16 Gerrath: thanks Nov 15 15:17:20 cat 'spare time' > /dev/zecke Nov 15 15:17:31 :-) Nov 15 15:18:44 ok.. lets take a look at fbgrab.. Nov 15 15:19:33 Jin^eLD, http://pastebin.ca/249222 its very LCD specific as I just through it together in a couple of minutes.. basically you will need to change it for your screen size. It is setup to take two ints and display half the screen with the first int and the second half with the second int. I had a color transistion problem I was tracking down :-) Nov 15 15:20:27 Jin^eLD, it generates a file called pixels.raw, just cat that file to your fb. Nov 15 15:20:46 got it, thanks Nov 15 15:21:47 but first.. lets look at what fbgrab produced.. Nov 15 15:23:30 hmm ok Nov 15 15:23:34 it looks noisy Nov 15 15:23:56 so one of the problems is that I have to somehow tell the device to display the fbdev stuff on video out Nov 15 15:24:34 but that is another problem.. not directfb related I guess.. because DFB exits Nov 15 15:24:47 you could test Gerrath's apps as well to see if you get 2 colours instead of noise Nov 15 15:25:23 true, let me try that.. Nov 15 15:29:01 seems to work Nov 15 15:29:06 I guess I medded up the resolution Nov 15 15:29:16 or maybe it needs ARGB or something Nov 15 15:29:30 but the colors are definetely the ones that I chose Nov 15 15:30:04 ok, let me figure how to bring this to the TV screen.. Nov 15 15:31:47 03gremlin 07org.oe.dev * r0479e731... 10/ (3 files in 3 dirs): Nov 15 15:31:47 linux-handhelds-2.6: Add defconfig for h3600. Nov 15 15:31:47 * Closes #1536. Nov 15 15:32:10 psokolovsky: discovered the --author switch? Nov 15 15:32:29 koen, ~2months ago ;-) Nov 15 15:32:49 and have been using it all since ;-) Nov 15 15:32:53 psokolovsky: what do you think about seperating h3600 into h3600 and h3800? Nov 15 15:33:14 they are 2 very different pieces of hardware Nov 15 15:33:27 koen, I have to say that I'm not yet familiar enough with SA models' guts Nov 15 15:33:34 asic2 vs h3600_micro Nov 15 15:33:53 psokolovsky: h36xx controls everything over a serial port, h38xx uses asic2 Nov 15 15:33:57 koen, but I would think that's grounded simply because h3800 has secondary storage, and h3600/h3700 doesn't. Nov 15 15:34:17 koen, yep, that's what I had in my dejavu's too ;-) Nov 15 15:34:43 koen, well, I guess, let gremlin[it] first revamp h3800 port, and then it can be done... Nov 15 15:34:46 I'm charging the h3780 right now via its dual pcmcia sleeve Nov 15 15:35:01 ;-) Nov 15 15:35:17 which I should return to the university sometime Nov 15 15:35:20 3870 rather? Nov 15 15:35:23 yep, I meant build secondary storage socket ;-) Nov 15 15:35:26 hrw|work: right Nov 15 15:35:44 friend has one with pcmcia+battery sleeve Nov 15 15:35:53 and 2200mAh battery inside of ipaq Nov 15 15:37:03 ha! I got some output :> fbset did the trick Nov 15 15:37:50 allthough the screen just turned blue :) still nothing displays when catting to fb0 Nov 15 15:41:02 psokolovsky: wanna task-base-ify h3600.conf? Nov 15 15:42:45 koen, sure. I'll make closing round on converting what's left of pocketpc's as soon as I flush hh.orh CVS pending patches queue. Nov 15 15:43:07 or if you're in mood now - ... ;-) Nov 15 15:43:37 h3800.conf is already using task base :) Nov 15 15:44:07 03koen 07org.oe.dev * r54e1ff98... 10/ (1 conf/machine/h3800.conf): h3800.conf: add the groundwork for splitting h38xx out of h3600.conf Nov 15 15:44:18 03koen 07org.oe.dev * r762382ad... 10/ (1 conf/machine/h3800.conf): h3800.conf: add task-base stuff Nov 15 15:47:46 Jin^eLD, you can set the fb parameters in your kernel config file or in your boot loader by passing the kernel the options at boot. Nov 15 15:48:24 Gerrath: at first I'd be happy to figure out which ones work :) Nov 15 15:48:43 fbset -init did at least something on the screen Nov 15 15:48:53 but still, nothing changed when cat'ing to fb0 Nov 15 15:56:59 03pfalcon 07org.oe.dev * rad5c6d61... 10/ (1 packages/tsclient/tsclient_0.140.bb): tsclient 0.140: Clean DESCRIPTION. Nov 15 16:08:25 03pfalcon 07org.oe.dev * rf2a6713d... 10/ (1 packages/tsclient/tsclient_0.140.bb): tsclient 0.140: One more typo, run thru oe-stylize. Nov 15 16:31:32 a+ Nov 15 16:40:24 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Nov 15 16:41:22 Jin^eLD: what device are you on with DFB? Nov 15 16:43:07 likewise: a micronas dev board, wischip Nov 15 16:43:33 Jin^eLD: the H.264/WM9 decoder one? Nov 15 16:43:59 uhm.. the 7x/8x (Cypher/Decypher) Nov 15 16:44:14 go7x8x Nov 15 16:44:35 Jin^eLD: lemme check, they also have a HDTV decoder only chip. Nov 15 16:44:39 it has HD Nov 15 16:44:43 so maybe it is the one you mean Nov 15 16:45:03 I have both here, analog out and HD out Nov 15 16:45:26 I managed to play a video on my analog tv from the board, but that seems to go via v4l Nov 15 16:45:48 Jin^eLD: what dev board do you have exactly? Nov 15 16:45:53 go7x8x is the name of the dev board? Nov 15 16:46:04 lemme find it online.. wait Nov 15 16:46:45 http://www.micronasusa.com/homeusa/index.html Nov 15 16:48:14 http://www.micronas.com/products/application/IPTV/index.html Nov 15 16:48:17 that's the one Nov 15 16:49:07 Jin^eLD: this is the one I mean. It does not have an encoder