**** BEGIN LOGGING AT Sun Feb 06 02:59:57 2011 Feb 06 04:10:05 * kergoth tests bitbake tweaked to use color for debug/error/warning/fatal rather than prefixing with levelname Feb 06 04:19:15 http://grab.by/8Nwp .. hmmmm Feb 06 05:27:54 hi Feb 06 05:27:58 I am trying build OE Feb 06 05:28:01 on ubuntu Feb 06 05:28:11 I am facing some issues Feb 06 05:28:30 I am not able to see the binary imagesw Feb 06 05:28:38 can someone help me? Feb 06 06:43:56 exit Feb 06 06:44:04 #c Feb 06 06:44:15 exit Feb 06 08:37:57 XorA|gone, I don't understand what the wiki is doing to you...need more information Feb 06 09:28:29 morning Feb 06 10:45:25 hi ericben Feb 06 13:02:20 he stefan schmidt Feb 06 13:02:23 hi klgimer Feb 06 13:02:55 hi woglinde Feb 06 13:03:10 stefan I fixed the libtool problem Feb 06 13:03:38 woglinde: yeah, I've seen the commit Feb 06 13:03:41 woglinde: great Feb 06 13:04:03 and I tested cacao-hg Feb 06 13:04:10 and thats working Feb 06 13:04:25 woglinde: what does this buy us? Feb 06 13:04:37 so we can build jamvm-initial -> jamvm-native -> cacoa-hg-navtive Feb 06 13:04:46 only one package which can fail Feb 06 13:04:58 ah, thats good :) Feb 06 13:42:18 WARNING: oestats: error sending task (exception , value HTTP Error 500: Internal Server Error), disabling stats Feb 06 13:42:23 hehe Feb 06 13:43:33 hehe Feb 06 17:17:36 ls Feb 06 17:18:53 ls nof found Feb 06 17:20:51 nof not found\nfound not found Feb 06 17:24:00 hi jama Feb 06 17:24:09 jama did you openjdk build in the meantime? Feb 06 17:28:52 I've tried but was hit by some strange icedtea-native issue (probably not related to icedtea itself) Feb 06 17:30:03 woglinde: what about that angstrom-jalimo.inc? can you integrate still needed parts to sane-toolchain-java? Feb 06 17:32:52 woglinde: retrying, will ping you with tinderbox log Feb 06 17:38:34 woglinde: http://tinderbox.openembedded.org/builds/122635/ Feb 06 17:38:49 hm icedtea worked here nice Feb 06 17:39:12 but maybee its jamvm-native again Feb 06 17:39:29 do you have 64b host? Feb 06 17:39:45 only at home Feb 06 17:39:48 just build again Feb 06 17:41:07 what the hell Feb 06 17:41:47 libc6-dev: /usr/include/gnu/stubs.h Feb 06 17:49:44 woglinde: that's most likely host issue as I've droped multilib about a week ago, so maybe stubs.h still thinks I have 32b headers Feb 06 17:57:19 Jay7, any progress? Feb 06 18:12:44 03Eric Bénard  07org.openembedded.dev * ra4108af45c 10openembedded.git/recipes/busybox/busybox_1.18.2.bb: Feb 06 18:12:44 busybox-1.18.2: add latest fixes Feb 06 18:12:44 modprobe : revert commit 1396221d5a741ef8e1e8abca88836b341a3cab84 Feb 06 18:12:44 bb git commit : 8ae386bf195c6ea53232bacd2eb8cf26676962e4 Feb 06 18:12:45 libbb: spawn should remove child which failed to exec Feb 06 18:12:45 bb git commit : c5d0f15dbd8087f60b377fa9fc18d08698429189 Feb 06 18:12:46 Signed-off-by: Eric Bénard Feb 06 19:17:42 woglinde: looks like icedtea-native starts building with defined __x86_64__ and later gets __i386__ or something, because icedtea6-1.7.6/generated/sun/awt/X11/generator/sizer.32.c is already wrong and sizer.64.c should be used instead Feb 06 19:18:02 jama ah Feb 06 19:18:16 hm we fixed this for crosscompile Feb 06 19:18:30 woglinde: and this in the end makes bits/wordsize.h thinks it's doing 32b build and include non-existent gnu/stubs-32.h instead gnu/stubs-64.h Feb 06 19:18:33 and never yeat occurs on the native build on 64bit Feb 06 19:18:57 or maybee we didnt saw of 32bit compat Feb 06 19:18:58 it worked fine few months agao Feb 06 19:19:07 on 64b host Feb 06 19:25:09 woglinde: log.do_configure is full of rejected patches, is it normal? Feb 06 19:27:29 jama hm normaly not Feb 06 19:27:32 it