**** BEGIN LOGGING AT Tue Oct 20 02:59:57 2009
Oct 20 06:22:42 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:22:48 <mayhem> wat to do in this?
Oct 20 06:22:51 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:22:52 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:22:52 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:23:10 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:23:18 <mayhem> this is the erro m facing
Oct 20 06:23:23 <mayhem> hw to resolve this??
Oct 20 06:28:11 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:28:12 <mayhem> ???
Oct 20 06:28:20 <mayhem> hw to resolve this dependency problem
Oct 20 06:36:30 <mayhem> Task /net/wintermute/vol1/home/anish/openembedded/recipes/update-rc.d/update-rc.d_0.7.bb (do_package) has circular dependency on /net/wintermute/vol1/home/anish/openembedded/recipes/glibc/glibc_2.9.bb (do_package)
Oct 20 06:36:36 <mayhem> hw to solve this?
Oct 20 06:46:54 <zecke> mayhem: when did this happen?
Oct 20 06:53:27 <cokebottles> hello all
Oct 20 06:54:33 <cokebottles> i have a few questions; i know this may be bias but... i'm using buildroot currently to build toolchain with uclibc 0.9.29 and having problems; then..
Oct 20 06:54:57 <cokebottles> i started looking into OpenEmbedded and i know a few ppl say it's a learning curve (which i don't mind)
Oct 20 06:56:10 <cokebottles> what's the best way to build x86 using uclibc and future creation of x86 small custom image?
Oct 20 06:56:51 <cokebottles> i find makefiles very hard to understand as they are complex and many dependancies on other variables and such
Oct 20 06:57:22 <mayhem> now @ zecke
Oct 20 07:01:02 <ynezz> cokebottles: I think, that you can find answer on the wiki, but simply it's setting MACHINE="x86"; DISTRO="angstrom-2008.1"; TARGET_OS = "linux-uclibc" variable in local.conf and running "bitbake minimal-image"
Oct 20 07:04:18 <cokebottles> ynezz: do i have to use angstrom-2008.1 ?  is this just an arbitrary name; so i could say "cokebottle-x86-rev1" ??
Oct 20 07:11:22 <ynezz> read wiki pls
Oct 20 07:14:06 <cokebottles> ynezz: will do. thank you
Oct 20 07:43:20 <mckoan> good morning
Oct 20 08:06:50 <mike_cw> is there a "right" way to build something that depends on Qt private headers?
Oct 20 08:44:43 <mickey|office> imp.org
Oct 20 08:48:29 <Eric33> hi, I'm fighting to build a meta-toolchain-qte form armv7 (OMAP3530). I started with "plain standard" angstrom which is using glibc-2.9 and the sdk is properly built but fails to link applications (undefined reference to `__libc_csu_fini'). Then, following an advice I got on IRC yesterday, I started again with eglibc and now eglibc-2.10-r7.1 fails to compile. 
Oct 20 08:49:59 <Eric33> Do you have any advice on how to get a qte sdk using OE ?
Oct 20 08:50:12 <Eric33> (a working qte sdk ;-)
Oct 20 08:55:21 <florian> good morning
Oct 20 09:37:21 <recalcati> good morning
Oct 20 09:38:17 <eFfeM> hi recalcati
Oct 20 09:38:33 <recalcati> eFfeM: Hi
Oct 20 09:39:25 <CIA-1> 03Koen Kooi <koen@openembedded.org> 07org.openembedded.dev * re82669abcf 10openembedded.git/ (11 files in 3 dirs): pixman: add some new versions, update git recipe
Oct 20 09:43:25 <recalcati> Anybody knows a linux hackable ethernet router?
Oct 20 09:43:39 <recalcati> with monitor port, obviously
Oct 20 09:47:03 <eFfeM> if it is hacakble you can add the monitor function
Oct 20 09:47:27 <eFfeM> i'd suggest look at the devices supported by openwrt
Oct 20 09:47:35 <eFfeM> most hacked router is probably linksys wrt54g
Oct 20 09:47:46 <eFfeM> make sure you get the linux version not the vxworks one
Oct 20 09:51:24 <recalcati>  WRT54G seems to be a little old
Oct 20 09:53:56 <recalcati> anyway I take a better look
Oct 20 09:54:27 <mckoan> recalcati: WRT54G is always the best choiche if you use OpenWRT
Oct 20 09:55:28 <mckoan> recalcati: BTW we ported OpenWRT on Koala PC http://www.koala.it/it/n33.htm 
Oct 20 09:55:46 <mckoan> recalcati: new 2 lan model :-D
Oct 20 09:56:05 <recalcati> mckoan:  it means that I can use it as a switch/router ?
