**** BEGIN LOGGING AT Mon Sep 06 02:59:57 2010 Sep 06 03:18:06 grg: your patch for gdb looks ok Sep 06 03:23:26 khem, excellent. I was slightly concerned that there was a lib64 dir there at all... in the meta toolchain tarball that results, libiberty.a is the only thing in that lib64/ dir, everything else is in lib/ Sep 06 03:39:27 grg: yeah someone needed it Sep 06 03:39:30 grg: Sep 06 03:39:35 I have to dig the ml Sep 06 03:39:51 * khem is watching US open Sep 06 03:40:20 khem, don't get up off the couch if you don't want to. I can live with not knowing about the lib64 dir :) Sep 06 04:39:53 03Jason Kridner  07org.openembedded.dev * r53382f4aa2 10openembedded.git/recipes/esc/esc-node-demo_git.bb: Sep 06 04:39:53 esc-node-demo: add missing index.html file Sep 06 04:39:53 Signed-off-by: Jason Kridner Sep 06 04:39:53 Signed-off-by: Khem Raj Sep 06 04:39:57 03Martin Jansa  07org.openembedded.dev * rc6a72ee2b9 10openembedded.git/recipes/matchbox2/matchbox-panel-2-icon-themes_0.0.1.bb: Sep 06 04:39:57 matchbox-panel-2-icon-themes: use same SRCREV as matchbox-panel-2 Sep 06 04:39:57 * battery/data was removed in 1919 http://svn.o-hand.com/view/matchbox/trunk/matchbox-panel-2/applets/battery/Makefile.am?rev=1919&view=auto Sep 06 04:39:57 * prevent downgrade/upgrade cycle in matchbox-panel-2 checkout ending with (svn: Failed to add directory 'applets': an unversioned directory of the same name already exists) Sep 06 04:39:57 Signed-off-by: Khem Raj Sep 06 04:41:44 03Anatoly Shipitsin  07org.openembedded.dev * r1540070798 10openembedded.git/recipes/sysvinit/sysvinit_2.86.bb: (log message trimmed) Sep 06 04:41:44 sysvinit_2.86.bb: change sysvinit-utils for compatible with upstart Sep 06 04:41:44 * move shutdown and runlevel to sysvinit package to halt and poweroff, upstart include this utils Sep 06 04:41:44 * move killall5 to sysvinit-utils, it used by scripts when system goes shutdown or halt, upstart not provide it Sep 06 04:41:45 * change PR = "r59" Sep 06 04:41:45 Signed-off-by: Anatoly Shipitsin Sep 06 04:41:45 Reviewed-by: Paul Menzel Sep 06 05:33:49 Hi all Sep 06 05:40:54 I need to clarify one thing Sep 06 05:41:23 when building native packages is there any scenarion that some taget packages are also being build Sep 06 05:41:54 My understanding is this that target packagaes will not build when we are building native package Sep 06 05:42:01 please correct me if I am wring Sep 06 05:42:04 wrong Sep 06 05:49:43 Noor, that is correct Sep 06 05:50:20 I am building seeing the ooposite in the llvm-native.inc file Sep 06 05:50:52 this file is included in llvm2.7-native_2.7.bb Sep 06 05:51:16 and in the depends list binutils-cross is there Sep 06 05:52:54 Noor, i suppose there is either a very good reason for that, or it is in error. Dunno, sorry. Sep 06 05:54:45 grg: OK. Thanks . I can see the same behaviour when I build sip-native. it also builds alot of target packages. But anyways thanks for ur reply. Sep 06 06:09:58 Hi eFfeM_work Sep 06 07:13:40 good morning Sep 06 07:34:14 hej , can I somehow access raw LDFLAGS? ld doesn't like the wrapped ones (and the Makefile that is used adds some other flags to the ld command), Sep 06 07:42:13 gm everyone Sep 06 07:53:20 morning Sep 06 07:56:29 good morning all Sep 06 07:56:54 HI everybody Sep 06 07:58:54 ka6sox, any news on the network issues (been using repo.or.cz for read only pulls since friday) Sep 06 07:59:30 morning Sep 06 07:59:46 can't find anything here on this side of the pond.... Sep 06 08:00:21 noor, bitbake -g sip-native will give you two .dot files in your working dir that explain all dependencies Sep 06 08:00:50 ka6sox strange, I thought you also got the low bitrates from your place so i kinda assumed the problem would