**** BEGIN LOGGING AT Tue Aug 08 02:59:57 2006 Aug 08 07:26:10 quick question: to use my own crosscompile toolchain, do i put stuff in build/tmp/cross and do some assume_provideds? Aug 08 07:26:14 or is there more to it? Aug 08 07:28:13 giel: what abou reading documentation Aug 08 07:28:23 giel: and ask questions to improve the documentation? Aug 08 07:28:52 sadly we have little documentation, and the bits we have are bad Aug 08 07:29:03 but if people don't read+commen it will stay tht bad :) Aug 08 07:29:09 ofcourse Aug 08 07:29:10 okay Aug 08 07:29:16 1 mom, i'll rephrase ;) Aug 08 07:29:27 http://www.openembedded.org/user-manual&dpage=ch05s05 Aug 08 07:29:52 I have written this, and I really suck Aug 08 07:30:05 first, let me say i'm sorry about not reading documentation, point is, this appeared after i last looked at it ;) Aug 08 07:30:12 thanks for pointing it out to me Aug 08 07:31:13 morning Aug 08 07:33:02 morning Aug 08 07:35:54 giel: I really suck at writing documentation. So read it, comment it, ask questions :) Aug 08 07:36:24 also the -B option for gcc could be interesting... Aug 08 07:36:37 zecke: slightly rephrased question: i see the method in the documentation, but isn't it also possible to just dump my whole toolchain in tmp/build/cross and set the assume_provided, so all the paths are taken care of? Aug 08 07:37:09 zecke: i will give some more constructional comments on the documentation as is when i go through the steps Aug 08 07:38:55 giel: that would be possible Aug 08 07:39:09 giel: if you use the -B option with the toolchain Aug 08 07:39:36 giel: gcc refuses to find libstdc++ if moved around (unless you use -B) Aug 08 07:40:22 giel: you could even change the path to tmp/cross within your local.conf Aug 08 07:41:23 zecke: now that would be interesting Aug 08 07:42:04 zecke: what is the difference between $HOST_TOOLS and $TOOLCHAIN? Aug 08 07:42:04 giel: see conf/bitbake.conf for the name of the variable Aug 08 07:43:01 I don't see a variable $HOST_TOOLS Aug 08 07:43:26 e.g. setting CROSS_DIR will allow you to store 'cross' somehwere else Aug 08 07:43:39 oh, i mean in your documentation Aug 08 07:43:50 TOOLCHAIN=/where/is/toolchain/bin Aug 08 07:43:50 HOST_TOOLS=/where/is/hosttools/bin Aug 08 07:43:56 export PATH=$BITBAKE_PATH:$TOOLCHAIN:$HOST_TOOLS:$PATH Aug 08 07:44:02 giel: ah :) Aug 08 07:44:05 let me see Aug 08 07:45:01 host-tools like make, sed, uic (for Qt), gtk-doc which are used to generate/build/prepare stuff Aug 08 07:45:22 I thik I meant TOOLCHAIN the compiler itself Aug 08 07:45:36 HOST_TOOLS tools needed to compile userspace software Aug 08 07:47:01 pb__: you are everywhere! Aug 08 07:47:12 "net-tools are maintained by Phil Blundell and me" Aug 08 07:52:10 moin koen Aug 08 07:54:33 hey zecke Aug 08 08:01:42 koen: this gcc bug is funny Aug 08 08:02:01 which gcc bug? Aug 08 08:02:17 (and glibc-intermendiate fails to build on my powerbook) Aug 08 08:02:37 koen: the one posted by Daniel on linux arm kernel Aug 08 08:02:47 morning all Aug 08 08:02:57 ah, the % bug Aug 08 08:03:01 hey ade|desk Aug 08 08:03:23 koen: post your config.log, your log._do_configure and log.do_compile Aug 08 08:03:26 zecke: okay, i get it Aug 08 08:03:37 zecke: i'll play with this and probably come up with some more docs Aug 08 08:04:45 cool Aug 08 08:05:35 yop Aug 08 08:11:21 morning all Aug 08 08:13:22 hey RP Aug 08 08:14:02 RP: when you get alsamixer to create static, where are you running it, X/Opie or ssh? Aug 08 08:18:58 zecke: http://linuxtogo.org/~koen/index.php?path=OE%2FOSX/ Aug 08 08:20:41 checking whether arm-angstrom-linux-gnueabi-g++ -march=armv5te -mtune=xscale accepts -g... no Aug 08 08:20:45 interesting Aug 08 08:21:40 checking for --as-needed option... no Aug 08 08:21:45 that should be wrong as well Aug 08 08:21:51 our binutils should support that Aug 08 08:25:50 XorA: on the console or via ssh. No X/opie involved Aug 08 08:26:21 koen: how did you do that? Aug 08 08:26:22 XorA: Its not alsamixer that creates the static, it just gets turned on at the end of playback of any sound Aug 08 08:26:31 koen: upload your local.conf as well Aug 08 08:27:20 zecke: done Aug 08 08:27:46 koen: what did you build? Aug 08 08:28:10 koen: virtual/libc should be enough Aug 08 08:28:11 zecke: I did 'bitbake nano' and hit the bug that it didn't build a toolchain, so I did a 'bitbake glibc' Aug 08 08:28:11 RP: ah right, I got the impression from your email that you only heard it when alsamixer was loaded Aug 08 08:28:26 RP: Im trying to make the c7x0 make the noise, its irritating me Aug 08 08:28:34 XorA: No, its at the end of any playback Aug 08 08:28:52 RP: exactly as I first saw it then, but now it wont bloody do it Aug 08 08:29:02 ~lart dissapearing bugs Aug 08 08:29:03 * ibot readies the nuke launcher and fires some rounds at dissapearing bugs Aug 08 08:29:16 morning Aug 08 08:29:37 RP: gpe-image on poodle suxx Aug 08 08:29:48 use opie-eye Aug 08 08:30:25 hrw|work: I know. I'm hoping someone will fix that now we have a decent kernel ;-) Aug 08 08:30:38 koen: MacForge Aug 08 08:31:16 god gpe on 32M is going to be panful Aug 08 08:31:31 RP: suspend/resume does not work Aug 08 08:31:39 koen: BTW: is your filesystem case sensitive? Aug 08 08:31:51 hrw|work: How did you suspend it? Aug 08 08:31:56 zecke: it's HFS+ Aug 08 08:32:09 there are four versions of HFS+ Aug 08 08:32:15 journaled/not journaled Aug 08 08:32:21 hrw|work: It just worked fine here... Aug 08 08:32:26 combined with case sensitive/case insensitive but case aware Aug 08 08:32:32 battery charging is a different matter :-/ Aug 08 08:32:36 RP: "apm -s" or cancel hold Aug 08 08:33:00 hrw|work: apm -s worked. I'm not sure the cancel button is setup for it Aug 08 08:33:25 RP: maybe we lack something in .oz354x Aug 08 08:33:48 hrw|work: or have something causing problems. This is a bootstrap-image Aug 08 08:35:29 I can try it later Aug 08 08:37:55 RP: I also need kernel modules fixing - poodle lack many Aug 08 08:38:20 since my last update of the .bb files bitbake is trying to compilge glibc_2.3.5+cvs20050627. I don't know why it's doing it. If I enter bitbake glibc it builds nothing because the version 2.3.2 is already builded. I also entered a PREFERRED_VERSION statement. Does exist a command that I see wich package depends on glibc_2.3.5? Aug 08 08:39:25 hrw|work: Agreed, its just not been important until now :) Aug 08 08:39:50 RP: I plan to move modules info to distro config Aug 08 08:40:38 hrw|work: That idea worries me a little. How are you planning to do that? Aug 08 08:41:16 zecke: indeed I am Aug 08 08:41:31 RP: extra bootstrap rdepends = 'wifi modules common modules' thing - machine related ones will stay in machine configs Aug 08 08:42:15 hrw|work: Ah, I thought you meant the writing of the defconfig. Did you see my ideas I posted to oe@ about that? Aug 08 08:43:06 pb__: hehe Aug 08 08:43:14 pb__: feel like adding net-tools to OE? Aug 08 08:43:32 pb__: BTW: g2c is a FORTRAN library? Aug 08 08:44:07 RP: 27.07.2006: Beyond BOOTSTRAP_EXTRA_RDEPENDS? Aug 08 08:44:27 hrw|work: RRECOMMENDS, not RDEPENDS Aug 08 08:44:40 hrw|work: yes, that;s the one Aug 08 08:45:08 koen: sure Aug 08 08:45:19 RDEPENDing on modules is almost always a bug Aug 08 08:46:28 I don't think anyone replied to that email... Aug 08 08:51:52 RP: I will Aug 08 08:52:03 pb__: should i get net-tools from tazenda.demon or berlios? Aug 08 08:52:05 morning all Aug 08 08:52:14 zecke: berlios, I think Aug 08 08:52:22 or maybe even debian Aug 08 08:52:35 pb__: I prefer the fedora patch over the debian one Aug 08 08:52:54 I will go for berlios mirror then Aug 08 08:55:54 righto Aug 08 08:55:56 florian_kc: good morning Aug 08 08:56:03 good morning Aug 08 08:56:19 * zecke doesn't like this 'dump' everything into one patch thing... Aug 08 08:56:40 it is hard to track what parts are for what kind of issues... Aug 08 08:57:21 zecke: right Aug 08 08:57:35 florian_kc: gpe-conf 0.2.1 released? Aug 08 08:57:48 03hrw 07org.oe.oz354x * r0ec106e9... 