**** BEGIN LOGGING AT Wed Jun 20 02:59:58 2007 Jun 20 06:51:01 good morning Jun 20 06:54:21 morning Jun 20 06:54:41 hey, what's the ratio of our alsa-supported devices vs. only-oss-supported ones? Jun 20 07:01:38 good morning Jun 20 07:03:40 mickey|fic: that boils down to 2.6 supported devices versus 2.4 supported devices Jun 20 07:05:25 koen: figured that. which is... (been out of the loop, you know) Jun 20 07:06:01 * * OE Bug 2539 has been RESOLVED (FIXED) by Jun 20 07:06:03 * *  Wrong place for patch Jun 20 07:06:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2539 Jun 20 07:07:08 mickey|fic: btw, you shouild really use insane.bbclass: http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-June/006217.html Jun 20 07:07:23 that spots x86 binaries creeping into packages Jun 20 07:09:28 mmh Jun 20 07:09:33 working on a laptop here Jun 20 07:09:41 trying to get all the performance Jun 20 07:09:48 so disabled all nonneeded classes Jun 20 07:09:56 when im back Jun 20 07:10:00 ill do Jun 20 07:11:51 the reason why i'm asking is, i have a patch for flite that enables alsa Jun 20 07:11:58 however it's a either-or Jun 20 07:12:02 so we can't have both Jun 20 07:12:20 and i'm slightly reluctant to adding flite-alsa and flite-oss Jun 20 07:12:22 heh Jun 20 07:13:29 and oss emulation just doesn't cut it Jun 20 07:14:15 is that because of performance issues with the oss emulation? Jun 20 07:14:39 yeah Jun 20 07:14:53 and the oss emulation prevents dmix from working correctly Jun 20 07:16:17 i wish we could use dmix on the greenphone Jun 20 07:16:43 mickey|fic: http://hughsient.livejournal.com/28753.html Jun 20 07:17:37 interesting Jun 20 07:45:12 , hello Jun 20 07:55:05 hey step2back Jun 20 07:55:53 do you want to continue our discussion about palmld.conf? Jun 20 08:04:31 step2back: I'm working on other stuff right now Jun 20 08:05:01 mickey|fic: http://www.teemuharju.net/2006/04/19/my-nokia-770-can-speak/ (sadly the patch can't be found anymore and the author has been swallowed by nokia) Jun 20 08:05:44 koen, ok. I'll wait..... Jun 20 08:05:54 This thing is cool, but it shouldn't run Windows CE: http://www.jadeintegration.com/jackpc.php Jun 20 08:07:00 step2back: basically remove all non essential wifi modules (e.g. kernel-module-arc4) and only leave drivers Jun 20 08:07:22 task-base-wifi already lists all needed modules for wep and wpa Jun 20 08:07:35 ok... I did it. Jun 20 08:09:48 What should I write in of MACHINE_EXTRA_RRECOMMENDS? If I have a list of "module_autoload_" Jun 20 08:11:40 MACHINE_EXTRA_RRECOMMENDS should only have *drivers* for hardware present in the device Jun 20 08:11:55 %{!dynamic-linker:-dynamic-linker /lib/ld-uClibc.so.0} \ <- that shouildn't be in gcc sources Jun 20 08:17:07 It's right for PalmLD MACHINE_EXTRA_RRECOMMENDS should only have *drivers* for hardware present in the device Jun 20 08:19:33 k, the angstrom-console-image actually runs on h3600 Jun 20 08:19:54 bad news is, the x11-image is some 6MB too big... Jun 20 08:25:56 , how to filling /etc/modules? If not use BOOTSTRAP_EXTRA_RDEPENDS += "update-modules" Jun 20 08:26:52 your distro takes care of that for you, no need to put stuff in the machine.conf for that Jun 20 08:27:16 * koen gets a little sick when looking at http://rafb.net/p/dU9Z5K45.html Jun 20 08:27:19 gm, all! Jun 20 08:27:23 polyonymous, hi! Jun 20 08:27:52 polyonymous, sorry, went offline acidentally yesterday and fell asleep :( Jun 20 08:34:00 anyone used directfb w/ tslib recently? Jun 20 08:36:32 slapin, the fix I've done actually fixes half-a-problem. That is, it lets me put cyrillic into rxvt when it's running on my host's display, but not with your matchbox thingie. Jun 20 08:37:24 slapin, and gpe-editor doesn't seem to save cyrillic either. Jun 20 08:38:00 polyonymous, I found that, yes. Jun 20 08:38:33 polyonymous, seems that locale is indeed broken but where - glibc or libx11? Jun 20 08:38:55 also xev gives quite different results with mb* and the proper xkb. Jun 20 08:39:07 Not sure, maybe everywhere :) Jun 20 08:39:19 Like glibc, libx11 and glib :) Jun 20 08:39:21 well, events are properly sent Jun 20 08:40:18 xev's XLookupString (or whatever that was) doesn't really give sensible output with mb* Jun 20 08:40:42 running on device with DISPLAY set to host it does, though. Jun 20 08:41:07 polyonymous, I don't like locale.alias on xlib, I don't like locale.alias in gpe-dm dir. All this looks suspicious. Jun 20 08:41:26 but! gpe-editor, when running on host's display doesn't even let me enter cyrillic. Jun 20 08:41:40 polyonymous, worst thing among these is that it used to work when I last checked a few months ago. Jun 20 08:41:54 That doesn't help much at the moment :) Jun 20 08:42:26 And it's also a matter of how extensive your checks were. Sometimes it _looks_ as if it works, but it doesn't really. Jun 20 08:42:53 polyonymous, I think you could build with MACHINE=qemuxx (arm or x86) so you will be able to test properly. Jun 20 08:43:23 No, I can't, qemu doesn't work properly on amd64 ;-) Jun 20 08:43:27 polyonymous, from start, there's need for locale testing sw. Jun 20 08:44:00 polyonymous, even with 32bit libs? Jun 20 08:44:18 polyonymous, what distro you have in host? Jun 20 08:44:24 morning Jun 20 08:44:37 hrw, morning! Jun 20 08:44:50 slapin, I was thinking about building a 32bit one, but haven't done it yet. I have gentoo Jun 20 08:44:55 and morning, hrw Jun 20 08:45:14 polyonymous, it should be able to build 32bit version Jun 20 08:45:39 slapin, theoretically, yes. I haven't tried, though. Jun 20 08:45:44 polyonymous, though I never used it but people there use it on amd64 boxes and all is fine. Jun 20 08:46:05 Use what? OE with qemu? Jun 20 08:46:24 