**** BEGIN LOGGING AT Thu Nov 05 02:59:56 2009 Nov 05 04:01:17 kergoth: how are CONFFILES used Nov 05 04:01:34 kergoth: is it similar to debian conffiles Nov 05 04:13:46 khem: yep Nov 05 04:20:34 cool Nov 05 06:16:22 03Martin Jansa  07shr/merge * r824a1cb1cc 10openembedded.git/recipes/ (gypsy/gypsy.inc navit/navit.inc): navit: fix depends Nov 05 07:24:54 03Sebastian Spaeth  07org.openembedded.dev * rb084808fca 10openembedded.git/ (conf/checksums.ini recipes/vala/vala_0.7.8.bb): Nov 05 07:24:54 vala: add version 0.7.8 and checksums Nov 05 07:24:54 Signed-off-by: Sebastian Spaeth Nov 05 07:26:21 03Sebastian Spaeth  07shr/merge * r3da5e2c22c 10openembedded.git/ (conf/checksums.ini recipes/vala/vala_0.7.8.bb): Nov 05 07:26:21 vala: add version 0.7.8 and checksums Nov 05 07:26:21 Signed-off-by: Sebastian Spaeth Nov 05 07:33:43 03Sebastian Spaeth  07shr/merge * r80ccb1c805 10openembedded.git/recipes/vala/vala-native_0.7.8.bb: Nov 05 07:33:43 vala-native: add 0.7.8 Nov 05 07:33:43 Signed-off-by: Sebastian Spaeth Nov 05 07:34:46 03Sebastian Spaeth  07org.openembedded.dev * rce66333fb4 10openembedded.git/recipes/vala/vala-native_0.7.8.bb: Nov 05 07:34:46 vala-native: add 0.7.8 Nov 05 07:34:46 Signed-off-by: Sebastian Spaeth Nov 05 07:40:28 good morning Nov 05 07:55:24 mckoan: good morning Nov 05 08:04:47 hi recalcati_ Nov 05 08:11:52 03Koen Kooi  07org.openembedded.dev * rcbda7eff82 10openembedded.git/recipes/mplayer/mplayer_svn.bb: angstrom: raise default pref in mplayer recipe as well Nov 05 08:22:54 03Koen Kooi  07org.openembedded.dev * r6d51f2acd4 10openembedded.git/conf/checksums.ini: checksums: sort and add missing checksum Nov 05 08:22:55 03Martin Jansa  07shr/merge * r26d5a16ea7 10openembedded.git/: Merge branch 'shr/merge' of git.openembedded.org:openembedded into shr/merge Nov 05 08:22:56 03Martin Jansa  07shr/merge * r5855f30d42 10openembedded.git/ (conf/checksums.ini recipes/tangogps/tangogps_0.9.8.bb): Nov 05 08:22:56 tangogps: add 0.9.9 bugfix release, drop 0.9.8 Nov 05 08:22:58 Signed-off-by: Martin Jansa Nov 05 08:53:30 DISTRO_PR has changed (.3 to .4) : I have to rebuild all. Nov 05 08:59:21 recalcati_: yeah, there was a change that will impact all packages Nov 05 09:00:44 zecke: I know it was important, just to complain .. Nov 05 09:01:03 recalcati_: Nov 05 09:01:11 recalcati_: complain about what? Nov 05 09:08:43 recalcati_: this happened almost a week ago and was announced on the mailing list (maybe not the oe one, maybe the angstrom one) Nov 05 09:09:38 eFfeM: it's good. I now that there are many think. Also there was eabi changing before . no pb . Nov 05 09:09:45 eFfeM: it's good. I now that there are many things. Also there was eabi changing before . no pb . Nov 05 09:09:50 eFfeM: and the actual change was developed on the OE ml... I love how people that only take complain Nov 05 09:10:37 I have many setups. I don't have to discourage!!! Nov 05 09:13:46 morning Nov 05 09:15:49 I just bumped ABI in tmp/ files Nov 05 09:23:14 god some people just dont know when to stop sending emails Nov 05 09:25:58 03Klaus Kurzmann  07shr/merge * rd3e271f9dc 10openembedded.git/ (3 files in 2 dirs): Nov 05 09:25:58 linux-openmoko.inc: move autoload of sound modules to the appropriate machine config Nov 05 09:25:58 Signed-off-by: Klaus Kurzmann Nov 05 09:26:32 btw Nov 05 09:26:42 http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/zaurus-devel Nov 05 09:26:47 fsck.. Nov 05 09:26:49 hm, anyone an idea how automake gets its defaults for cross test? somehow for mysql5 stack_direction has no value whereas it has for mysql4. apparently I'm lost in automake hell Nov 05 09:26:59 ML for zaurus development coordination :) Nov 05 09:26:59 how to tell update-rc.d to apped not to $PN package? Nov 05 09:27:34 eFfeM: last argument was for cross? Nov 05 09:28:03 eFfeM: you can look into patches for fvwm recipe Nov 05 09:28:13 hrw: yeah I noticed you broke transmission packages :-D Nov 05 09:28:19 XorA|gone: yep ;( Nov 05 09:28:27 there I've fixed some cross-awareness Nov 05 09:28:39 XorA|gone: thinking about doing s/PN-daemon/PN and s/PN/PN-gui/ Nov 05 09:28:47 hrw, don't understand your comment, Jay7 will do (actually looked at the mysql files but didn't find there but there is a big change between 4 and 5) Nov 05 09:28:54 hrw: I was just going to say that Nov 05 09:29:00 hrw: we think alike Nov 05 09:30:10 eFfeM: recipes/fvwm/files/upstream-configure.ac.patch Nov 05 09:30:13 XorA: I have to say that transmission web is nice Nov 05 09:30:15 eFfeM: autoconf isnt that bad Nov 05 09:30:24 XorA: lacks some things but usable Nov 05 09:30:28 eFfeM: pretty easy to work out how the tests work Nov 05 09:30:39 XorA: can you share your settings for transmission? Nov 05 09:30:58 hrw: I use the defaults Nov 05 09:31:02 XorA: never got to understand autoconf, sorry, maybe need to spend some time in it Nov 05 09:31:05 hrw: apart from whitelist Nov 05 09:31:27 I'm using transmission-daemon + remote + web + qt gui at home :) Nov 05 09:31:29 Jay7: you lost me, my recipes dir does not have a fwm subdir (puzzled?) Nov 05 09:31:30 but on x86 Nov 05 09:31:57 an i am at yesterdays head Nov 05 09:32:15 Jay7: are you on a branch ? Nov 05 09:32:45 eFfeM: in that patch I've fixed test results for cross-compiling Nov 05 09:33:02 but may be I've musunderstand your first question.. Nov 05 09:33:29 Jay7: i don't even have a recipes/fwm dir ! Nov 05 09:33:39 would love to have a look at your patch Nov 05 09:33:46 eFfeM: hm.. it should be in oe.dev Nov 05 09:33:54 earthquake Nov 05 09:33:59 cd recipes; ls | wc gives 1989, ls fwm gives not found Nov 05 09:34:09 f v w m :) Nov 05 09:34:10 03Robert P. J. Day  07org.openembedded.dev * r810df5f14c 10openembedded.git/docs/usermanual/chapters/introduction.xml: Nov 05 09:34:10 usermanual: Aesthetic fix in chapter 1. Nov 05 09:34:10 A simple tidy of a comma-separated list in Chapter 1. Nov 05 09:34:10 Signed-off-by: Robert P. J. Day Nov 05 09:34:10 Acked-by: Holger Hans Peter Freyther Nov 05 09:34:12 03Robert P. J. Day  07org.openembedded.dev * raa6bc11505 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: (log message trimmed) Nov 05 09:34:15 oopks 1898 Nov 05 09:34:16 User manual typoes and aesthetics, chapter 2 only. Nov 05 09:34:18 Some minor cleanups for Chapter 2 of the User Manual: Nov 05 09:34:20 * add some DocBook tags for filenames Nov 05 09:34:22 * update BitBake version to 1.8.16 Nov 05 09:34:24 * add missing space after "wget" command Nov 05 09:34:26 * correct misspelled "local/conf" to "local.conf" Nov 05 09:34:28 03Robert P. J. Day  07org.openembedded.dev * rf2b43d21cf 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: (log message trimmed) Nov 05 09:34:31 User Manual: Warn readers about mmap and SELinux issue. Nov 05 09:34:33 On some Linux distros, if you set the kernel /proc file Nov 05 09:34:35 /proc/sys/vm/mmap_min_addr to zero, you will probably get an SELinux Nov 05 09:34:36 wow that was a long one Nov 05 09:34:37 violation if you're running SELinux in enforcing mode. Explain to the Nov 05 09:34:39 reader how to get around that. Nov 05 09:34:41 Signed-off-by: Robert P. J. Day Nov 05 09:35:40 03Sebastian Spaeth  07org.openembedded.dev * r56b84b4f4c 10openembedded.git/recipes/vala/ (vala-native_0.7.8.bb vala_0.7.8.bb): Nov 05 09:35:40 vala vala-native: Make use of GNOME_MIRROR in SRC_URI Nov 05 09:35:40 Signed-off-by: Sebastian Spaeth Nov 05 09:35:51 03Sebastian Spaeth  07org.openembedded.dev * ra3ee457e82 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.net/openembedded into org.openembedded.dev Nov 05 09:36:12 Jay7: oops sorry just found out after browsing the git tree on the web Nov 05 09:36:16 http://cgit.openembedded.net/cgit.cgi/openembedded/tree/recipes/fvwm/files/upstream-configure.ac.patch Nov 05 09:36:18 here :) Nov 05 09:36:42 time to eat something.. Nov 05 09:36:45 afk Nov 05 09:37:25 Jay7 enjoy :-) Nov 05 09:41:23 hrw: are you going to make that name change for transmission? I think thats the best option shortterm Nov 05 09:42:06 yes Nov 05 09:42:53 fixing it now Nov 05 09:43:15 Hi Guys. A little while back I think we decided that ccache was no longer a requirement (in fact it was causing random compile problems). Shall I remove the remaining mentions of it from the wiki on http://wiki.openembedded.net/index.php/Required_software & http://wiki.openembedded.net/index.php/Getting_started ? Nov 05 09:55:07 ~hail commit --amend Nov 05 09:55:09 * ibot bows down to commit --amend and chants, "I'M NOT WORTHY!!" Nov 05 09:55:44 03Marcin Juszkiewicz  07org.openembedded.dev * r7f845d2618 10openembedded.git/recipes/transmission/transmission.inc: Nov 05 09:55:44 transmission: fix breakage of "r3" version - GUI in PN-gui, daemon in PN Nov 05 09:55:44 update-rc.d.bbclass adds only to PN package so I moved PN-daemon into PN Nov 05 09:55:44 and PN became PN-gui now Nov 05 09:58:07 MWelchUK_work__: yes, go ahead Nov 05 09:58:39 brb Nov 05 10:01:38 03Koen Kooi  07org.openembedded.dev * rac8425fce9 10openembedded.git/recipes/squashfs-tools/ (squashfs-tools-native_4.0.bb squashfs-tools_4.0.bb): squashfs-tools: fix up native version Nov 05 10:07:41 pb__, I've left the mention on http://wiki.openembedded.net/index.php/Getting_started as it's mentioning a partial work around if it is installed. I've extended the text to suggest it should be uninstalled. Nov 05 10:10:11 I sent patch to oe-devel list a few days ago. I don't know customs there, shall I ask for ack and commit to repo directly? Nov 05 10:10:28 jest: got a patchwork URL? Nov 05 10:10:51 http://patchwork.openembedded.org/patch/1192/ Nov 05 10:11:59 this applicable to both dev and stable branches Nov 05 10:12:32 jest: oh boy... boost... atomic operations Nov 05 10:12:39 yeah :) Nov 05 10:13:11 do you have any idea why atomic_count_gcc_atomicity.patch is only applied in 1.33.1? Nov 05 10:13:11 zecke: this is a patch resulting from reading sources of newer version of Boost Nov 05 10:13:25 zecke: no Nov 05 10:13:42 zecke: I can try if any other version compiles without it for ARM Nov 05 10:14:11 jest: can I ask you to find that out? if you do that within the next ~3 hours I can land your patch (otherwise I'm off to bed) Nov 05 10:14:23 sure, I'll take a look Nov 05 10:19:05 ola all Nov 05 10:19:32 df Nov 05 10:20:51 In my custom Angstrom, I want an empty /upgrade folder created in my rootfs. Is there a recipe / script that I could use for this? Nov 05 10:21:19 tasslehoff: you could define a rootds post process command Nov 05 10:21:24 zecke: I'll find out that it depends on GCC version Nov 05 10:22:25 zecke: for GCC < 4.2 it should be bits/atomicity.h (just as is), for GCC >= 4.2 ext/atomicity.h should be used Nov 05 10:22:56 zecke: I think I should correct the patch, just copying the code from boost 1.40 Nov 05 10:23:27 zecke: how can I do that with git? Nov 05 10:27:06 XorA: sounds like what I want. any such processes defined already that I can "learn from"? Nov 05 10:27:22 tasslehoff: I think in angstrom.conf Nov 05 10:31:57 jest: copying or what? Nov 05 10:32:01 XorA: can't find it, but then I don't understand much of that file yet :) Nov 05 10:32:37 zecke: shall I just post another patch, or relate the new one with the old one (somehow)? Nov 05 10:34:01 jest: fixing all boost versions in one go would be great Nov 05 10:34:20 good morning Nov 05 10:34:55 zecke: I have no possibility to check, but it's just '#if __GCC__' stuff, so should work Nov 05 10:35:27 jest: why? you could do bitbake -b boost_version.bb (just compile them... or is this header file not used during compile?) Nov 05 10:36:00 tasslehoff: http://pastebin.com/dcca9723 should give you an idea (it's how I generate a version file for images I build for the Pandora) Nov 05 10:36:31 zecke: the problem is it's a header file and should be included somewhere. I'll try Nov 05 10:37:49 tasslehoff: ROOTFS_POSTPROCESS_COMMAND += "mkdir ${IMAGE_ROOTFS}/upgrade" should do it to be honest, add that to your image file. Not checked that will work but it should. Nov 05 10:38:38 03Martin Jansa  07shr/merge * r6c1ed757c1 10openembedded.git/recipes/update-alternatives/ (3 files in 2 dirs): Nov 05 10:38:38 update-alternatives-cworth: if there is more alternatives with same preference use latest Nov 05 10:38:38 Signed-off-by: Martin Jansa Nov 05 10:38:48 03Martin Jansa  07shr/merge * rd20ff32699 10openembedded.git/: Merge branch 'shr/merge' of git.openembedded.org:openembedded into shr/merge Nov 05 10:45:31 03Martin Jansa  07shr/merge * r791907dd12 10openembedded.git/recipes/tasks/task-shr-feed.bb: Nov 05 10:45:31 task-shr-feed: remove some unneeded apps Nov 05 10:45:31 Signed-off-by: Martin Jansa Nov 05 10:46:50 florian: good morning Nov 05 10:46:56 zecke: good afternoon Nov 05 10:47:30 