10/ (3 files in 3 dirs): hostap-conf: added TrendNet TEW PC16 card from http://cablemodem.fibertel.com.ar/lateral/weblog/2006/08/03.html#P427 Aug 08 08:57:54 03hrw 07org.oe.dev * r22a005b8... 10/ (4 files in 3 dirs): hostap-conf: synced with .oz354x (added two cards) Aug 08 08:58:42 ops.. looks like I need to sync .dev too Aug 08 08:58:45 hrw|work: not yet, but soon Aug 08 09:05:55 oh well, oz needs a goe maintainer :-) Aug 08 09:05:58 eh gpe Aug 08 09:06:29 florian_kc: CoreDump was kind of such Aug 08 09:06:54 pb__: will you ever merge the +45 patches Fedora has for net-tools? Aug 08 09:07:07 hrw|work: hmm.. "was" sounds bad :-( Aug 08 09:07:34 florian_kc: he is MIA recently Aug 08 09:08:22 if someone renamed OpenZaurus to somethign more generic it would become more popular :-) Aug 08 09:09:02 florian_kc: ipaq support again? Aug 08 09:10:35 hrw|work: it would be the perfect distribution for all sorts of devices. even the image for the netbook was close to be perfect... much better than any Familiar image ever. Aug 08 09:11:29 thx Aug 08 09:13:16 hrw|work: That's a fact... don't ask me why, but doing two minor things OE didn't know about (creating a link and one chmod) everything was fine. Aug 08 09:13:56 RP: something from kernel modules discussion: Aug 08 09:13:57 11:35 hrw@work:conf$ mtn dif .|diffstat distro/openzaurus-3.5.4.2.conf | 53 +++++++++++++++++++++++++++++++++++++ machine/poodle-2.6.conf | 29 -------------------- machine/zaurus-clamshell-2.6.conf | 54 -------------------------------------- Aug 08 09:14:07 fsck.. ugly paste Aug 08 09:14:56 florian: creating that link means you can't install extra modules Aug 08 09:15:20 florian: I'd rather have some firstboot script copy the modules into the rootfs Aug 08 09:15:32 koen: right Aug 08 09:15:49 s/OpenZaurus/OpenZebedee/ Aug 08 09:16:04 OpenPDA is already taken.. Aug 08 09:16:07 OpenPalmtop? Aug 08 09:16:16 * XorA hopes he remebers rght that Zebedee is a Jack in the Box Aug 08 09:18:10 Angstrom? Aug 08 09:18:32 hrw|work: there is a long wiki dicussion about distribution names in hh.org wiki Aug 08 09:18:40 hrw|work: I like Open Palmtop. That remember me of Open Palmtop Integrated Environment Aug 08 09:18:44 if someone deleted it there is always the history. Aug 08 09:19:06 koen: binutils-cross failed for me Aug 08 09:19:06 zecke: I also want angstrom available... but its for 2.6 machines only Aug 08 09:19:18 that means we could have OPXE as well, X Environment :-) Aug 08 09:20:02 we could have Opie and Opee Aug 08 09:20:17 Open Palmtop Integrated Environment and Open Palmtop Extragated(?) Env Aug 08 09:20:19 X Open Revolutional Assistant Aug 08 09:20:33 koen: for some reason I like that Aug 08 09:20:53 s-cross-2.17-r0/binutils-2.17/gas/read.c: In function 's_comm_internal':/Users/koen/Projects/OpenEmbedded/build/tmp/angstrom/work/armv5te-darwin/binutils-cross-2.17-r0/binutils-2.17/gas/read.c:1430: warning: 'stopc' may be used uninitialized in this function Aug 08 09:20:57 OPSE Aug 08 09:20:57 haha Aug 08 09:21:00 -Werror Aug 08 09:24:00 florian: you could always go the sharprom cloners way Aug 08 09:24:17 morning all Aug 08 09:24:38 florian_kc: mount the jffs2, sed s/openzaurus/superflorianROMROMROM/g, and distribute it :) Aug 08 09:24:40 hey lrg Aug 08 09:24:51 hi Liam Aug 08 09:26:08 hi lrg Aug 08 09:26:26 koen: uh no, i don't think i want to do this... Aug 08 09:26:28 koen: do you know why apache on ewi sends http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.2-test/rc0/tosa/gpe-install.tar.gz.md5 as application/x-tar instead of text/plain? Aug 08 09:26:33 hey lrg Aug 08 09:26:42 hrw|work: no idea Aug 08 09:26:50 do13: lrg mentioned something about you having some code from Sharp for the Tosa for the touchscreen Y bug Aug 08 09:32:47 zecke: copying some stuff in staging and pointing CROSS_DIR somewhere else, setting an ASSUME_PROVIDED seems to work nicely, except the compiler i try to use seems broken :( Aug 08 09:32:52 zecke: but that's not our fault :) Aug 08 09:32:59 hrw|work: i found some more missing updates, i'll push them as soon as the image is known to work. Aug 08 09:33:02 giel: well Aug 08 09:33:21 giel: if it is broken in regard to not finding binaries do man gcc and search for -B Aug 08 09:33:41 zecke: it's broken in regard to "error: insn does not satisfy its constraints:" Aug 08 09:33:48 lol Aug 08 09:33:52 on OpenSSL? Aug 08 09:33:52 zecke: internal compiler error: in copyprop_hardreg_forward_1, at regrename.c:1549 Aug 08 09:33:57 uh. dropbear Aug 08 09:33:59 so, more or less Aug 08 09:34:12 I had that yesterday with a gcc3.3.6 toolchain :) Aug 08 09:34:18 not that instruction though Aug 08 09:34:21 but this is codesourcery fuckery Aug 08 09:34:43 s/-O2/-Os/ Aug 08 09:34:47 let me guess iwmmxt? Aug 08 09:34:58 hm, will try Aug 08 09:35:23 koen: workarounded by mv *.md5 *.md5.txt Aug 08 09:35:24 armv5te Aug 08 09:36:03 zecke: anyway, a prebuilt toolchain seems to work with two "cp"s and 4 lines in local.conf Aug 08 09:38:59 awesome :) Aug 08 09:39:34 but use OE to create your toolchain :) Aug 08 09:39:40 morning Aug 08 09:39:43 Aug 08 09:51:57 wtf, net-tools start to annoy me Aug 08 09:52:54 what package is ifconfig in? Aug 08 09:53:03 busybox Aug 08 09:53:21 zecke why? Aug 08 09:54:21 because it loves to revert its config Aug 08 09:54:37 what? Aug 08 09:56:05 haha, I have outsmarted me Aug 08 09:56:20 I had the fedora packag net-tools-1.60 in my way Aug 08 09:56:31 lol Aug 08 09:56:42 how can that happen? Aug 08 09:56:43 it took the patches from SRPM :} Aug 08 09:57:01 fedora is no good on my opionion Aug 08 09:57:08 executing fakeroot alien .rpm from within packages/net-tools Aug 08 09:57:20 woglinde: at least they are capable of splitting up packages Aug 08 09:57:29 ???? Aug 08 09:57:30 and patches Aug 08 09:57:36 zecke: doesn't OE have an srpm.bbclass? Aug 08 09:57:49 koen: I think it has it Aug 08 09:58:06 but I wanted to learn some spec Aug 08 09:58:20 woglinde: net-tools is a good example about Debian's issues Aug 08 09:58:43 woglinde: you have one big diff with securiy fixes, patches, features, buildsystem stuff all chewed/mixed Aug 08 09:59:15 woglinde: there is no easy to identiy which hunk is for what, if security issues where fixed, and which part of the big patch is responsible... Aug 08 09:59:36 zecke did you apt-get source net-tools? Aug 08 09:59:46 yes Aug 08 09:59:55 and it has only one big patch Aug 08 09:59:59 ? Aug 08 10:00:08 than the maintainer suckz Aug 08 10:00:18 yes Aug 08 10:01:13 woglinde: one big diff. Aug 08 10:04:06 I need to imrpvoe my GNU make knowledge Aug 08 10:04:27 zecke: yes! Aug 08 10:04:34 zecke: and teach it to mickeyl Aug 08 10:04:39 heh? Aug 08 10:04:43 why mickeyl? Aug 08 10:04:47 so he can stop using qmake Aug 08 10:05:09 koen: qmake rocks. It generates Visual Studio and XCode projects Aug 08 10:06:20 koen: I will fix GNU make, just to make Schilling shut up Aug 08 10:06:34 koen: next time he will claim GNU make is broken, he will get sued ;) Aug 08 10:07:33 zecke whats shilling concern with gnu make? Aug 08 10:07:41 zecke: start the OpenSchilly project ;) Aug 08 10:08:04 he need to develop his own tar because the tar of sun und bsd suckz Aug 08 10:08:31 woglinde: well GNU make has 10 year old grave issues Aug 08 10:08:35 woglinde: so use smake ;) Aug 08 10:08:49 he is right that GNU make is basicly not maintained Aug 08 10:09:33 RP: pushed changeset which can interest you a bit Aug 08 10:10:34 03hrw 07org.oe.oz354x * r071f1cc9... 10/ (4 files in 3 dirs): Aug 08 10:10:34 Moved common kernel modules list from machine configs into distro config. Aug 08 10:10:34 OZ 3.5.4.2-rc0 lacked many modules in poodle images... Aug 08 10:10:39 03hrw 07org.oe.oz354x * r71e359e5... 