polyonymous, they run qemu with something other inside. Jun 20 08:46:25 polyonymous: qemu + OE built image works on 64bit Jun 20 08:46:41 polyonymous: qemu + locale generation failed on my 64bit Jun 20 08:46:55 hrw, how come it can't even generate locale, but works? Jun 20 08:47:15 OE-built image with disabled locale-generation is supposed to generate it on first boot, hmm? Jun 20 08:47:20 yes Jun 20 08:47:30 and it works? Jun 20 08:47:44 hey hrw Jun 20 08:47:47 oom on localegen ofcourse on 64M machines Jun 20 08:47:48 polyonymous, only on 128Mb+ devices Jun 20 08:48:00 slapin: 128MB+ you mean? Jun 20 08:48:07 it needs close to 60MB of ram Jun 20 08:48:08 polyonymous, 32Mb and 64Mb fail for me. Jun 20 08:48:08 on 128Mb it fails Jun 20 08:48:24 hrw, yes, MB Jun 20 08:48:29 slapin, I have no problems generating locales on the device (with swap file), I'm talking about qemu Jun 20 08:48:33 I've run it succesfully on 64MB devices with no services running at boot Jun 20 08:48:56 polyonymous, you can make it swap, too :) Jun 20 08:49:07 what about generating locales before other services? Jun 20 08:49:12 polyonymous, but why care since you have enough RAM on host Jun 20 08:49:39 so my question was - locale generates normally on qemu on 64bit, but just not when building OE? Jun 20 08:49:45 and in some way which could consume less memory? Jun 20 08:59:01 ~seen zecke Jun 20 08:59:04 zecke was last seen on IRC in channel #openmoko, 11h 10m 59s ago, saying: 'ubuntu'. Jun 20 09:06:01 * * OE Bug 2540 has been created by  Jun 20 09:06:03 * * Asterist 1.4.x recipe Jun 20 09:06:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2540 Jun 20 09:06:22 asterist? Jun 20 09:20:38 psokolovsky, ping Jun 20 09:21:38 hi florian Jun 20 09:22:09 hi all Jun 20 09:23:59 ahah, connection at last, damn dain Jun 20 09:24:05 s/dain/rain/ Jun 20 09:40:46 hi everybody Jun 20 09:41:10 hi Thomas Jun 20 09:44:36 hrw: I tried to use the new mmc-spi patches from spi-devel and compiled rc4-mm2. I used our old flash driver. But the device is realy unstable. It sometimes boots the kernel and freezes in the init process and sometimes freezes while booting the kernel. Jun 20 09:45:07 hrw: I didn't have time, to investigate what caused this. Maybe some mm2 patch. Jun 20 09:47:03 hi thesing Jun 20 09:49:03 thesing: I would avoid mm tree for development Jun 20 09:50:01 hrw: the mmc-spi patches need the mmc-patches in there. But maybe I should just extract the needed patches. Jun 20 09:50:12 exactly Jun 20 09:52:45 hi Jun 20 09:53:16 ih Jun 20 10:13:12 shit Jun 20 10:13:24 blackfin needs the blackfin for of uclibc Jun 20 10:13:41 http://blackfin.uclinux.org/gf/project/toolchain/scmsvn/?action=browse&path=%2Ftrunk%2FuClibc%2F Jun 20 10:14:13 * koen is getting quite sick of buildroot makefiles Jun 20 10:18:49 sorry logged in with wrong nick Jun 20 10:27:13 * koen bangs head against desk Jun 20 10:27:24 blackfin also have their own gcc fork..... Jun 20 10:27:36 http://blackfin.uclinux.org/gf/project/toolchain/scmsvn/?action=browse&path=%2Ftrunk%2Fgcc-4.1%2F Jun 20 10:27:57 no wonder I get cc1: error: unrecognized command line option "-mleaf-id-shared-library" Jun 20 10:29:56 koen, why the blackfin obsessoin? Jun 20 10:30:04 auć Jun 20 10:30:55 Crofton: I want bfin support in OE :) Jun 20 10:31:15 this would be cool Jun 20 10:31:42 you planning on doing some signal processing on one? Jun 20 10:31:51 Crofton: yes Jun 20 10:31:58 even cooler Jun 20 10:32:14 I blame buildroot inability to handle patches for all the forks in the embedded world Jun 20 10:32:14 I'll ask you which board when I am at school and I'll see what hasan has Jun 20 10:32:40 Crofton: I trying to get a 537-stamp board donated Jun 20 10:33:29 blackfin.org has "Phorums" Jun 20 10:33:34 at least accorsing to goolge Jun 20 10:33:39 http://www.analog.com/en/prod/0%2C2877%2CBF537%25252DSTAMP%2C00.html Jun 20 10:34:09 I found out that there are more atrocious things as geek forums: guinea pig owner forums Jun 20 10:34:18 someone should kill them for such links Jun 20 10:34:28 heh Jun 20 10:34:40 koen: nothing wrong with guinea pigs Jun 20 10:34:56 We have been threatening to start a community davinci email list becasue TI's has been down for two weeks Jun 20 10:34:57 XorA: indeed, I'm getting 2 after returning from guadec Jun 20 10:35:08 Are they good eating? Jun 20 10:35:18 koen: got 6 next to me Jun 20 10:35:25 XorA: :) Jun 20 10:35:42 Crofton: http://www.alobbs.com/photolog/p1020028 Jun 20 10:35:51 they are being suspiciously quiet though, I shall have to find out what mischeif they are upto Jun 20 10:36:06 they probably sense my cats :) Jun 20 10:36:12 koen: BTW IME the people on the guinea pig forums are even stupider that AOL me tooers Jun 20 10:37:22 Crofton: I put them in the garden once, a cat came to investigate, decided them mice are too fscking big and legged it :-D Jun 20 10:37:56 Crofton: http://www.rowetel.com/blog/ Jun 20 10:38:22 http://www.flickr.com/photos/32615155@N00/345385199/ Jun 20 10:38:25 * koen will try to get OE into his master courses Jun 20 10:39:23 http://www.flickr.com/photos/32615155@N00/405229713/ <- note the absense of dark rum Jun 20 10:39:46 I have the echo cancellation guy tagged in del.icio.us Jun 20 10:54:59 Crofton: can you recommend any good books on signal processing? Jun 20 10:55:21 hmm Jun 20 10:55:27 for dummies? Jun 20 10:56:12 all my textbooks suck Jun 20 10:56:16 but Jun 20 10:56:19 I'm looking for something more advanced Jun 20 10:56:39 http://www.amazon.com/Understanding-Digital-Signal-Processing-2nd/dp/0131089897/ref=pd_bbs_sr_1/103-0794516-4346263?ie=UTF8&s=books&qid=1182336986&sr=8-1 Jun 20 10:56:47 define more advanced? Jun 20 10:57:00 this book was recommended to me by the gnu radio guys Jun 20 10:57:04 I know what LTI and FFT mean :) Jun 20 10:57:12 It has a chpater on practical things Jun 20 10:57:59 don't we all :) Jun 20 10:58:13 but you have no clue how to actually use this knowledge Jun 20 10:58:35 right Jun 20 10:59:02 take a look at that book Jun 20 10:59:15 * koen puts that book on his wishlist Jun 20 10:59:24 itis going to rehash some stuf, but he seems to try to make it related to things you might do Jun 20 11:01:42 * hrw know FFT but not LTI Jun 20 11:04:11 Linear Time Invariant system Jun 20 11:12:46 RP: take care :) Jun 20 11:21:14 ~seen justinp Jun 20 11:21:17 justinp was last seen on IRC in channel #openzaurus, 22d 14h 37m 2s ago, saying: 'having problems with it?'