pb__: hey! Nov 05 10:57:44 Is anyone actively building OE on MacOSX? I've found Koen's darwin-fixes branch, but there's no commits in it since March ... Nov 05 10:58:23 I think kergoth is/was working on OS X. Nov 05 10:59:57 morning all Nov 05 11:00:46 hi rp Nov 05 11:04:29 DJWillis: thanks. (late reply because of lunch) Nov 05 11:33:16 anyone got any tips for getting icecc working? at the moment the helper machines don't seem to be doing any work, although iceccd on those machine uses up a few % CPU when a build is in progress elsewhere Nov 05 11:33:59 DJWillis: worked like a charm Nov 05 11:34:30 03Martin Jansa  07shr/merge * r565f3559ef 10openembedded.git/conf/distro/shr.conf: Nov 05 11:34:30 shr: few more preferred providers for task-shr-feed, add testlab Nov 05 11:34:30 Signed-off-by: Martin Jansa Nov 05 11:52:04 * * OE Bug 5332 has been created by ospite(AT)studenti.unina.it Nov 05 11:52:06 * * building ncurses triggers an internal compiler error Nov 05 11:52:08 * * http://bugs.openembedded.org/show_bug.cgi?id=5332 Nov 05 12:05:46 hm, doing something stupid again or don't understand things: Nov 05 12:05:54 my msql5 recipe contains a line: FILES_${PN}-leftovers = "/" Nov 05 12:06:28 i get a QA issue on /work/armv7a-angstrom-linux-gnueabi/mysql5-5.1.40-r0/install/mysql5-leftovers/usr/lib/plugin/.debug/ha_innodb_plugin.so.0.0.0 Nov 05 12:06:35 whiich is understandable Nov 05 12:07:05 added FILES_${PN}-leftovers-dbg = "${libdir}/plugin/.debug/" Nov 05 12:07:05 to the recipe but still I get the error Nov 05 12:07:16 what am I doing wrong here???? Nov 05 12:07:23 eFfeM: add an '*' to the end Nov 05 12:07:29 ah ok Nov 05 12:08:03 as the leftovers package contained only a dir , thought that would be ok Nov 05 12:08:10 yeah, I think that should be ok without the * Nov 05 12:08:15 what is the value of ${PACKAGES}? Nov 05 12:09:04 zecke, pb__ with * does not matter Nov 05 12:09:18 packages contains ${PN}-leftovers not the -dbg Nov 05 12:09:23 should that be there too ? Nov 05 12:09:26 yes Nov 05 12:09:34 and, it needs to be before ${PN}-leftovers Nov 05 12:10:00 mickey|office: good morning Nov 05 12:10:02 ah ok Nov 05 12:11:00 03Thomas Zimmermann  07shr/merge * r9d555403fe 10openembedded.git/recipes/xorg-lib/ (libxext-native_1.1.1.bb libxfont-native_1.4.1.bb): Nov 05 12:11:00 libx*-native: fix PROVIDES for some packages Nov 05 12:11:00 Signed-off-by: Klaus Kurzmann Nov 05 12:11:02 03Thomas Zimmermann  07shr/merge * rd589342d65 10openembedded.git/recipes/matchbox2/matchbox-panel-2_svn.bb: Nov 05 12:11:02 matchbox-panel-2: add -Wno-error to TARGET_CFLAGS to build with gcc 4.4 Nov 05 12:11:04 Signed-off-by: Klaus Kurzmann Nov 05 12:11:06 03Thomas Zimmermann  07shr/merge * r08d79665e9 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Nov 05 12:11:09 task-shr-minimal: include matchbox-keyboard-im again in task-shr-minimal-gtk Nov 05 12:11:11 Signed-off-by: Klaus Kurzmann Nov 05 12:11:18 hmm meta-toolchain is broken in more ways :( Nov 05 12:11:32 oh dear Nov 05 12:11:35 what's wrong with it? Nov 05 12:11:57 pb__: the libtool-cross-sdk (or such package) stopped installing config.guess install-sh... Nov 05 12:12:05 so an autoreconf --install is failing Nov 05 12:12:20 drat Nov 05 12:16:28 pb__, thanks that worked Nov 05 12:16:36 one issue left: what is the policy to resolve Nov 05 12:16:38 NOTE: Multiple libraries (libmysqlclient.so.16, libmysqlclient_r.so.16) found and LEAD_SONAME not defined Nov 05 12:17:03 put those two in separate packages Nov 05 12:17:52 ok, mysql4 did nod seem to have this Nov 05 12:18:40 possibly mysql4 didn't ship both those libraries Nov 05 12:21:01 03Klaus Kurzmann  07shr/merge * r8f88d5e6e1 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Nov 05 12:21:01 task-shr-minimal.bb: remove notifier and libframeworkd-phonegui-efl Nov 05 12:21:01 Signed-off-by: Klaus Kurzmann Nov 05 12:21:41 morning pb__ Nov 05 12:23:42 zecke: I checked my patch on 1.33 and it works Nov 05 12:24:00 zecke: on boost 1.34 it is not applicable Nov 05 12:33:58 if I have a recipe that has 2 git repos in SRC_URI how can I specify a different SRCREV for each? Nov 05 12:35:28 hrw: You about Nov 05 12:35:31 ? Nov 05 12:35:53 Was just wanting to check wrt email that by vendor page you mean on the wiki? (or is there some other list?) Nov 05 12:46:38 th1_: I don't think you can. Nov 05 12:46:59 in that kind of situation I don't think SRCREV really works. Nov 05 12:49:12 I'm trying to keep my own application outside the OE tree. I have a myimage.bb that picks modules from oe/recipes and myown/recipes. It works well, but how can I make my applications and scripts start automatically? Could I have a myown/recipe/startup.bb that has a script that starts all my stuff, put that in some magic directory where it will automatically be started? Nov 05 12:51:06 03Robert Nelson  07org.openembedded.dev * r6f0e47264d 10openembedded.git/recipes/linux/ (2 files in 2 dirs): linux-omap 2.6.29: fix FIFO issues with large transfers on the MUSB controller Nov 05 12:51:10 03Koen Kooi  07org.openembedded.dev * rf389ecae02 10openembedded.git/ (3 files in 2 dirs): linux-omap-pm 2.6.29: cleanup some touchbook stuff Nov 05 12:51:20 03Koen Kooi  07org.openembedded.dev * r2b0e95fd5d 10openembedded.git/recipes/gnome/ (gdm-2.28.0/gdm gdm_2.28.0.bb): gdm: tweak initscript and add grep to depends since busybox grep doesn't do -w Nov 05 12:53:00 tasslehoff: yes, you can put anything you like in /etc/init.d Nov 05 12:53:04 that will do what you want for most distros Nov 05 12:55:58 jic23_: in machine config I mean Nov 05 12:56:02 pb__: that simple? great :). how can I control when it runs? I think I want it to run as the last initscript. Nov 05 13:03:03 more or less that simple, yes. the actual running is controlled by the symlinks in /etc/rcN.d, which are usually managed by update-rc.d. Nov 05 13:03:26 in straightforward cases, you can use update-rc.d.bbclass to take care of all that stuff for you. Nov 05 13:04:22 to make it be the last thing started, you can set INITSCRIPT_PARAMS = "default 99" Nov 05 13:04:27 er, "defaults 99" Nov 05 13:05:14 pb__: I think I have a straightforward case on my hands :) Nov 05 13:07:21 hrw: Ah. Fair point. Nov 05 13:13:01 pb__: about oedem, AFAYK there is a meeting somewhere in Cam to go at the Red Lion? Nov 05 13:13:31 03Koen Kooi  07org.openembedded.dev * r12c4514ab0 