be in CA Sep 06 08:01:02 , could you give me a couple of trace routes? Sep 06 08:01:12 yes Sep 06 08:02:34 s/yes/no/ Sep 06 08:02:46 oops Sep 06 08:02:52 our firewall or so seems to block all, I only get * * * Sep 06 08:03:11 I can try from home tonight Sep 06 08:03:20 or i need to talk to our sysop Sep 06 08:03:48 okay whenever you get me something I'll look @ it Sep 06 08:03:54 ok Sep 06 08:04:30 sysop is not online, will try to see him later today Sep 06 08:05:33 new to oe, but the build seems to detect bash in ubuntu as dash, is this a known issue? Sep 06 08:05:52 disabled sanity checking to get around it, but a little worried about the error message Sep 06 08:07:25 foobar___: it does not detect bash as dash it detects that /bin/sh is dash, not bash Sep 06 08:07:47 oh, i see Sep 06 08:07:55 if you are root you can just change the symlink for /bin/sh, guess there are other more decent ways Sep 06 08:08:25 and if you do not have root access you could copy dash to e.g. ~/bin/sh and put ~/bin first in your searchpath Sep 06 08:08:36 nah, got it Sep 06 08:08:37 thanks Sep 06 08:09:03 eFfeM_work: I have built llvm2.7-naitve successfully after removing binutils-cross dependency from inc file Sep 06 08:10:18 just curious why binutils-cross was added in llvm-native dependency Sep 06 08:11:51 so, how often do these 550 step "bitbake gcc-cross" things fail in the middle of the night? ;) Sep 06 08:12:07 Noor: rebuilding after removing a dep does not tell anything on whether the dep is needed or not, if the binutils-cross package is there the build will find the data it needs Sep 06 08:12:17 and it will still build Sep 06 08:12:40 I have cleared it as well Sep 06 08:12:51 cleaned Sep 06 08:13:01 cleaned gcc-cross ? Sep 06 08:13:08 or llvm Sep 06 08:13:26 foobar___: does not fail here Sep 06 08:14:01 :) Sep 06 08:14:22 kudos on the project, it seems like this is gonna be really useful as a generic framework for cross-compiling stuff Sep 06 08:15:05 first i cleaned binutils-cross Sep 06 08:15:10 then cleaned llvm Sep 06 08:15:22 remove binutils-cross depoendcy Sep 06 08:15:32 and rebuilt llvm-native Sep 06 08:15:42 ah ok, that should be safe (assuming binutils-cross is not rebuild) Sep 06 08:15:50 yeah Sep 06 08:16:37 foobar___: it is once you get the hold of it, the learning curve is sometimes a little bit steep, but the reward is substantial Sep 06 08:16:41 but ofc I am biased Sep 06 08:18:34 eFfeM_work: if binutils-cross is required the that build should break. isnt it Sep 06 08:20:09 yes, if you cleaned it first it should Sep 06 08:20:28 good catch! Sep 06 08:20:41 please submit patch Sep 06 08:21:50 while looking at the log when this change was added the commit message says Sep 06 08:22:25 llvm.inc: Accomodate to changes made to cmake.bbclass which broke the LLVM builds. Sep 06 08:22:30 llvm-native.inc: Dito. Sep 06 08:22:51 from what date is that one ? Sep 06 08:23:22 Apr 30 2010 Sep 06 08:23:33 ah ok, so fairly recent Sep 06 08:23:41 have link to the patch ? Sep 06 08:23:42 yeah Sep 06 08:24:16 6da4ed714f269373271a14149b131bf1ff7a6def Sep 06 08:27:57 got it, maybe ask on the list with a cc to Robert Schuster who authored the patch, it is needed since you are referring to Sep 06 08:28:26 this does not look good I'd say, it might depend on binutils-native, but on cross Sep 06 08:28:31 but please ask on the list Sep 06 08:28:57 OK Sep 06 08:42:01 eFfem_work send me email when you have information. Sep 06 08:54:53 ok, will do Sep 06 08:55:30 likewise: saw you added a 1020 box, did gcc4.5 work for you? i have problems with busybox Sep 06 08:57:45 eFfeM_work: you using angstrom or minimal? Sep 06 08:57:58 eFfeM_work: I can do a few tests, just would like to know what you