should break Feb 06 19:27:41 rick-home: yeah.. Feb 06 19:27:44 WARNING: oestats: error sending task (exception , value HTTP Error 500: Internal Server Error), disabling stats Feb 06 19:28:54 woglinde: ok, I'll check again tomorrow, now gtg.. thanks Feb 06 19:29:06 bye jama Feb 06 19:29:08 I will check too Feb 06 19:29:10 at work Feb 06 19:33:33 anyway, I'll commit fix Feb 06 19:33:50 at least it will work as before Feb 06 19:33:56 may be even better sometimes Feb 06 19:37:24 Jay7, excellent Feb 06 19:44:45 03Mike Westerhof  07org.openembedded.dev * r42136288cb 10openembedded.git/recipes/initscripts/ (initscripts-1.0/slugos/reboot initscripts-slugos_1.0.bb): Feb 06 19:44:46 SlugOS: initscripts - simplify kexec handling in reboot, fix option ordering. Feb 06 19:44:46 At some point, kexec became sensitive to the order of the --exec and --force Feb 06 19:44:46 options, so kexec was no longer booting a previously-loaded kernel at reboot. Feb 06 19:44:50 should I bump up recipes including an .inc file in a patch modifying this .inc file? Feb 06 19:44:58 I mean PR Feb 06 19:47:15 "landley: RT @Pres_Bartlet: Hey Republicans, you know that bit in the Bible about worshiping false idols? I'm pretty sure that includes Reagan as well" Feb 06 19:47:17 * kergoth chuckles Feb 06 19:48:44 he might get shoot in texas Feb 06 19:49:17 That's true. It would really tick off God to see people NOT worshipping his Son, Obama. Feb 06 19:49:18 woglinde, that was a Democrat... Feb 06 19:50:12 got is vaporware Feb 06 19:50:17 s/got/god/ :P Feb 06 19:50:39 but Duke Nukem Forever is no longer. Feb 06 19:50:50 so, any suggestions about bumping up PR in case of changes in .inc files? Feb 06 19:51:24 03Yuri Bushmelev  07master * rda89a83b1d 10openembedded.git/classes/oestats-client.bbclass: Feb 06 19:51:24 oestats-client.bbclass: Fix oestats-client Feb 06 19:51:24 * Relax empty data checking (replace asserts) Feb 06 19:51:24 * Replace direct path manipulations with os.path.join Feb 06 19:51:24 * Replace bb.note() with bb.warn() where appropriate Feb 06 19:51:24 * Improve diagnostic messages Feb 06 19:53:08 well.. I've pushed oestats-client fix Feb 06 19:53:10 good job jay7 Feb 06 19:53:56 * Jay7 hope that kergoth will have time to prototype new oestats-client ;) Feb 06 19:55:08 kergoth: I've changed bb.note() to bb.warn() because new bitbake have disabled bb.note() from tasks Feb 06 19:55:42 if they're really warnings, that's appropriate Feb 06 19:55:58 well.. that messages about disabling oestats :) Feb 06 19:56:11 I've thought about bb.error() :) Feb 06 19:56:29 but fear that it may break build Feb 06 19:56:55 only warn Feb 06 19:56:57 please Feb 06 19:57:02 it is now Feb 06 19:57:13 user can decide to kill the build Feb 06 19:59:24 error is a bit overused Feb 06 20:01:25 well.. let's remove tinderclient as well then.. Feb 06 20:01:38 I'll RFC it Feb 06 20:05:00 oh.. my testbuilder is using tinderbox.oe.net instead of oe.org Feb 06 20:05:29 eFfeM's testscript.sh have .net as well Feb 06 20:05:31 03Mike Westerhof  07org.openembedded.dev * r2529fdc41f 10openembedded.git/ (3 files in 3 dirs): Feb 06 20:05:32 SlugOS: kernel -- make 2.6.37 the preferred kernel for SlugOS Feb 06 20:05:32 and introduce the SlugOS linux-ixp4xx kernel recipe to make it so. Feb 06 20:05:32 (preference set to -1 so that nothing picks this up until ready.) Feb 06 20:05:35 Wheee! Feb 06 20:06:38 the lowly pre is *still* on 2.6.24 ;( Feb 06 20:09:21 Yeah, but at this point the NSLU2 has all the patches pushed upstream. There's only one patch -- and that's actually for the silly via-velocity GB ethernet in the DSM-G600, not the NSLU2. Feb 06 20:09:33 So it was pretty easy to