AFAIK. Nov 15 16:49:48 Jin^eLD: what's the dev board price like? Nov 15 16:51:09 cu all Nov 15 16:52:24 likewise: not sure about the encoder, I thinkg you are right, it does not seem to have one Nov 15 16:52:27 I have no idea about the price Nov 15 16:52:36 I just got the board :) Nov 15 16:53:07 Jin^eLD: for lend, or did you(r company) buy it? Nov 15 16:53:47 our company has it, I do not know the details on where it came from Nov 15 16:54:17 Jin^eLD: we needed an encoder as well, so we did not buy it. Nov 15 16:54:33 what kind of architecture is that? Nov 15 16:54:38 Jin^eLD: I can lend it from the distributor though, so let me know if you a nice demo :-) Nov 15 16:54:39 koen: mipsel Nov 15 16:54:43 koen: mips core Nov 15 16:54:45 cute Nov 15 16:55:09 likewise: you may want to take a look at this one http://www.nxp.com/applications/video_broadcast/stb/stb810/index.html Nov 15 16:55:21 but I do not remember if it has an encoder or not Nov 15 16:55:23 :) Nov 15 16:55:52 hmmm Nov 15 16:56:04 #oe is starting to look like a philips conspiracy Nov 15 16:56:14 huh, why that? :) Nov 15 16:56:25 did you have philips guys asking around here? Nov 15 16:56:32 we have some :) Nov 15 16:56:40 interesting :) hehe Nov 15 16:56:44 koen: do I have to do anything mtn special to pull & upgrade my local rep with your mx31 changes yesterday, no changes appeared in the pull & update messages Nov 15 16:56:46 I worked for Philips for about 2-3 years Nov 15 16:57:16 lrg: you shouldn't need anything special Nov 15 16:57:25 * lrg ponders Nov 15 16:57:25 lrg: 'mtn sync ; mtn up' Nov 15 16:57:36 koen: thanks, will try Nov 15 16:58:44 lrg:Would you like me to test your alsa fixes for poodle? Nov 15 16:59:23 hvontres|poodle: yes please. I think RP has updated the bb's accordingly Nov 15 16:59:27 lrg: the changes should be transparent, though Nov 15 16:59:52 koen: 188 revs in Nov 15 17:00:20 lrg: looks like you were a few days behind :) Nov 15 17:00:41 koen: yep, I'm using our new mirror as well. Using your scripts Nov 15 17:00:49 cool Nov 15 17:02:13 * koen spots asoc 12.3 Nov 15 17:04:46 koen: that did the trick :) Nov 15 17:06:33 lrg: are they in the .dev branch? Nov 15 17:07:59 lrg, hvontres|poodle: I might not have done with the latest ones. I've a few things queued up OZ wise :-( Nov 15 17:08:09 RP: np Nov 15 17:08:30 RP, hvontres|poodle: XorA tested on terrier and corgi. Nov 15 17:08:30 RP: okay.. my fix works for now :) Nov 15 17:08:47 hi, how can i force gnu ld to place symbols according to their appearance in code ? I have c file containing register definitions which where placed at the right address by the linker. Unfortunately the linker reorders them. The object file itself maintains the right order. Nov 15 17:08:51 hvontres|poodle: what is your fix ? Nov 15 17:08:58 lrg: isn't animal testing banned? Nov 15 17:09:14 koen: heh :) Nov 15 17:09:19 koen: :-) Nov 15 17:09:49 poodle# cat Nov 15 17:10:01 lrg: http://bugs.openembedded.org/show_bug.cgi?id=1553 Nov 15 17:10:09 poodle# cat mouse >/dev/watchdog Nov 15 17:10:20 ~corgi Nov 15 17:10:21 corgi is, like, sharp sl-c700, or a dog Nov 15 17:10:55 lrg: if it works on corgi, it should on poodle too... corgi = poodle+w100+VGA Nov 15 17:11:08 hvontres|poodle: ah, now I see. Patch in alsa HG as of last week. Nov 15 17:11:32 hvontres|poodle: renamed to use master keyword Nov 15 17:11:50 lrg : alsa HG ??? Nov 15 17:12:01 hvontres|poodle: mercurial repo Nov 15 17:12:02 mercurial Nov 15 17:12:23 lrg: ok.. so fix is now upstream ? Nov 15 17:12:40 hvontres|poodle: yep and in latest patch Nov 15 17:12:43 I got output!!! Nov 15 17:12:48 Gerrath: I figured out the mode :) Nov 15 17:12:51 Jin^eLD: cheers Nov 15 17:13:03 yay! :) Nov 15 17:13:10 koen: I get the nice noise from urandom now :> Nov 15 17:13:21 Jin^eLD: now you can get back your original problem ;) Nov 15 17:13:27 yes ;) Nov 15 17:13:35 lrg: cool. now the only remaining 2.6 niggle on poodle is "suspend with cancel button" :) Nov 15 17:13:55 * hvontres|poodle needs to build a serial cable... Nov 15 17:17:43 dfbg works! that's something Nov 15 17:23:35 Jin^eLD: good! are you planning to use OE on that device. Would be cool. Nov 15 17:24:55 likewise: we have several linux boards here, it makes sense to use OE because one could easily replace the local.conf and work with the .bb files Nov 15 17:25:14 couple of days ago I ported my opensource UPnP server to an ARM platform in about 30 minutes Nov 15 17:25:29 including all support libraries like sqlite3, libexif, etc. - they all were already there Nov 15 17:25:35 OE is really cool! Nov 15 17:25:44 ~hail OE Nov 15 17:25:49 oh yeah! Nov 15 17:25:52 * ibot bows down to OE and chants, "I'M NOT WORTHY!!" Nov 15 17:26:41 likewise: the difficulty is always, that lots of manufacturers ship prebuilt stuff, compiler, glibc, kernel Nov 15 17:26:58 some use rootfs images, some use other stuff Nov 15 17:27:14 so I am still not sure if I should try integrating all that stuff into the environment or work in parallel Nov 15 17:27:35 Jin^eLD: I know.... and it's always outdated. I have to bringup a Freescale Coldfire board still, MCF5474. I am not sure if upstream GCC has the proper support yet. Nov 