10/ (1 conf/distro/openzaurus-3.5.4.2.conf): oz 3.5.4.2: bump to rc1 Aug 08 10:10:44 03hrw 07org.oe.oz354x * r27bf219d... 10/ (1 packages/meta/task-bootstrap.bb): task-bootstrap: bump PR due to module changes, set maintainer to me Aug 08 10:15:17 * RP waits for the servers to catchup Aug 08 10:16:31 RP: which servers? Aug 08 10:17:01 anyone has interest in net-tools? Should I push it to OE? Aug 08 10:17:12 zecke: push it Aug 08 10:17:33 koen: The commit list and the viewmtn on oe.org seem to be lagging atm Aug 08 10:17:53 RP: viewmtn is refreshed every 30 minutes Aug 08 10:17:57 RP: It can lack up to 30 minutes Aug 08 10:18:04 commits should be instantly though Aug 08 10:18:08 RP: the commit list suffers from th enormal hh.org brokenness Aug 08 10:18:28 koen: shut up Aug 08 10:18:46 I'll just have to wait then :) Aug 08 10:18:51 RP: the mails are sent at the same time as the cia notification, so if you see it in IRC, the mail has been sent Aug 08 10:19:12 koen: I know. That doesn't mean I can read them yet though :) Aug 08 10:19:26 koen: or we have enormal breakage on openembedded.org and the mail is stuck Aug 08 10:20:20 zecke: no Aug 08 10:21:50 zecke: with -Os it worked, now i only have a couple of the "NOTE: Couldn't find shared library provider for ...", but luckily you have documented how to fix that ;) Aug 08 10:21:59 zecke: 2006-08-08 12:35:01 1GAOvN-0006AN-UA <= oe-commits@handhelds.org H=localhost [127.0.0.1] P=esmtp S=10531 Aug 08 10:22:05 hehe Aug 08 10:23:15 koen: when mirgation to lists.ltg or lists.oe? Aug 08 10:23:16 I wouldn't want to configure a mail server like hh.org Aug 08 10:23:30 hrw|work: ask mickeyl and kergoth Aug 08 10:23:43 hrw|work: lists.oe.org iirc Aug 08 10:23:52 CIA-9: ping Aug 08 10:24:05 http://lists.openembedded.org/ Aug 08 10:24:05 03hrw 07org.oe.oz354x * r1eff48b4... 10/ (4 files in 2 dirs): zaurus machines: RRECOMMEND kernel modules instead of RDEPENDing on them Aug 08 10:24:45 CIA-9: it was 3 files in 1 dir Aug 08 10:24:52 http://www.linuxtogo.org/sponsors Aug 08 10:25:01 hrw|work: _MTN/ probably Aug 08 10:25:18 maybe Aug 08 10:25:43 koen: mickey|zzZZzz : woglinde : hrw|work : I just noticed. If we get-to-gether in Berlin we could found OE e.V. as wel Aug 08 10:26:03 zecke: that'd be nice Aug 08 10:26:39 zecke: Are there details of the proposed get together anywhere? Aug 08 10:26:54 zecke: Such as timescales, purpose etc? Aug 08 10:27:02 RP: no Aug 08 10:27:24 we do not have a location as well Aug 08 10:27:34 worst case. We all crash at mine and al's place Aug 08 10:27:59 woglinde: BTW: ROAD offered their rooms for the weekend Aug 08 10:28:32 zecke: lists.oe.org now redirect to oe.org/contact Aug 08 10:28:45 RP: timescale: 5th,6th,7th,8th october Aug 08 10:29:00 RP: puporse: talk about OE and its' and our future(s) Aug 08 10:29:25 zecke: and we'll only eat soup and chinese food Aug 08 10:29:40 zecke: we don't need forks ;) Aug 08 10:30:23 I have a tent, enough forks at home Aug 08 10:30:30 zecke: Thats more like the 1st week in October :) (Thur-Sun?) Aug 08 10:30:33 (wow my mom will love that...) Aug 08 10:30:57 RP: I'm not good at counting. The 1st week is the 1st october? Aug 08 10:31:07 RP: so basicly it is the beinning of October Aug 08 10:31:23 That's easier :) Aug 08 10:31:31 It depends when you class a week as starting as I guess Aug 08 10:31:32 *puh* Aug 08 10:32:02 ~lart google Aug 08 10:32:02 * ibot takes a large goose feather pillow and swings it wildly in google's direction, hitting google and sending google flying into the closet Aug 08 10:32:42 I could have made the google joke, but now it is too late Aug 08 10:33:23 RP: I think the purpose is in between: Having fun, Creating Code, Talking about issues, addressing them Aug 08 10:33:58 zecke: sounds good Aug 08 10:34:11 I think there are a fair few issues would could address :) Aug 08 10:34:19 morning all Aug 08 10:34:27 hi Dirk Aug 08 10:34:43 hey Richard Aug 08 10:34:49 larer guys Aug 08 10:36:30 cya zecke Aug 08 10:37:17 hey Dirk Aug 08 10:37:27 hi Koen Aug 08 10:50:16 hrw|work: did you build the new images from scratch? Aug 08 10:50:46 giel: out of interest. What are you building? Aug 08 10:53:18 no Aug 08 10:54:20 zecke: in the end, i'm trying to build our own dvb-h video player application, using ffmpeg Aug 08 10:54:23 zecke: but! Aug 08 10:54:31 lol Aug 08 10:54:44 zecke: i'm trying to hack some routines in ffmpeg to link against some libraries that do hw accelerated h264 Aug 08 10:55:05 zecke: and these libraries are built with the codesourcery 3.4.0-vfp toolchain, and i don't have the source Aug 08 10:55:05 giel: using Linux DVB? dvbnet? Aug 08 10:55:33 zecke: no, using a proprietary dvb-h layer that injects in the tcp-ip stack :( Aug 08 10:55:50 dibcom? ati? Aug 08 10:55:55 zecke: philips Aug 08 10:55:59 oh Aug 08 10:56:01 hehe Aug 08 10:56:12 but the dvb part is mostly working Aug 08 10:56:19 and our app as well Aug 08 10:56:23 giel: which FLUTE are you using if any? Aug 08 10:56:24 giel: same division as eFfeM-log ? Aug 08 10:56:30 it's just that i want to link against this hw-accel stuff Aug 08 10:56:32 RP, zecke: also signing opengpg keys Aug 08 10:56:39 koen: DVB is not done in netherlands... Aug 08 10:56:48 koen: otherwise I would have spent this summer in Eindhoven... Aug 08 10:56:58 zecke: at the moment we don't do flute, but we plan to do in the near future, probably madflute based Aug 08 10:57:01 hrw|work: GnuPG Aug 08 10:57:05 zecke: or, S3 stack Aug 08 10:57:21 zecke: S3 does do an ESG stack to which i have access Aug 08 10:57:24 giel: MADfcl sucks, mcl even more... Aug 08 10:57:28 zecke: pbly also flute in there Aug 08 10:57:31 giel: hehe Aug 08 10:57:43 bleh Aug 08 10:57:46 BTW: I'm sitting in the BMCO Office, doing student projects Aug 08 10:57:47 I hate writing docs Aug 08 10:57:51 zecke: cool! Aug 08 10:58:17 zecke: i've got all the streams from that first bmco trial still somewhere Aug 08 10:58:27 not doing them for BMCO but at their office Aug 08 10:58:31 zecke: and the stuff that got broadcasted at 3gsm Aug 08 10:58:37 hehe Aug 08 10:58:48 that is funny Aug 08 10:59:15 giel: the S3 ESG uses mcl as flute? Aug 08 10:59:38 up till now i could do with a fairly clean OE build running on our arm device, but it's the hw accelleration and the hardheadedness of the people who won't give me the source Aug 08 10:59:59 zecke: s3 stack is doing something of themselves Aug 08 11:00:19 giel: annoy them, at one point it is more easy for them to give the source away Aug 08 11:00:30 giel: are you doing that at philiips? Aug 08 11:00:48 zecke: yes, especially as this is some philips semiconductors division in france, and i'm philips researcg Aug 08 11:00:52 research! Aug 08 11:01:00 so, why can't they just give me the source! Aug 08 11:01:18 hehe Aug 08 11:01:34 it is a corporate desease Aug 08 11:01:42 I totally blame MontaVista for this mess Aug 08 11:01:46 hehe yes Aug 08 11:01:49 well, it's like this Aug 08 11:01:50 completely even... Aug 08 11:02:07 philips semiconductors hired philips apptech (applied technologies) to implement this Aug 08 11:02:13 zecke: gnupg is just one of implementations of openpgp Aug 08 11:02:17 I hope to have completed my BSD licensed FLUTE by the end of this month Aug 08 11:02:21 apptech won't give the source, because they want to sell it to sony ericsson as well Aug 08 11:02:33 zecke: keep me posted! Aug 08 11:02:55 ghehe Aug 08 11:02:55 florian_kc: I build only machine related stuff, generate machine feeds and build images from feeds Aug 08 11:03:06 BTW: hacking ffmpeg gives legal trouble Aug 08 11:03:25 zecke: oh, i'm fully aware of that, this is nothing that will ever end up in any product, ever Aug 08 11:03:25 hrw|work: ah okay Aug 08 11:03:26 one of the reasons they won't give you the source, even if they are forced to... Aug 08 11:03:40 zecke: i have a fair degree of freedom, being research Aug 08 11:03:41 giel: I have heard that plenty of times :) Aug 08 11:04:16 anyway, i can't seem to link against the only binary i have with anything else but the 3.4.0-vfp from codesourcery Aug 08 11:04:45 giel: don't know your skills but... Aug 08 11:05:16 florian_kc: this way I know that all stuff is available and do not have to duplicate packages