Oct 20 09:56:44 <recalcati> because I was searching a switch/router with monitor port and I thought, why not buy an hackabkle one ?
Oct 20 09:58:52 <mckoan> recalcati: I mistake, it is not a router, and I never used a Linksys router with OpenWRT
Oct 20 09:59:09 <mckoan> recalcati: BTW WRT54G is not a router
Oct 20 09:59:38 <recalcati> but its nice your product, I'm looking
Oct 20 09:59:53 <eFfeM> mckoan: why not, to me it routes between wan and lan
Oct 20 10:00:48 <eFfeM> recalcati: wrt54g is indeed old
Oct 20 10:02:46 <mckoan> eFfeM-lunch: usually router means routing to WAN
Oct 20 10:05:16 <mckoan> recalcati: BTW I'm using one WRT54GL at home and at office both hacked with OpenWRT
Oct 20 10:05:50 <mckoan> eFfeM-lunch: I mistake again, you're right I meant a Gateway like AG241 ;-)
Oct 20 10:06:30 <recalcati> mckoan: it is possible to send upnp multicast to one port and not to another
Oct 20 10:06:32 <recalcati> ?
Oct 20 10:07:38 <CIA-1> 03Thomas Zimmermann <zimmermann@vdm-design.de> 07shr/import * r59d1dea095 10openembedded.git/recipes/xorg-driver/ (7 files in 2 dirs): 
Oct 20 10:07:38 <CIA-1> shr/import: xf86-input-tslib: patch for xserver-1.7 + additional patches from open-wrt
Oct 20 10:07:38 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 10:08:02 <mckoan> recalcati: you have the sources ;-)
Oct 20 10:08:23 <recalcati> nice! thx!
Oct 20 10:08:30 <mckoan> recalcati: LOL
Oct 20 10:10:36 <recalcati> last LOL: can you give me an idea of a Mini PC 2PCI and 4 LAN GIGABIT ? ready to use (with Hard disk)  
Oct 20 10:11:07 <mckoan> recalcati: we're OT here let's switch to private
Oct 20 10:11:21 <recalcati> ok, sorry
Oct 20 10:11:58 <recalcati> I send you an email
Oct 20 10:13:09 <mckoan> recalcati: I already sent you one ;-)
Oct 20 10:27:28 <pb__> hi lrg
Oct 20 10:40:04 <lrg> hey pb__
Oct 20 10:44:57 <florian> hi pb__ 
Oct 20 10:45:02 <recalcati> eFfeM: I found a lot of Linux ethernet router: for example ADM5120, but http://www.linux-mips.org/wiki/Adm5120_devices means that each include adm5120?
Oct 20 10:45:16 <florian> !
Oct 20 10:48:45 <jbrouault> khem: thanks for your help last night, I hadn't seen the colorgcc error :)
Oct 20 10:50:03 <pb__> re
Oct 20 10:51:04 <florian> pb__: wb
Oct 20 10:51:22 <pb__> stupid computer locked up there
Oct 20 10:51:24 * pb__ stabs linux
Oct 20 10:52:04 <florian> all this evil open source stuff ;)
Oct 20 10:53:07 <florian> I have to admit that its quite some time ago I had this on a PC here other than the laptop who seems to fail ~once a month returning from suspend.
Oct 20 10:54:00 <pb__> yeah, it doesn't happen to me often either.
Oct 20 10:54:28 <pb__> I did use to have trouble with opengl apps (particularly xscreensaver) making my video driver lock up, but that doesn't seem to be happening anymore.
Oct 20 10:54:39 <pb__> fairly rare for it to just freeze in the middle of typing an email like it did then.
Oct 20 10:54:46 <pb__> oh well
Oct 20 11:19:50 <hrw> morning
Oct 20 11:21:09 <florian> hi hrw 
Oct 20 11:26:50 <blindvt`> i must be missing some crucial idea, i fear. Can anyone enlight me? E.g.: hildon-lgpl is "nonworking", i.e.  filtered-out via the default BBMASK. Now osso-application-installer (which is not filtered out) depends on hildon-lgpl. On bitbake world i get an error (AFAICS via task-maemo) since it sees unbuildable dependencies. I would have expected that filtered-out chains are just (verbosely) dropped, no? How would i go about fixing/bypassing such unb
Oct 20 11:26:50 <blindvt`> uildable dependencies automagically?