15 17:31:34 likewise: yep.. I always wondered how they decide to take a particular compiler or glibc version Nov 15 17:31:45 if there is some reason behind the choice, or if it is just out of the blue Nov 15 17:32:22 Jin^eLD: probably some developer decides early in the process and never changes. Then after the product is released (one or two years later) you get their toolchain and kernel. Nov 15 17:32:42 Freescale is at 2.6.10 with their coldfire support. Nov 15 17:32:46 likewise: http://gnumonks.org/~laforge/weblog/2006/10/30#20061030-gpl-devices Nov 15 17:34:19 likewise: my first thought is always to dump everything except the kernel and create an own distro Nov 15 17:34:36 Jin^eLD: dump everything. Nov 15 17:34:43 but well. that's usually not an option because you get deliverables based on the old stuff and so on Nov 15 17:34:57 likewise: hmm.. everything? but they usually patch the kernel and add their drivers for the hardware to work? Nov 15 17:35:04 Jin^eLD: No, I always assume the upstream DOES NOT WORK. Nov 15 17:35:13 I'm an application level guy, not much of a kernel hacker :) Nov 15 17:35:16 s/upstream/given stuff/ Nov 15 17:36:08 Jin^eLD: Yes, ok, you get the patches, so you can verify them against their datasheets Nov 15 17:36:29 likewise: well - that's the thing - datasheets, hardware... not my domain :) Nov 15 17:36:37 people get some stoneage binary-only kernel-module,libs and/or apps from the vendor, and thus get stuck to the only combination of kernel/gcc/libc that the vendor provided.. Nov 15 17:37:00 suihkulokki: don't talk about mvista in here ;) Nov 15 17:38:00 * lrg -> train Nov 15 17:38:05 koen: they are hardly the only company doing that.. Nov 15 17:38:11 suihkulokki: isn't the vendor obliged to release the kernel source? Nov 15 17:38:25 after all it's GPL, so they have to give out the source Nov 15 17:38:30 suihkulokki: yeah, that's the sad truth Nov 15 17:38:53 Jin^eLD: I hope you have the time and money to take them to court Nov 15 17:39:05 doh... Nov 15 17:39:53 Jin^eLD: sure, you get all the *gpl* piece sources. but not for the stuff they created themself. ie wifi and other driver, "multimedia" libraries etc Nov 15 17:40:07 hmm that is true indeed.. Nov 15 17:42:08 case examples: broadcom wireless, intel ipp library Nov 15 17:43:56 anyone much familiar with hostap/wifi interior wonkery? Nov 15 17:44:00 well. nevertheless I am happy to see Linux becomming more and more present in the embedded world.. Nov 15 17:49:15 * mrz80 is back (gone 01:08:51) Nov 15 17:54:25 I really don't get it.. DFB debug messages seem pretty normal, and then somewhere out of nothing: (!) [ 911: 20.504] --> Caught signal 11 (sent by the kernel) <-- Nov 15 17:56:01 I guess I need get gdb on the platform Nov 15 17:57:26 hmm.. gdb package seems to be broken.. Nov 15 18:04:30 ~nslookup download.epson-europe.com Nov 15 18:04:33 download.epson-europe.com is 62.41.72.156 Nov 15 18:25:40 03cbrake 07org.oe.dev * rf4893d86... 10/ (1 packages/linux/logicpd-pxa270_2.6.17-rc5.bb): logicpd-pxa270: do not include any files in the kernel package Nov 15 18:25:51 03cbrake 07org.oe.dev * r2a7d899b... 10/ (1 packages/linux/compulab-pxa270_2.6.16.bb): compulab-pxa270: do not include any files in the kernel package Nov 15 18:33:24 btw, has anyone heard of ptxdist? I learned that one can create a small distro from scratch with it; but same can be done with OE, right? Nov 15 18:33:50 and more! Nov 15 18:34:28 a 4 line recipe in OE takes a gazillion lines in ptxdist Nov 15 18:34:41 and ptxdist doesn't track shared libraries for you like OE does Nov 15 18:34:56 ptxdist also doesn't create toolchains Nov 15 18:35:20 I see.. heh :) well, I never heard about ptxdist before, just learned today that one of the distros for some hardware that we received was created with that Nov 15 18:36:03 managers like it because it has the same config interface as the linux kernel Nov 15 18:36:12 'OMG!!! KConfig!!!!' Nov 15 18:36:27 hehe Nov 15 18:37:00 the creators (penguitronix) are doing nice work on the linux kernel, though Nov 15 18:37:07 well, I have to admit it takes time to really understand and setup the OE environment :) I am still not sure if I did everything right Nov 15 18:37:13 Jin^eLD: OE takes longer to learn and comprehend than simpler systems like buildroot and ptxdist, but OE can do a lot more. Just look at the package count :-) Nov 15 18:37:28 cbrake: yes, that was actually what I was going to say :) Nov 15 18:37:40 so at our company I tought people to use OE Nov 15 18:38:02 and I guess we will stick with that, but we are not creating distros from scratch... Nov 15 18:38:15 so we still have to live with stuff that gets delivered Nov 15 18:38:23 * chouimat is thinking about an OE template for kdevelop ... Nov 15 18:38:56 cbrake: also, look who's using each system Nov 15 18:39:05 Jin^eLD: OE is pretty easy to use with external toolchains/glibc. Are you doing that now? Nov 15 18:39:21 cbrake: yes I am Nov 15 18:39:42 cbrake: but as I understood it is better to creata metadata packges out of external distros so they also go to the staging dir Nov 15 18:39:48 and that is not always trivial Nov 15 18:40:06 because you