on disk Aug 08 11:05:19 giel: that's probably because OE doesn't know how to generate hardvfp toolchains for gcc 3.x :( Aug 08 11:05:33 koen: might be it indeed Aug 08 11:05:39 giel: you use the csr toolchain? and you pass -march=iwmmxt and -mcpu=iwmmxt Aug 08 11:05:46 to the compiler or nothing? Aug 08 11:05:59 vfp and iwmmxt together? Aug 08 11:06:00 leet Aug 08 11:06:10 koen: either the linker was complaining about linking non-vfp code to vfp code, or about mixing EABI 0 and EABI 4 Aug 08 11:06:11 koen: I have included the ezx-cross-dev (gcc3.3.6) Aug 08 11:06:21 oh Aug 08 11:06:41 giel: take a look at the CSR 'spec' and either make sure OE is not passing -mcp or -march flags Aug 08 11:07:00 or use the right CPU settings Aug 08 11:07:11 zecke: uh, dunno. I haven't looked at those patches. Aug 08 11:07:13 also objdump can be a nice tool Aug 08 11:07:14 well, it seems that i am able to compile most of my stuff now with the 3.4.0-vfp anyway Aug 08 11:07:18 maybe bernd will merge them one day Aug 08 11:07:25 pb__: hehe Aug 08 11:07:53 zecke: but thanks for this pointers, i will try them somewhere this week... i'm not fulltime on this Aug 08 11:08:06 these* Aug 08 11:08:10 good luck Aug 08 11:08:56 koen: private flags = 602: [ACPS-32] [VFP float format] [software FP] Aug 08 11:09:02 koen: is that good? Aug 08 11:09:29 that's softvfp Aug 08 11:09:33 which is nice Aug 08 11:09:43 gcc 3.3.6 Aug 08 11:09:59 taken from ezx-cross-dev Aug 08 11:10:54 uhm wait a minute, maybe a stupid question... what is the CSR toolchain? i mis-interpreted the whole time for csl Aug 08 11:11:05 zecke: softvfp enables you to use iwmmxt Aug 08 11:11:38 CodeSourcey I think Aug 08 11:11:45 at least this is what I meant to say Aug 08 11:12:24 ah okay Aug 08 11:12:32 then pbly csr == csl ;) Aug 08 11:12:48 right :) Aug 08 11:13:02 I'm confused, too many terminals and screens... Aug 08 11:13:10 :) Aug 08 11:13:33 zecke: may I reuse some of the wiki in my SoCDOC? Aug 08 11:14:56 # PREFERRED_VERSION_gcc-cross = "3.4.4+csl-arm-2005q3" Aug 08 11:14:57 that one? Aug 08 11:15:50 hi CoreDump|home Aug 08 11:15:57 koen: sure Aug 08 11:16:01 giel: that's a csl toolchain, but I don't now how to tell it to generate EABI0 instead of EABI4 Aug 08 11:17:03 CoreDump|home: welcome back Aug 08 11:22:41 koen: i might just try to build the 3.4.0 csl 2004q1 from source with OE Aug 08 11:32:15 03xora 07org.oe.dev * rade4046c... 10/ (3 files in 3 dirs): gtk+_2.10.1.bb : rediff and re-add the no-demos patch Aug 08 11:35:57 CoreDump|home: can you fix poodle keyboard config in gpe? Aug 08 11:36:18 koen: let us migrate to launchd for angstrom Aug 08 11:38:42 zecke: that's actually a good idea Aug 08 11:39:05 what is launchd? Aug 08 11:39:27 apple's answer to initscripts Aug 08 11:39:31 ah Aug 08 11:39:40 koen: that gpe-conf bb is bad - or something else. it doesn't pull in things it needs for installing Aug 08 11:40:03 florian_kc: that .bb is made for .dev Aug 08 11:40:40 afaik .oz doesn't have BUILDALL_DEPS Aug 08 11:40:47 * hrw|work decides to not add gpe stuff into .oz354x with own hands Aug 08 11:41:47 koen: ah Aug 08 11:42:31 heh Aug 08 11:42:44 hrw|work: that's why i build from the scratch :-) Aug 08 11:43:32 koen: any more differences i should consider? Aug 08 11:44:34 hehe -DPOSIX_MISTAKES Aug 08 11:44:39 hehe -DPOSIX_MISTAKE Aug 08 11:44:40 even Aug 08 11:44:45 I love OpenBSD Aug 08 11:55:20 re Aug 08 11:58:22 hrw|work: wb Aug 08 12:06:47 hrw|work: the kernel patch broke? Aug 08 12:06:53 http://ewi546.ewi.utwente.nl/tinderbox/showbuilds.pl?tree=OpenZaurus-354x Aug 08 12:07:55 zecke: qemu-native Aug 08 12:08:42 TINDERBOX Task do_populate_staging failed (FAILURE) Aug 08 12:08:45 weird Aug 08 12:09:05 and linux-openzaurus is failing as well Aug 08 12:09:11 * zecke checks if the machine burns... Aug 08 12:09:43 zecke: which build is kernel failing? Aug 08 12:09:45 morning Aug 08 12:10:03 RP: poky fails as well Aug 08 12:10:36 when ich try to make a image I get Cannot find package task-bootstrap. Aug 08 12:11:16 RP: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=58&logfile=20060808043204.log Aug 08 12:12:06 hrw|work: http://ewi546.ewi.utwente.nl/tinderbox/showlog.pl?machine_id=80&logfile=20060808121831.log Aug 08 12:12:14 zecke: A fix for that was checked in recently ;-) Aug 08 12:12:49 RP: is this service useful for you? (tinderbox) Aug 08 12:12:59 zecke: can we have a tinderIRCbot? Aug 08 12:13:09 sure Aug 08 12:13:31 zecke: Partly :). I just wish the output was more readable :) Aug 08 12:13:43 even more readbale? Aug 08 12:14:22 "<--- TINDERBOX Task do_patch failed (FAILURE)" doesn't help me much Aug 08 12:15:00 RP: that is a bitbake bug Aug 08 12:15:08 Its a step the right way, we just need to build upon it. OH has plans to run our own automated builds at some point and we'll need a reporting infrastructure Aug 08 12:15:35 * RP -> food Aug 08 12:15:49 if possible please coordinate with me Aug 08 12:19:41 zecke: can tinderbox report also PR? Aug 08 12:19:46 ---> TINDERBOX Package linux-openzaurus-2.6.17 started Aug 08 12:19:52 ---> TINDERBOX Package linux-openzaurus-2.6.17-r15 started Aug 08 12:20:02 hrw|work: yes! Aug 08 12:20:09 hrw|work: tinderclient.bbclass needs to send it Aug 08 12:20:10 second version is usefull, first is not Aug 08 12:20:49 feel free to add it now Aug 08 12:20:56 I plan to do it tonight... Aug 08 12:21:34 zecke: can look but first have to finish e24 club, my VW and few other websites... Aug 08 12:21:46 sure! Aug 08 12:21:51 same applies here Aug 08 12:26:40 hi Cliff Aug 08 12:43:11 zecke: As soon as we know what we're thinking of doing, I'll discuss it with you :) Aug 08 13:00:23 how do I add the modules to the gpe-image? they come in an externel .tgz file that I don't think the Tosa flasher uses Aug 08 13:00:37 mikearthur: which one you lack? Aug 08 13:00:48 mikearthur: add them into BOOTSTRAP_EXTRA_RRECOMENDS Aug 08 13:01:36 hrw|work: well, it generates a file with all the modules, but they don't end up in the jffs2 Aug 08 13:02:05 mikearthur: modules*.tgz is extra bonus. all modules are packaged anyway Aug 08 13:02:20 mikearthur: I ask again: Which module you lack? Aug 08 13:02:22 name Aug 08 13:02:37 hrw|work: any, there are no modules on the device Aug 08 13:02:46 none at all? Aug 08 13:02:50 mikearthur: .dev or .oz354x? Aug 08 13:02:56 oz354x Aug 08 13:03:10 mikearthur: mtn pull;mtn update and rebuild Aug 08 13:07:57 mikearthur: 68 kernel modules installed in just built gpe-image for tosa Aug 08 13:08:08 ok Aug 08 13:08:18 mikearthur: That .tgz is just for help with kernel development. You don't actually have to use it ;-) Aug 08 13:08:29 RP: just checkin ;) Aug 08 13:09:02 sadly, you can't make half the tosa stuff modules anyway :( Aug 08 13:09:15 because then the pwm won't build Aug 08 13:09:34 pwm? Aug 08 13:09:56 i'm probably blind because i've been doing this all day, but why is the preferred version of glibc 2.4 not available while there really is a glibc_2.4.bb Aug 08 13:10:03 hrw|work: the image is ready but i'll test it before pushing Aug 08 13:10:09 florian_kc: cool Aug 08 13:10:19 florian_kc: h3900 one? Aug 08 13:10:24 hrw|work: yes Aug 08 13:12:50 RP: on tosa we have audio inside of kernel? Aug 08 13:13:13 hrw|work: We might have to have for the touchscreen/battery, I'm not sure Aug 08 13:13:35 RP: yes. inside of kernel Aug 08 13:13:40 just checked Aug 08 13:13:49 iirc ts/battery like you told Aug 08 13:17:51 giel: some bitbake bug, which he not yet addressed/searched/fixed Aug 08 13:19:18 hrw|work: only the alsa core is in the kernel. oss emulation are modules Aug 08 13:19:49 zecke, giel: I'm not sure it is an acutal bug - its to do with glibc-initial Aug 08 13:20:50 do13_: I was suprised a bit when saw only few sound modules Aug 08 13:21:41 It's correct, tosa needs ac97 to read the battery Aug 08 13:22:59 btw, can you remove pxaficp-ir from openzaurus.conf, these needs to stay in the machine config, because collie doesn't have this module Aug 08 13:23:41 