. Jun 20 11:21:21 ~summon justinP Jun 20 11:21:24 apt takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so he gets delivered a phonelist containing justinP's coordinates Jun 20 11:21:47 ~summon ibot Jun 20 11:21:48 apt takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so he gets delivered a phonelist containing ibot's coordinates Jun 20 11:27:59 * koen_ is debating wether to patch uclibc to work with vanilla gcc or to add bfin-gcc Jun 20 11:38:52 goodbye ATI X200-onboard, welcome NVidia 6600 GT Jun 20 11:42:19 * polyonymous welcomes hrw's board. Jun 20 11:44:58 grrrr Jun 20 11:45:03 bitbake -b bfin-gcc-cross_svn.bb -c patch Jun 20 11:54:36 Question? How dose bitmake match RDEPENDS = "kernel", what inside another bb is linked to "kernel" Jun 20 11:55:27 pelle2004: open your mail client Jun 20 11:55:34 pelle2004: check the answers you got to the mail Jun 20 11:55:34 OK! Jun 20 12:06:55 Suppose I damaged the partition table of a borzoi. Is there any other way to repartition the disc without having to flash cacko, partition the hd and flash back to angstrom? Jun 20 12:07:24 which partition table? Jun 20 12:07:33 polyonymous, could you please comment-out on my posting? Jun 20 12:07:36 for the hd you just need fdisk Jun 20 12:07:36 of internal HD. Jun 20 12:08:10 slapin, I did. Michael Krelin is my name, just in case :) Jun 20 12:08:11 * koen hacks a bit on feedbrowser Jun 20 12:08:18 koen: Well the device doesn't boot with damaged partition table. Jun 20 12:08:30 http://www.angstrom-distribution.org/repo/?name=&action=search&arch=avr32 <- all packages that have 'avr32' as arch Jun 20 12:09:04 pelle2004: found an answer? Jun 20 12:09:21 polyonymous, ok, nice to meet you Jun 20 12:09:31 :) Jun 20 12:09:54 slapin, but I'm not interested in where that discussion is headed, I'm more interested in technical aspect of making it work. Jun 20 12:15:47 "If you get errors from this file, comment out the entire file by putting #if 0 at the beginning and #endif at the end. It works. " Jun 20 12:16:24 heh Jun 20 12:19:39 re Jun 20 12:20:47 hrw: wb Jun 20 12:21:04 OpenGL renderer string: GeForce 6600 GT/PCI/SSE2 Jun 20 12:21:46 zecke, no not jet. Funny Hotmail has still no replaies even that they are in the archives.... Jun 20 12:21:51 zecke, But I have added in my build/conf/machine/image.bb 'PREFERRED_PROVIDER_virtual/kernel = "pc1119-linux-2.6.21"' Jun 20 12:22:02 zecke, Still get a long list of 'multiple providers are available for runtime kernel' in the verbose log Jun 20 12:22:22 pelle2004: pc1119-linux is what you need Jun 20 12:22:31 pelle2004: but rename it to linux-pc1119 please Jun 20 12:22:42 zecke, is it the name of the file or? Jun 20 12:22:52 * koen suggest adding it to linux_2.6.21.bb instead Jun 20 12:23:12 and a mtn pull && mtn update Jun 20 12:25:25 pelle2004: image.bb? Jun 20 12:26:02 pelle2004: This PREFERRED_PROVIDER is a global setting, it must be read/set when conf/bitbake.conf is parsed. Check the bottom of conf/bitbake.conf to see which files are parsed Jun 20 12:26:48 pelle2004: these include conf/${MACHINE}.conf conf/${DISTRO}.conf conf/local.conf conf/auto.conf... and bitbake is using BBPATH to find the config files Jun 20 12:28:57 zecke, OK! that probaly what I need to understand the way it works. Just looking in bitbake.comf now. Thanks Jun 20 12:29:09 you are welcome :) Jun 20 12:32:22 03Laibsch 07org.oe.dev * rd67d31a3... 10/ (1 packages/mozilla/minimo_cvs.bb): Jun 20 12:32:22 minimo: declare missing dependency on zip-native. Jun 20 12:32:22 Looks like there are other dependencies missing since compilation from scratch Jun 20 12:32:22 still fails for me. zip-native is certainly needed, though and thus I commit Jun 20 12:32:22 early and often. Jun 20 12:32:28 03Laibsch 07org.oe.dev * rce858583... 10/ (1 packages/jikes/jikes-native_1.21.bb): Jun 20 12:32:28 jikes: make jikes-native PROVIDE virtual/javac-native Jun 20 12:32:29 c.f. http://www.openembedded.org/wiki/AdvancedConfiguration Jun 20 12:32:33 03Laibsch 07org.oe.dev * r6f04c15a... 10/ (1 packages/classpath/classpath-gtk_cvs.bb): Jun 20 12:32:35 classpath: add virtual/java-native to DEPENDS Jun 20 12:32:37 this package now becomes unbuildable until bug 2323 is fixed which also lists a workaround Jun 20 12:32:39 Reasoning: the compilation of java requires the java command being available Jun 20 12:32:41 java is not among the RequiredSoftware as listed in the wiki Jun 20 12:32:43 There is no package providing virtual/java-native yet which is the aim of 2323 Jun 20 12:32:47 03Laibsch 07org.oe.dev * ra88190de... 10/ (1 packages/classpath/classpath-gtk_cvs.bb): classpath: unify the gtk_cvs version and leave a note on how to compile it. Jun 20 12:32:52 03Laibsch 07org.oe.dev * raeb3d1a3... 10/ (3 files in 2 dirs): Jun 20 12:32:54 classpath: fix compilation failure of classpath-gtk_cvs. Closes 2519. Jun 20 12:32:56 * add libxtst to DEPENDS Jun 20 12:32:58 * fix packaging Jun 20 12:33:00 * move stuff to *.inc (unify) Jun 20 12:33:02 03Laibsch 07org.oe.dev * r86eacc43... 