**** BEGIN LOGGING AT Tue Apr 15 02:59:56 2008 Apr 15 03:47:46 03xjqian 07org.oe.dev * r88f857b2... 10/ (3 files in 2 dirs): Apr 15 03:47:46 libtomcrypt: fix SRC_URI (close oebug 4159), bump to 1.17, fix staging Apr 15 03:47:46 * no idea why the previous maintainer leave do_install empty, but assuming there is a legit reason. Apr 15 03:49:04 * * OE Bug 4159 has been RESOLVED (FIXED) by xjqian(AT)gmail.com Apr 15 03:49:06 * * libtomcrypt URL changes Apr 15 03:49:08 * * http://bugs.openembedded.org/show_bug.cgi?id=4159 Apr 15 05:13:04 * * OE Bug 4174 has been RESOLVED (FIXED) by xjqian(AT)gmail.com Apr 15 05:13:06 * * lib-sdl-qpe cannot be build Apr 15 05:13:09 * * http://bugs.openembedded.org/show_bug.cgi?id=4174 Apr 15 05:15:04 03xjqian 07org.oe.dev * rae71623a... 10/ (3 files in 2 dirs): libsdl-qpe: add skipped patch, replace qte with qte-mt thoroughly (close oebug 4174) courtesy of Michal Hrusecky Apr 15 06:56:19 morning Apr 15 07:55:36 hi Apr 15 07:59:36 which task do I need to clean to be sure an images get the correct set of ipks? Apr 15 09:02:00 what do I need to do build a local feed? Apr 15 09:12:24 morning Apr 15 09:34:09 good morning Apr 15 09:42:52 morning all Apr 15 10:23:54 re Apr 15 10:41:58 03thesing 07org.oe.dev * r9bafa1b7... 10/ (5 files in 2 dirs): Apr 15 10:41:58 klibc: add wc to tools Apr 15 10:41:58 * for use with uniboot Apr 15 10:42:44 hrw : can you remind your naming proposal for the p4 celleron package arch ? Apr 15 10:44:50 i686p4 Apr 15 10:45:25 steliosk: or i686p4c if you want to mark celeron tuning Apr 15 10:45:52 but i686p4 should be ok for -mtune=celeron or -mtune=pentium4 Apr 15 10:46:23 not all celerons are pentium4 Apr 15 10:46:35 so we would have i686 i686p3 i686p4 i686c2d etc Apr 15 10:46:48 XorA: -march=pentium4 -mtune=celeron are steliosk's settings Apr 15 10:47:50 when -march is pentium4 I bet the mtune is a NOP Apr 15 10:48:05 rschuster: Sorry if you're the wrong person to ask, but if cacao fails to compile, should I file a bug at cacaojvm.org or oe.org? Apr 15 10:48:39 According to the recent gcc -march defines the instruction set and mtune the "schedulling" of the commands Apr 15 10:49:01 yeah, but when you say, use latest you cant then tune for an older cpu :-) Apr 15 10:49:18 yes Apr 15 10:49:27 things like -march=i386 -mtune=celeron would make code run everywhere, but best on celeron Apr 15 10:49:35 or at least i think so Apr 15 10:49:51 in oe we need i486 Apr 15 10:51:03 the thing is that there are so many 'weird' cores with or without certain instructions or will instructions that operate differently depending on the schedulling Apr 15 10:51:06 is x86 in oe handled like any other arch or is there a shortcut (e.g. no toolchain build) ? Apr 15 10:51:30 it builds like eveyrhing else Apr 15 10:52:17 having code that runs on all machines is one things, but having code that runs better for a specific board/cpu is another Apr 15 10:53:43 btw. did anybody try to run oe on not x86 arch? Apr 15 10:56:55 thesing : what do you mean ? Apr 15 10:57:09 i am using i686 and x86 a lot Apr 15 10:57:48 steliosk: I mean building images using oe on a ppc, mips or arm host. Apr 15 10:58:13 03thebohemian 07org.oe.dev * r7845f20b... 10/ (7 files in 2 dirs): swt-gtk: Do not install/stage swt-pi.jar and swt.jar any more. Apr 15 11:00:28 thesing : i have been told that someone was building OE on ppc g3 or g4. I have a g4 machine laying around and intend to try it, but no time yet Apr 15 11:00:49 was building OE -> some time back Apr 15 11:00:55 not sure if .dev builds now Apr 15 11:02:54 03thesing 07org.oe.dev * racd40ed8... 