10openembedded.git/recipes/gnome/ (gdm-2.28.0/gdm gdm_2.28.0.bb): gdm: add tcp-wrappers to DEPENDS, fix typo in initscript Nov 05 13:19:20 pb__, any other way I can get a similar effect? Nov 05 13:22:43 03Thomas Zimmermann  07shr/merge * r5c2aac7c18 10openembedded.git/recipes/e17/elmdentica_svn.bb: Nov 05 13:22:43 elmdentica: add inherit pkgconfig and some cleanup Nov 05 13:22:43 Signed-off-by: Martin Jansa Nov 05 13:23:26 pb__: this simple? : http://pastebin.com/d457637f5 Nov 05 13:25:56 th1_: what exactly is the effect you're trying to achieve? Nov 05 13:26:07 tasslehoff: yep, that should do it Nov 05 13:26:36 oh, might need to arrange to install mystartup into /etc/init.d; I don't remember whether update-rc.d.bbclass will do that for you. Nov 05 13:26:40 probably not I guess Nov 05 13:28:13 pb__: I had that at first, but it gave an error so I just removed it :) I'll try again if it's missing Nov 05 13:33:37 pb__, I have a package that is built from two local git repositories, and I need to specify the revision ID for both the repositories so I can be sure I have a consistent build (and can update the recipe with "known-working" combinations) Nov 05 13:33:56 th1_: if all you need is to specify the revision IDs in your recipe, you can do that directly in the URLs. Nov 05 13:34:13 so omit SRCREV altogether? Nov 05 13:34:17 right Nov 05 13:37:18 pb__, how do I specify the revision ID in the URL? ;) Nov 05 13:38:03 git:///;tag=012345 I think Nov 05 13:38:43 yes, embed branch and tag Nov 05 13:38:49 hi pb__ Nov 05 13:38:52 or might be "rev=" not "tag=" Nov 05 13:38:56 ready for heavy weekend? Nov 05 13:38:59 I'm not sure if git distinguishes them Nov 05 13:39:01 hi ant_work Nov 05 13:39:22 yeah, more or less ready Nov 05 13:39:58 pb__, from looking at git.py it looks like it must be tag but I don't quite understand that file.. Nov 05 13:40:06 th1_: see also the fine manual: http://bitbake.berlios.de/manual/ch03s07.html Nov 05 13:40:45 pb__, yes I've seen that but it doesn't have a REV only tags and branch .. anyway I'll try to put the ID in tag and see if it worsk Nov 05 13:41:33 th1_: I don't think git makes any distinction between tags and revs at that level. Nov 05 13:41:40 jest: I commented on your GNU_HASH fix. When touching a .inc file you will need to bump the PR on every of its users Nov 05 13:42:01 zecke: I did Nov 05 13:42:05 jest: in the last year we have done something like FILE_PR/INC_PR you could convert the boost recipes to that Nov 05 13:42:12 th1_: I guess patches for the manual are welcome :-) Nov 05 13:42:13 jest: the .inc is only used by 1.33.1.bb? Nov 05 13:42:17 yes Nov 05 13:42:46 zecke: 1.34 has everything "embedded", 1.36 uses another .inc Nov 05 13:42:54 pb__, after we make the switch there'll be lots of patches hopefully :) Nov 05 13:42:55 pb__: I'd ike to debug a non-booting kernel and was thinking about have early output. I found this new patch Nov 05 13:42:57 http://lists.infradead.org/pipermail/linux-arm-kernel/2009-November/003452.html Nov 05 13:43:03 so far it's a struggle because I'm the only one maintaining the OE stuff Nov 05 13:43:15 wuestion: does it work on framebuffer? Nov 05 13:43:26 *question Nov 05 13:43:31 (damned qwerty) Nov 05 13:43:35 ant_work: no Nov 05 13:43:39 that would be for serial only Nov 05 13:43:41 jest: oh, e.g. writing this in the commit message would be good :) Nov 05 13:43:48 ah, ok, still fine Nov 05 13:44:21 the kernel isn't in control of the framebuffer at the point where that runs. Nov 05 13:44:33 zecke: in fact, I'd vote for deprecation policy in this case... :) Nov 05 13:44:35 yep, even console is not there yet..silly question Nov 05 13:44:38 obviously you could move the fb initialisation earlier, but that starts to defeat the object of the "early" printk Nov 05 13:44:48 jest: deprecation as in throwing this file out? Nov 05 13:44:49 atm I have no output on serial Nov 05 13:45:06 since, obviously, the more code that you have to move to the "early" stage, the greater the chance that one of those things will be what is causing your crash Nov 05 13:45:09 I have doubt about the decompression Nov 05 13:45:31 I don't even see 'Uncompressing...' Nov 05 13:45:44 ah, that doesn't sound good Nov 05 13:45:54 wait, this only booting from nand Nov 05 13:45:57 he Nov 05 13:46:04 all fine from SD Nov 05 13:46:09 (kexec) Nov 05 13:46:10 zecke: more or less. In cases where nobody uses old versions, as newer ones are better. But it is a matter of attitude and kind of philosophical. Nov 05 13:46:50 pb__: anyway that 'Uncompressing..' msg comes *after* console is up Nov 05 13:47:07 or? Nov 05 13:47:23 pb __: nope, it doesn't work: fatal: Couldn't find remote ref 04087d7838830880a5bc52207b13bf22d6a8bfa2 Nov 05 13:47:33 so it has to be a tag there apparently :( Nov 05 13:47:34 ant_work: no, before Nov 05 13:47:50 ah, is logged and displayed later Nov 05 13:48:12 how could it go to serial without console? Nov 05 13:48:19 no, that particular message is displayed by special code in the first stage bootstrap Nov 05 13:48:25 I see Nov 05 13:48:29 see arch/arm/boot/compressed/head.S Nov 05 13:48:48 this is completely separate to the "real" kernel, which can't be running at that point since it hasn't been uncompressed yet Nov 05 13:49:04 is a bit like that earlyprintk patch Nov 05 13:49:09 extra Nov 05 13:49:13 th1_: ah, that is a shame. I guess you should look at what git command the fetcher is running and compare that to what you are trying to get it to do. Nov 05 13:49:35 zecke: Considering commit style: I can excuse myself by watching others patches' commits -- sometimes they are really sparse :) Nov 05 13:49:48 pb__, yes that's what I'm doing but it's using git commands I wouldn't normally use so taking me a bit of time to figure out :) I'm sure I'll get it in the end though Nov 05 13:51:38 pb__: "defaults 99" will make it start last and stop first? Nov 05 13:52:04 pb__: yes, arm boot is a long story :) Nov 05 13:52:09 jest: this argument is weak, today morning I had to explain a ~2 year old patch of mine and my commit message had grammar issues and was crap :) Nov 05 13:52:13 tasslehoff: I think that will make it start last and stop first. Nov 05 13:52:17 er, sorry, start last and stop last Nov 05 13:52:27 if you want start last, stop first, you need "defaults 99 01" Nov 05 13:56:30 zecke: I understand all the care of commit