Oct 20 11:33:24 <mickey|office> you can't. bitbake is instructed to report errors for unbuildable dependencies rather than warnings. any (buildable) recipes that depend on recipes in nonworking are considered to be bugs, so i would not mind moving osso-application-installer to nonworking as well until the dependencies have been fixed
Oct 20 11:36:09 <blindvt`> mickey|office, i see. That's a pity, but ok
Oct 20 11:38:21 <mickey|office> well
Oct 20 11:38:25 <mickey|office> i consider this being a feature
Oct 20 11:38:37 <mickey|office> otherwise we would get lots of complaints of packages being filtered out automatically
Oct 20 11:38:43 <blindvt`> mickey|office, and is there _any_ way to express "soft" dependencies, for example for some specific flavour? I've seen that there are base_contains that i could use to specify e.g. DEPENDS. But AFAICS this doesn't help me to depend e.g. skype-static.bb on some specific LIBC, does it?
Oct 20 11:39:35 <mickey|office> this touches a long standing controversy in OE
Oct 20 11:39:51 <mickey|office> until recently the common opinion was that builds should be deterministic
Oct 20 11:40:07 <mickey|office> which means we don't want packages that have the same name but could be built in different configurations
Oct 20 11:40:19 <mickey|office> (which is essentially what your "soft" dependency would lead to)
Oct 20 11:40:34 <mickey|office> on the next OEDEM USE_FLAGS will be talked about again
Oct 20 11:41:00 <mickey|office> you can always have multiple recipes though
Oct 20 11:41:13 <mickey|office> that depend on certain specific recipes
Oct 20 11:41:21 <mickey|office> see the qt dirs e.g.
Oct 20 11:41:27 <mickey|office> or libsdl
Oct 20 11:41:36 <mickey|office> there's also virtual/foo as provider
Oct 20 11:41:54 <mickey|office> which might be handy at times
Oct 20 11:42:19 <blindvt`> mickey|office, perhaps i described it in the wrong way. Suppose i use LIBC = libmyc. How can i get rid of e.g. skype-static (which depends on glibc-gconv-utf-16 which is not provided by libmyc on purpose)?
Oct 20 11:42:57 <mickey|office> what do you mean by 'get rid' ?
Oct 20 11:43:03 <mickey|office> just don't include it in your image
Oct 20 11:43:23 <blindvt`> mickey|office, "get rid of" as in i don't want to hear any complains about skype-static
Oct 20 11:43:31 <mickey|office> just don't attempt to build it
Oct 20 11:43:40 <blindvt`> mickey|office, neither do i want to (attempt to) build it
Oct 20 11:43:41 <mickey|office> i.e. do not include it in DEPENDS for your image
Oct 20 11:43:58 <blindvt`> hm
Oct 20 11:45:22 <Crofton> um, lack of deterministic builds are still a big concern
Oct 20 11:45:29 <blindvt`> ok, so i'm doing something horribly wrong, somehow. I'm attempting to build a "micro" distro, like:
Oct 20 11:45:30 <blindvt`> $ egrep -v "^(#|$)" conf/local.conf
Oct 20 11:45:30 <blindvt`> DL_DIR = "${HOME}/src/oe/sources"
Oct 20 11:45:30 <blindvt`> BBFILES := "${HOME}/src/oe/openembedded/recipes/*/*.bb"
Oct 20 11:45:30 <blindvt`> BBMASK = "(maemo|skype)"
Oct 20 11:45:30 <blindvt`> PREFERRED_PROVIDERS = "virtual/qte:qte virtual/libqpe:libqpe-opie"
Oct 20 11:45:32 <blindvt`> PREFERRED_PROVIDERS += " virtual/libsdl:libsdl-x11"
Oct 20 11:45:34 <blindvt`> PREFERRED_PROVIDERS += " virtual/${TARGET_PREFIX}gcc-initial:gcc-cross-initial"
Oct 20 11:45:36 <blindvt`> PREFERRED_PROVIDERS += " virtual/${TARGET_PREFIX}gcc-intermediate:gcc-cross-intermediate"
Oct 20 11:45:38 <blindvt`> PREFERRED_PROVIDERS += " virtual/${TARGET_PREFIX}gcc:gcc-cross"
Oct 20 11:45:40 <blindvt`> PREFERRED_PROVIDERS += " virtual/${TARGET_PREFIX}g++:gcc-cross"
Oct 20 11:45:42 <blindvt`> PREFERRED_PROVIDERS += " virtual/clutter:clutter-gtk_git.bb"
Oct 20 11:45:44 <blindvt`> MACHINE = "i386-generic"
Oct 20 11:45:44 <mickey|office> err
Oct 20 11:45:45 <mickey|office> ~pastebin
Oct 20 11:45:47 <ibot> [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste
Oct 20 11:45:48 <blindvt`> TARGET_OS = "linux-uclibc"
Oct 20 11:45:50 <blindvt`> DISTRO = "micro-uclibc"
Oct 20 11:45:52 <blindvt`> IMAGE_FSTYPES = "jffs2 tar"
Oct 20 11:45:54 <blindvt`> PERSISTENT_DIR = "${HOME}/src/oe/persist"
Oct 20 11:45:56 <blindvt`> BBDEBUG = "yes"
Oct 20 11:45:58 <blindvt`> BBINCLUDELOGS = "yes"
Oct 20 11:46:00 <blindvt`> erm, sorry
Oct 20 11:46:16 <mickey|office> i see no skype-static in there, what's your problem?