changed to move to 4.5 Sep 06 08:58:29 eFfeM_work: The box is leaving me in an hour (for a few days), so I have too little time to test. Sep 06 08:58:50 eFfeM_work: but will mark that on my todo (as always....) Sep 06 08:58:59 likewise: hi Sep 06 08:59:09 mckoan: hi there Sep 06 09:01:54 minimal Sep 06 09:02:04 eh that was: Sep 06 09:02:18 likewise, I use minimal, had problems with busybox and urjtag Sep 06 09:02:51 i have urtjag binaries, size change is remarkably Sep 06 09:03:53 -rwxr-xr-x 1 root root 15972 2010-09-03 15:05 jtag_4.3.3 Sep 06 09:03:54 -rwxr-xr-x 1 root root 10748 2010-09-03 15:05 jtag_4.5 Sep 06 09:04:02 the gain of 4.5 is too big to look good Sep 06 09:04:08 afk Sep 06 09:07:51 eFfeM_work: the linker might be throwing more unused things out, but that's speculation on my part and needs further inspection. Sep 06 09:08:05 hi Jin^eLD Sep 06 09:08:27 hey likewise Sep 06 09:29:43 likewise: didn't really have time to look into it, I can compare the symbol tables and fuction sizes Sep 06 10:20:21 ls Sep 06 10:28:31 good morning Sep 06 10:28:43 eFfeM_work: wrong window ;) Sep 06 10:29:00 what? nothing to list here ? Sep 06 10:29:04 & gm florian Sep 06 10:32:32 eFfeM_work: then it was the wrong command... /who * is what you might want to use in oder to list things in here :-) Sep 06 10:33:15 no, the window still had focus, don't use focus follows cursor, ofc thought I was typing elsewhere Sep 06 10:37:45 florian: gm (maybe I should make that a Unix command too) Sep 06 10:38:02 The program 'gm' is currently not installed. You can install it by typing: Sep 06 10:38:04 sudo apt-get install graphicsmagick Sep 06 10:38:08 hmm, nm Sep 06 10:38:10 likewise, like gnu make? Sep 06 10:38:25 leon@sideways:~/sandbox/ccm/topha/openembedded$ nm Sep 06 10:38:25 (shortcut for gnu make) Sep 06 10:38:26 nm: 'a.out': No such file Sep 06 10:38:31 lol Sep 06 10:38:39 nm = never mind Sep 06 10:38:46 ? Sep 06 10:38:52 leon@sideways:~/sandbox/ccm/topha/openembedded$ lol Sep 06 10:38:54 No command 'lol' found, did you mean: Sep 06 10:38:55 Command 'col' from package 'bsdmainutils' (main) Sep 06 10:38:57 Command 'sol' from package 'aisleriot' (main) Sep 06 10:39:00 nm = never mind, indeed Sep 06 10:39:50 heh Sep 06 10:39:54 hi likewise Sep 06 11:47:50 hi, i've got a problem: Sep 06 11:47:50 if I boot a image it will hang at the end of rcS (after the last one) and wont continue with next rc5. (Ive checked this by placing markes in scripts). Sep 06 11:47:51 in /etc/init.d/rcS it would run scripts in /etc/rc.boot if any, but this never happens. so, it hangs somewhere in /etc/init.d/rc I think. Sep 06 11:47:54 distribution: angstrom, machine: looxn420/pxe270, images: base-image, console-image, ... Sep 06 11:47:55 no more special modifications. Sep 06 11:48:00 does anyone have a idea? Sep 06 11:59:47 dnjl: this seems more a question for the angstrom channel Sep 06 12:00:09 #oe is not a distro support channel Sep 06 12:09:14 eFfeM_work: k Sep 06 12:23:08 the keyword for ${datadir} expands to /usr/share, i was wondering if there is one for /usr/local/share ? Sep 06 12:23:22 anyone? Sep 06 12:23:26 turanmo: there is no /usr/local/ in oe packages Sep 06 12:24:51 if i need to install any binary into /usr/local/sbin then, i must specify the entire path name in my do_install() ? Sep 06 12:25:56 generally, your DISTRO would set a variable to the path and you would then use that in your do_install(). Sep 06 12:26:01 why do you want to install into /usr/local/sbin specifically? Sep 06 12:26:18 a binary called twprint Sep 06 12:26:35 and several others for that matter Sep 06 12:27:02 right, but why? Sep 06 12:27:51 pb_: because the application install script (install.sh) does so Sep 06 12:28:21 thought