10/ (5 files in 2 dirs): classpath: some more fixes to packaging, etc and clean-up Jun 20 12:33:05 03coredump2 07org.oe.dev * r8e0b4688... 10/ (1 packages/xchat/xchat_2.6.1.bb): xchat: Fix QA error over .debug directry in /usr/lib/xchat/ Jun 20 12:33:10 03koen 07org.oe.dev * ra70445de... 10/ (7 files in 2 dirs): zaurus 2.4 kernels: add COMPATIBLE_HOST = "arm.*-linux" Jun 20 12:34:12 a flood, I'm sinking Jun 20 12:34:48 koen: zaurus 2.4 kernels rather need comp_machine Jun 20 12:35:08 hrw: no, that doesn't work Jun 20 12:35:26 hrw: since they set MACHINE="" in the recipe Jun 20 12:35:38 wow, where did that come from? I did not push for at least 12 hours. Jun 20 12:38:05 the mirror sync script was offline Jun 20 12:38:19 oh, that explains it Jun 20 12:39:43 although technically ewi546 isn't a mirror :) Jun 20 12:43:11 heh Jun 20 13:06:18 someone use compiz? Jun 20 13:07:20 no, it locks up X on my efika Jun 20 13:29:29 03koen 07org.oe.dev * ref18f00d... 10/ (3 files in 3 dirs): Jun 20 13:29:29 angstrom: fix OABI package archs Jun 20 13:29:29 * downside: now there are 3 places for people to edit for people with oabi machines Jun 20 13:34:46 ~botmail for hopsnbarley: http://article.gmane.org/gmane.linux.ports.arm.kernel/32820 Jun 20 13:41:40 koen: What are those three places? Jun 20 13:42:10 angstrom-2007.1.conf, angstrom-glibc.inc, angstrom-uclibc.inc Jun 20 13:42:26 I am trying to build angstrom with uclibc for arm, in the log when uclibc is configured I see > 1. alpha (TARGET_alpha) (i.e. choice 1 is made), that can't be right? Jun 20 13:42:58 i.e. everywhere where teh not in machine ['collie', 'simpad' ,etc line is Jun 20 13:43:06 Jin^eLD: did you add a config for your board? Jun 20 13:43:28 koen: I added a machine config file for the board if thats what you mean Jun 20 13:43:48 and I have TARGET_ARCH = "arm" there Jun 20 13:43:59 you also need one in packages/uclibc/uclibc-0.9.29/ Jun 20 13:44:13 ooh ok, I did miss that one, thanks! Jun 20 13:44:20 koen: Oh, you mean to add new machines. I thought that was for those who want to compile for supported oabi-machines. Jun 20 13:45:27 Laibsch: for the people that are stuck with a crappy vendor kernel that can't do EABI (or strongarm CPUs, but those are rare) Jun 20 13:45:34 koen: wait.. but there is a uclibc-0.9.29/arm thing, why would that not be enough? Jun 20 13:46:35 koen: Will that affect DISTRO=sharprom-compatible? Jun 20 13:46:36 there is ? Jun 20 13:46:42 Laibsch: no Jun 20 13:46:55 I just committed it like 10 seconds ago Jun 20 13:47:23 koen: yes.. there is, subdirs for arm, armeb, mipsel, sh* and a few others Jun 20 13:47:43 that's .28 Jun 20 13:48:13 oh crap, I looked in the wrong version directory Jun 20 13:48:28 you win :) Jun 20 13:49:34 if your board is armv4t, use the arm/ one I just added, if it's armv5te, use the c7x0/ one, if it's armv6 use the c7x0 one as well Jun 20 13:50:01 morning Jun 20 13:50:05 koen: thanks Jun 20 13:50:25 in the armv6 case you could consider turning on hardfloat as well Jun 20 13:50:57 I have to check what I actually have Jun 20 14:02:07 03koen 07org.oe.dev * r2e808e7d... 10/ (4 files in 4 dirs): uclibc 0.9.29: add generic config for arm (where 'generic', read 'armv4t') Jun 20 14:03:47 mickey|fic: did you hear that hildon is going the be put in gnome svn? Jun 20 14:04:49 no, that's news to me. interesting development. Jun 20 14:05:25 koen: i thought they start with using the gnome bugzilla? Jun 20 14:07:29 zecke: yes and put hildon in gnome svn as well Jun 20 14:08:33 03Laibsch 07org.oe.dev * rf8a5fa4d... 10/ (5 files in 2 dirs): freetype: update to 2.3.4. Closes 2218. Jun 20 14:15:51 polyonymous, could you please provide something technical for discussion please? Jun 20 14:18:32 ~botmail for sirfred: ipkg install expedite-themes to fix expedite (4MB) Jun 20 14:19:11 good morning folks Jun 20 14:19:35 hey chris144 Jun 20 14:19:45 hi koen Jun 20 14:20:00 I saw the e.v. plans on the mailing list Jun 20 14:20:32 is there any chance to participate even if one cannot make it to Birmigham? Jun 20 14:20:36 in July? Jun 20 14:26:02 is uboot-utils broken? complains about not being able to stat fw_env.config Jun 20 14:26:44 mhm Jun 20 14:29:39 ok, do_unpack fails, because it searches for this file in uboot/files but it's actually located in uboot/u-boot-1.2.0/ Jun 20 14:29:59 any ideas? Jun 20 14:30:17 SRC_URI just lists it as file://fw_env.config Jun 20 14:31:04 Good evening Jun 20 14:31:19 hey sirfred Jun 20 14:31:27 koen: Just read your message. Jun 20 14:31:30 koen: Thanks Jun 20 14:31:33 http://www.angstrom-distribution.org/~koen/expedite3.png Jun 20 14:32:00 koen: Cool, with the x11 engine? Jun 20 14:32:54 yes Jun 20 14:33:35 I've read something about the x11-16 engine, that seems to be focused in 16bpp hardware. It seems there's some effort related with maemo to improve it. Jun 20 14:35:19 koen: Thanks for the hint about the size. Installing on SD Jun 20 14:35:35 slapin: just ignoring PACKAGE_ARCH doesn't make it go away Jun 20 14:36:14 slapin: e.g. if distributions allow binary uploads you will end up with packages having the same name, version but different RRECOMMENDS Jun 20 14:39:09 zecke, bui it is ok for own custom images, am I right? Jun 20 14:39:45 zecke, and for source distros? Jun 20 14:40:05 you are trying to solve an issue at the wrong position Jun 20 14:40:31 zecke, I try to research all variants a come up to. Jun 20 14:40:57 zecke, probably I miss something, that's why I [RFC] Jun 20 14:42:08 zecke, if I understand correctly, anyway, I described it completely enough, sincenobody suggested anything in addition. Jun 20 14:42:34 koen: It works Jun 20 14:42:57 I