Oct 20 11:46:41 <mickey|office> you don't build a distro btw.
Oct 20 11:46:43 <mickey|office> you build an image
Oct 20 11:46:49 <mickey|office> with a distro being configured
Oct 20 11:47:00 <mickey|office> (that's an important difference)
Oct 20 11:48:02 <blindvt`> so why does micro apparently pull in (or at least parse and complain about) all this madness like maemo and skype and whatnot?
Oct 20 11:48:23 <pb__> what exactly are you trying to build?
Oct 20 11:48:28 <hrw> blindvt`: it is not micro but rather your BBFILES?
Oct 20 11:48:29 <blindvt`> rafb.net/paste was unfortunately turned off some months ago, btw.
Oct 20 11:48:29 <pb__> i.e. what is the argument you are giving to bitbake?
Oct 20 11:49:18 <blindvt`> BBPATH="$OEMAIN/build:$OEMAIN/openembedded" bitbake world (with the config/local.conf erroneously pasted above)
Oct 20 11:49:49 <pb__> ah, you are trying to bitbake world?  that would explain it.
Oct 20 11:50:05 <pb__> I guess there is some package in world which depends on the things you have filtered out.  bitbake doesn't have any automatic logic to fix that up.
Oct 20 11:50:24 <blindvt`> pb_, i'm trying to build a small uClibc based system. Small as in ultimately i only want uClibc and busybox and a kernel in it
Oct 20 11:50:28 <Crofton|work> maybe we should remove bitbake world?
Oct 20 11:50:34 <Crofton|work> it seems to create lots of trouble
Oct 20 11:50:39 <pb__> blindvt`: in that case, "world" is probably not the appropriate target.
Oct 20 11:50:44 <pb__> try "bitbake micro-base-image"
Oct 20 11:51:04 <blindvt`> pb_, oh. Let me try that..
Oct 20 11:52:37 <blindvt`> pb_, looks much better, yes. Thanks!
Oct 20 11:53:01 <blindvt`> it tries to download coreutils. Where's my chainsaw
Oct 20 11:53:10 * blindvt` goes to play with it
Oct 20 11:54:20 <Crofton|work> blindvt, somethings things get built to satisfy dependencies but do not get installed
Oct 20 11:54:38 <Crofton|work> when the image is done, look in the testlab directory for details on what is in the image
Oct 20 11:56:40 <blindvt`> i don't want to build coreutils, not even as a dependency. All coreutils are provided by busybox (at least optionally)
Oct 20 12:03:39 <Crofton> blindvt, let the build complete and see what gets in the image
Oct 20 12:03:45 <Crofton> then we can work on correcting problems
Oct 20 12:04:03 <CoRfr> if I want to write a task just before do_configure(), how should I call it ?
Oct 20 12:04:38 <Crofton> look in the gnnuradio recipes
Oct 20 12:04:55 <Crofton> we add a task to build libusb-0.12 there and run it b4 configure
Oct 20 12:07:05 <CoRfr> thx Crofton
Oct 20 12:11:33 <pb__> blindvt: it is probably trying to build the cross version of coreutils (i.e. to run on your build machine)
Oct 20 12:11:40 <pb__> that might be annoying, but it is almost certainly harmless. 
Oct 20 12:11:52 <pb__> I would suggest you just grit your teeth and let it get on with it, it only takes a couple of minutes
Oct 20 12:12:59 <blindvt`> i despise such conceptual brokenness way too much to ignore it
Oct 20 12:13:52 <blindvt`> i could give up on it if it turns out to be too intrusive which it will not be
Oct 20 12:14:09 <pb__> in that case, you should be able to set "ASSUME_PROVIDED += coreutils-native" to defeat that particular thing.
Oct 20 12:15:38 <blindvt`> completely unrelated, but is the commented line in recipes/tasks/task-sugar-sucrose.bb ok? (didn't look how bitbake treats comments ending with a line continuation)
Oct 20 12:16:29 <pb__> mickey|office: did you reply to that guy?  he is on my case again.