get this to happen. Feb 06 20:09:51 I spent more time figuring out why kexec wouldn't work than actually working on the kernel. Feb 06 20:10:12 hm.. what was seppuku? Feb 06 20:10:24 angstrom.inc and kaelios.inc have some setup for it Feb 06 20:11:53 recipes/bitbake/bitbake.inc have python-netclient and note that it is needed for tinderclient.bbclass Feb 06 20:12:28 are not any parts of BB/OE depends on netclient? Feb 06 20:14:44 ah.. well.. python-netclient provides urllib2 which is used by oestast as well Feb 06 20:19:43 * mwester wonders if there's any real reason to support ext4 with SlugOS? Feb 06 20:22:10 would any advantage be eaten up in USB overhead? Feb 06 20:24:17 I have no idea. I don't know enough about ext4 (all my customers are on older RHEL systems, so no reason to mess with until now) Feb 06 20:25:12 I'm afraid I dont' know enough about ext4 to say. Feb 06 20:25:32 I use ext3 for rootfs' and jfs for /home Feb 06 20:25:43 ok. then i'll just see how bit the module is, and see what makes sense. Feb 06 20:25:58 BTW, what is the advantage of jfs? Feb 06 20:26:09 for BIG files is fast. Feb 06 20:26:25 hmmm... Feb 06 20:26:33 I'm looking @ what is fast for lots of little files (like builds take) Feb 06 20:26:47 Not something I need for my slug, then. But jfs might be interesting on my big file server. Feb 06 20:26:51 (mythtv) Feb 06 20:26:54 xfs is slow at deleting Feb 06 20:27:40 woglinde, we need something fast for builds...I might just use ext2 and forget the journal. Feb 06 20:27:50 since its scratch space. Feb 06 20:27:52 * mwester can see the advantage of Microsoft's NTFS solution ("You can have any color you like, so long as it's black.") Feb 06 20:28:09 mwester, that was Henry Ford's solution too. Feb 06 20:28:18 Can you imagine if Windows users had choices?? Feb 06 20:29:39 mwester: I'm sorry, but what purpose is behind SlugOS? Feb 06 20:29:40 Bah. :D I painted myself into a corner, so to speak... by switching to a new kernel, I can't install any modules easily to test this, until my autobuilder catches up and populates my feeds. Feb 06 20:29:45 NAS? Feb 06 20:30:03 Jay7, ya, mostly Feb 06 20:30:16 * Jay7 is looking some starting point for iSCSI storage Feb 06 20:30:23 Jay7, I use it for controlling my observatory. Feb 06 20:30:29 even redundant iSCSI storage Feb 06 20:30:34 Jay7, General Purpose micro-server. NAS is not so good, but acceptable. Many people use it as a platform for small media servers, and some use it for control systems (Dallas one-wire, etc). Feb 06 20:30:48 mwester: ah, now I know it :) Feb 06 20:31:17 imho we should describe all known distros on our wiki Feb 06 20:31:25 (of not done already) :) Feb 06 20:31:27 Jay7, I now have 3 slugs controlling my observatory...when the Fat Slug arrives it will replace 3 regular slugs. Feb 06 20:31:31 s/of/if/ Feb 06 20:31:45 I have one that is a file server, one that is my network DHCP/PXEboot device, and one that monitors all aspects of my heating/cooling system. Feb 06 20:32:44 BTW, my new favorite small device for iSCSI server is the Synology. Feb 06 20:33:12 I have one (CameraSLUG) that runs the camera with gphoto, 1 (mountSLUG) that runs the mount and a final one(enviroslug) that runs the 1wire, dome control and general monitoring of the solar/battery system. Feb 06 20:37:24 jo effem Feb 06 20:39:13 hi woglinde, all Feb 06 20:39:21 hope you had a nice day Feb 06 20:40:30 yes, it was quite interesting Feb 06 20:42:07 actually didn't see everyone who was on the wiki (if I recall florian was only there on saturday), Feb 06 20:42:52 re Feb 06 20:43:06 hi florian Feb 06 20:43:08 hi rp Feb 06 20:43:08 hi florian Feb 06 20:43:09 yes eFfeM is