would 'fix' GPE/Opie and ipkg, rpm, dpkg (or whatever) Jun 20 14:43:08 sirfred: 10.37 , EVAS SPEED Jun 20 14:43:12 on my hx4700 Jun 20 14:43:34 koen: What test is that? Jun 20 14:43:42 All tests? Jun 20 14:43:44 "all tests" the final result Jun 20 14:45:05 The xserver doesn't receive any accelerated request. Jun 20 14:45:56 koen: Sure the zaurus will lack behind your hx4700 Jun 20 14:46:16 less bandwidth Jun 20 14:46:19 hello Jun 20 14:46:41 anyone: does default angstrom kernel provides console connection? Jun 20 14:49:46 hi Jun 20 14:50:18 hi bipolar Jun 20 14:53:23 zecke, what are you think about variant 3? Jun 20 14:54:18 Laibsch: please don't remove old versions of recipes without checking if they are being used Jun 20 14:54:51 Laibsch: ./angstrom-2007.1.conf:139:PREFERRED_VERSION_freetype = "2.3.3" Jun 20 14:55:05 koen: I think it is easy enough to restore them Jun 20 14:55:17 that's not the point Jun 20 14:55:18 Or upgrade preferred version Jun 20 14:55:27 I understand yourpoint Jun 20 14:55:39 The thing is that it is impractical Jun 20 14:55:42 people are now staring at a build of freetype 2.1.10 Jun 20 14:55:58 slapin: well, you can recreate this information easily Jun 20 14:56:00 koen: 8.40 , EVAS SPEED Jun 20 14:56:26 koen: the right solution is to upgrade angstrom-2007.1.conf I guess? Jun 20 14:56:30 I can commit that Jun 20 14:56:34 slapin: and the tools (Language Selector) and package manager need to know this. So there is no use of storing them in OE Jun 20 14:56:40 Laibsch: yes Jun 20 14:56:43 koen: I had built 2.3.3 but with a twist Jun 20 14:57:13 koen: I will fix up my mess Jun 20 14:57:29 the problem is that freetype 2.1.10 isn't really compatible with 2.3.x Jun 20 14:57:35 so you get things like this: Jun 20 14:57:39 /data/build/koen/OE/build/tmp/angstrom/staging/arm-angstrom-linux-gnueabi/lib/libevas.so: undefined reference to `FT_Get_TrueType_Engine_Type' Jun 20 14:57:48 Sorry for disturbing. I can rgrep through packages in the future. Jun 20 14:58:18 But I don't think it is feasible to always go through an RFC when upgrading stuff without preserving the old version Jun 20 14:59:15 koen: The test on x11-16 engine segfaulted after 10 tests. :-( Jun 20 15:00:00 03koen 07org.oe.dev * r50739801... 10/ (1 packages/efl/evas-x11_0.9.9.038.bb): evas-x11: enable x11-16 engine Jun 20 15:00:20 Laibsch: grepping through conf/distro should be enough Jun 20 15:00:36 koen: Anyway, some of the tests where getting really better numbers, Image Quality Scale got 14,95 against 6,24 using x11 engine. Jun 20 15:00:54 koen: Did you remember to add the x11-16 engine header ? Jun 20 15:01:04 ehm, I forgot Jun 20 15:01:16 grrr Jun 20 15:01:17 koen: THat way, expedite is not going to activate that engine Jun 20 15:01:19 koen: :) Jun 20 15:01:27 why doesn't it use autotools_stage_all? Jun 20 15:01:38 koen: No idea. Jun 20 15:02:09 koen: Perhaps it gots more headers than needed installed. Jun 20 15:02:34 what's the correct header? Jun 20 15:02:40 Hmm, a second Jun 20 15:03:36 * steliosk wonders how much hotter is going to be tomorrow Jun 20 15:04:00 koen: Evas_Engine_Software_16_X11.h Jun 20 15:04:23 ../modules/engines/software_x11/Evas_Engine_Software_16_X11.h ? Jun 20 15:04:47 No Jun 20 15:04:48 just heard on the news that because of the heat, there are going to be rolling black-outs as there is big demand for electricity Jun 20 15:05:05 ../modules/engines/software_16_x11/Evas_Engine_Software_16_X11.h Jun 20 15:05:43 koen: Anyway, I'm a little disappointed with all this. I expected some more cooperation with the xserver. Jun 20 15:05:44 zecke, I think this info is not easily recreatable, because you could have locale packages which do not belong to your image. Jun 20 15:06:10 Perhaps implementing some xrender primitives in the xserver could help better Jun 20 15:06:33 sirfred: k-s[WORK] in #openmoko is working on x11-16 Jun 20 15:06:42 slapin: well, ipkg knows what it is going to install, and we have a common names for locales (same applies for rpm, dpkg, etc.) Jun 20 15:07:05 koen: If we have an accelerated card, we should use it. Jun 20 15:07:22 slapin: the question is how to transport this information to the user, creating an image is one thing but people WILL install software Jun 20 15:07:53 koen: x11-16 is targeted to dumb framebuffers, imo Jun 20 15:08:39 help please Jun 20 15:08:45 Some times I think it should be better that the zaurus had such a dumb framebuffer, it would be less frustrating. Jun 20 15:09:14 should I see anything after I use rootfs on my device? how can I tell its working? Jun 20 15:09:55 zecke, thanks for enlightenment! Jun 20 15:10:15 is the oe viewmtn down? it returns acces forbidden Jun 20 15:10:24 slapin: sorry? Jun 20 15:13:23 zecke, I got another idea thanks to you. Jun 20 15:23:05 03koen 07org.oe.dev * r785af4fb... 10/ (1 packages/efl/evas-x11_0.9.9.038.bb): evas-x11: also stage x11-16 header Jun 20 15:25:28 koen: Have you already tried expedite on x11-16 ? Jun 20 15:25:35 not yet Jun 20 15:25:42 I'm adding it to the feeds right now Jun 20 15:25:55 koen: ok Jun 20 15:26:29 koen you patched the uclibc stuff i sent yesterday Jun 20 15:26:48 but you miss the bb file? or wasn't correct? Jun 20 15:26:58 I did? Jun 20 15:27:02 hummm Jun 20 15:27:10 you appear at the log Jun 20 15:27:11 I just commit my own work for uclibc Jun 20 15:27:19 aaaaam Jun 20 15:27:43 the ixp4xxbe/uClibc.machine too? Jun 20 15:28:02 it is exactly like mine now so i suposed you saw my bug report Jun 20 15:30:00 i patched too the uclibc_0.9.29.bb file because without a FILESPATH correctly set it doesn't use the correct config files Jun 20 15:31:49 koen ? Jun 20 15:32:17 uclibc 0.9.29 works for me, so I see no need to patch