Oct 20 12:22:08 <mickey|office> pb__: argh, i think i forgot that. shame on me. will do it when i'm home today, sorry.
Oct 20 12:24:03 <pb__> okay, no problem, I will apologise to him and tell him to stand by
Oct 20 12:24:11 <mickey|office> thanks
Oct 20 12:25:50 <ka6sox> cbrake, ping
Oct 20 12:28:04 <Crofton> ka6sox, might be early for him
Oct 20 12:28:26 <ka6sox> Crofton, I didn't know if he was an early starter or not.
Oct 20 12:28:30 <cbrake> ka6sox: hello
Oct 20 12:28:47 <ka6sox> morning cbrake. did you get my message about the upgrade today?
Oct 20 12:29:03 * cbrake is an early starter, but not always doing development early :-)
Oct 20 12:29:04 <Crofton> cbrake, btw, there is an OEDEM agenda item on hosting
Oct 20 12:29:15 <cbrake> ka6sox: yes, thanks for the notification
Oct 20 12:29:29 <Crofton> ka6sox, I posted something to our list based on your message in irc last night
Oct 20 12:29:49 <ka6sox> it should be a quick down and up. just enough time to install another 16GB of RAM.
Oct 20 12:29:55 <Crofton> yay
Oct 20 12:29:57 <cbrake> ka6sox: sounds good
Oct 20 12:30:08 <ka6sox> okay thanks...
Oct 20 12:30:45 <sgh> Hi
Oct 20 12:30:59 <sgh> How do I add startup scripts i an image ?
Oct 20 12:31:17 <sgh> I mean rc-scripts in init.d
Oct 20 12:32:15 <sgh> Neer mind .... I found it in the manual I think...
Oct 20 12:42:04 <blindvt`> khem, do you mind if i CC you on patchlets for oe (so i can trick you into reviewing and applying them for me ;) or do you prefer to just pick the stuff you want to respond to yourself?
Oct 20 12:43:58 <Crofton> blindvt, just send them to the list, I'm sure he will pick up the pieces he is interested in
Oct 20 12:52:58 <tasslehoff> I'm want to have multiple swimages installed on my beagle? If the OS on Partition-A is active, I want to scp in a OS-package, untar it to Partition-B, and update bootarguments to boot from Partition-B. Anyone seen solutions for this?
Oct 20 12:57:00 <hrw> re
Oct 20 13:03:33 <blindvt`> is there some existing way (apart from adding the extension to be recognized) to rename a SRC_URI uri? like SRC_URI="http://download2.berlios.de/gsmmux/${P}.tb2;replace=(.tb2,tar.bz2)"
Oct 20 13:03:40 <sgh> how do I avoid that package_qa_hash_style is run for at specific recipe ?
Oct 20 13:04:29 <blindvt`> there's uri_replace in the fetcher, but that seems to be used for handling mirrors only, right now
Oct 20 13:07:44 <blindvt`> reasoning is that someone upstream named the gsmmux.tar.bz2 file gsmmux.tb2 and that sounds like it will not work too well out of the box (not that i'd care)
Oct 20 13:13:35 <czr> hmm. using a very old oe, with glibc_2.5bb, I'm having trouble finding a version of libc6.so with debugging symbols
Oct 20 13:13:48 <czr> (a lot of packages have -dbg versions, but can't find one for libc6)
Oct 20 13:19:36 <tasslehoff> I tried starting out with console-base-image.bb, and doing IMAGE_INSTALL += omapfbplay. I seem to have gotten omapfbplay in place, but avcodec_open fails with "[h264 @ 0x22020]codec type or id mismatches". Did I add omapfbplay in a not-so-good way?
Oct 20 13:36:27 <sgh> How can I suppress "No GNU_HASH in the elf binary:"
Oct 20 13:36:52 <florian> sgh: fix LDFLAGS
Oct 20 13:37:22 <sgh> florian: ehm .... its a binary only lib... So I am kind of stuck.
Oct 20 13:38:12 <pb__> You can set INSANE_SKIP_pkg to disable the zecke checks for a particular package. 
Oct 20 13:38:13 <thebohemian> sgh: then support the FSF ;)
Oct 20 13:38:43 <florian> *g*
Oct 20 13:39:02 <pb__> or, alternatively, it is possible at least in theory to postprocess a binary to add the appropriate hash table to it, although I don't know of any existing tool that can do this.
Oct 20 13:39:25 <sgh> thebohemian: sure, but it does not help here... Just having OE warn instead of making the recipe fail would be nice.