correct Feb 06 20:43:16 did you have a good trip back ? Feb 06 20:43:28 so the tables was empty today? Feb 06 20:43:51 were Feb 06 20:44:05 eFfeM: yes... wasn't too bad. Feb 06 20:44:33 nope, I've seen Crofton, RP, Robert and Ulf behind the table Feb 06 20:45:07 crofton had his gnuradio, ulf his arm board, and there was a pandan and one or two other things Feb 06 20:45:17 ah okay Feb 06 20:45:27 not as many gadgets than wen you are here florian :-) Feb 06 20:45:35 woglinde: not really... quite some people are there as well as a pile of hardware :) Feb 06 20:45:39 heh Feb 06 20:45:46 and I think the guys from hackable used part of the 2nd table Feb 06 20:45:49 i even left one there :) Feb 06 20:45:50 sounds like enough hardware. Feb 06 20:46:18 * florian is going to leave for the next event in a few hours :-/ Feb 06 20:46:50 not as much as when florian is there, he has an amazing set of gadgets, still recall the projected kbd from 2 years ago that drew quite a lot of attention Feb 06 20:47:05 florian where if I may ask ? Feb 06 20:56:01 eFfeM: A kind if in-house training and exhibition at a customer. Feb 06 20:56:56 ah ok, thought it was yet anoher open source event Feb 06 21:16:18 fosdem is great fun Feb 06 21:16:25 except brussels Feb 06 21:28:18 Does OE support IMake/xmkmf style packages? I don't see any classes for it. Feb 06 21:28:32 or recipes for the native tool Feb 06 21:29:43 hm.. I've seen something Feb 06 21:33:44 zenlinux: seems no.. grep by xmkmf show me only patches Feb 06 21:33:58 Jay7, ok, thanks for the confirmation Feb 06 21:35:03 grep imake show nothing interesting as well.. Feb 06 21:35:25 zenlinux: have poky/yocto any? Feb 06 21:35:44 that would be good to have, imho Feb 06 21:35:58 Jay7, I just noticed that Debian packages xmkmf as part of the xutils-dev package...that gives me some hope Feb 06 21:37:06 I've remember that some years ago I've tried to port some afterstep/windowmaker applets and they was dependent on imake/xmkmf.. Feb 06 21:37:18 that was blocker Feb 06 21:37:40 oh I used to use windowmaker for a couple of years... Feb 06 21:37:44 I liked it Feb 06 21:38:36 I've tried to create fvwm-based environment for my Zaurus :) Feb 06 21:39:18 Jay7: oh man, the configuration is crazy Feb 06 21:39:44 but it may be configured as you wish :) Feb 06 21:40:06 Jay7: I remember being pissed of by some limitations Feb 06 21:41:06 ah, the clock implementation was sick afair Feb 06 21:41:20 hehe.. Feb 06 21:41:56 well.. I'll prefer to have wm that may be configured by some well-known language like lua :) Feb 06 21:42:06 and it was very hard to create stuff like applets Feb 06 21:42:45 it somehow didn't cooperate properly with things like tray icons, something about the tray size maybe... don't remember Feb 06 21:43:24 there just was no tray, only swallowing buttons :) Feb 06 21:46:45 Jay7: there was an utility - xtray, but it sucked at trasparency and couldn't dynamically resize Feb 06 21:47:15 geez, I've done a lot of weird stuff during these years :P Feb 06 21:47:42 there is stalonetray in recipes ;) Feb 06 21:55:13 for a package that generates binaries and symlink aliases to those binaries....should those symlinks be created during do_install and packaged, or should they be part of a pkg_postinstall? Feb 06 21:55:50 zenlinux depends Feb 06 21:56:12 zenlinux: kexecboot using first approach Feb 06 21:56:57 The package I'm dealing with is opensp, that provides various sgml tools Feb 06 21:58:02 http://www.linuxfromscratch.org/blfs/view/stable/pst/opensp.html Feb 06 21:58:26 ^ this shows the various symlinks I'm referring to Feb 06 22:01:33 zenlinux hm you could look at the debian package Feb 06 22:02:47 