often get a bunch of stuff that is just somewhere Nov 15 18:40:41 I figured out the main thing though, how to use the prebuilt gcc, they all use crosstool so that's at least somehow a standard Nov 15 18:40:48 Jin^eLD: yes, I've never went that far. All I've done is build the extra packages I need and untar them on the target. I can see that would be a challenge. Nov 15 18:41:30 yep.. Nov 15 18:41:54 well, so far I do it the risky way, probably missing some rdepends because my staging is not filled up correctly Nov 15 18:42:01 there was some warning about it in the docs Nov 15 18:42:08 but all seems to work Nov 15 18:45:33 enough work for today.. time to code my own project :) and time for some food :> l8r Nov 15 18:59:29 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Nov 15 19:24:50 Jin^eLD, good to hear you got it working.. Nov 15 19:36:52 CosmicPenguin: any chanche for getting an OLPC for my QA 'studies'? Nov 15 19:37:18 zecke: I'm surprised that jg didn't respond to you Nov 15 19:37:23 :( Nov 15 19:37:40 probably just too busy, I consider mailing him again Nov 15 19:37:52 yeah - since the build happened, things are much quieter Nov 15 19:40:14 hi zecke Nov 15 19:40:39 hey Nov 15 19:43:12 Gerrath: well.. I am still not any further with DirectFB Nov 15 19:44:37 the only app that works is dfbg :) everything else crashes with signal 11 Nov 15 19:49:55 Jin^eLD, I got a lot of sig 11's in directfb do to input devices as well. Nov 15 19:49:59 heh http://dev.laptop.org/ticket/249#comment:5 Nov 15 19:50:19 Jin^eLD, strace it again and see what is right before the sig 11. Nov 15 19:50:23 Gerrath: I tried disabling all of them, because I also had an issue with that Nov 15 19:50:26 pb_: I need some autoconf help/counseling Nov 15 19:50:46 Gerrath: strace did not help me, for some reason strace did hang way before the sig11, it hangs in some nfs call Nov 15 19:50:48 my rootfs is mounted via nfs Nov 15 19:50:53 zecke: oh dear Nov 15 19:51:09 Gerrath: so strace did not tell me much at all.. Nov 15 19:51:41 Gerrath: I also tried it with gdb but did get any meaningful results Nov 15 19:54:22 pb_: http://www.openembedded.org/filebrowser/org.openembedded.dev/contrib/site-conf Nov 15 19:54:41 pb_: can I somehow avoid this aclocal.m4? Nov 15 19:54:44 lo Nov 15 19:56:12 hey gerwinin Nov 15 19:56:50 zecke: I'll take a look Nov 15 19:57:23 Koen : how do I get my patches config files and stuff in oe ? Should I just submit them to bugzilla ? Nov 15 19:57:56 gerwinin: yes, attach them to seperate bugs Nov 15 19:59:25 Koen : okay Nov 15 19:59:29 sorry i got strange error from mtn ... >>>mtn: misuse: path _MTN/tmp/1 already exists<<< doing mtn update ... any clue ? Nov 15 20:00:12 gremlin[it]: rm _MTN/tmp -rf ; mtn up Nov 15 20:00:22 zecke: can't you just put the contents of that file in configure.ac directly? Nov 15 20:00:23 Koen: maybe we should upgrade util-linux Nov 15 20:00:49 pb_: no, I do not want this Nov 15 20:01:04 'morning' Nov 15 20:01:15 >>> mtn: misuse: path 'contrib/site-conf' already exists, cannot create <<< Nov 15 20:01:15 zecke: I guess I don't understand what you do want Nov 15 20:01:16 gerwinin: aren't we using the latest util-linux already? Nov 15 20:01:20 grrrr Nov 15 20:01:21 hey hrw Nov 15 20:01:38 Koen : Let me checl Nov 15 20:01:40 pb_: We have this site cache, it generated results from different projects builtin autoconf tests, dbus, ... Nov 15 20:01:40 let test opie/amd64 Nov 15 20:01:55 pb_: and I want one m4 file for each project/program Nov 15 20:02:01 gerwinin: I known it fails for x86 Nov 15 20:02:15 pb_: so you can just call autoreconf -I m4, execute ./configure on a target system and have a working site cache Nov 15 20:02:16 gerwinin: applying the patch from debian/ubuntu didn't help Nov 15 20:06:27 Koen: I couldn't find that one Nov 15 20:06:43 Koen: I made a patch so far myself :( Nov 15 20:07:19 zecke: okay. I still don't really understand what your aclocal.m4 file is about though. Nov 15 20:08:39 pb_: okay. Imagine we have some results of samba in i686-linux Nov 15 20:08:57 pb_: now I want to copy and paste the autoconf to samba.m4 Nov 15 20:09:06 pb_: and somehow include it in configure.ac and execute it Nov 15 20:09:21 zecke: and then run it on wrt54 which is mipsel? Nov 15 20:12:38 ~lart collie Nov 15 20:12:38 * ibot decapitates collie conan the destroyer style Nov 15 20:19:22 03justinp 07org.oe.oz354x * r99721413... 10/ (35 files in 4 dirs): efl, e17: update to newest versions Nov 15 20:19:27 03justinp 07org.oe.dev * r7d8e0f1f... 10/ (35 files in 4 dirs): efl, e17: update to latest versions Nov 15 20:19:33 03justinp 07org.oe.dev * r5030fa4f... 10/ (1 contrib/mtnpatch.py): mtnpatch: update for newer monotone versions, move patch command to end to allow for rename or add and patch Nov 15 20:24:20 03justinp 07org.oe.oz354x * r05aac212... 10/ (1 conf/local.conf.sample): local.conf.sample: fix typo Nov 15 20:24:27 03hrw 07org.oe.dev * rf4cb521c... 10/ (5 files in 2 dirs): debianutils: updated to 2.17.3, dropped older Nov 15 20:28:26 03justinp 07org.oe.dev * r1c5b1996... 10/ (1 packages/musicpd/mpd_0.11.5.bb): mpd: add faad2 support Nov 15 20:28:28 03justinp 07org.oe.dev * r0eb8efc4... 