quality and totally support it. However, this requirement may pose problems for unexperienced OE users, as they are not so knowledgeable to understand fully the whole machinery, dependencies, consequences, and so on. So there is a risk of losing potential patches. Nov 05 13:58:29 jest: it is more about putting thought into the change, no need to be perfect... One difference from user to maintainer is... the user fixes his problem (e.g. in boost version X), the maintainer has to ask if it is in other versions too Nov 05 13:58:30 guys, we yesterday discussed the naming of the mysql5 recipe, it it now mysql5...., should the pacakges also get a mysql5 prefix or just mysql_5.... ? Nov 05 13:59:03 zecke: I definitely a user here :) Nov 05 13:59:11 zecke: I -> I'm Nov 05 13:59:12 jest: e.g. a commit message telling why other boost recipes do not need to be changed is great for everyone Nov 05 13:59:39 zecke: I guess. It just didn't cross my mind. Nov 05 14:00:09 zecke: Shall I correct it somehow and send a new patch? Nov 05 14:02:36 jest: please :) Nov 05 14:03:01 zecke: which one? both, or just GNU_HASH Nov 05 14:03:05 pb__: thanks Nov 05 14:03:07 zecke: ? Nov 05 14:04:41 eFfeM: do what debian does Nov 05 14:05:05 XorA: will check Nov 05 14:05:46 eFfeM: they have mysql-4.0 mysql-5.0 mysql-5.1 packages Nov 05 14:05:54 ok Nov 05 14:05:56 so they obviously saw some ABI/API problems Nov 05 14:06:33 jest: try to write a less brief message Nov 05 14:06:44 btw any idea why I would get libmysqlclient16_5.1.40-r0.4_armv7a.ipk (and especially the number 16) my recipe just has FILES_libmysqlclient = "${libdir}/libmysqlclient.so.*" Nov 05 14:07:09 eFfeM: debian renaming Nov 05 14:07:45 ah ok Nov 05 14:08:03 * eFfeM learns something new every day (and sometimes even every hour ;-) ) Nov 05 14:09:07 XorA: the issue with the client name is that if I just postfix with a 5 to distinguish between 4 and 5 the name will become something like libmysqlcient516... Nov 05 14:09:22 it wont Nov 05 14:09:34 ok Nov 05 14:10:17 I think you could name it "randomlibraryinhere666" and it would still change name to libmysqlclient16_5.1.40-r0.4_armv7a.ipk Nov 05 14:14:15 SOVERSION is used for name not number in PACKAGES Nov 05 14:19:20 still one more thing: when installing mysql-server wants libmysqlclient-dev not the regular package, is this something wrong in the recipe ? Nov 05 14:21:41 eFfeM: yes, means you stuck the libraries in the -dev Nov 05 14:22:08 eFfeM: normally anyway Nov 05 14:27:09 shell question I have: how should "if [ -e filename ]" react to filename being broken symlink? Nov 05 14:29:17 false Nov 05 14:29:27 thx Nov 05 14:29:36 and how to make it return true? Nov 05 14:30:11 -h Nov 05 14:30:33 -h FILE True if file is a symbolic link. Nov 05 14:30:40 but it can be also a file sometimes Nov 05 14:31:12 -h FILE -o -e FILE Nov 05 14:31:21 thx Nov 05 14:38:16 recompiled: still this error: http://pastebin.com/f7c9b47f3 . any idea? Nov 05 14:40:31 meh Nov 05 14:41:00 03Marcin Juszkiewicz  07org.openembedded.dev * r4270cb2928 10openembedded.git/classes/java.bbclass: (log message trimmed) Nov 05 14:41:00 java.bbclass: fixed oe_jarinstall for broken symlinks Nov 05 14:41:00 Code checked for existance of JAR file and if it was present then it had Nov 05 14:41:00 to be removed before symlinking to it. But it was not working sometimes Nov 05 14:41:02 so we forced symlinking. Nov 05 14:41:04 Proper fix is to check "does file exists or is a symlink" which we do Nov 05 14:41:06 now. Nov 05 14:41:54 XorA: ok, thanks (was afk) just copied the packaging from mysql 4. Nov 05 14:42:04 guess this needs a review once it it done :-) Nov 05 14:47:03 03Koen Kooi  07org.openembedded.dev * rf54128714e 10openembedded.git/ (conf/checksums.ini recipes/sqlite/sqlite3_3.6.20.bb): sqlite3: add 3.6.20 Nov 05 14:47:04 03Koen Kooi  07org.openembedded.dev * r38396f8aa5 10openembedded.git/conf/machine/omap3-touchbook.conf: omap3-touchbook: ship both omapfb and fbdev drivers Nov 05 14:47:05 03Koen Kooi  07org.openembedded.dev * r2b1a41cbb8 10openembedded.git/recipes/xorg-xserver/ (2 files in 2 dirs): xserver-xorg-conf: switch to fbdev driver for touchbook till the vrfb bug gets solved Nov 05 15:01:08 03Otavio Salvador  07org.openembedded.dev * r0da71f8df5 10openembedded.git/recipes/chicken/chicken.inc: Nov 05 15:01:08 chicken.inc: fix install and compilation issues Nov 05 15:01:08 Signed-off-by: Otavio Salvador Nov 05 15:05:00 kergoth: good morning Nov 05 15:05:04 hey Nov 05 15:22:38 hi kergoth Nov 05 15:25:50 join #reddit Nov 05 15:39:41 good morning Nov 05 15:40:14 I'm trying to build the gumstix-basic-image but it's failing almost immediately when trying to unpack libtool Nov 05 15:40:24 I was following the instructions on the wiki: http://www.gumstix.net/wiki/index.php?title=Build_Environment_Ubuntu_9.04 Nov 05 15:43:17 after rebasing OE from upstream I'm getting several of these: ERROR: Multiple .bb files are due to be built which each provide XXXX (e.g. xfvwm4) (/path/to/XXXX_4.6.1.bb /path/to/XXXX_4.4.1) Nov 05 15:43:38 I don't know why it still wants to build the old version of those Nov 05 15:53:15 th1_: is the Gumstix setup building Xfce? I did not realise many people used 4.4.1 still (when I asked on the mailing list) when I pushed the upgrade to 4.6. What's in the gumstix-basic-image? Xfce seems OTT for a 'basic' image. Nov 05 15:55:28 DJWillis, sorry it was 4.4.2 Nov 05 15:55:39 but it is not gumstix, it's XenClient our own custom project. Nov 05 15:56:13 I've tried replacing my dependency on task-xfce-base with task-xfce46-base hoping it might be that which pulls in the old version Nov 05 15:56:51 th1_: Sorry, mixed you and tzanger up there ;-). Anyway, 4.4.2 is still old, I am suprised you have a prefered version in there somewhere, that would normally cause that sort of issue. Nov 05 15:57:06 I don't have it in my custom tree anywhere Nov 05 15:57:22 th1_: something old and Xfce 4.4 in there like xffm or xcalender? Nov 05 15:57:35 but after replacing task-xfce-base with task-xfce46-base in my image recipe it's not showing it anymore for xfwm (yet at least) Nov 05 15:57:57 no there's nothing like that in there actually Nov 05 15:58:03 I am using the task-xfce46-* for a few builds without any issues with hte 4.4.