woglinde, good point. I'll check it out Feb 06 22:13:55 nice, I am compiling cairo to build bootstrap-image. something is wrong here.. Feb 06 22:16:15 hm Feb 06 22:16:27 I thought I solved the deps Feb 06 22:16:39 lets see Feb 06 22:19:01 woglinde: it pulls in a lot of X stuff (gtk, cairo, X libraries, freetype and others) Feb 06 22:20:19 yeah I am looking at it mom Feb 06 22:21:28 filip is woglinde's mom? Feb 06 22:21:31 ! Feb 06 22:22:20 ;) Feb 06 22:25:56 haha Feb 06 22:26:11 * mwester has completed most of the fun things to fix for SlugOS, and is nearly at the point where he will have to tackle that least favorite part: updating opkg. Feb 06 22:26:34 hm "pango-native" -> "pango-module-basic-x" [ Feb 06 22:29:13 "python-pygobject-native" -> "python-textutils" Feb 06 22:30:55 mwester, opkg? you're still using that old thing? Its all ball bearings these days... Feb 06 22:31:19 * grg is feeling silly Feb 06 22:31:28 grg: what package mgr would you suggest? Feb 06 22:31:44 i dont like any Feb 06 22:31:55 but if you're using oe, opkg it is Feb 06 22:32:21 hm Feb 06 22:32:33 that need to be solved in .class files Feb 06 22:32:34 grg: I am fixing dpkg support in OE Feb 06 22:32:40 or bitbake Feb 06 22:33:22 filip, nice. Good to have options. And good to ensure that OE doesn't have an unhealthy dependence upon opkg Feb 06 22:33:58 * mwester has had great success with bronze bearings, and does not look forward to ball bearings. Feb 06 22:34:09 grg: it depends whether I will manage to complete it before it makes me mad Feb 06 22:34:38 complete it after you go mad. It will probably work better that way Feb 06 22:40:17 woglinde: are you narrowing it down? Feb 06 22:42:35 filip hm all RRECOMMENDS_${PN} and RDEPENDS_${PN} drags in the stuff for native package when using virtclass Feb 06 22:43:50 woglinde: you've just confused the hell out of me ;p Feb 06 22:46:13 hm saw now its in all native packages Feb 06 22:48:38 hm tasks-base drags in avahi Feb 06 22:50:02 filip thats the problem Feb 06 22:50:13 avahi drags in all gtk plunder Feb 06 22:52:16 woglinde: is avahi needed in a bootstrap image? Feb 06 22:52:22 filip are you using angstroem? Feb 06 22:52:37 woglinde: nope, I'm a jlime guy Feb 06 22:52:55 hm Feb 06 22:52:58 okay Feb 06 22:53:07 angstroem is dragging in task-distro-base Feb 06 22:53:13 and task-distro-base has avahi Feb 06 22:53:15 crap Feb 06 22:54:22 I will commit three patches Feb 06 22:54:35 that might solve your problem Feb 06 22:55:22 thanks :) Feb 06 22:55:33 jlime is mostly based on angstrom iirc Feb 06 22:58:38 hm Feb 06 22:58:46 the patches are universal anyway Feb 06 23:01:01 03Henning Heinold  07org.openembedded.dev * rfa3c3599d7 10openembedded.git/classes/update-rc.d.bbclass: update-rc.d.bbclass: set RDEPENDS_${PN} to nothing for virtclass-native Feb 06 23:01:01 03Henning Heinold  07org.openembedded.dev * re317ed0f8b 10openembedded.git/recipes/pango/pango.inc: pango: set RRECOMMENDS_${PN} to nothing for virtclass-native Feb 06 23:01:03 03Henning Heinold  07org.openembedded.dev * r0146b25d16 10openembedded.git/recipes/python/python-pygobject-native_2.20.0.bb: python-pygobject-native: set RDEPENDS to nothing Feb 06 23:02:19 hm ah Feb 06 23:02:29 # Additional content II (can be masked with BAD_RECOMMENDATIONS) Feb 06 23:02:30 DISTRO_EXTRA_RRECOMMENDS += " \ Feb 06 23:02:52 thats from minimal Feb 06 23:03:34 but thats not in jlime Feb 06 23:13:26 ok off to bed, bye. [d] Feb 06 23:13:47 hrms Feb 06 23:13:55 its only for rootfs Feb 06 23:14:10 so its getting to tasks-base anyway Feb 06 23:14:12 *sigh* Feb 06 23:14:19 nite filip Feb 06 23:14:37 I am going too **** ENDING LOGGING AT Mon Feb 07 02:59:56 2011