10/ (1 packages/gstreamer/gst-plugins-base_0.10.7.bb): gst-plugins-base: fix PROVIDES Nov 15 20:28:47 hrw: would you have a problem switching ENABLE_BINARY_LOCALE_GENERATION to use ?= so it can be overridden? Nov 15 20:28:55 hrw: qemu-native is failing build on my machine... Nov 15 20:29:03 JustinP: 0.8.2 fails? Nov 15 20:29:08 03hrw 07org.oe.oz354x * r3e901f95... 10/ (3 files in 2 dirs): Nov 15 20:29:08 debianutils: updated to 2.17.3, dropped older (from .dev) Nov 15 20:29:08 applied changes from 5030fa4fd12eecce406862803a77d7b278ac71d7 Nov 15 20:29:08 through f4cb521c91e69fd20d48512754141dbffe32d446 Nov 15 20:32:27 hrw: I'm trying again right now to make sure Nov 15 20:35:41 drw: still working on IOP stuff? Nov 15 20:36:35 Koen: found out why it fails Nov 15 20:36:46 hrw: hmmm...perhaps something changed....qemu-native built. I'll re-try glibc Nov 15 20:36:51 koen: IOP? Nov 15 20:37:06 drw: the intel nas stuff Nov 15 20:37:12 utils is looking for asm/io.h but that should be asm-i386/io.h Nov 15 20:37:18 drw: probably another Dan Williams :) Nov 15 20:37:28 koen: not part of that Nov 15 20:37:43 yeah the other one (kernel hacker) Nov 15 20:38:06 drw: that explains a lot :) Nov 15 20:38:17 Koen: now I need to see how I can make utils-linux to this directory Nov 15 20:38:37 koen: I'll take that in a good way... ;) Nov 15 20:39:27 drw: :) Nov 15 20:50:19 ~lart libiconv Nov 15 20:50:20 * ibot brandishes Excalibur! "With this sword, I vanquish thee, libiconv!" and lops off libiconv's head Nov 15 20:50:36 hehe Nov 15 20:52:06 it's just a flesh wound Nov 15 20:55:39 just to start flame again monotone ... is possible to get (checkout) only a part of a tree ? Nov 15 21:06:34 is an admin of OESF-Forums here? Nov 15 21:06:48 patriot: unlikely, OE has nothing to do with OESF Nov 15 21:07:51 * chouimat|bored wonders if update to monotone 0.31 worth it Nov 15 21:08:01 12h14mmmmm ive been now waiting for one week, to get validated with my account on oesf Nov 15 21:08:09 ~oesf Nov 15 21:08:11 i heard oesf is rumour has it, oesf is newer name of ZUG - it provides forum for Zaurus/Ipaq/Simpad/etc users (see zug). It has nothing in common with OpenEmbedded (except for the first two letters), see http://www.oesf.org/ Nov 15 21:08:30 *ouch* Nov 15 21:08:32 patriot: oesf has nothing to do with OE Nov 15 21:08:53 glibc-2.3.5+cvs20050627-r3/temp/run.do_prep_locale_tree: cp: Argument list too long Nov 15 21:09:07 now how do I fix *that* one Nov 15 21:09:10 i thought so, because this channel is posted on there site Nov 15 21:09:56 JustinP: xargs is your friend Nov 15 21:09:59 * v8jlene is back Nov 15 21:10:19 CosmicPenguin: yes, but that's not my code...this is glibc from oz354x Nov 15 21:10:29 then xargs is their friend... :) Nov 15 21:10:34 its a very friendly tool Nov 15 21:10:41 I suppose I'll try to xargs it if it fails again... Nov 15 21:10:47 yes and no Nov 15 21:11:00 for example....how the hell do I make find and xargs work together consistently Nov 15 21:11:02 Morning everyone Nov 15 21:11:21 hey v8jlene Nov 15 21:11:26 it tends to try to run one long command....and it won't work with spaces Nov 15 21:11:45 Koen: problem solved Nov 15 21:12:01 gerwinin: cool Nov 15 21:12:15 Koen: I will add it to bugzilla Nov 15 21:12:28 gerwinin: keigaaf ;) Nov 15 21:12:43 JustinP: if you use FreeBSD recompile the kernel with a longer command line length Nov 15 21:12:52 JustinP: alternatively use GNU Hurd :) Nov 15 21:13:12 Koen: Jazeker Nov 15 21:13:39 Koen : Als ik mirror voor oe wil zijn hoe regel ik dat ? Nov 15 21:14:04 gerwinin: source tarballs or monotone? Nov 15 21:14:24 Koen : source to start with Nov 15 21:14:28 zecke: are you telling me you all are using OE on FreeBSD or HURD? Nov 15 21:14:36 Because sometimes oesources is so slow Nov 15 21:14:39 zecke: I've never heard of anyone actually using HURD.... Nov 15 21:14:52 gerwinin: hrw or zecke should be able to help you Nov 15 21:15:01 JustinP: hehe, right Nov 15 21:15:03 zecke: I'm on Linux anyway Nov 15 21:15:14 Koen: we can discuss this issues on t-dose Nov 15 21:15:24 JustinP: I had the glibc issue on FreeBSD and then stopped there and I know that Hurd hasn't any such artificial limitations Nov 15 21:15:31 JustinP: IOError: [Errno 2] No such file or directory: '/home/koen/OE/monotone/org.openembedded.dev/packages/efl/ecore/configure-abstract-sockets.patch' Nov 15 21:16:13 Koen: you got enough $ to come overhere ? Nov 15 21:16:24 gerwinin: OV jaarkaart :) Nov 15 21:16:37 Koen: weekend or week ? Nov 15 21:16:42 weekend Nov 15 21:17:01 koen: ack....I'll fix that Nov 15 21:17:25 Koen: Do we have some merchandise for on t-dose Nov 15 21:17:49 gerwinin: maybe some OE shirts Nov 15 21:18:01 koen: if I come I have some OE shirts ;) Nov 15 21:18:11 Koen: bring them and sell them , I want one for sure :) Nov 15 21:18:32 Zecke/Koen: If I can help let me know Nov 15 21:18:55 Koen: do you think you can arrange free bandwith for an oe server ? Nov 15 21:19:13 Koen: I mean at the uni or something ? Nov 15 21:19:23 gerwinin: sort of Nov 15 21:19:45 Koen: I will ask gigabyte and jetway to sponsor us a server Nov 15 21:20:03 gerwinin: the server I have is under 'reconsideration' Nov 15 21:20:21 Koen: The more servers the better :) Nov 15 21:20:44 gerwinin: I think I have some credits left to plugin a 1U server in the rack Nov 15 21:20:50 03justinp 07org.oe.oz354x * r0ef5f4c1... 