* stuff still in tree. Nov 05 15:58:09 ok Nov 05 15:58:42 How many recipies do it other than XFWM4? Nov 05 15:58:48 just gconf Nov 05 15:58:59 probably down to network manager applet Nov 05 15:59:09 which I have a custom hacked version for... Nov 05 15:59:46 I would expect any odd vesion stuff to call in more then just XFWM4, hmmm, what DISTRO if any? Nov 05 16:00:10 ahh yes xfce4-panel and sundy others as well :) Nov 05 16:00:14 but that has gone with using the 46 task Nov 05 16:00:39 DISTRO="angstrom-2008.1" Nov 05 16:05:01 th1_: do you have your image file or task file somewhere public? Nov 05 16:06:40 Hello... I'm tring to remove a dependency under the angstrom distro that is confounding me: Missing or unbuildable dependency chain was: ['virtual/kernel', 'u-boot-mkimage-openmoko-native'] Nov 05 16:07:08 I'm not using openmoku, and I've set up my own linux kernel and u-boot (local overlay) recipes Nov 05 16:12:22 DJWillis, let me find out ... :) Nov 05 16:16:46 ahhhhh... nevermind... i found the hard dependency under kernel.bbclass. don't know why my grep didn't pull this up earlier.. Nov 05 16:20:11 * DJWillis wonders what was wrong with the existing mkimage-native that OM needed to override it? Nov 05 16:20:52 DJWillis: other way round Nov 05 16:21:17 DJWillis: the OM version was first, people didnt like seeing the name openmoko in build logs and created the other one Nov 05 16:22:04 and openmoko one is ugly due to lot of patches applied Nov 05 16:22:14 uboot-utils-native do the same iirc Nov 05 16:22:27 XorA: Mea culpa! Nov 05 16:23:04 the OM version could go away now, but no-one can be bothered :-D Nov 05 16:23:10 when a bitbake script fails, is there somewhere I can look to see the actual compiler output? Nov 05 16:23:16 or a way to tell bitbak to put that up on the screen for me? Nov 05 16:23:46 tzanger: it will tell you where the log you want to look at is ;-) Nov 05 16:26:20 DJWillis: hmm, I'm not seeing it but I do see the compiler errors Nov 05 16:26:35 stupid "error out on warnings" is fine, but for output of asprintf? gimme a break Nov 05 16:27:51 wtf Nov 05 16:27:59 (void)asprintf(...) should be enough to shut that warning up Nov 05 16:34:27 hmm very weird errors, maybe I really should have downgraded the compiler :-) Nov 05 16:34:38 gcc-4.4.1 doesn't like it, gcc-4.1 is supposed Nov 05 16:34:41 er supposed to Nov 05 16:45:15 03Koen Kooi  07org.openembedded.dev * re9c3ba01c8 10openembedded.git/recipes/gnome-mplayer/ (gecko-mediaplayer_0.9.8.bb gnome-mplayer_0.9.8.bb): Nov 05 16:45:15 gecko-mediaplayer: fix packaging Nov 05 16:45:15 gnome-mplayer: bump PR Nov 05 16:45:16 The symlinking should be done in a postinst, but I'm too low on caffeine right now to make that work Nov 05 16:49:04 hmm, is there a way to tell bitbake to use a specific gcc/g++? I'd really, REALLY rather not have to have the entire computer using the older compiler Nov 05 16:49:27 as your native compiler, you mean? Nov 05 16:51:12 correct Nov 05 16:51:20 I'd like to continue to use gcc 4.4.1 for everything else on the system Nov 05 16:51:28 see ${BUILD_CC} in bitbake.conf Nov 05 16:51:32 but I'd like to tell OE to use gcc 4.1 to build the cross compiler and tools Nov 05 16:53:35 aha, that looks like what I want to modify. perfect Nov 05 16:53:51 now that I've done that though, what's the command to clean out everything I've built so far so I can start over? Nov 05 16:54:09 rm -rf tmp? Nov 05 16:54:13 bye all Nov 05 16:54:21 see you on oedem Nov 05 16:54:24 really, that simple? groovy. Nov 05 16:54:26 oh Nov 05 16:54:32 as you can see I'm a little new to OE and bitbake :-) Nov 05 16:54:34 bye hrw Nov 05 16:54:41 yeah, everything in tmp is expendable Nov 05 16:54:53 the downloads aren't in there are they? Nov 05 16:55:02 tzanger, no Nov 05 16:55:03 shouldn't be. they usually go to ~/sources/ or some such Nov 05 16:55:16 you can configure that in local.conf though so it might be worth checking what you have set. Nov 05 16:55:38 seems unlikely that you would have accidentally set DL_DIR to be a path inside tmp, but... :-} Nov 05 17:01:37 :-) Nov 05 17:01:39 thanks again guys Nov 05 17:08:48 see you tomorrow in UK Nov 05 17:25:45 03Jeremy Lainé  07org.openembedded.dev * r56db8bc2ab 10openembedded.git/conf/bitbake.conf: Nov 05 17:25:45 bitbake.conf: add support for "cpio.lzma" images Nov 05 17:25:45 * add support for LZMA-compressed cpio images for use as Linux Nov 05 17:25:45 RamDisk images. Nov 05 17:26:07 hmm, that didn't seem to change the compielr used, bugger. Nov 05 17:26:17 03Brijesh Singh  07org.openembedded.dev * r0816c962bd 10openembedded.git/recipes/ti/ti-dmai.inc: Nov 05 17:26:17 ti-dmai_svn: bumup svn rev Nov 05 17:26:17 This has proper CPU detection, so no need to manually add platforms anymore Nov 05 17:28:14 hi Nov 05 17:28:32 is somebody knows how to launch a QT embedded application ? Nov 05 17:28:58 is there a framebuffer manager to launch before running the application ? Nov 05 17:29:08 qws ? Nov 05 17:33:18 you can run it with -nodaemon or so Nov 05 17:33:26 cant remember exactly Nov 05 17:35:18 okay bye till later Nov 05 17:45:58 ok I manage to run a qte application : ./qte-example -qws Nov 05 17:46:06 all is good ! Nov 05 18:16:06 hi guys, have an Angstrom install on Zaurs c760, booting from SD, it used to boot fine. I have installed a whole host of packages and rebooted and now it hangs at usb0: RNDIS ready. odd thing is the version i flashed to the internal flash does this too, but always did from the first time i flashed it Nov 05 18:16:42 any ideas what casues that? Nov 05 18:17:02 as it boots from SD i shoul din thoery be ablel to fix the issue form PC Nov 05 19:16:05 has anyone experienced an issue in compiling connman_git.bb with "signal" being redefined in bluetooth.lo? Nov 05 19:47:06 03Artur Wronowski  07org.openembedded.dev * r69026f3da9 10openembedded.git/ (conf/checksums.ini recipes/libpar2/libpar2_0.2.bb): libpar2: new recipe; multi-part archives library Nov 05 19:47:16 03Artur Wronowski  07org.openembedded.dev * r9a06361338 10openembedded.git/recipes/nzbget/ (nzbget.inc nzbget_0.6.0.bb nzbget_0.7.0-testing.bb): nzbget: new recipe; a command-line based binary newsgrabber for nzb files Nov 05 19:47:17 re Nov 05 19:56:41 hi florian Nov 05 20:06:25 good evening everyone. i'm trying to extract angstrom's linux-rp kernel .config i thought i could make it with bitbake -c menuconfig virtual/kernel Nov 05 20:06:42 but this starts menuconfig