RP: I have never looked at it, but it is another issue of "we need better error reporting capabilities" Aug 08 13:23:47 do13_: As long as its a RRECOMMENDS, it shouldn't be an issue Aug 08 13:24:22 zecke: Right, agreed. I'm still working on the dependency code rewrite and it should make this kind of thing much easier Aug 08 13:24:51 RP: thx, right I haven't read the whole line Aug 08 13:25:46 yipee Aug 08 13:25:58 I try to teach my body to prepare for the exams... Aug 08 13:26:07 sadly it is freaing/chickening out again Aug 08 13:26:19 getting really depressed :} Aug 08 13:26:27 zecke: :-/ Aug 08 13:26:46 * RP is suffering stress of a different form :-/ Aug 08 13:26:50 NAbyss: hey, progress on the survey? Aug 08 13:28:48 I think I should rent a cave for one month+internet access Aug 08 13:28:53 and just play cave man Aug 08 13:34:19 do13_: few other modules need such cleanup too but current situation is better then was anyway Aug 08 13:35:15 NOTE: Applying patch 'hrw-pcmcia-ids-r4.patch' Aug 08 13:35:16 ERROR: function do_patchcmd failed Aug 08 13:35:24 on building a gpe-image, just did a monotone update Aug 08 13:35:53 o. Aug 08 13:36:25 checking.. Aug 08 13:37:12 hrw|work: cheers Aug 08 13:39:54 hrw|work: agreed. any plans to do this in .dev? Aug 08 13:39:58 hrw|work: I seem to have r3 of the patch Aug 08 13:41:49 do13_: Did you see ny comments in the mail "beyond BOOTSTRAP_EXTRA_RDEPENDS" to oe@ ? Aug 08 13:41:54 hrw|work: changed to -r3 in the .bb, works fine Aug 08 13:42:05 mikearthur: pushed r4 Aug 08 13:42:14 hrw|work: ok Aug 08 13:42:18 mtn update again? Aug 08 13:42:28 mikearthur: pull and update :) Aug 08 13:42:29 mikearthur: pull/update Aug 08 13:42:40 I have to resend this patch to pcmcia list Aug 08 13:43:46 03hrw 07org.oe.oz354x * r6c655233... 10/ (2 files in 2 dirs): linux-openzaurus 2.6.17: added missing patch Aug 08 13:45:40 RP: sorry, which mail? Hmm I overlooked these mail :) Aug 08 13:46:00 do13_: .dev is a bit harder as there are many distros which use zaurus machines Aug 08 13:46:08 do13_: The subject was the quote Aug 08 13:46:40 do13_: Basically I was proposing a rework for the machine files and task-bootstrap - I've already applied to poky Aug 08 13:48:18 RP: sounds good. I'll dig later in mail folder :) Aug 08 13:49:12 If Zaurus people are agreeable, we could apply to the zaurus and maybe angstrom and others can convert if they want Aug 08 13:50:13 hrw|work: Is the modules stripping fixed in .354x? Aug 08 13:51:12 do13_: what was error? Aug 08 13:51:31 RP: angstrom would welcome it Aug 08 13:51:49 RP: although I'm pretty happy with what packages/angstrom is doing Aug 08 13:51:51 hrw|work: Hmm from memory: modules aren't stripped into the image Aug 08 13:53:01 hrw|work: http://bugs.openembedded.org/show_bug.cgi?id=1182 Aug 08 13:54:10 and glibc doesn't package utf8 stuff Aug 08 13:54:50 hrw|work: do you know of any program that can re-indent a php file? Aug 08 13:54:58 koen: This would basically be a solution to the problem of the machine speciifc modules which task-angstrom doesn't currently cover Aug 08 13:55:18 koen: vim? Aug 08 13:55:20 koen: vim can re-indent almost every programming language Aug 08 13:56:51 do13_: I did not touched it Aug 08 14:00:10 hrw|work: works fine now Aug 08 14:03:05 hey all Aug 08 14:03:34 could someone confirm that sbagen is not in OE, please ? Aug 08 14:04:42 alan|philo: is 'find' or 'ls' broken on your computer? Aug 08 14:05:25 koen: i don't have enough space for now, so i don't have any local copy anymmore Aug 08 14:05:48 and afaik, no bot here can answer that question... Aug 08 14:06:10 alan|philo: http://www.openembedded.org/filebrowser/org.openembedded.dev/packages/ Aug 08 14:06:17 hrw|work: works... apart from some ipaq issues. pushing now. Aug 08 14:06:48 koen: wonderfull ! i had forgotten that page... :/ Aug 08 14:06:50 florian_kc: great Aug 08 14:07:01 * hrw|work will push #524 into .oz354x Aug 08 14:07:26 hehehe ! it's in ! Aug 08 14:08:44 alan|philo: http://hrw.one.pl/tmp/PICT0002.JPG - Zaurus pile Aug 08 14:09:09 ~lart people who paste patches inline Aug 08 14:09:09 * ibot gives people who paste patches inline a "free" copy of Windows and then charges double for "Upgrades" Aug 08 14:09:27 koen: normal on lkml Aug 08 14:11:40 hrw|work: (O_o) ... (T_T) ... (^_^) Aug 08 14:12:16 ~lart monotone Aug 08 14:12:17 * ibot send killer squirrels to attack monotone Aug 08 14:12:18 ~lart monotone Aug 08 14:12:18 ~lart monotone Aug 08 14:12:48 mtn: misuse: 17 missing files; use 'mtn ls missing' to view Aug 08 14:14:42 koen: I think we should compile all start scripts into a binary Aug 08 14:15:01 koen: time one run, run a News Story that distris created using OE Aug 08 14:15:11 start up to 200% faster than MVista... Aug 08 14:15:16 and then publish a news story Aug 08 14:15:38 zecke: mickeyl's plan was to replace altboot with his boot-manager Aug 08 14:15:53 zecke: so all we need is to integrate launchd into that Aug 08 14:16:13 koen: we can do it as hacky as MVistas start scripts Aug 08 14:16:29 what a crap.. not even the options mtn itself tells me work Aug 08 14:16:30 zecke: fwiw, intel cancelled its contract with mv and is using timesys tools now Aug 08 14:16:32 just to win the record... Aug 08 14:16:39 florian_kc: mtn list missing Aug 08 14:16:49 koen: oh Aug 08 14:17:08 florian_kc: pull/update/commit/push Aug 08 14:17:18 zecke: lennert and I were drooling over the dual-core xscales with 2x512kB cache Aug 08 14:17:33 timesys has nice presentation 'why to choose us' Aug 08 14:17:43 hrw|work: that's what i did - pull&merge ran, up failed Aug 08 14:18:03 florian_kc: mtn revert --missing? Aug 08 14:18:07 hrw|work: it complained about files i removed (and checked in) as well as files someone else removed. Aug 08 14:18:18 florian_kc: hm.. Aug 08 14:18:30 hrw|work: that's what fails, it gives me the help Aug 08 14:18:33 hrw|work: the presentation was nice Aug 08 14:18:33 03florian 07org.oe.oz354x * r4a7bc9b7... 10/ (1 packages/gpe-edit/gpe-edit_0.31.bb): gpe-edit: add 0.31 Aug 08 14:18:39 03florian 07org.oe.oz354x * r7e639a13... 10/ (6 files in 3 dirs): gpe-edit: drop some ancient versions Aug 08 14:18:40 koen: how is this dual core thing named? Aug 08 14:18:47 03florian 07org.oe.oz354x * rab8d79c2... 10/ (3 files in 2 dirs): gpe-bluetooth: add 0.55, drop 0.51 Aug 08 14:18:56 03florian 07org.oe.oz354x * r864dd012... 10/ (6 files in 2 dirs): gpe-what: add 0.43, drop ancient versions. Aug 08 14:19:04 03florian 07org.oe.oz354x * rb2a74eb5... 10/ (1 conf/distro/preferred-gpe-versions-2.8.inc): preferred-gpe-versions-2.8.inc: New file - this is not a release, but we start fixing the releases. Aug 08 14:19:11 03florian 07org.oe.oz354x * r11646fa3... 10/ (6 files in 3 dirs): gpe-aerial: include fix for compatibility with newer libgpewidget. drop old releases. Aug 08 14:19:16 03florian 07org.oe.oz354x * rb62b9572... 10/ (18 files in 3 dirs): gpe-login: Add 0.87 supporting new keylaunch. Drop a huge pile of ancient releases. Aug 08 14:19:21 03florian 07org.oe.oz354x * rf0ae8360... 10/ (1 conf/distro/preferred-gpe-versions-2.8.inc): preferred-gpe-versions-2.8.inc: Use newer gpe-login. Aug 08 14:19:25 koen: we need a few of them for ANgstrom Aug 08 14:19:28 03florian 07org.oe.oz354x * re490a656... 10/ (1 packages/gpe-conf/gpe-conf_0.2.0.bb): gpe-conf: Fix dependencies. Aug 08 14:19:33 03hrw 07org.oe.oz354x * r715f6de1... 10/ (4 files in 4 dirs): {module-init-tools,modutils}-cross, kernel.bbclass: stage depmod-2.