10/ (1 packages/efl/ecore/configure-abstract-sockets.patch): ecore: add missing patch Nov 15 21:20:55 03justinp 07org.oe.dev * rd30814db... 10/ (1 packages/efl/ecore/configure-abstract-sockets.patch): ecore: add missing patch Nov 15 21:21:30 JustinP: fwiw, that abstracts socket thing should be able to get 'fixed' via site files Nov 15 21:21:49 Koen: I will ask for the server , maybe we should place in every continent a server Nov 15 21:22:05 gerwinin: that's is my plan as well Nov 15 21:22:17 Koen: I will try to help you Nov 15 21:22:21 koen: I figured....but I had no idea whatsoever how to set it.... Nov 15 21:22:30 oesources is in california and oe.org in berlin Nov 15 21:22:45 JustinP: I needed that result for some other app Nov 15 21:22:46 Koen : I got some good links with hardware manufacturers Nov 15 21:22:47 oe.org is in berlin? Nov 15 21:22:55 JustinP: I'll fix ecore as well if it works Nov 15 21:23:02 zecke: yep, strato.de Nov 15 21:23:48 Koen: let me send a mail to gigabyte and jetway Nov 15 21:24:28 gerwinin: once there is a physical server, it shouldn't be hard to find hosting Nov 15 21:24:50 Koen: Well I do not have too much contact in that world Nov 15 21:25:32 koen: thanks Nov 15 21:25:58 Koen: would a sempron64 be enough ? Nov 15 21:26:09 CosmicPenguin: how do I get xargs to work....? The examples are simply not working. It always treats the input as one entry. Nov 15 21:26:16 gerwinin: I think any hardware as long as relocatable is fine Nov 15 21:26:17 gerwinin: a p100 would be enough Nov 15 21:26:38 koen: a p100 could run a msp430 emulation at half the speed :) Nov 15 21:26:51 koen: this would make use debian style :) Nov 15 21:26:58 Koen: If i give you a server than you the datacenter thing ? Nov 15 21:27:04 re Nov 15 21:27:27 Zecke: I made a bluetooth swiss knife with msp430 Nov 15 21:27:43 gerwinin: we use them to build ScatterWeb Nov 15 21:27:48 we will have source mirror in Oregon, USA if everything will go ok Nov 15 21:27:52 ......what is this line for? "cp -a /home/papercrane/oe/oz354.build/tmp/staging/arm-linux/lib/* $treedir/lib" Nov 15 21:28:04 Zecke: At my work I am hit by the pic microcontroller with this increadable stupid mplab Nov 15 21:28:04 v8jlene: hi Nov 15 21:28:30 v8jlene: can you help me to get perl-native built on x86_64? Nov 15 21:28:36 JustinP: http://ewi546.ewi.utwente.nl/tmp/fix-debug-packaging-for-e.diff Nov 15 21:28:39 gerwinin: they plan to use .NET micro vm on these boards until M$ pays the redesign we use the GNU CC :) Nov 15 21:28:41 Zecke: With those "nice" registers Nov 15 21:28:43 JustinP: any thoughts on that? Nov 15 21:28:53 gerwinin: they were schocked when they saw gdb-remote Nov 15 21:29:10 hrw|tv: Yeah, I see that. It worked when I last built it a few days ago. I'll look it at later today... after I finish getting these job applications finished and sent off. Nov 15 21:29:30 v8jlene: ok, you have access to such machine? Nov 15 21:29:40 zecke: I used pikdev under kde , at my work they also bought .net but after somebody made an alarm application with it we threw it away too dangerous :) Nov 15 21:30:06 koen: the reason, IIRC, that I didn't use += in those was because of ordering and duplicates issues Nov 15 21:30:26 Zecke: .net can never be cheaper , you need to spend a fortune on calming tablets and paracethamol Nov 15 21:30:30 JustinP: right, there's still the duplicates issues, but =+ prepends stuff Nov 15 21:30:34 hrw|tv: Yes, my main system here is an AMD X2 running x86_64. Nov 15 21:30:40 ok Nov 15 21:30:40 JustinP: my goal is to get the -dbg packages working Nov 15 21:30:57 v8jlene: my main system is X1 only Nov 15 21:31:01 gerwinin: can't you just replace the developers with zombies? Nov 15 21:31:09 (i think you'd save money that way, too) Nov 15 21:31:27 koen: suppose I could test it and see what ends up in the packages... Nov 15 21:31:30 raduga: helmets can be expensive :) Nov 15 21:31:34 koen: will have to be later, though Nov 15 21:31:51 Raduga : I was obligated to work with .net and I threw it out of the window and went back to my favourite vi Nov 15 21:32:11 Raduga: After that I felt much better Nov 15 21:32:21 i feel much better to hear that, too Nov 15 21:33:21 back to matrix.. Nov 15 21:33:37 JustinP: http://ewi546.ewi.utwente.nl/tmp/fix-debug-packaging-for-e-the-easy-way.diff :) Nov 15 21:40:30 koen: ;-) that works Nov 15 21:40:46 koen: feel free to check it in (or I can do it if you like) Nov 15 21:41:37 JustinP: committed Nov 15 21:42:10 03koen 07org.oe.dev * rd65177cb... 10/ (1 classes/e.bbclass classes/efl.bbclass): e classes: include dbg packages Nov 15 21:43:38 koen: does oz354x need that fix? Nov 15 21:43:55 JustinP: no, oz doesn't know about -dbg packages Nov 15 21:44:24 koen: ok, thanks Nov 15 21:46:41 anyone here know how to properly use xargs? I used to be able to use it.... Nov 15 21:47:53 gerwinin: could you attach the patch instead of pasting it inline? Nov 15 21:48:06 gerwinin: and bugs are only