Oct 20 13:39:49 <CIA-1> 03Jeremy Lainé <jeremy.laine@m4x.org> 07org.openembedded.dev * ra1444ed29c 10openembedded.git/recipes/linux/linux-2.6.31/boc01/ (008-091005-spi.patch defconfig): linux-2.6.31: update defconfig and SPI patch for boc01
Oct 20 13:39:55 <Crofton|work> what is GNU_HASH?
Oct 20 13:40:05 <thebohemian> sgh: sure, that why there was a ";-)" you should follow pb__ advice using INSANE_SKIP
Oct 20 13:40:28 <thebohemian> Crofton|work: special kind of symbol table
Oct 20 13:40:29 <pb__> but, be warned, the error is there for a reason: you may find that the dynamic linker refuses to load your library if it doesn't have a gnu_hash.
Oct 20 13:41:14 <pb__> depends on what dynamic linker you are using and how it was configured.
Oct 20 13:41:45 <thebohemian> Crofton|work: http://people.redhat.com/drepper/dsohowto.pdf
Oct 20 13:42:09 <sgh> thebohemian: I got it :D ,,,, thanks will try INSANE_SKIP
Oct 20 13:45:46 <Crofton|work> thebohemian, for an annoying guy, he writes ok
Oct 20 13:52:04 <thebohemian> Crofton|work: :) nobody said he's a bad coder ...
Oct 20 13:53:23 <ant_work> he, fellow Gentoo guy did some check about size on disk...http://tinyurl.com/ygq27me
Oct 20 13:55:53 <ant_work> I don't have a separate partition anymore...would be interesting to see if OE gains with smalerl page-size
Oct 20 13:56:27 <mickey|office> thebohemian: report!
Oct 20 13:56:33 <mickey|office> thebohemian: (please send around)
Oct 20 14:00:54 <eFfeM> recalcati: missed your message, i think the page that you gave are devices with that chip, but they are not all routers
Oct 20 14:02:40 <chouimat> morning
Oct 20 14:08:22 <recalcati> eFfeM: yes, I have understood, thx. Anyway, by now, I'll by a general purpose pc and after it I'll integrate in an embedded platfrom such bridging and routing concepts
Oct 20 14:22:48 <hrw> http://www.linuxfordevices.com/c/a/News/Marvell-Armada-100-500-600-and-1000/
Oct 20 14:23:00 <hrw> dual core armv5te sounds strange
Oct 20 14:25:25 <pb_> dual core doesn't necessarily mean SMP, it might just be two independent cores.
Oct 20 14:25:41 <hrw> sure
Oct 20 14:26:03 <pb_> that wouldn't be too unusual, people have been making dual- or triple- core chips ever since armv3.
Oct 20 14:26:27 <hrw> I wonder when ST-Ericsson will get their u8500 (dualcore cortex-a9) available on boards
Oct 20 14:27:02 <pb_> I worked on a project back in 1998 or so which had two arm7tdmis plus a third arm core that I forget the details of.
Oct 20 14:30:15 <eFfeM> hrw, it is the kirkwood core which they duplicated, kirkwood is also what is in sheevaplug
Oct 20 14:31:10 <hrw> eFfeM: I know. I just wonder how it will work
Oct 20 14:33:27 <eFfeM> as pb_ said they could be independent cores, with shared memory access
Oct 20 14:33:50 <eFfeM> actually i worked with something like that back in 81 as graduation project (two LSI 11's with shared mem)
Oct 20 14:34:24 <eFfeM> that would also fit the use case where one cpu e.g. could do blue playback and the other one could do everything else
Oct 20 14:39:23 * mwester recalls the Prime Computer Inc. "P850" -- the Attached Processor Extension.
Oct 20 14:40:19 <mwester> And even before that, IBM had the I/O Processors on the mainframes -- the loosely-coupled processor idea has been around for a long time, and makes a lot of sense for embedded devices.
Oct 20 14:59:38 <CIA-1> 03Steffen Sledz <sledz@dresearch.de> 07org.openembedded.dev * r1a78669beb 10openembedded.git/recipes/linux/linux-2.6.24/hipox/defconfig: linux-2.6.24: enable POSIX message queues for hipox machine
Oct 20 15:03:47 <CIA-1> 03Klaus Kurzmann <mok@fluxnetz.de> 07shr/import * r18db1e48eb 10openembedded.git/ (3 files in 2 dirs): 
Oct 20 15:03:47 <CIA-1> elementary-theme-gry: new recipe
Oct 20 15:03:47 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 15:32:56 <hrw> ~hail Robert for sending GA report
Oct 20 15:32:58 * ibot bows down to Robert for sending GA report and chants, "I'M NOT WORTHY!!"