{4,6} instead of depmod - close #524 Aug 08 14:20:20 florian_kc: pulling.... Aug 08 14:21:19 merged/pushed Aug 08 14:22:05 XorA: a few of what? Aug 08 14:22:11 florian_kc: building Aug 08 14:22:14 (caffeine is wearing off) Aug 08 14:22:18 koen: dualcore xscales Aug 08 14:22:47 hrw|work: i dropped the files i removed again and restored the missing files, pulling and merging now. Aug 08 14:22:50 NOTE: couldn't find pkgconfig module 'sqlite' in any package Aug 08 14:22:57 sqlite need update? Aug 08 14:23:01 http://www.em.avnet.com/trv/home/0,1876,RID%253D0%2526CID%253D30175%2526CCD%253DUSA%2526SID%253D32214%2526DID%253DDF2%2526LID%253D32235%2526BID%253DDF2%2526CTP%253DTRV,00.html Aug 08 14:23:09 hrw|work: use sqlite3 Aug 08 14:23:28 zecke: gpe use sqlite2 still Aug 08 14:23:34 koen: dual cores Aug 08 14:23:37 florian_kc: did you use the '-e' option? Aug 08 14:23:53 XorA: more like "freaking fast coprocessor" Aug 08 14:23:59 re Aug 08 14:24:08 hrw|work: huh? unlikely... Aug 08 14:24:10 XorA: I was playing with your potential new machine today ;) Aug 08 14:24:26 koen: no Aug 08 14:24:35 woglinde: http://nopaste.php-q.net/231133 look at that crap Aug 08 14:24:36 now i get: mtn: misuse: path _MTN/tmp/1 already exists Aug 08 14:24:50 koen: wince on one core + linux on another... interesting Aug 08 14:25:04 x29a yeah your flash broken Aug 08 14:25:08 this happens Aug 08 14:25:41 so i have to solder in a new one? Aug 08 14:25:42 dont know if flashmem can exclude broken blocks Aug 08 14:25:55 why the heck with me? Aug 08 14:25:59 I think yes because that are early blocks Aug 08 14:26:35 bbl Aug 08 14:26:48 aaar shit Aug 08 14:26:59 hrw|work: now after the second merge the files it compained about were removed (e.g. some emacs stuff) Aug 08 14:28:43 re Aug 08 14:28:56 mikearthur: cool Aug 08 14:29:13 ok... here we go Aug 08 14:29:41 woglinde: erasing block 8 works tho Aug 08 14:29:50 woglinde: are u 100% positiv on a broken flash? Aug 08 14:30:01 no Aug 08 14:30:18 03florian 07org.oe.oz354x * rd5df64d4... 10/ (15 files in 2 dirs): gpe-login: drop old releases again Aug 08 14:35:02 florian_kc: using 'mtn drop -e ' usually avoids troubles like that Aug 08 14:35:26 whats the openzaurus default dhcp client? Aug 08 14:36:12 hm udhcpc I guess Aug 08 14:36:32 koen: hum well, from the fact that these files were missing i guess that this actually happened. Aug 08 14:54:12 http://www.handhelds.org/hypermail/oe-commits/0/0022.html Aug 08 14:54:18 I miss my very own special branch Aug 08 14:55:13 Having issues compiling libx11-X11R7, seems like some include issues. any preferred provider I should set? Aug 08 14:55:16 zecke: I think short-lived branched would do wonders for OE qa Aug 08 14:55:17 zecke: :D Aug 08 15:00:09 Kristoffer: see conf/distro/include/ Aug 08 15:00:25 morning Aug 08 15:00:33 hey chouimat Aug 08 15:03:44 koen: thx Aug 08 15:04:11 zecke: where short-lived means a < 1 month Aug 08 15:07:04 koen I just did something I should have done more than 2 years ago when I moved in ... plug the soundsystem :) Aug 08 15:07:28 chouimat: yay! Aug 08 15:07:43 chouimat: I hope your neighbours share your choice of music :) Aug 08 15:10:57 koen: actually I was tired of listening to spanish pop, arab pop and asian pop ... so now I playing something more robust: Dream Theater and Fate Warning ;) Aug 08 15:13:04 chouimat: :-) Aug 08 15:22:32 oooh, Dream Theater Aug 08 15:38:05 koen: they pwn Aug 08 15:41:46 mikearthur: they sure do Aug 08 15:43:07 sadly they aren't playing at lowlands Aug 08 15:47:11 cu Aug 08 15:47:55 koen: don't worry, good band, but terrible live Aug 08 15:49:58 off home, slaters all Aug 08 15:53:04 koen: I wouldn't say terrible live....James LaBrie is...not good, but it's awesome to see the rest playing :-) Aug 08 15:53:42 hi andersee Aug 08 15:54:14 koen you are travelling to lowland? Aug 08 16:07:40 woglinde: morning Aug 08 16:10:34 going home how Aug 08 16:10:35 cya Aug 08 16:14:56 woglinde: actually I'm going to work there Aug 08 16:15:44 koen as rowdy? Aug 08 16:16:30 woglinde: site-crew Aug 08 16:16:42 I'm the one with the gaffa tape and tie-wraps Aug 08 16:16:59 hehe Aug 08 16:22:22 he renoo Aug 08 16:22:35 hi woglinde Aug 08 16:23:54 hi all Aug 08 16:24:42 ping x29a_ Aug 08 16:27:32 morning Aug 08 16:28:03 hey mickeyl Aug 08 16:28:12 I sent you my key a week or so back Aug 08 16:28:16 he mickeyl Aug 08 16:28:31 RP: sorry for not having replied to your feature-based reorginzation of task-bootstrap. I fully agree with this - it's much better than all we have right now, so lets go for it Aug 08 16:28:38 Crofton: did you? which key address? Aug 08 16:28:45 philip@balister.org Aug 08 16:28:47 things are very hectic here lately Aug 08 16:28:50 heh Aug 08 16:28:51 yeah Aug 08 16:29:43 ah right. i see it in my inbox. would it be possible for you to regenerate the key with a @openembedded.org address? I'm afraid that's our recent key policy Aug 08 16:30:37 I suppose Aug 08 16:31:01 excellent. in that case I'll add it right after I get it today :) Aug 08 16:31:42 * mickeyl balancing between thesis polishing, freelance work and open source stuff Aug 08 16:31:56 how do I get an openembedded.org address? Aug 08 16:31:56 private life? nah, who needs this :D Aug 08 16:32:00 right Aug 08 16:32:49 Crofton: you just use it as the key address. oe org mailing list forwarders are on the todo eventually. for now, key adresses are not necessarily valid e-mail adresses. Aug 08 16:33:11 ah Aug 08 16:33:11 ok Aug 08 16:33:33 mickeyl: Thanks. As that's ok with you, I'll rearrange the Zaurii to start with and change OZ in .dev to use it. Hopefully others will then follow the lead :) Aug 08 16:33:55 RP: i think so Aug 08 16:34:46 The oe list appears to be stalled for me - is anyone else getting mails? Aug 08 16:35:09 twice today. "Re: Help a Newb" or so Aug 08 16:35:37 I got those two but nothing since 10am ish (i.e. no commits) :-/ Aug 08 16:36:57 * koen will follow the RP Aug 08 16:38:11 I expect I'll do it one night this week, maybe tomorrow :) Aug 08 16:38:33 * RP must spend some time on bitbake core soon though as I really want to get that done... Aug 08 16:38:57 y'ello! Aug 08 16:39:14 is mr rp present? Aug 08 16:39:22 hi dmd Aug 08 16:39:29 Hi RP! Aug 08 16:39:40 hmm Aug 08 16:39:55 can I do TARGET_ARCH_xscale = "arm" inside a .bb? Aug 08 16:40:07 or will that disturb the time/space continuum? Aug 08 16:40:20 Is xscale an override? Aug 08 16:40:37 it's the TARGET_ARCH I'm building for Aug 08 16:40:38 bye Aug 08 16:40:55 apparently gcc enables some intrinsics when using that Aug 08 16:41:32 I'd try and stick to arm but change it for gcc's benefit when needed if possible Aug 08 16:41:51 otherwise the arm overrides are going to fail Aug 08 16:41:54 RP: do you know what I'm going to ask (this is danboid)? Aug 08 16:42:08 its that time again :) Aug 08 16:42:12 dmd: overlay support? Aug 08 16:42:16 yup :) Aug 08 16:42:48 no change in status and my free time is none existent atm :-( Aug 08 16:42:56 Its still planned though Aug 08 16:43:17 * RP -> food Aug 08 16:43:18 darn! one day, one day! Aug 08 16:43:25 we will get there Aug 08 16:43:37 patches welcome Aug 08 16:43:47 I'm going to organise a party or summat when you get it done Aug 08 16:44:06 send you some beer Aug 08 16:44:49 any news on GPE 2.8? Aug 08 16:45:03 dmd: try #gpe Aug 08 16:45:11 yeah! Aug 08 16:45:14 good idea! Aug 08 16:46:18 oh yeah, koen, forgot you in the geek comp. le geek, c'est chic! Yr probably not mightily bothered tho I'd imagine Aug 08 16:48:09 dmd: not at all Aug 08 16:49:15 just a laaff! Aug 08 16:58:50 RP: Hello. I'm starting to understand the odd results of alpha blending on the w100. Aug 08 17:00:19 RP: A little question about it: Empirically, it seems that the framebuffer pixels are RGB565. On the other hand, the "supported" working modes are ARGB4444 and ARGB1555 as they are the only 16 bpp known values (constants). Aug 08 17:01:44 RP: Alpha blending arithmetic is assuming ARGB1555 or ARGB444, and that's the reason colors get distorted. Do you know why the screen works in RGB565? Aug 08 17:02:02 Is it something physical, not configurable? Aug 08 17:08:53 hehe Aug 08 17:09:03 koen|away: Mr. Drepper qualifies OSX as insane Aug 08 17:10:12 zecke: heh Aug 08 17:10:23 zecke: he also says glibc can't be crosscompiled Aug 08 17:10:56 koen: and I say "his mother is so huge...." Aug 08 17:11:20 ... she has her own zipcode Aug 08 17:11:26 hehe Aug 08 17:12:03 "------- Additional Comments From drepper at redhat dot com 2006-08-08 17:24 ------- Aug 08 17:12:06 You need a sane environment. OS X does not qualify. Aug 08 17:12:08 " Aug 08 17:12:14 * koen searches for some stroopwafels Aug 08 17:12:20 oh Aug 08 17:12:21 zecke: is your bike still functional? Aug 08 17:12:43 koen: It is still non-functional Aug 08 17:13:37 my mon threw the quick release skewers away... Aug 08 17:13:54 and I considered it too dangerous to drive without them... Aug 08 17:15:19 need to reboot now Aug 08 17:15:51 koen: so drepper says: To compile glibc 2.4 on linux 2.6.16 for x86 you need linux 2.6.16, x86 and glibc2.4 running Aug 08 17:33:57 hi sirfred Aug 08 17:34:14 RP: Hello! Aug 08 17:34:15 sirfred: Isn't there a bit in the magic debug register that has something to do with 565? Aug 08 17:34:53 sirfred: I'm not here for long but did read your email with interest. 