FILESPATH Jun 20 15:32:44 ok, i'll check better Jun 20 15:39:57 koen, in my case it gets before the uClibc.machine from files/uClibc.machine than from uClibc-0.9.29/ixp4xxbe/uClibc.machine, that is why i changed, any idea? Jun 20 15:40:22 nope Jun 20 15:40:54 what distro are you compiling? Jun 20 15:41:30 DISTRO="angstrom-2007.1" with ANGSTROM_MODE = "uclibc" Jun 20 15:41:54 ok, i'll check Jun 20 15:45:02 [ 4] 0.0- 5.0 sec 0.00 MBytes 0.00 MBytes/sec Jun 20 15:45:08 thats the speed ;D Jun 20 15:47:33 hrw: I had the same on the at91 board Jun 20 15:47:58 common speed on devboards Jun 20 15:48:11 but this is on 802.11b/g card Jun 20 15:54:48 koen : Can you have a look at this ? http://www.pocketinfo.nl/nieuws/software/5447/linux-op-een-htc-universal-i-mate-jasjar-qtek-9000-mda-pro.html Jun 20 15:55:39 it's a translation of http://www.linuxdevices.com/news/NS3647920562.html Jun 20 15:56:06 we do have HTC targets in OE right ? Jun 20 15:56:13 yes Jun 20 15:56:34 has anyone test them ? :) Jun 20 15:57:56 pH5 and florian have them Jun 20 15:58:07 steliosk: there's active htc devices porting community which keeps up with new devices too Jun 20 15:58:40 steliosk: maintenance of older ports is on worse level, but it exists, and improves together with all hh.org ports Jun 20 15:59:26 psokolovsky : I have a qtek 9000 i am using Jun 20 15:59:53 steliosk: do you know original machine name? Jun 20 15:59:59 steliosk: most universal users hang out at #htc-linux Jun 20 16:00:53 psokolovsky : No idea... As model number it lists PU10 Jun 20 16:01:31 9k, that's a universal, right? Jun 20 16:01:39 PXA250 @520Mhz 64MB ram 128MB flash M-systems lcd 460x640 Jun 20 16:01:44 steliosk: aha that appears to be htcuniversal Jun 20 16:01:45 HTC Universal (Qtek 9000/MDA Pro/SPV M5000/i-mate JASJAR) Jun 20 16:02:02 yep, the best-supported htc port ;-) Jun 20 16:02:15 * steliosk feels lucky Jun 20 16:02:21 steliosk: universal is pxa27x afaik Jun 20 16:02:41 koen: Interesting, the evas xrender engine calls the PrepareBlend kdrive hook masively. Jun 20 16:03:42 pH5: do you have defconfigman patches for latets mag refactors? Jun 20 16:03:54 anybody has links on getting linux on it ? Jun 20 16:04:16 So, implementing xrender in the kdrive server is expected to help it. Jun 20 16:05:58 psokolovsky: defconfigman should be up to date. the only thing missing is the pxa27x overlay support, but that isn't in hh.org anyway. Jun 20 16:06:28 pH5: egpio stuff? dropped magician_cpld? Jun 20 16:07:46 psokolovsky : Is this up to date ? http://www.handhelds.org/moin/moin.cgi/UniversalStatus Jun 20 16:08:20 steliosk: generally yes, afaik. Jun 20 16:10:16 psokolovsky: ow right, I have selected HTC_EGPIO in the Kconfig, so I didn't notice :-) Jun 20 16:10:18 http://en.pastebin.ca/579245 Jun 20 16:10:43 pH5: thanks! Jun 20 16:14:14 koen: that sl811 thing, did you test it with both h2200 and hx4700? Jun 20 16:14:46 hi Jun 20 16:15:05 woglinde: Hello Jun 20 16:19:02 steliosk: you added DISTRO_EXTRA_RDEPENDS to task-boot? Jun 20 16:20:58 steliosk, the status site at hh.org for universal is up2 date. Only need to update the latest kernel source status. Jun 20 16:24:01 * * OE Bug 2527 has been RESOLVED (FIXED) by arkaitzj(AT)gmail.com Jun 20 16:24:03 * *  Ucslugc does not compile and must be fixed to compile with uclibc0.9.29 Jun 20 16:24:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2527 Jun 20 16:25:51 koen i've marked that bug as fixed since with that commit you did seems to be resolved Jun 20 16:26:20 steliosk: does 57M cpio sounds good for minimal-image? or rather 5.2M cpio one? Jun 20 16:27:31 psokolovsky: with 2.6.16, yes Jun 20 16:28:03 he koen Jun 20 16:28:13 hey woglinde Jun 20 16:28:27 koen: ok, so I assume it should be there for any devices with builtin pcmcia Jun 20 16:28:35 koen I will send you my new monotone key today is woglinde2@openembedded.org okay? Jun 20 16:28:45 woglinde: send it to mickeyl Jun 20 16:28:51 koen okay Jun 20 16:42:21 hey people, /win 4 Jun 20 16:45:50 sirfred: http://www.rasterman.com/files/n800-engines.pdf Jun 20 16:46:58 koen: Interesting Jun 20 16:47:00 psokolovsky: yes Jun 20 16:47:42 koen: The x11-16 performs better. Jun 20 16:48:20 The xrender is the worst, but that is logic as all the xrender operations are probably made in software in the N800 xserver, Jun 20 16:48:51 The pitty is that the imageon should be able to perform some xrender acceleration, but for the pixmaps being 32bpp Jun 20 16:49:35 I've enabled some dummy callbacks in the xserver, and the composite callback is called to draw text antialiased, but the source format is 32bpp Jun 20 16:50:35 sirfred: did you see http://www.openembedded.org/repo/org.openembedded.dev/packages/xorg-xserver/xserver-kdrive-1.2.0/fbcompositesrc8888revnpx0565.patch already? Jun 20 16:50:41 please help someone: I've built angrstom all -images, how do I load it on my device? Jun 20 16:50:50 Humm, no. Jun 20 16:52:08 which device? Jun 20 16:52:35 compulab-x270 Jun 20 16:53:04 I'm trying to load kernel and it can't find RAMDISK... Jun 20 16:53:14 how do I define it then? Jun 20 16:53:21 hm did you compile the kernel with ramdisks on? Jun 20 16:53:31 lokk at the .config file Jun 20 16:54:07 probably... Jun 20 16:54:09 * djay-il checks Jun 20 16:54:23 if so, how do I say which ramdisk to load? Jun 20 16:54:34 but are you sure that kernel the image flash from the ramdisk? Jun 20 16:54:42 sirfred: on the n800 it gives a 3x speedup iirc Jun 20 16:54:43 djay-il: your vendor should have documentation on that Jun 20 16:54:45 isnt there an other tool for that? Jun 20 16:55:04 koen: I'm reading it. It's a software conversion enhancement. Jun 20 16:55:39 koen: The good way to go, imo, would be to