resolved if the patch is in oe :) Nov 15 21:56:20 ....hmmm....it's working differently now.... Nov 15 21:59:11 great....I built qemu-native again finally but now "Unable to load interpreter" Nov 15 22:07:28 koen : okay Nov 15 22:10:26 psokolovsky: what do you think about using LAB as 3rd stage on hx4700? Nov 15 22:10:42 psokolovsky: I really hate it to digg out a cf card to flash the kernel Nov 15 22:15:08 damn it.... Nov 15 22:15:17 why is qemu doing this to me? Nov 15 22:15:49 JustinP: ${PN}-dbg needs to go before ${PN} Nov 15 22:17:02 RP: koen checked it in ;-) I'll fix it, then Nov 15 22:17:34 RP: .dev hasn't been able to build up to efl for me lately as it is Nov 15 22:17:49 RP: do you know why qemu might not be able to find its interpreter? Nov 15 22:18:26 JustinP: ah, yes. I'm getting the commits/authors confused. Not my day today :) Nov 15 22:18:54 RP: he authored it as me? Nov 15 22:18:56 03koen 07org.oe.dev * r2bca6e25... 10/ (1 classes/e.bbclass classes/efl.bbclass): Nov 15 22:18:56 e classes: fix -dbg Nov 15 22:18:56 RP says: "${PN}-dbg needs to go before ${PN}" Nov 15 22:19:08 JustinP: I've never seen that before with qemu :-/ Nov 15 22:19:42 I've seen it before....I don't recall what the fix is, though... Nov 15 22:19:54 blah.... Nov 15 22:19:56 this should be fun Nov 15 22:21:46 koen: check the diff of that last commit :-/ Nov 15 22:22:05 koen: missing space Nov 15 22:22:08 RP: already fixed Nov 15 22:22:13 koen: already fixed Nov 15 22:22:18 03justinp 07org.oe.dev * r3ab6e538... 10/ (1 classes/e.bbclass classes/efl.bbclass): efl.bbclass, e.bbclass: move -dbg before Nov 15 22:22:37 heh....shell escaped... Nov 15 22:23:44 03koen 07org.oe.dev * r60961ec4... 10/ (1 packages/images/e-image.bb): e-image: the e dudes have gotten a clue and are using .desktop files nowadays, so the convert script isn't needed anymore Nov 15 22:25:04 JustinP: all the e files are still using the old DEPENDS and RDEPENDS format, btw Nov 15 22:25:17 koen: ? Nov 15 22:25:22 old format? Nov 15 22:25:37 listing stuff in DEPENDS and RDEPENDS Nov 15 22:25:44 like entrance Nov 15 22:25:56 DEPENDS = "bash", REPENDS = "bash" Nov 15 22:26:30 JustinP: putting it only in rdepends is enough in .dev Nov 15 22:26:34 ok Nov 15 22:26:46 well....perhaps I'll update that after I get this damn glibc problem worked out Nov 15 22:26:56 or perhaps I'll just turn off binary locale genration again Nov 15 22:27:10 you do need .deb Nov 15 22:27:12 ehm Nov 15 22:27:14 .dev Nov 15 22:27:37 .oz354x is quite 'old' Nov 15 22:27:51 "need"? for what? Nov 15 22:27:59 for the new RDEPENDS stuff Nov 15 22:28:07 yeah, I know that Nov 15 22:28:26 I'm still running my Z on oz354x, though as .dev doesn't work for me Nov 15 22:28:47 if you're building for OZ, you can just emerge qemu and ASSUME_PROVIDED += "qemu-native" Nov 15 22:29:20 RP: having fun with ncurses? Nov 15 22:30:15 koen: I had some problems with that before....guess I'll try it again Nov 15 22:30:19 re Nov 15 22:30:49 koen: Thinking about it atm. Its going to need some planning Nov 15 22:31:02 Do we need to maintain a bitbake commandline mode? Nov 15 22:31:15 RP: why does bitbake execute do_clean on a failed build nowadays? Nov 15 22:31:34 * RP blames tinderbox Nov 15 22:31:43 NOTE: package epsilon-0.3.0.007-r0: task do_compile: failedERROR: TaskFailed event exception, abortingNOTE: package epsilon-0.3.0.007-r0: task do_clean: started Nov 15 22:31:59 koen: You're using tinderbox? Nov 15 22:32:02 yep Nov 15 22:32:08 * koen disables tinderbox Nov 15 22:32:52 I find that really annoying but I can understand why zecke did it :) Nov 15 22:33:23 I usually have all QA stuff enabled Nov 15 22:33:52 /data/build/koen/OE/build/tmp/angstrom/work/armv5te-angstrom-linux-gnueabi/epsilon-0.3.0.007-r0/epsilon-0.3.0.007/src/common/epsilon_thumb_common.c:56: error: conflicting types for 'epsilon_ipc_server_send' Nov 15 22:33:56 * koen stabs epsilon Nov 15 22:33:58 koen: you can disable auto cleaning Nov 15 22:34:16 at least I hope I finished that task Nov 15 22:38:55 RP: -c devshell rocks Nov 15 22:39:59 koen: Someone noticed at last :) Nov 15 22:40:36 koen: You can add variables to export into the shell in devshell.bbclass fwiw Nov 15 22:41:08 We should also add something to export oe's shell functions into the env, then it would really really rock :) Nov 15 22:41:16 RP: I think I have a misfeature in bitbake Nov 15 22:41:52 RP: packages/e17/entrance_0.9.0.007.bb Nov 15 22:42:23 koen: What about it? Nov 15 22:42:25 koen: building e images for angstrom? Nov 15 22:42:49 RP: that inherits e, so it should have PN-themes in PACKAGES, but bitbake refuses to build it since nothing provices entrance-themes Nov 15 22:42:56 JustinP: yes Nov 15 22:43:52 koen: bitbake doesn't build things in RDEPENDS space, just DEPENDS space Nov 15 22:44:55 Arguably we probably could support it now but it would get extremely confusing Nov 15 22:45:14 RP: I don't think that is the problem Nov 15 22:45:41 koen: Are you talking about typing "bitbake entrance-themese" or something else which pulls that in as a dep? Nov 15 22:45:54 'bitbake entrance' Nov 15 22:45:57 that bails out Nov 15 22:46:08 entrance has entrance-themes in