and the generated.conf seems to be a default one Nov 05 20:06:49 am i missing something here? Nov 05 20:09:26 03Michael 'Mickey' Lauer  07org.openembedded.dev * r97ae177996 10openembedded.git/conf/machine/htcdream.conf: conf/machine/htcdream.conf: mention commercial device names in description Nov 05 20:09:37 03David Lanzendörfer  07org.openembedded.dev * r1d86318cd9 10openembedded.git/recipes/linux/ (12 files in 10 dirs): linux-replicant: new recipe; kernel tree for HTC devices Nov 05 20:24:12 gm all Nov 05 20:28:37 urg. Nov 05 20:28:52 unifdef-native redefines getline() Nov 05 20:29:08 mickey|zzZZzz: I need to talk to you about python and these staging changes at some point ;-) Nov 05 20:29:31 hi khem Nov 05 20:30:03 RP: I will cherry pick your branch into my local branch Nov 05 20:30:07 this weekend Nov 05 20:30:35 khem: ok, sounds good. My testing in Poky has gone well Nov 05 20:30:49 My OE testing is less complete but also showed no problems Nov 05 20:31:06 * kergoth should try it in mvl6, too Nov 05 20:31:31 tzanger: hmmm it must be fixed getline is a POSIX function. May be unifdef should rename it to mygetline or something else Nov 05 20:31:48 Coverage wise for poky, 205/972 recipes to convert to new staging. 59 -native recipes not using BBCLASSEXTEND remain Nov 05 20:32:00 cool Nov 05 20:32:24 and of that -59, a load only have a -native version Nov 05 20:35:01 tzanger: hmmm OE version of unifdef is fixed on .dev Nov 05 20:35:25 khem: that's exactly what I did, was make it mygetline Nov 05 20:36:05 tzanger: thats right approarch Nov 05 20:41:08 thanks Nov 05 20:56:08 for anyone who might be interested, there seems to be a variable name conflict between connman_0.42 and /usr/include/signal.h in the connman/plugins/bluetooth.c file. Both define a variable "signal", but of different types. Upstream, connman changed this variable in their package to "watch" -- see version 0.46. Unfortunately, the "sane-rev" for connman in OE is stuck at 0.42, so you have to grab the new GIT tag and override it in your Nov 05 20:56:08 local.conf. (This message is mostly for the IRC sniffers for those who may Google search for the problem.) Nov 05 21:10:33 you can post a patch to ml Nov 05 21:11:43 khem: yeah, i have a number of patches to submit. i'm queueing them up for when we release our product. Nov 05 21:12:42 kg4ysn: nice Nov 05 21:13:31 khem: does OE have their patch submission guidelines anywhere? I'm used to the kernel's process, but wasn't sure if there were differences preferred by the OE team. Nov 05 21:13:53 its same almost Nov 05 21:14:15 there is a wiki link Nov 05 21:14:38 khem: ok i've seen that. it was fairly sparse, so i'll just do my best and hope it works. :-) Nov 05 21:14:47 khem: thanks for confirming Nov 05 21:15:59 yeah, git-format your patches with signed-offs and git send-email them to openembedded-devel ml Nov 05 21:16:43 they will then appear in the patchwork.openembedded.org and reviewed and applied Nov 05 21:16:47 i'll try to make them an early christmas present then Nov 05 21:17:08 heh that might be a time when activity is low Nov 05 21:22:12 has anybody done any performance tests with qt4-embedded vs qt4-x11 ? Nov 05 21:28:37 qsp: you are right, there is a problem with menuconfig Nov 05 21:29:24 jest: and there is some workaround? Nov 05 21:29:41 qsp: I'm trying with linux-2.6.28 and have the same effect as you did Nov 05 21:30:01 as i've been telling in the last days here, i'm configuring a kernel for c700 cause angstrom one wont boot Nov 05 21:30:34 so i wanted to get current config and then try to meddle with the options :-) Nov 05 21:31:16 so it took me 2 days to compile a kernel with bitbake :-) many obstacles, and this is the latest one :-) that i cant extract the .config :-) Nov 05 21:31:17 qsp: I just tried 'bitbake virtual/kernel' and now I have two files, .config and .config.old. Of course the same, but this allows me to copy .config and reuse later Nov 05 21:31:49 mm, i'll try then Nov 05 21:33:17 qsp: try this: Nov 05 21:33:17 1. bitbake virtual/kernel -c configure Nov 05 21:33:17 2. bitbake virtual/kernel -c menuconfig Nov 05 21:33:17 3. bitbake virtual/kernel Nov 05 21:33:39 qsp: I try it as well right now, I'll let you know Nov 05 21:33:52 thank you! :-) starting the process :-) Nov 05 21:40:04 bitbake virtual/kernel -c listtasks will show you available tasks Nov 05 21:40:18 for given recipe Nov 05 21:40:26 yes, i was inspecting the tasks Nov 05 21:40:31 and i tried unpack one Nov 05 21:40:35 03Stanislav Brabec  07org.openembedded.dev * r3cc58ee273 10openembedded.git/recipes/ncurses/ncurses.inc: ncurses: Always use --enable-static to work-around possible linking problems. Nov 05 21:40:39 thought it would just unpack the sources Nov 05 21:40:43 unpack just creates src tree Nov 05 21:40:46 03Stanislav Brabec  07org.openembedded.dev * r11f33f70f6 10openembedded.git/recipes/gdb/gdb_7.0.bb: gdb-7.0: Always use --enable-static to work-around possible linking problems. Nov 05 21:41:21 thats it.. it didnt touch .config Nov 05 21:41:33 i'm now trying as per jest suggestion Nov 05 21:41:58 :-) should open champagne when i finally can boot my zaurus again :-D Nov 05 21:43:49 03Stanislav Brabec  07org.openembedded.dev * r21cb1d0889 10openembedded.git/recipes/gnome/gnome-icon-theme_2.28.0.bb: gnome-icon-theme-2.28.0: Set it as arch-independent package. Nov 05 21:46:04 qsp: looking at do_menuconfig, it simply does 'make menuconfig'. So, no special magic regarding even creating .config Nov 05 21:46:44 qsp: and do_configure is different, as it creates .config based (somehow) on defconfig Nov 05 21:47:01 its making the configure task Nov 05 21:47:07 just, so that i learn it as well Nov 05 21:47:16 how do you see the details of the task? Nov 05 21:47:39 qsp: because both do_configure and do_menuconfig goes after do_patch, I think it should work to do 'configure' and then 'menuconfig' Nov 05 21:48:18 mm... it failed :-/ Nov 05 21:48:20 qsp: I run a specified task, and then look at run.do_... files in work Nov 05 21:48:55 qsp: this way you see exactly what is going on, after any magic of bitbake, inheritance, etc. Nov 05 21:49:27 ah, so i should tail them :-) Nov 05 21:51:15 Melo (git server) will be down for about 20 minutes over the next 1.5 hours Nov 05 21:51:17 03Koen Kooi  07org.openembedded.dev * r25904bc071 