Oct 20 15:37:54 <cbrake> hrw: heh, so are we finally members?
Oct 20 15:38:04 <hrw> cbrake: officially even
Oct 20 15:38:10 <cbrake> hrw: wow :-)
Oct 20 15:46:57 <mckoan> hrw: what? where?
Oct 20 15:53:52 <mckoan> hrw: ok OE ML
Oct 20 16:12:44 <thesing> morning all
Oct 20 16:13:26 <florian> hi thesing 
Oct 20 16:35:25 <hrw> bye
Oct 20 17:08:26 <khem> blindvt`: I am subscribed to the ml so no worries if your patches are something I can review and apply I will do it 
Oct 20 17:13:00 <Crofton> khem, gm
Oct 20 17:13:28 <khem> Hey Crofton 
Oct 20 17:13:53 <khem> Crofton: will you be at OEDEM 
Oct 20 17:13:58 <Crofton> yeah
Oct 20 17:14:59 <khem> Can you get a OEDEM T-shirt for me :)
Oct 20 17:15:19 <khem> I will mail you the check 
Oct 20 17:15:32 <Crofton> yes
Oct 20 17:15:44 <Crofton> be sure to send size info to pb__ 
Oct 20 17:16:05 <Crofton> I should likely do this for all .us contributors
Oct 20 17:16:12 <khem> cool 
Oct 20 17:16:24 <khem> I wear Medium 
Oct 20 17:29:50 <blindvt> khem, TIA
Oct 20 17:34:08 <blindvt> Dependency loop #1 found: ... do_configure) (depends: Set([88, 282, 92, 22, 87]))
Oct 20 17:34:21 <blindvt> printing a set there isn't too helpful, isn't it? ;)
Oct 20 17:39:24 <CIA-1> 03Frans Meulenbroeks <fransmeulenbroeks@gmail.com> 07org.openembedded.dev * r7bc8adad26 10openembedded.git/ (conf/checksums.ini recipes/samba/samba_3.3.9.bb): samba: added 3.3.9 recipe
Oct 20 17:50:42 <kergoth> blindvt: but it is, actually.  it tells you what tasks those numbers correspond to.  you could certainly make it print a space separated list rather than an str() of  the set, but either way those numbers are what you'll see
Oct 20 18:16:08 <CIA-1> 03Michael Lippautz <michael.lippautz@gmail.com> 07org.openembedded.dev * r4319ab27f4 10openembedded.git/recipes/ralink/rt3070_2.1.1.0.bb: 
Oct 20 18:16:08 <CIA-1> rt3070_2.1.1.0.bb: Install missing etc/Wireless/RT2870STA
Oct 20 18:16:08 <CIA-1> Signed-off-by: Khem Raj <raj.khem@gmail.com>
Oct 20 18:36:29 <blindvt> kergoth, ah. s/Set/Task-ID/ and it would be easier to grok; like (or however that's expressed in python)
Oct 20 18:36:33 <blindvt> -                            msgs.append("  Task %s (%s) (depends: %s)\n" % (dep, self.get_user_idstring(dep), self.runq_depends[dep]))
Oct 20 18:36:33 <blindvt> +                            msgs.append("  Task %s (%s) (dependent Task-IDs %s)\n" % (dep, self.get_user_idstring(dep), [x for x in self.runq_depends[dep]]))
Oct 20 18:44:53 <florian> re
Oct 20 18:50:40 <dth> wb
Oct 20 19:04:30 <CIA-1> 03Klaus Kurzmann <mok@fluxnetz.de> 07shr/import * rd3279013c1 10openembedded.git/recipes/shr/shr-splash-theme-logo_git.bb: 
Oct 20 19:04:30 <CIA-1> shr-splash-theme-logo_git.bb: new recipe
Oct 20 19:04:30 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 19:05:57 <CIA-1> 03Tom <tom@stosb.com> 07shr/import * r75683d9629 10openembedded.git/recipes/shr/ (3 files in 2 dirs): 
Oct 20 19:05:57 <CIA-1> fixed the gtk themes
Oct 20 19:05:57 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 19:13:29 <CIA-1> 03Klaus Kurzmann <mok@fluxnetz.de> 07shr/import * r48257bc2bd 10openembedded.git/conf/distro/include/shr-autorev-unstable.inc: 
Oct 20 19:13:29 <CIA-1> shr-autorev-unstable.inc: set shr-splash-theme-logo to AUTOREV
Oct 20 19:13:29 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 19:59:38 <CIA-1> 03Klaus Kurzmann <mok@fluxnetz.de> 07shr/import * r91e3c2cafc 10openembedded.git/recipes/shr/phoneuid_git.bb: 
Oct 20 19:59:38 <CIA-1> phoneuid_git.bb: install new phoneui-wrapper.sh
Oct 20 19:59:38 <CIA-1> Signed-off-by: Klaus Kurzmann <mok@fluxnetz.de>
Oct 20 20:06:04 <Crofton> cbrake, looks like dns for docs has switched
Oct 20 20:06:21 <Crofton> should I see a directly listing at docs.oe.org
Oct 20 20:06:26 <cbrake> yes
Oct 20 20:06:30 <Crofton> ok
Oct 20 20:07:25 <Crofton> that leaves patchwork and the wiki :)
Oct 20 20:08:04 <cbrake> Crofton: yup.  I updated the nightly doc generation script to point to the new server.