10/ (1 packages/klibc/files/wc.patch): klibc: add missing wc.patch Apr 15 11:24:25 steliosk, the mainstream distros seem to not generate custom packages for every possible x86 CPU Apr 15 11:24:45 Crofton: but we are Apr 15 11:25:03 Crofton: i486 mplayer is much much slower then sse/mmx/3dnow mplayer Apr 15 11:25:36 can't mplayer run time detect available extensions? Apr 15 11:27:31 Crofton: mplayer yes, but not all apps have this functionality. btw mplayer cpudetection costs lot of memory that application need to allocate Apr 15 11:27:35 ok, it can but it was example Apr 15 11:27:50 :) Apr 15 11:28:14 * Crofton suspects mainstream distros assume infinite disk space also Apr 15 11:28:46 but I don't think building 50 different versions of mplayer is sane either Apr 15 11:29:00 * NineX is an gentoo user - all my applications are optimized for my hardware Apr 15 11:29:24 Crofton: why not ? Apr 15 11:29:25 :> Apr 15 11:31:47 Crofton: /dev/sda6 123T 39T 85T 31% /home Apr 15 11:33:26 hrw: load 1023.0 1101.0 1800.0 <- i consider hard reset Apr 15 11:33:37 Hehe Apr 15 11:33:40 hmm, how to enforce an OABI build with angstrom? setting the ARM_ABI variable does not do the trick Apr 15 11:34:56 steliosk: please do not steal threads ;( Apr 15 11:35:19 [oe] union fs -> [oe] union fs -> [oe] union fs Apr 15 11:35:31 [oe] union fs -> [oe] union fs -> [oe] union fs -> [oe] stelios about archs Apr 15 11:35:33 argh Apr 15 11:37:18 mhm, so any ideas about oabi setting? or is that rather a question for the ml? using .stable Apr 15 11:45:27 ANGSTROM_MODE maybe Apr 15 11:45:33 maybe that was libc tough Apr 15 11:45:35 though Apr 15 11:45:40 Crofton: nah, that's uclibc/glibc/eglibc selection Apr 15 11:46:07 angstrom has an ARM_ABI thing in the configs and it sets it itself depending on machines, I tried overriding with := but it did not do the trick Apr 15 11:46:55 in linux.inc the selection is controlled by TARGET_OS Apr 15 11:46:55 I managed to derive an own distro where I could produce EABI/OABI stuff on demand, but I hoped to be able to do it with Angstrom too Apr 15 11:47:22 I think you need TARGET_OS=linux, not linux-gnueabi Apr 15 11:47:37 maybe that's what I missed Apr 15 11:47:41 I will try that, thanks Apr 15 11:47:55 np, I happened to be looking in linux.inc atm Apr 15 12:00:34 hrw : opps did a reply and forgot to change he subject Apr 15 12:01:44 steliosk: NEVER use reply to start thread Apr 15 12:01:48 hi mickeyl Apr 15 12:01:53 hi hrw Apr 15 12:02:11 steliosk: each normal mail application has thing called "new mail" Apr 15 12:02:12 steliosk: you bad bad man, you forgot the In-Reply-To: headers Apr 15 12:02:39 XorA: not me... its the outlooks fault ;) Apr 15 12:03:00 ~boltfromheavens steliosk Apr 15 12:03:01 which i am not sure or worst thing to claim Apr 15 12:03:17 mickeyl : hi ! Apr 15 12:06:44 hi steliosk Apr 15 12:33:58 to produce a working local feed, do I need to run anything extra? at