565 is a property of the physical screen :-/ Aug 08 17:35:23 RP: :( Aug 08 17:35:43 RP: But now that you mention it, I remember something like 565_for_non2d Aug 08 17:36:43 RP: dst565_for_not_2d. Good hint. Aug 08 17:39:15 sirfred: There isn't a dst565_for_2d is there? :) Aug 08 17:39:43 RP: Well, I think that the arithmetic operations (stretchblt, alpha blending) are the ones they call non2d Aug 08 17:40:03 RP: When the E2_ARITHMETIC register is involved. Aug 08 17:40:22 RP: I'm looking at the altcore code, and it seems that the mmDEBUG1 is involved. Aug 08 17:41:26 It smells like a workaround for not supporting RGB565 directly or something so. Aug 08 17:47:17 sirfred: It does - maybe an afterthought... Aug 08 17:48:21 RP: It worked! Aug 08 17:48:58 RP: OUTREG(mmDEBUG1, 0x70000), and red is not "migrating" to "green" bits anymore Aug 08 17:49:19 RP: Thanks a lot Aug 08 17:50:21 sirfred: excellent! :) Aug 08 18:26:00 RP: I wonder where this name came from : lpgDrvState->dst565_for_not_2d Aug 08 18:27:36 RP: It seems too specific for reverse engineered stuff, perhaps from some log string ? Aug 08 18:29:45 mr_nice_slacker: pong Aug 08 18:31:29 woglinde: i erased blocks 0-128 and that worked just fine Aug 08 18:31:38 x29a hm Aug 08 18:31:43 woglinde: u think i should just try to reflash the image? Aug 08 18:31:43 than only one block is defekt Aug 08 18:31:52 hm why not Aug 08 18:31:54 test it Aug 08 18:31:55 but it erased block 8 as well Aug 08 18:32:03 just takes a shitload of time Aug 08 18:32:11 have no cristal bowl Aug 08 18:32:28 aww, bummer Aug 08 18:33:11 mr_nice_slacker: http://nopaste.php-q.net/231133 Aug 08 18:35:33 mickeyl, new key otw Aug 08 18:41:48 wookey, you the caving wookey? Aug 08 18:48:19 x29a:hi Aug 08 18:49:11 mr_nice_slacker: hi Aug 08 18:49:12 x29a: sorry i have no guess :(. you can try to check the flash with jtag tools? Aug 08 18:49:28 mr_nice_slacker: dont know the command, think its not capable of checking Aug 08 18:49:37 mr_nice_slacker: i erased block 0-128, worked just fine Aug 08 18:50:33 woglinde: may I use these servers for OE http://www.mi.fu-berlin.de/tec/serv_compute_server.shtml? Aug 08 18:50:36 x29a: well if it will be really the flash - what is not clear at the moment the - good news are that you will geta new chip for about 20 Euros Aug 08 18:50:52 x29a: but we have to figgure out how to check it first Aug 08 18:50:55 mr_nice_slacker: im not too excited about soldering them Aug 08 18:51:38 x29a: well as i sayed the state of your chips it is unclear. jtag should be able to check it Aug 08 18:51:55 not that i know of Aug 08 18:52:37 x29a_: did you tried it a few times or only one time? Aug 08 18:52:51 just once Aug 08 18:53:00 took already 4h to get to block 8 Aug 08 18:53:08 i think it must have been almost done Aug 08 18:53:26 and tonight a need a goodnights sleep (lots to do tomorrow at work) so i cant leave my computer running Aug 08 18:53:33 any gurus on the git fetcher here? Aug 08 18:54:01 I need to fethc a git tagged version of uboot for the OMAP5912OSK ... Aug 08 18:54:39 x29a_: try it a few times - i have heard from simon about errors while flashing and he tried it a second and it worked. Aug 08 18:55:12 k, ill get up earlier tomorrow to start the flashprogram Aug 08 18:55:38 x29a_: if another try will fail on another adress it could be only a connection problem Aug 08 18:55:53 mr_nice_slacker: hmm Aug 08 18:56:37 x29a_: It is worth a try I think Aug 08 18:57:05 fo shizzle on that one Aug 08 19:01:07 x29a_:hopefully you will be able to bring it back to life Aug 08 19:01:19 yeah Aug 08 19:01:29 would suck if flashsoldering is included Aug 08 19:01:47 x29a_: you are one of them with the most articles ;) Aug 08 19:02:31 x29a_: well i want to buy a hotair solder station which is exactly designed for such things Aug 08 19:02:41 x29a_: but it is a bit expensive Aug 08 19:02:45 where do u live? Aug 08 19:02:54 x29a_:wiesbaden Aug 08 19:03:01 ah right Aug 08 19:03:06 500km orso Aug 08 19:03:25 x29a_:well that is a bit to far away Aug 08 19:05:24 x29a_: I got respond for the bluetoot / wlan chips Aug 08 19:05:48 x29a_: they want to know a bit more about my company *lol Aug 08 19:06:17 x29a_: I have to think about the right words I have to send them. Aug 08 19:06:55 x29a_: can't find the chips elswhere and all are vpe > 1000 Aug 08 19:07:54 x29a_: we will enter the market and then buy you a new simpad - if necessary. Aug 08 19:12:28 wow....well.... Aug 08 19:12:41 my system is so much faster now Aug 08 19:12:50 justinp wich? Aug 08 19:13:51 my server/router/build box Aug 08 19:14:13 it went belly up on me yesterday so I went out and bought a new motherboard, CPU, and ram Aug 08 19:15:04 from an older AMD (1Ghz) with SDRAM to a new Athlon64 (3200) with DDR RAM....such a difference ^_^ Aug 08 19:15:11 building will be much less painful now :-) Aug 08 19:15:23 *g* yes Aug 08 19:18:00 of course I wasn't prepared to spend the money on it......but oh well Aug 08 19:22:14 ... just found the screencast how to setup OE on the web-page: very nicely done, thanks to who ever did this. Aug 08 19:30:22 screencast? Aug 08 19:30:45 justinP: the flash tutorial Aug 08 19:38:56 where is the code for the git fetcher? Aug 08 19:47:29 pgfeller: where is this? Aug 08 19:49:14 03hrw 07org.oe.oz354x * rb18eee0b... 10/ (1 packages/meta/gpe-image.bb): gpe-image: do not ship sylpheed for poodle - it does not fit on screen Aug 08 19:49:15 any git fetcher gurus awake? Aug 08 19:49:57 justinP: http://www.openembedded.org/ => Documentation => Tutorial Aug 08 19:50:12 justinP: http://www.openembedded.org/screencasts/org.openembedded.htm Aug 08 19:50:34 hmmm, hadn't seen that Aug 08 19:50:42 thx Aug 08 20:06:53 Crofton: iirc it's in bitbake/ Aug 08 20:07:54 koen Aug 08 20:08:07 I need to make a bb file for uboot based on a git tag Aug 08 20:08:08 meh Aug 08 20:08:12 I suck at php Aug 08 20:08:50 koen: I rock at php Aug 08 20:08:58 * JustinP bites his tongue Aug 08 20:11:49 justinp jesus Aug 08 20:11:52 php Aug 08 20:11:54 oh my god Aug 08 20:12:26 how do you figure out a git path from the gitweb cgi script ????? Aug 08 20:12:44 pgfeller: ... what is wrong with php? Aug 08 20:14:14 woglinde: what is wrong with PHP? I realize it's not as tried and true some othe rlanguages, but it works very well for the web and scripting Aug 08 20:14:47 and my parser for Packages.gz is ~10x faster as the python one Aug 08 20:15:12 not a fair competition since I'm using sqlite and the python one mysql Aug 08 20:15:13 justinp security Aug 08 20:15:23 and rails makes more fun Aug 08 20:16:41 koen? Aug 08 20:16:58 Packages.gz and mysql? Aug 08 20:17:05 woglinde: http://www.angstrom-distribution.org/repo/repo.php Aug 08 20:17:35 mysql is the second thing I dont like Aug 08 20:17:55 woglinde: I'm writing a new ipkgfind type of app Aug 08 20:18:08 hm? Aug 08 20:18:08 with uploading and gpg signing support Aug 08 20:18:23 whats ipkgfind exactly doing? Aug 08 20:19:30 ah, I have cloned the u-boot git repo Aug 08 20:20:28 crofton hehe Aug 08 20:21:47 woglinde: think packages.debian.org, but on steroids Aug 08 20:23:33 ah