Oct 20 20:09:11 <cbrake> Crofton: I could start looking at the wiki.
Oct 20 20:09:30 <Crofton> if you have time, that would be good
Oct 20 20:09:56 <cbrake> Crofton: last time I checked, the version of Ubuntu on melo has a slightly older packaged version of wikipedia than amethyst
Oct 20 20:09:56 <Crofton> we should get this item closed :)
Oct 20 20:10:05 <cbrake> Crofton: so we need to do some testing first
Oct 20 20:10:31 <cbrake> Crofton: Laibsch insists on using packaged sw when possible, which I agree with
Oct 20 20:10:54 <Crofton> as long as someone else is doing it, and it gets done, I don't care :)
Oct 20 20:12:58 <cbrake> above should have been packaged version of mediawiki on ...
**** ENDING LOGGING AT Tue Oct 20 20:16:14 2009

**** BEGIN LOGGING AT Tue Oct 20 22:16:15 2009
Oct 20 22:17:41 <otavio> woglinde: thx by the hint
Oct 20 22:18:45 <ant_home> ok, git can pull again :)
Oct 20 22:20:08 <otavio> ant_home: thx :-D hahah
Oct 20 22:44:34 <ka6sox> anybody please check the oe webserver to see that its back up?
Oct 20 22:45:34 <Crofton|work> webserver is elsewhere :)
Oct 20 22:45:48 <ka6sox> k
Oct 20 22:45:52 <Crofton|work> however cgit is up
Oct 20 22:46:09 <ka6sox> k
Oct 20 22:46:09 <ka6sox> thanks
Oct 20 22:46:14 <Crofton|work> git looks up
Oct 20 22:46:15 <Crofton|work> thanks
**** ENDING LOGGING AT Tue Oct 20 22:56:54 2009

**** BEGIN LOGGING AT Tue Oct 20 23:36:15 2009
Oct 21 00:57:55 <kg4ysn> anyone experienced a "command variables / bad substitution" error with bitbake 1.9 and the latest OE tree?
Oct 21 01:06:59 <grg> opkg makes me cry
Oct 21 01:07:26 <grg>   opkg = calloc (1, sizeof (opkg_t));
Oct 21 01:07:27 <grg>   opkg->args = calloc (1, sizeof (args_t));
Oct 21 01:08:15 <grg> segfault central
Oct 21 01:13:05 <mwester> opkg is undoubtedly the poorest-written package in OE, yet it is one of the most critical packages for OE to work.
Oct 21 01:14:22 <mwester> About a year ago I had access to a commercial software analysis tool, and I fed it opkg -- the results where amazing... the diagnostics were larger than the source code itself. :D
Oct 21 01:15:44 <mwester> Unfortunately, I didn't have access to actually do anything to fix the issues. :(
Oct 21 01:30:54 <martinmeba> I have a OE/Python question.  I was using the dev branch but was forced to switch to the stable branch.  When using the dev branch, I made a recipe for python and python-modules to build what would be a desktop equivalent install of python
Oct 21 01:31:21 <martinmeba> This does not work on the stable branch.  How can I get the image to include all of the python modules?
Oct 21 01:38:53 <martinmeba> and by not work, bitbake throws an error that there are no providers for python-modules
Oct 21 01:59:50 <Crofton> martinmeba, there may be python updates that have not been backported to stable
Oct 21 02:01:16 <Crofton> mwester, can you share the output?
Oct 21 02:02:10 <Crofton> martinmeba, remind me of this in the morning an dmaybe I can poke around some
Oct 21 02:07:21 <martinmeba> Crofton: that would be great - thanks
**** ENDING LOGGING AT Wed Oct 21 02:59:59 2009