the moment it complains (on the device) that it can't download the Packages.sig Apr 15 12:35:04 gpg -sign Packages Apr 15 12:35:54 --detach-sign is needed I think Apr 15 12:36:38 XorA: And -a too, I'd imagine Apr 15 12:36:55 I dont think it needs to be ascii armored Apr 15 12:41:23 the official angstrom feed don't have a sig file... Apr 15 12:42:03 cedric_: file a bug then Apr 15 12:42:40 XorA: is this a requited thing for opkg? Apr 15 12:42:49 cedric_: use opkg-nogpg instead of opkg? Apr 15 12:42:58 hrw: nice Apr 15 12:44:17 better to file a bug, because Angstrom would benefit from signed feeds Apr 15 12:44:31 ok I'll Apr 15 12:46:05 XorA: but who will sign them? Apr 15 12:46:18 hrw: the feed update scripts can Apr 15 12:46:27 hrw: can make an angstrom official key Apr 15 12:47:03 then we make an anstrom-file-bug command that detects 3rd party broken feeds and packages :-) Apr 15 12:47:25 hrw: can I have opkg and opkg-nopgp coexist in the same system? Apr 15 12:47:33 no Apr 15 12:47:51 ah that's why I can install it... Apr 15 13:16:37 ok now how to import my public key to the device? gpg is not a OE package Apr 15 13:16:48 cedric_: gnupg? Apr 15 13:17:12 :-) Apr 15 13:28:18 XorA, hrw, NAByss: thx it works... Apr 15 13:29:06 cedric: sorted Apr 15 13:33:48 XorA: yep Apr 15 13:41:59 the Packages* file are generated every time I'm building a new package? or is there any thing to update these? Apr 15 13:44:23 bitbake package-index Apr 15 13:48:57 03crofton 07org.oe.dev * r101eec50... 10/ (7 files in 5 dirs): Apr 15 13:48:57 davinci-sffsdr machine : Update kernel config, add board patch, Apr 15 13:48:57 update machine file to work with linux.inc, Apr 15 13:48:57 delete patches that are upstream. Apr 15 14:20:08 * ScaredyCat taps florian Apr 15 14:21:04 uh? :) Apr 15 14:21:06 * XorA taps ScaredyCat with a 150lbs lump hammer Apr 15 14:21:35 www.angstrom-distribution.org in on the same box as monotone ? Apr 15 14:21:47 ScaredyCat: nope Apr 15 14:21:56 mmm Apr 15 14:22:07 ScaredyCat: monotone is port forwarded to a different machine Apr 15 14:22:16 gahhh Apr 15 14:22:19 * ScaredyCat slaps people Apr 15 14:22:45 well www.angstrom-distribution.org is down anyway... Apr 15 14:22:52 whoever needs to know.. Apr 15 14:23:02 * ScaredyCat slaps XorA a few times Apr 15 14:26:48 ~lart ScaredyCat Apr 15 14:26:48 * ibot throws a AN/M-8 smoke grenade at ScaredyCat Apr 15 14:27:18 * ScaredyCat farts in dcordes general direction Apr 15 14:27:31 * chouimat|work dies Apr 15 14:28:10 ~lart Apache Apr 15 14:28:11 * ibot farts in Apache's general direction Apr 15 14:28:24 heh Apr 15 14:29:11 * ScaredyCat thinks that machine runs on a zx81 Apr 15 14:30:22 ScaredyCat: Unlikely... it servers ~800 GB a month. Apr 15 14:30:47 slowly Apr 15 14:31:12 Donations welcome :-) Apr 15 14:31:16 seems to have woken up now... Apr 15 14:31:36 * ScaredyCat donates XorA's legs Apr 15 14:33:08 I dont think some ugly hairy legs will help florian optimise apache :-) Apr 15 14:34:00 they might scare him into fixing it ;) Apr 15 14:34:57 creepy Apr 15 14:44:19 ScaredyCat: we can take any spare rack machine from you Apr 15 14:45:00 spare? can you have spare machines? Apr 15 14:45:02 :) Apr 15 14:46:30 ScaredyCat: then do not complain ;D Apr 15 14:46:51 I wasn't