be able to intercept the attempts of kdrive to upload 32bpp pixmaps to the offscreen memory. Jun 20 16:55:49 zecke: well, I am sorta of vendor Jun 20 16:55:55 I help them create it Jun 20 16:56:01 koen: Converting them before being uploaded, and them, be able to composite directly with the imageon hardware Jun 20 16:56:11 djay-il oehm than you should know how to flash the device Jun 20 16:56:12 But it seems that kdrive doesn't allow to do so. Jun 20 16:56:30 woglinde: no, I use nfs mount, I don't want to flash it yet Jun 20 16:56:48 It just upload to the card and them, it shamelessly give me the address of the pixmap: "Here you are", "But it's 32bpp!!", "Ooops" Jun 20 16:56:55 djay-il are I see Jun 20 16:57:10 djay-il hm but for this you dont neet a ramdisk Jun 20 16:57:37 only when you have to setup special thinks which the kernel cannt do alone Jun 20 16:57:41 koen: Intercepting that upload, we could blend in hardware, the imageon is capable of Jun 20 16:57:43 like configure pcmcia cards Jun 20 16:57:54 woglinde: I know, but kernel complains that it can't find one... Jun 20 16:58:01 Even with different mask values per colour component. Jun 20 16:58:27 I have images in /deploy/images, and I will try them, I just don't know how to pass the correct parameter to a kernel Jun 20 16:58:42 The problem is that the mask is only 4 bits per colour. Intercepting would be the key Jun 20 16:58:46 djay-il which is your network card? Jun 20 16:59:03 hm, I think realtek Jun 20 16:59:54 no, dm9000 says anything to you? Jun 20 17:00:10 djay-il -> http://pastebin.ca/579360 Jun 20 17:00:10 well, its Davicom Jun 20 17:00:19 this shows nfsroot for the simpad Jun 20 17:00:29 but you can figure out the parts Jun 20 17:00:49 root=/dev/nfs rw nfsroot=192.168.2.1:/dev Jun 20 17:00:49 el/arm/tmp/openzaurus/rootfs ip=192.168.2.2:192.168.2.1:192.168.2.1:255.255.255. Jun 20 17:00:50 0:simpad:usb0:off,v3 Jun 20 17:01:11 hmmm Jun 20 17:01:18 german spammers are sending pdfs now Jun 20 17:01:27 koen hu? Jun 20 17:01:42 woglinde: thanks, so I think I just need noinitrd part Jun 20 17:01:55 djay-il maybee Jun 20 17:02:40 the usb0 parameter shoul be your networkinterface Jun 20 17:02:52 :simpad: is only the hostname Jun 20 17:03:04 koen: Koen_report.pdf? Jun 20 17:03:18 woglinde: yea, I have ip::::cm-x270:eth0: Jun 20 17:03:36 I get an IP with dhcp Jun 20 17:04:00 hrw: yes Jun 20 17:04:07 second one today Jun 20 17:04:58 djay-il are okay Jun 20 17:05:30 woglinde: ok, it now doesn't show this, so I guess noinitrd worked Jun 20 17:05:54 djay-il hehe good look it shows the login prompt Jun 20 17:06:01 it doesn't Jun 20 17:06:12 it stuck on "Freeing memory" Jun 20 17:06:15 hm look at your host Jun 20 17:06:23 I mean, its not stuck, but no login Jun 20 17:06:26 if something goes wrong Jun 20 17:06:38 maybee you console parameter is wrong Jun 20 17:06:47 if I connect keyboard, I see usb log reports, but no login Jun 20 17:07:24 djay-il the other think which could bee is that the glibc is broken and login isnt loaded Jun 20 17:08:39 woglinde: I think its not loaded Jun 20 17:09:42 or getty is broken Jun 20 17:09:48 hm Jun 20 17:09:50 and has the wrong console parameter Jun 20 17:09:58 that might be Jun 20 17:10:08 how can I see the default console settings? Jun 20 17:10:11 you could try init=/bin/bash Jun 20 17:10:37 look at inittab Jun 20 17:10:43 init=/bin/busybox sh Jun 20 17:10:53 koen args sorry Jun 20 17:10:58 I am little whacky Jun 20 17:10:59 yes Jun 20 17:11:03 no bash on the image Jun 20 17:12:47 koen: "init=/bin/busybox sh" ? Jun 20 17:13:10 http://pastebin.ca/579387 Jun 20 17:13:16 objections? steliosk? Jun 20 17:13:18 Shouldn't /bin/sh be a link to busybox? Jun 20 17:14:48 mwester: it is Jun 20 17:19:46 hrw : Just for the MACHINE_EXTRA_RRECOMENTS Jun 20 17:20:18 hrw : if a machine specifies something it should be included Jun 20 17:20:41 hrw : For the distro its not a big problem Jun 20 17:30:33 steliosk: if machine need something to boot it set it in MACHINE_ESSENTIAL_EXTRA_RDEPENDS MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS Jun 20 17:30:37 djay-il busbyox is mutltiple binary which decide how to act the link name which was given or with the parameter Jun 20 17:31:17 mwester: but init=/bin/sh does not give argc/argv for busybox so it does not know what to call (something like that) Jun 20 17:31:32 mwester: so "init=/bin/sh sh" is proper way iirc Jun 20 17:32:39 steliosk: task-boot is to be as small as possible. x86 machines usually have "kernel-modules" package in MACHINE_EXTRA_RRECOMMENDS to get all modules in rootfs. but they need only few (if at all) to boot to shell Jun 20 17:34:27 woglinde: ok, cool, I suddenly have a problem with system unable to mount nfs, so still checking Jun 20 17:34:39 later Jun 20 17:34:47 djay-il: how much does cm-270 costs? Jun 20 17:35:14 djay-il /etc/exports isnt right? Jun 20 17:35:23 or kernel-daemon not loaded? Jun 20 17:35:29 woglinde: testing on other computer is working Jun 20 17:35:37 kernel-daemon? Jun 20 17:35:57 hrw: I don't really know. I can find out of you like :-) Jun 20 17:36:12 close to $300 for a usuable config Jun 20 17:36:40 iirc Q1 is twice the price of Q100 Jun 20 17:36:52 (I checked last year) Jun 20 17:37:10 koen: yea, something like that Jun 20 17:37:18 I just got one for free Jun 20 17:37:33 hrw: iirc mallum has one Jun 20 17:37:59 have to go, sorry. Jun 20 17:38:08 koen: I know ;) Jun 20 17:45:52 hrw: I was unaware of that issue with init and busybox; that's good to know. Jun 20 17:47:48 mwester: you have slug. did you used netconsole? Jun 20 17:48:45 argh.. /me need to really switch slug to something working instead of openslug 3.10 Jun 20 17:48:56 where working == with