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.org:openembedded into org.openembedded.dev Nov 05 21:51:20 03Koen Kooi  07org.openembedded.dev * rb8d29529be 10openembedded.git/recipes/linux/ (2 files in 2 dirs): linux-omap-pm 2.6.29: add hack to get asoc to register on the touchbook Nov 05 21:51:24 khem: yes, just its not always obvious what each task do Nov 05 21:51:26 qsp: slowly... after 'configure' I have .config which is exactly the same as 'defconfig', with exception of CONFIG_LOCALVERSION Nov 05 21:51:29 we need to replace some RAM Nov 05 21:52:07 jest:then its exactly what i need :-) Nov 05 21:52:15 cbrake: how much are you putting on Nov 05 21:52:19 maybe my machine is slower here, cause its still running Nov 05 21:52:41 qsp: so, you don't need to add manual changes to .config? Nov 05 21:52:55 yes, i will Nov 05 21:53:04 after i get the default ones Nov 05 21:53:24 khem: I'm not actually sure Nov 05 21:53:38 jest: i'll compare with kexecboot kernel, which is stripped down, and see what might be causing it not to boot in my zaurus Nov 05 21:53:55 khem: we're hoping to upgrade the Melo VM to 1.5G, but the machine itself has something crazy like 24G Nov 05 21:54:04 cbrake: it will be faster afterward I hope Nov 05 21:54:09 jest: as everyone told me so far, it might be a mem issue, cause the c700 is only 32mb Nov 05 21:54:14 khem: have you noticed it being slow? Nov 05 21:54:28 cbrake: I think its been good Nov 05 21:54:38 not that I clone often Nov 05 21:54:44 but pulls are ok Nov 05 21:54:50 khem: nod, neither do I Nov 05 21:55:13 someone cloning frequently could tell better Nov 05 21:56:13 qsp: what 'configure' does is 'make oldconfig'; what is interesting, it calls this target *twice* :O Nov 05 21:57:30 i did try a bitbake -c oldconfig before, but it isnt a task on its own, so you just found exactly what i was looking for :-) thank you. can't wait for it to finish Nov 05 21:58:05 03Koen Kooi  07org.openembedded.dev * rc1ff4bb5b8 10openembedded.git/ (conf/checksums.ini recipes/webkit/arora_0.10.1.bb): Nov 05 21:58:05 arora: add 0.10.1. Arora is a qtwebkit based browser. Nov 05 21:58:05 This recipe is ugly because neither I nor the upstream authors know how to use qmake Nov 05 21:58:05 screenie: http://scap.linuxtogo.org/files/25c0d644dfabb484382183269e8af286.png Nov 05 22:03:22 qsp: looks like the procedure I gave you works, i.e. it creates default, correct .config for my machine and let me change anything with menuconfig, and then build the kernel Nov 05 22:04:04 qsp: I can't confirm fully, as I don't know how to check contents of the kernel image file without booting it Nov 05 22:04:52 hmm Nov 05 22:05:21 libtool-cross is erroring out on decompress with a broken pipe... any ideas on how to get around that? it decompresses just fine by hand but it still tries to do it again when I re-run bitbake Nov 05 22:05:47 sounds like you're using an oe tree that doesn't have the subprocess SIGPIPE fix Nov 05 22:06:14 possible, it's the gumstix-oe tree Nov 05 22:09:25 cd upstreamoe; git log -SPIPE org.openembedded.dev -- classes/base.bbclass Nov 05 22:13:48 I'll have to pull it manually, they use svn and git won't let me do what you're suggesting :-) Nov 05 22:14:45 i said "cd upstreamoe" for a reason. upstream oe isn't svn. Nov 05 22:15:15 khem: hi, any idea how to circumvent this (new kernel headers brought this issue): kernel/qpixmapcache.cpp:125:57: error: asm/page.h: No such file or directory Nov 05 22:19:04 ant__: search for this include and delete it in the sources Nov 05 22:19:19 ah :) Nov 05 22:19:29 provided by modern libc? Nov 05 22:19:56 I had found it in klibc Nov 05 22:21:13 oh Nov 05 22:21:16 and bluelightning found out this: http://www.nntp.perl.org/group/perl.perl5.porters/2007/02/msg121370.html Nov 05 22:21:22 klibc is probably not so modern Nov 05 22:21:48 nkhem: was only found by grepping, no use of klibc in this project :) Nov 05 22:23:07 kergoth: that fixed it, many thanks. Nov 05 22:23:47 np Nov 05 22:24:03 ant__: which packge is this Nov 05 22:24:54 he qte-mt_2.3.10.bb Nov 05 22:25:04 (opie-image) Nov 05 22:26:24 ant__: you should try to get rid of including this file and replacing the functionality **** ENDING LOGGING AT Thu Nov 05 22:35:36 2009 **** BEGIN LOGGING AT Thu Nov 05 22:46:16 2009 Nov 05 22:46:54 Jay7: we still miss Pavel Nov 05 22:47:39 ant__: I see.. Nov 05 22:47:44 we are working on this :) Nov 05 22:48:28 btw I fear that Laibsch did NOT yet send a serial cable to Pavel :/ Nov 05 22:48:28 florian: I suspect some problems.. Nov 05 22:48:55 florian: I've sent some test mails to zaurus-devel@ but w/o any visible success Nov 05 22:49:00 Jay7: oh no :-( Nov 05 22:49:13 what is graylisting timeout? Nov 05 22:49:21 (if any) Nov 05 22:50:11 ant__: do Laibsch still have zaurus? Nov 05 22:50:33 think so, but he has some cables Nov 05 22:50:42 iirc Nov 05 22:50:59 seems we should catch him then :) Nov 05 22:51:27 Jay7: the default ;) Nov 05 22:51:39 one month? Nov 05 22:51:43 :D Nov 05 22:51:50 :)) Nov 05 22:51:58 well... Nov 05 22:52:05 let's wait then.. Nov 05 22:57:06 Jay7: hooray \°/ Nov 05 22:57:21 florian: congrats Nov 05 22:59:30 ~20 min is gone from last mail to list was sent :( Nov 05 22:59:31 Jay7: okay Nov 05 22:59:39 hm.. Nov 05 22:59:44 lol Nov 05 22:59:51 smart bot :) Nov 05 22:59:57 ibot: what is 20 min Nov 05 22:59:58 well, 20 min is gone from last mail to list was sent :( Nov 05 23:00:02 hahaha Nov 05 23:00:09 hmpfff Nov 05 23:00:13 good teaching there Nov 05 23:00:25 good english too Nov 05 23:00:30 ibot: forget 20 min Nov 05 23:00:31 i forgot 20 min, bluelightning Nov 05 23:00:54 ~botsnack Nov 05 23:00:54 well.. i should replace ~ with about then :) Nov 05 23:00:55 bluelightning: :) Nov 05 23:01:17 bluelightning: don't harrass the pet! Nov 05 23:01:26 ~lart ant__ Nov 05 23:01:26 * ibot gives ant__ an extra strength ACME sleeping pill, sending ant__ to sleep for 150 years, and awakening to seven strange dwarfs and a large apple Nov 05 23:01:42 hmm... Nov 05 23:01:58 omg Nov 05 23:09:34 Jay7: its supposed to be 10 minutes Nov 05 23:10:07 florian: hm.. Nov 05 23:10:35 Jay7: it does not seem to work as expected, disabling it now Nov 05 23:12:25 florian: can you look into smtp-server log for mail from jay4mail at gmail.com? Nov 05 23:13:32 * florian takes a look Nov 06 00:41:08 * Jay7 -> sleep **** ENDING LOGGING AT Fri Nov 06 02:59:57 2009