okay Aug 08 20:23:40 woglinde: please restate your problem. All apps are insecure if you write them incorrectly. Aug 08 20:24:21 justinp but there is a higly rate of unsecure apps with php Aug 08 20:24:45 woglinde: that doesn't make PHP bad Aug 08 20:24:51 no? Aug 08 20:24:52 name one, i've been writing it for the last 2 weeks, havn't found a hole yet Aug 08 20:24:58 koen, do you know much about writing bb files that get source from a git tag? Aug 08 20:25:01 woglinde: PEAR, for instance, is high quality PHP code Aug 08 20:25:11 and yes, i have been executing programs with user input through it :) Aug 08 20:25:16 tries to close up as many security problems as possible Aug 08 20:25:17 I can't find an example in packages Aug 08 20:25:26 you can write crap in any language Aug 08 20:25:43 justinp yes, but php seems to force you Aug 08 20:26:00 woglinde: no, it doesn't. What are you talking about? Aug 08 20:26:10 JustinP is a PEAR developer Aug 08 20:26:18 I talk about I dont like php Aug 08 20:26:21 nothing perosnal Aug 08 20:26:42 woglinde: it has some misfeatures that people use incorrectly, yes (magic quotes is evil) but that doesn't mean there is a problem with the language. Aug 08 20:26:49 anyway, this is OT Aug 08 20:26:54 jepp Aug 08 20:26:58 'evening ! Aug 08 20:27:06 hi cyril Aug 08 20:27:43 hey woglinde Aug 08 20:28:10 *g* I remeber the times php didnt have for-loops Aug 08 20:28:30 heh Aug 08 20:28:50 for loops are just convenience syntax sugar for while loops anyway ;-) Aug 08 20:29:08 but at that time I didnt study cs, so I would know how, to convince btw. for and while loops Aug 08 20:29:20 wouldnt Aug 08 20:29:21 damnit Aug 08 20:29:41 *anyway* Aug 08 20:30:13 koen: if you'd like some help I'd be happy to help. DB=>Forms are my specialty (see DB_DataObject_FormBuilder) Aug 08 20:30:26 but I have of course done all sorts of things Aug 08 20:34:32 "Added 2955 packages in 0 days, 0 hours, 0 minutes and 23 seconds" Aug 08 20:36:02 git sucks Aug 08 20:38:45 rofl Aug 08 20:38:51 http://www.linux-mips.org/wiki/Git Aug 08 20:46:50 haha Aug 08 20:46:51 lol Aug 08 20:46:53 http://thedailywtf.com/forums/thread/85040.aspx Aug 08 20:55:03 lol, comments are even more funny that the original post Aug 08 21:03:08 JustinP: and now e17 switch to using .desktop files.... Aug 08 21:05:58 koen: it did? Aug 08 21:06:29 JustinP: see the mail to e-dev from david seikel Aug 08 21:09:21 well.... Aug 08 21:09:28 good, then :-) Aug 08 21:11:25 I hope they implement some db storage for caching them Aug 08 21:11:49 :-) Aug 08 21:11:51 Crofton: I am indeed a caving Wookey. Not sure about 'the', but probably :-) Aug 08 21:12:11 I use survex Aug 08 21:12:23 small world :-) Aug 08 21:12:25 Seen your name their Aug 08 21:12:26 yeah Aug 08 21:13:12 Where are you? Aug 08 21:13:25 Blacksburg VA Aug 08 21:13:39 (and is off-topic chat tolerated on this channel?) Aug 08 21:14:07 well Aug 08 21:14:19 Survex is not widely used inthe states SFAIK. But I suppose if you are a linux user... Aug 08 21:14:23 I think most of the guys are asleep now :) Aug 08 21:14:25 yeah Aug 08 21:14:48 the locals all bitch Aug 08 21:14:55 we don't have a survex IRC yet - just the mailing list Aug 08 21:15:06 that would be cool (irc) Aug 08 21:15:22 well it's trivial to do. Aug 08 21:15:22 You know a Mike Futrell? Aug 08 21:15:27 yeah Aug 08 21:15:37 I've met him once (at china caves thing last year in london) Aug 08 21:15:42 thought so Aug 08 21:15:52 nice guy - done some fine work Aug 08 21:16:02 yeah Aug 08 21:16:15 I've know him forever, since back in his "collegiate" days Aug 08 21:16:26 One day I might end up on a trip with him, which would be cool Aug 08 21:16:39 well if you are ever over here, look us up Aug 08 21:17:14 RIght-ho. Unfortuntely the missus doesn;t like caving whilst on holiday much these days ... Aug 08 21:17:25 heh Aug 08 21:17:36 So we have to do nice things instead Aug 08 21:17:36 Holidays are a vacation from caving Aug 08 21:18:21 ah - youre mr balister. Hi :[-) Aug 08 21:18:33 yeah Aug 08 21:18:35 Hi Aug 08 21:19:11 I I do find the IRC 'extra name' thing confusing Aug 08 21:19:16 yeah Aug 08 21:19:24 you are lucky to get yours Aug 08 21:19:42 Indeed (I don't on freenode - someone beat me to it Aug 08 21:21:13 we should see about a survex/therion irc channel Aug 08 21:22:26 I need to generate more data to play with Aug 08 21:23:28 you hang out here often? Aug 08 21:23:48 I just created #cavesurvey on oftc - that'll do for now Aug 08 21:23:57 oftc? Aug 08 21:24:10 irc.oftc.net Aug 08 21:29:54 03kristoffer 07org.oe.dev * r06626817... 10/ (7 files in 5 dirs): Aug 08 21:29:54 Some patches to add Jornada 6xx support and altera model 720 to 7xx. minor Aug 08 21:29:54 alterations of the patches to fix typos. Aug 08 21:30:00 03kristoffer 07org.oe.dev * r279de7b0... 10/ (1 conf/distro/jlime-donkey.conf): Aug 08 21:30:00 conf/distro/jlime-donkey.conf : Fix PREFERRED_PROVIDERS -> PROVIDER. Aug 08 21:30:00 * Minor fix, to set PREFFERED_PROVIDER instead of PROVIDERS. Aug 08 21:30:06 03kristoffer 07org.oe.dev * r3521a040... 10/ (1 conf/distro/jlime-donkey.conf): Aug 08 21:30:06 conf/distro/jlime-donkey.conf : Fix typo Aug 08 21:30:06 * PREFFERD_PROVIDER_virtual/libx11 = "libx11" and not "x11" Aug 08 21:30:11 03kristoffer 07org.oe.dev * rfbacb29c... 10/ (1 conf/distro/jlime-donkey.conf): Aug 08 21:30:11 conf/distro/jlime-donkey.conf : Add X11R7 include, fix typo Aug 08 21:30:11 * Add require conf/distro/include/preferred-x..versions.. Aug 08 21:30:14 * Fix virtal -> virtual in PREFERRED_PROVIDER..xlib11 Aug 08 21:53:28 what are typical boot times from OE-based installations? Aug 08 21:53:46 we have like 1min or so and i'd like to improve that Aug 08 21:53:52 does anybody has some hints? Aug 08 21:54:04 busybox' speed seems to be a rather limiting factor Aug 08 21:56:31 do you have specific examples? Aug 08 21:57:59 night Aug 08 21:58:01 tmbinc: zecke and I were discussing using launchd (recently opensource from apple) Aug 08 21:58:13 anyway Aug 08 21:58:16 'night all Aug 08 22:00:04 CosmicPenguin: "populate-volatile.sh" for example took in the range of 15s. I've replaced it with a tarfile. Aug 08 22:00:19 populate-volatile.sh isn't busybox, for the record Aug 08 22:00:26 and it is not well optimized Aug 08 22:00:37 hm, i've always thought that /bin/sh is busybox Aug 08 22:00:37 I'll agree - but don't blame the tools, blame the craftsman Aug 08 22:00:47 bash is twice as fast Aug 08 22:01:02 mainly because it has internal versions of "echo" etc. which are currently disabled in busybox Aug 08 22:01:04 becuase of some bugs Aug 08 22:01:17 (sorry, that was valid for 1.01 or so. I don't know if this has changed now.) Aug 08 22:01:52 so i blame /etc/init.d for being written as shellscripts, or alternatively i blame busybox for being slow when executing shell scripts Aug 08 22:02:02 he tmbinc Aug 08 22:02:07 your old problem Aug 08 22:02:16 he port smf form sun Aug 08 22:02:22 *duck and hide* Aug 08 22:02:29 aehm solaris Aug 08 22:02:44 i still didn't get it ;) Aug 08 22:03:04 so - the question is more like: Do we have an alternative to /etc/init.d, or is optimizing busybox the clue? Aug 08 22:03:24 I suppose a fork() to busybox takes some time, too much time for a simple "echo". So it sums up. Aug 08 22:04:02 tmbinc isnt another init system a solution? Aug 08 22:04:34 like what? Aug 08 22:04:41 minit Aug 08 22:04:47 some people suggested using a init system which executes in parallel Aug 08 22:04:53 yes Aug 08 22:05:01 http://www.fbunet.de/minit.shtml Aug 08 22:05:04 but i think cpu is at 100%, so it wouldn't help so much Aug 08 22:05:12 oh, fefe tools Aug 08 22:05:15 hehe Aug 08 22:06:11 i would try it Aug 08 22:06:25 because dietlibc linking statical Aug 08 22:06:30 okay Aug 08 22:06:34 have go to bed Aug 09 00:06:10 does anyone get http://www.openembedded.org/filebrowser/org.openembedded.dev/packages/ to sort by last-modified? it doesnt work for me **** ENDING LOGGING AT Wed Aug 09 02:59:56 2006