good feed Jun 20 17:55:27 hrw : Yes i know, but the "problem" with generic targets like i586-generic and i686-generic is that you don't know which modules they need for network for example Jun 20 17:55:53 so you have to get everything in Jun 20 17:56:05 or compile them in the kernel Jun 20 17:56:45 if its a specific machine i.e wrap you can get just the module it needs Jun 20 18:00:02 steliosk: for example let get maluch (dell d400). to boot it needs devicemapper, sha256, aes (as rootfs is on lvm). but to be usable it also needs tg3, bcm43xx and few others Jun 20 18:00:31 so I would add devicemapper, sha256, aes into MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS and tg3, bcm43xx and few others into MACHINE_EXTRA_RRECOMMENDS Jun 20 18:04:40 hrw : Assuming that you define a new machine that should not be a problem Jun 20 18:05:33 so we can keep the MACHINE_ESSENTIAL_EXTRA_RRECOMMENDS and MACHINE_EXTRA_RRECOMMENDS for these cases Jun 20 18:10:14 MACHINE_ESSENTIAL_EXTRA_RDEPENDS is used only by nokias now. Jun 20 18:13:33 * hrw goes out of computer Jun 20 18:13:38 see yo all Jun 20 18:15:21 "# Steve Jobs was still running NeXT when Duke Nukem Forever was announced." Jun 20 18:16:47 hmm Jun 20 18:16:52 has it been released Jun 20 18:17:00 nope Jun 20 18:17:02 see http://duke.a-13.net/ Jun 20 18:17:23 "Britney Spears' entire musical career as a pop star has taken place during Duke Nukem Forever's development." Jun 20 18:18:26 have any of you even played Duke Nukem? Jun 20 18:18:55 time to kick ass and chew bubblegum? Jun 20 18:19:05 * koen played duke nukem and duke3d Jun 20 18:25:11 duke nukem forever? Jun 20 18:25:13 *g* Jun 20 18:33:02 time to go home Jun 20 18:33:08 bye Jun 20 18:33:10 bye steliosk Jun 20 19:08:34 slapin, I would if I had any tech info ready and I will once I have something to say regarding the matter. Jun 20 19:53:20 03pfalcon 07org.oe.dev * r0b1741c2... 10/ (14 files in 14 dirs): Jun 20 19:53:20 linux-handhelds-2.6: Update defconfigs: Jun 20 19:53:20 * CONFIG_NLS_DEFAULT="utf8" (suggested by Koen Kooi) Jun 20 19:53:20 * For devices with builtin PCMCIA, enable driver for Ratoc USB Host Card. Jun 20 19:53:20 * htcuniversal: add supportf or ts-adc-debounce, modular, old driver, Jun 20 19:53:21 htcuniversal_ts, made modular too. Jun 20 19:53:23 * magician: Update for latest CVS. Jun 20 19:53:25 03pfalcon 07org.oe.dev * rc664220a... 10/ (1 conf/machine/htcuniversal.conf): Jun 20 19:53:29 htcuniversal.conf: Update for transition to ts-adc-debounce driver. Jun 20 19:53:31 * Now both odl and new drivers are modules, old one is loaded by default Jun 20 19:53:33 for now. Jun 20 19:53:35 03Laibsch 07org.oe.dev * r0725dca6... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom.conf: prefer version 2.3.4 of freetype Jun 20 19:53:40 03koen 07org.oe.dev * ra88acd45... 10/ (1 packages/efl/evas.inc): evas: inherit lib_package Jun 20 20:08:02 hi Jun 20 20:08:14 ho Jun 20 20:08:25 kermit the frog here? Jun 20 20:10:58 no just gonzo Jun 20 20:11:37 I always liked the Swedish chef... but that might not be politically correct anymore. Jun 20 20:11:47 oh come on! "hi ho, kermit the frog here" Jun 20 20:30:47 ~logs Jun 20 20:30:48 i heard logs is apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz Jun 20 21:17:58 someone with prism3 card? Jun 20 21:26:42 Good night Jun 20 22:06:59 hi Jun 20 22:07:33 hi Jun 20 22:09:46 Join #debian Jun 20 22:12:37 Laibsch: heh have fun Jun 20 22:13:26 Well, I sort of do Jun 20 22:13:37 I was helped fixing up the mess at ltg.org Jun 20 22:13:45 The server is not in such a good shape Jun 20 22:15:38 heh we all noticed Jun 20 22:15:46 hardware probs? Jun 20 22:16:07 it needs more ram too Jun 20 22:17:03 I'll be happy to provide mirroring for non-cpu heavy services which do not require user logins Jun 20 22:17:32 (forget about mirroring the OE sources tho...) Jun 20 22:17:50 I could provide some mirroring, too Jun 20 22:17:57 I don't think the machine is overloaded Jun 20 22:18:18 And I don't understand why memory is tight when 700M are cached Jun 20 22:18:27 What is in all that cache, I wonder. Jun 20 22:18:49 just saying Jun 20 22:19:24 Thanks, nice of you Jun 20 22:19:39 I want to get this in a bit of a cleaner state software wise. Jun 20 22:19:44 And then we'll see Jun 20 22:20:42 ;) Jun 20 22:28:11 CoreDump|home: rsync, or tarball mirror? Jun 20 22:29:30 zecke: ? Jun 20 22:29:56 CoreDump|home: providing mirroring Jun 20 22:30:22 preferably one I can keep synced via rsync or similar Jun 20 22:30:28 http://www.linuxdevices.com/news/NS6984164452.html Jun 20 22:30:58 zecke: however, mirroring > 10G is not possible unfortunately Jun 20 22:31:35 in traffic? Jun 20 22:31:59 in HDD space Jun 20 22:33:08 zecke: You guys need mirror space? Jun 20 22:33:22 hmm, I need to do more consumer oriented embedded work :) Jun 20 22:38:31 heh.. Jun 20 22:38:50 so-dimm ddr 512M is same price as so-dimm ddr2 1G ;( Jun 20 22:57:47 03Stefan 07org.oe.dev * r9d0ddeec... 10/ (4 files in 4 dirs): packages/memedit: Add memedit 0.7 a readline based tool like devmem2 but more convient Jun 20 23:50:54 ltg is back in a state where at least "apt-get install" returns without an error. Jun 20 23:52:49 Laibsch: ouch Jun 20 23:58:29 well, it was ouch Jun 20 23:58:35 Now it should be better Jun 20 23:58:49 It was and largely still is a hell of a mess Jun 20 23:59:43 03coredump2 07org.oe.dev * r58aad6e3... 10/ (3 files in 2 dirs): fluxbox: Fix .debug QA error Jun 21 00:58:35 03coredump2 07org.oe.dev * rac749382... 10/ (3 files in 3 dirs): gaim: Fix icon for pidgin Jun 21 00:58:39 03coredump2 07org.oe.dev * r6784a6b7... 10/ (3 files in 2 dirs): xchat: Update to 2.6.8 **** ENDING LOGGING AT Thu Jun 21 02:59:57 2007