the rigth way to install would be to follow what the install script does? Sep 06 12:28:45 and add the same functionality into do_install() Sep 06 12:29:26 hope i am making sense, i am new to all this Sep 06 12:42:17 while converting to new style staging is it wise to use autotools_do_install function in place of do_stage which is only calling autotools_stage_includes function Sep 06 12:43:27 Noor: that's for me? Sep 06 12:43:45 No I am asking my own question Sep 06 12:43:50 :) Sep 06 12:43:55 Noor: ah! ok :) Sep 06 12:44:39 noor, i think so, but i am not an expert Sep 06 12:44:51 if there is no additional install needed you can evencompletely omit it Sep 06 12:45:40 I looked around but there is not function which installs include dir in ${D} Sep 06 12:45:52 ah ok Sep 06 12:48:50 Noor: autotools_do_install is default value of do_install() when 'inherit autotools' is used Sep 06 12:51:58 hrw: yeah, I know. In one recipe do_stage is doing autotools_stage_includes. this recipe is inherited from autotools. so if we remove do_stage to convert it to new style staging I think the only possibility is to use autotools_do_install by not defining any do_install function in the recipe Sep 06 12:55:37 or we should define some funtions just like autotools_stage_includes. That function might be autotools_install_includes to get exact ipk files Sep 06 13:29:02 Hej! Our Makefiles use $(git describe) to build version numbers from git tags while compiling. This doesn't work while Sep 06 13:29:02 crosscompiling because S = ${WORKDIR}/git isn't a git repository. Sep 06 13:29:11 Is there a trick to get it working? Sep 06 13:35:30 Uh, that doesn't seem to follow... Sep 06 13:48:07 I beg your pardon? Sep 06 13:49:29 Your problem is entirely orthogonal to cross compilation. Sep 06 13:49:40 hej , anyone mind looking at http://patchwork.openembedded.org/patch/2865/ ; it's a stopper for igep0020 machine Sep 06 13:52:13 broonie: entirely? I can create a SDK, cross-compile the program locally and build the ipk from the binary. This would work, it's just not automatic Sep 06 13:53:05 vps: It's not the cross compilation that's got you into trouble, it's the out of tree build. Sep 06 13:55:37 Yes, that's the problem. Now how to solve it? I could pass the version from the recipe. Feels hacky Sep 06 13:58:37 03Koen Kooi  07org.openembedded.dev * rfdbbbd91ab 10openembedded.git/recipes/linux/multi-kernel.inc: multi-kernel: work around do_unpack breakage a bit Sep 06 13:59:28 The most obvious thing would seem to be do do releases to tarball. Sep 06 14:04:46 Hmm, yes, tarballs. We will see. Sep 06 14:48:26 meh, I'm not good at writing recipes. any idea what sort of obvious screwup would cause this: Sep 06 14:48:52 | LD [M] /mnt/space/openaos/build/angstrom-dev/work/archos5it-angstrom-linux-gnueabi/psfreedom-git-r46/git/psfreedom.ko Sep 06 14:48:56 | arm-angstrom-linux-gnueabi-ld: unrecognized option '-Wl,-rpath-link,/mnt/space/openaos/build/angstrom-dev/staging/armv7a-angstrom-linux-gnueabi/usr/lib' Sep 06 14:50:04 you are passing a wrapped LDFLAGS to ld Sep 06 14:50:17 '-Wl,...' is meant for gcc Sep 06 14:50:52 ok Sep 06 14:53:24 hmmm LD=${KERNEL_CC} makes it compile, but this looks like bad style to me. should I approach this from some other side? Sep 06 15:05:31 dm8tbr, kernel Sep 06 15:05:37 dm8tbr, LDFLAGS="" ? Sep 06 15:05:57 dm8tbr, I usually do that in devshell: Sep 06 15:06:24 make ARCH=arm CROSS_COMPILE=arm-oe-linux-gnueabi- LDFLAGS="" Sep 06 15:06:29 for compiling a kenrel Sep 06 15:06:42 the LDFLAGS fixes the issues Sep 06 15:12:50 GNUtoo|laptop: I'm writing a recipe for psfreedom Sep 06 15:13:15 which is a kernel module based on gadget/musb Sep 06 15:13:38 I don't know what to inherit for kenrel