RDEPENDS Nov 15 22:46:38 ah, hmmm Nov 15 22:46:51 I guess that typo wasn't fixed Nov 15 22:47:29 D'oh Nov 15 22:47:31 * koen mtn up Nov 15 22:47:35 cu Nov 15 22:47:58 -c devshell... have to remember Nov 15 22:48:01 RP: ${PN}${PN}-themes was the problem Nov 15 22:48:09 koen: :) Nov 15 22:48:20 koen: mtn update ;-) Nov 15 22:52:00 in ~70 minutes the angstrom armv5te feed should have the updated e packages Nov 15 22:53:41 * koen wonders what's keeping people from adding sun's GPL java bits to OE Nov 15 22:53:56 'night all Nov 15 22:54:00 koen: probably the same thing thats keeping them from adding OpenOffice Nov 15 22:54:07 so close Nov 15 22:56:51 BTW: I just found I have an unoccupied Dual Xeon system Nov 15 22:57:42 03rpurdie * r675 10bitbake/bin/bitbake: bin/bitbake: Improve -b error message Nov 15 22:58:24 Things like "bitbake -b linux-openz.*git" now work Nov 15 23:15:10 Koen: it seems gpe-image for x86(epia) compiles now Nov 15 23:16:07 Koen: tv-out needs some work Nov 15 23:18:47 Speaking of unoccupied Dual Xeon systems Nov 15 23:19:17 did any developers in southern california need one? Nov 15 23:19:37 ugh....qemu is still failing for me with an emerged copy with ASSUME_PROVIDED Nov 15 23:24:35 ggilbert_: the NSLu2 folks might be the only ones in california i know Nov 15 23:24:45 ah Nov 15 23:26:01 ggilbert_: I'm in northern CA, but I'm not a prodigious enough developer to need a dual xeon.... Nov 15 23:26:20 ggilbert_: you could always set up a tinderbox client on it and constantly compile ;-) Nov 15 23:27:55 ggilbert_: I have even scripts for that Nov 15 23:28:07 Yeah, but my power bill is high enough :) Nov 15 23:28:18 ggilbert_: hehe Nov 15 23:34:18 ggilbert: nslu2 has our own build farm in SoCal (SB), thanks. Nov 15 23:34:24 ah Nov 15 23:34:56 Well if you need more hardware, feel free to poke me. I actually work in the SB area part of the time so it wouldn't even be a pain for me to bring up. Nov 15 23:38:57 * JustinP wishes he could claim that hardware.... Nov 15 23:39:54 03justinp 07org.oe.oz354x * rfbab6d9b... 10/ (1 packages/glibc/glibc-package.bbclass): glibc: use xargs to avoid a long command-line Nov 15 23:40:03 ggilbert: I'll ask ka6sox to contact you - he looks after our hardware Nov 15 23:40:03 03justinp 07org.oe.dev * r96668be2... 10/ (1 packages/glibc/glibc-package.bbclass): glibc: use xargs to avoid a long command-line Nov 15 23:40:17 cool deal Nov 15 23:40:49 * JustinP turns off binary locale generation again Nov 15 23:40:55 * Jin^eLD burned his dual xeon when mounting ultra silent coolers :( Nov 15 23:41:37 those damn huge things.. touched the graphics card and produced a shortage Nov 15 23:41:54 Jin^eLD: lol silent and xeon? Nov 15 23:42:22 zecke: well yes :) there are huge coolers with heatpipes and whatever, and it is silent indeed Nov 15 23:43:27 ggilbert: are you still here? Nov 15 23:43:57 koen|away: http://macslow.thepimp.net/shots/cairo-clock_on_gtk-quartz.png cool Nov 15 23:44:21 ggilbert_: are you still here. Nov 15 23:44:22 yeah Nov 15 23:44:35 where in Socal is the server? Nov 15 23:45:02 I'm the infrastructure Manager for nslu2-linux.org Nov 15 23:45:23 03rwhitby 07org.oe.dev * r9fa712a9... 10/ (1 packages/apex/apex_1.4.7.bb): apex: Now works for LE too Nov 15 23:45:29 I'm Ventura, but I'm usually up at our santa barbara office once a week Nov 15 23:45:41 then we can do lunch Nov 15 23:45:58 I"m in Santa Barbara Nov 16 00:35:42 * gerwinin is fixing dbus compilation Nov 16 00:40:04 is there something that forces NPTL to require TLS or is that yet another configurable item? Nov 16 00:40:39 glibc-intermediate_2.5 fails with: ports/sysdeps/arm/nptl/tls.h:48:3: error: #error "TLS support is required." Nov 16 00:52:58 Hi I'm having problems with linux-hotplug. It fails on the install process. Here is the error log. http://pastebin.ca/248887 Nov 16 00:56:13 SonicvanaJr: what machine? Nov 16 00:56:22 SonicvanaJr: perhaps there are no rules set up for it or something.... Nov 16 00:56:29 Akita Nov 16 00:57:07 hmmm...then I have no idea Nov 16 00:57:32 what branch are you building from? Nov 16 00:57:42 oz354x Nov 16 00:58:25 yeah, that's building fine for me for spitz Nov 16 00:59:41 weird Nov 16 01:24:05 koen|away: cairo 1.3.2 is out, with the first round of performance improvements Nov 16 01:25:32 ugh. what is this horrid phpize thing? yet another hurdle to cross compilation? sigh. Nov 16 01:30:05 VoodooZ_, hold on, lemme look at my bb file... Nov 16 01:31:13 HopsNBarley: thanks. Nov 16 01:31:39 IF you can explain phpize (-; Nov 16 01:34:56 phpize? Nov 16 01:35:09 nicely done. Nov 16 01:39:07 VoodooZ, lemme see if i can send you my bb file. Nov 16 01:39:21 with DCC? Nov 16 01:41:28 i guess, i've never done this before. Nov 16 01:41:44 or simply email it Nov 16 01:42:01 i have an "offer" up. did you get anything? Nov 16 01:42:09 nope. Nov 16 01:42:22 I'm new to the irssi client though Nov 16 02:23:12 nite Nov 16 02:55:55 03lenehan 07org.oe.dev * r2948284e... 10/ (4 files in 2 dirs): Nov 16 02:55:55 quagga: Quote the rhs of all variable assignments since this will be Nov 16 02:55:55 required with future versions of bitbake. **** ENDING LOGGING AT Thu Nov 16 03:00:06 2006