complaining as much as trying to notify ... Apr 15 14:46:57 it was very slow... Apr 15 14:47:03 so needed a kick.. Apr 15 14:47:08 ScaredyCat: ltg machine is slow always Apr 15 14:47:22 I'm sure if I had a spare machine you could have it Apr 15 14:47:59 I have spare amd64 cpu Apr 15 14:48:29 just the cpu ? Apr 15 14:52:23 yes Apr 15 14:52:45 it was left from my amd64 3200+ -> amd64 x2 4200+ upgrade Apr 15 14:56:32 and its price is so low that I do not even want to sell it Apr 15 15:56:15 bye Apr 15 17:02:42 florian, tmut has simple account management now Apr 15 17:02:51 time to replace gpephone's email client? :) Apr 15 17:03:27 or perhaps just quickly rewrite it by making tmut having a compiletime option to build it as a library and subclassing all of tmut's high level types? :) Apr 15 17:27:02 re Apr 15 18:47:27 hrw : Have a look at this -> http://gentoo-wiki.com/Safe_Cflags#Geode_LX_.28AMD.29 Apr 15 18:48:38 hrw : not just for the geode, but for all the x86 arches Apr 15 19:08:53 steliosk: nice reference. The gentoo project does produce good documentation Apr 15 19:19:12 cbrake : yes it summarizes what i found searching the net over the last few days Apr 15 19:23:39 morning all Apr 15 20:08:38 03koen 07org.oe.dev * r93a15d8b... 10/ (1 conf/distro/angstrom-2008.1.conf): angstrom 2008: remove oz mirror from tarball stash Apr 15 20:11:59 damn, now I can't stalk mickeyl by ip number Apr 15 20:12:19 heh Apr 15 20:12:32 big cheater Apr 15 20:12:55 Crofton: just take control of Frankfurt CCTV nerwork instead :-) Apr 15 20:13:23 or we fix him up with an 810 that broadcasts his lat/lon Apr 15 20:13:31 03koen 07org.oe.dev * r9de6d1cc... 10/ (1 packages/linux/linux-davinci_2.6.x+git-davinci.bb): linux-davince: fix PV and anchor SRCREV inside the recipe to avoid reparse penalties Apr 15 20:14:17 Crofton: buy his wife shoes in turn for spying Apr 15 20:14:52 you guys have crazy ideas :D Apr 15 20:15:02 sabine is not a treater for shoes Apr 15 20:15:12 you will have to offer her different things Apr 15 20:15:31 chocolate always works. Apr 15 20:16:43 mickeyl: if we get crazy enough we get allowed to work in marketing :-) Apr 15 20:17:08 *cough* Apr 15 20:17:13 oh that's your plan? :D Apr 15 20:18:11 If it is, then we're going to have to ask you to leave Apr 15 20:37:53 03koen 07org.oe.dev * rb8d314e8... 10/ (1 packages/linux/linux-davinci_2.6.x+git-davinci.bb): Apr 15 20:37:53 linux-davinci: fix thinko, you can't use ?= with srcrev, since bitbake.conf still sets it Apr 15 20:37:53 * Mickey, how does that fit into your "no distro" proposal? Apr 15 20:56:41 03koen 07org.oe.dev * rba1cc115... 10/ (1 classes/kernel.bbclass): (log message trimmed) Apr 15 20:56:41 kernel bbclass: have modules.tgz have the same version and packages and deployed images (PV) instead of pulling it out of the makefile (KERNEL_VERSION) Apr 15 20:56:41 * before: Apr 15 20:56:41 836K modules-2.6.25-rc8-davinci1-r1-davinci-sffsdr.tgz Apr 15 20:56:41 1.4M uImage-2.6.24+2.6.25-rc8-r1-davinci-sffsdr.bin Apr 15 20:56:43 * after: Apr 15 20:56:47 836K modules-2.6.24+2.6.25-rc8-r1-davinci-sffsdr.tgz Apr 15 23:25:01 03tmbinc 07org.oe.dreambox * r2ce338f9... 10/ (3 files in 3 dirs): linux-dm800: fix memsize patch **** ENDING LOGGING AT Wed Apr 16 02:59:56 2008