modules Sep 06 15:13:43 GNUtoo|laptop: I'll try to do that in the recipe Sep 06 15:13:45 but you could look at exisiting ones Sep 06 15:14:20 there is something I inherited already for modules. but no idea if I'm doing it right. will just clean it up a bit and give it to koen :) Sep 06 15:16:25 ok Sep 06 15:16:27 let me look Sep 06 15:16:35 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rb28081b603 10openembedded.git/recipes/mplayer/ (mplayer-common.bb mplayer-common/htcdream/mplayer.conf): Sep 06 15:16:35 mplayer-common: fix buffer underruns with -ao sdl on htcdream Sep 06 15:16:36 Using -ao sdl on the htcdream fixes the buffer underruns with mplayer Sep 06 15:16:36 Without that a movie or big buck bunny can't be viewed,and audio Sep 06 15:16:36 suffers from bad quality. Sep 06 15:16:36 A comment that informs the user about -vo dga was added Sep 06 15:16:37 PR was bumped Sep 06 15:16:37 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r60f43a7db7 10openembedded.git/recipes/git/git-native_1.7.0.2.bb: (log message trimmed) Sep 06 15:16:38 git-native: bump PR (libssl issue) Sep 06 15:16:38 git-native failled to run because of libssl changes. Sep 06 15:16:39 bumping PR solve that. Sep 06 15:16:39 Note that if autorev is used,the user has to bitbke -b git-native Sep 06 15:16:40 (to bypass the parsing) else: Sep 06 15:16:41 it start parsing the recipes Sep 06 15:16:41 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r3aad2d9e2f 10openembedded.git/recipes/xkeyboard-config/xkeyboard-config_1.8.bb: Sep 06 15:16:42 xkeyboard-config: add 1.8 version Sep 06 15:17:14 dos1, ^^^ Sep 06 15:19:59 dm8tbr, inherit module Sep 06 15:20:12 try that Sep 06 15:30:30 GNUtoo|laptop: I'm inheriting that, yes Sep 06 15:30:56 tem brasileiro aqui ? Sep 06 16:13:47 03Eric BENARD  07org.openembedded.dev * r5fa3d15364 10openembedded.git/site/arm-common: (log message trimmed) Sep 06 16:13:47 postgresql-8.1.8 : fix configure Sep 06 16:13:47 without this patch, using angstrom 2010.x on armv6 I actually get : Sep 06 16:13:47 | checking alignment of short... configure: error: Sep 06 16:13:47 cannot compute alignment of short, 77 Sep 06 16:13:48 | ERROR: Function do_configure failed Sep 06 16:13:48 NOTE: package postgresql-8.1.8-r4: task do_configure: Failed Sep 06 16:39:45 is there a new testing branch? Sep 06 16:42:56 03Simon Busch  07org.openembedded.dev * re1bed8d165 10openembedded.git/recipes/ (5 files in 2 dirs): (log message trimmed) Sep 06 16:42:56 msmcomm: split msmcommd recipe into three parts as upstream layout has changed Sep 06 16:42:56 We have now three parts for the msmcomm daemon: Sep 06 16:42:56 * libmsmcomm: the actual protocol implementation of the msmcomm protocol Sep 06 16:42:56 * msmcommd: the daemon which manages the link layer + takes command from the clients and Sep 06 16:42:56 submits them to the modem Sep 06 16:42:57 * msmcommd-specs: provides the dbus API definition for the msmcomm daemon which lets the Sep 06 16:42:57 03Simon Busch  07org.openembedded.dev * r4d97fc0a0a 10openembedded.git/conf/distro/include/fso-autorev.inc: Sep 06 16:42:58 fso-autorev: add msmcommd-specs and libmsmcomm Sep 06 16:42:58 Signed-off-by: Simon Busch Sep 06 16:42:59 03Simon Busch  07org.openembedded.dev * r65106eba3e 10openembedded.git/recipes/freesmartphone/libfsotransport_git.bb: Sep 06 16:42:59 libfsotransport: add palmpre specific configure option Sep 06 16:43:00 On the palmpre machine we need the HsuartTransport libfsotransport provides but only Sep 06 16:43:00 offeres with an additional configure option. Sep 06 16:43:01 Signed-off-by: Simon Busch Sep 06 16:43:02 03Simon Busch  07org.openembedded.dev * r1e3df2fdc2 10openembedded.git/recipes/freesmartphone/fsogsmd_git.bb: Sep 06 17:05:24 eFfeM any luck with tracing? Sep 06 17:13:50 didn't get to it will do so now Sep 06 17:15:13 ka6sox: http://www.pastie.org/1141837 Sep 06 17:15:43 last I can make sense of is Seattle Sep 06 17:15:54 guess we're blocked by some windows machine Sep 06 17:18:04 okay that means 80% of the connection is over Level3 Sep 06 17:18:34 okay this is good to know...the washington-netherlands hop is the one I'll watch Sep 06 17:27:29 #2 and #3 are within my isp's network (up/chello) Sep 06 20:22:48 ericben one more thing wrt the asterisk patch: who does deliver that tinfo stuff? should there be something in DEPENDS ? Sep 06 20:23:18 eFfeM: it's ncurses Sep 06 20:24:18 and yes it should be added to DEPENDS (I misread it because I was quite sure I saw it there !) Sep 06 21:10:06 ericben: do you have commit rights ? Sep 06 21:13:39 eFfem: yes Sep 06 21:14:30 I will push asterisk once Arigead tell us it works on its target Sep 06 21:15:16 ok, thought you had commit rights but wanted to make sure that it didn't remain dangling at this point Sep 06 21:15:37 calling it a day! nite all Sep 06 21:15:49 bye eFfeM Sep 06 21:16:25 03Eric BĂ©nard  07org.openembedded.dev * r27c47106ed 10openembedded.git/lib/oe/unpack.py: (log message trimmed) Sep 06 21:16:25 lib/oe/unpack.py: fix uncompressing .gz .bz2 or .xz Sep 06 21:16:25 * Actually, the uncompressed file is written into the source Sep 06 21:16:25 path instead of the destination path. Sep 06 21:16:25 * By also removing the source path, this patch should fix this problem. Sep 06 21:16:26 * An other fix could be to not uncompress to stdout but let gunzip|bunzip2| Sep 06 21:16:27 xz uncompress directly to the file. Sep 06 21:41:56 khem: ping Sep 06 21:52:16 03Leon Woestenberg  07org.openembedded.dev * re0e7d53a3f 10openembedded.git/recipes/ltp/ltp_20100831.bb: Sep 06 21:52:16 ltp-20100831: New release of the Linux Test Project. Sep 06 21:52:17 Not run-tested yet. Specific targets may need target-dependent patches. Sep 06 21:52:17 Signed-off-by: Leon Woestenberg Sep 06 22:18:10 hey Sep 06 22:18:28 how to add param to configure qt/e ? Sep 06 22:19:44 angelox_123: modify recipes/qte/qte-common_2.3.10.inc Sep 06 22:19:49 look for EXTRA_OECONF Sep 06 22:20:23 and to change original param's ? Sep 06 22:21:41 done Sep 06 22:21:43 angelox_123: well, for testing you should just be able to edit what's there... Sep 06 22:21:46 i changed..ghow to re-compile Sep 06 22:21:54 how* Sep 06 22:22:11 well I would recommend bumping PR in qte-mt_2.3.10.bb Sep 06 22:22:17 then you can just bitbake opie-image Sep 06 22:22:49 what i do? Sep 06 22:23:34 angelox_123: edit qte-mt_2.3.10.bb and increase the PR number... it's r29 now, make it r30 Sep 06 22:23:59 ah. it's revision number correct? Sep 06 22:24:20 yep Sep 06 22:24:38 then just do bitbake opie-image ? Sep 06 22:24:56 yep that should do it Sep 06 22:25:33 thank you Sep 06 22:26:45 np Sep 07 00:49:04 how to unpack IPK file? Sep 07 00:49:29 it's an 'ar' archive with two tar.gz files inside. Sep 07 00:49:31 ar x foo.ipk Sep 07 00:49:39 it's the same file format as .deb Sep 07 00:50:02 oh thanks! Sep 07 00:50:36 np Sep 07 00:52:58 what is debian-binary file? Sep 07 01:02:03 it's just part of the deb file format. you should be able to find the documentation on the format in the debian docs somewhere Sep 07 01:02:23 ok. thanks Sep 07 01:02:28 * kergoth doesn't recall offhand, might be the version of the format or something Sep 07 01:07:36 http://www.debian.org/doc/debian-policy/ Sep 07 01:07:46 that has info on all the .deb stuff Sep 07 01:07:53 cheers Sep 07 01:08:36 .ipk is only a subset of the control files Sep 07 01:09:51 :) thanks for digging up the link **** ENDING LOGGING AT Tue Sep 07 02:59:57 2010