**** BEGIN LOGGING AT Thu Mar 26 02:59:57 2009 Mar 26 06:27:16 03Jeremy Lainé  07org.openembedded.dev * r1b045e71ce 10openembedded.git/recipes/python/python-pyqt_4.4.3.bb: python-pyqt: only apply patches for qreal on arm/mips machines Mar 26 07:23:13 morning Mar 26 07:45:19 good morning :) Mar 26 07:52:24 morning Mar 26 07:52:37 i have built a image with enlightment as my window manager, when i try to play mplayer, i get a enlightment loading message and nothing comes on the screen, when i try to play mplayer from command line it works Mar 26 08:32:17 Hello, I try to remove --hash-style=gnu from my compilation. I try to set TARGET_LINK_HASH_STYLE in local.conf or export it in the shell, but it's not taken. Mar 26 09:11:37 why doesnt open embedded create a /usr/share/application/mplayer.desktop file when i select mplayer with a x11 Mar 26 09:12:51 vsr: is a bug, you have to add a patch ;-) Mar 26 09:13:58 good morning Mar 26 09:14:07 morning Mar 26 09:14:24 anybody been trying to build latest boost (1.38.0 or even svn) in OE? Mar 26 09:14:39 I made some feeble attempts... but the boost build procedure is rather complicated :-/ Mar 26 09:15:15 (I mean... there's no recipe for 1.38.0, I tried to hack one) Mar 26 09:17:40 mckoan: i will try Mar 26 09:19:27 does someone know how Angstrom its IMAGE_CMD_ set? Mar 26 09:19:51 I'd like to try squashfs instead of jffs2 Mar 26 09:20:42 or maybe a better question: where does Angstrom choose its IMAGE type? Mar 26 09:21:49 i think these .desktop files are generated on the fly Mar 26 09:21:58 is there any place i should look? Mar 26 09:23:31 seems to be in classes/image.bbclass Mar 26 09:23:54 mpoullet|work: can't you just add it to IMAGE_FSTYPES in you local.conf? Mar 26 09:24:57 tsjsieb: you're right get_imagecmfs looks at this variable, thanks, will try this then :) Mar 26 09:25:21 vsr: recipes/mplayer Mar 26 09:39:21 mckoan: i am not sure. since most gui packages like gxine, pidgin dont have a .desktop file in thier respective reciepe diretories Mar 26 09:45:16 what does this mean: NOTE: base_set_filespath usage is deprecated, squashfs-lzma-tools-native-3.1r2 should be fixed ? Mar 26 09:45:33 vsr: if there isnt one in the tarball we normally just make one and copy it in manually in the recipe Mar 26 10:12:11 03Marcin Juszkiewicz  07org.openembedded.dev * ra7be854977 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 26 10:12:12 03Marcin Juszkiewicz  07org.openembedded.dev * r4921c7b07a 10openembedded.git/recipes/linux/ (linux_2.6.29-rc5.bb linux_2.6.29-rc6.bb linux_2.6.29.bb): linux: moved vortex86sx to 2.6.29 Mar 26 10:12:12 03Marcin Juszkiewicz  07org.openembedded.dev * re2718269c5 10openembedded.git/conf/checksums.ini: checksums.ini: added some new entries Mar 26 10:13:22 morning Mar 26 10:13:36 hi hrw Mar 26 10:14:41 hey hrw jumping on the 2.6.29 band wagon I see :-D Mar 26 10:14:56 XorA: I tend to use latest ;) Mar 26 10:15:13 XorA: 2.6.29 for vortex I had when 2.6.29 appeared - forgot to push. Mar 26 10:15:20 so I had to solve merge conflicts Mar 26 10:15:28 hrw: hehe Mar 26 10:15:54 tosa is working well with 2.6.29 so far :-) Mar 26 10:16:41 my desktop waits for reboot to have 2.6.29 Mar 26 10:16:46 rc8-something now Mar 26 10:17:01 ok, time to fetch some beagleboard images/kernels Mar 26 10:17:47 * XorA needs to finish that defconfig for eee Mar 26 10:43:21 03Jan Lübbe  07jan/master * read2bc340c 10openembedded.git/ (3 files in 2 dirs): tangogps: upgrade to 0.9.6 Mar 26 10:43:22 03Jan Lübbe  07jan/master * r2d3dbdee87 10openembedded.git/recipes/images/fso-image.inc: fso-image: add cellhunter Mar 26 10:43:22 03Jan Lübbe  07jan/master * r03b733eb9e 10openembedded.git/recipes/freesmartphone/ (menu-freesmartphone.bb menu-freesmartphone/applications.menu): menu-freesmartphone: create a simple XDG menu for illume which shows all applications Mar 26 10:43:24 03Jan Lübbe  07jan/master * r45fa1f32b3 10openembedded.git/recipes/e17/e-wm_svn.bb: e-wm: split out XDG menu definition into e-wm-menu Mar 26 10:43:27 03Jan Lübbe  07jan/master * r80550a6ea6 10openembedded.git/recipes/images/fso-image.inc: fso-image: use menu-freesmartphone and drop the .desktop file hack Mar 26 10:43:30 03Jan Lübbe  07jan/master * r3685c3d034 10openembedded.git/recipes/freesmartphone/frameworkd_git.bb: frameworkd: depend on pyalsaaudio for ADC/DAC configuration Mar 26 10:43:35 03Angus Ainslie  07jan/master * rffe44fe293 10openembedded.git/recipes/openmoko-projects/openmoko-qa-scripts.bb: Add openmoko QA testing scripts Mar 26 10:43:38 03Angus Ainslie  07jan/master * r50d59a42e9 10openembedded.git/recipes/tasks/task-openmoko-feed.bb: (log message trimmed) Mar 26 10:43:41 Add some new packages to the feeds Mar 26 10:43:43 - i2c-tools Mar 26 10:43:45 - minicom Mar 26 10:43:47 - cu Mar 26 10:43:49 - abiword Mar 26 10:43:51 - leafpad Mar 26 10:43:53 03Jan Lübbe  07jan/master * r3284518f7d 10openembedded.git/ (conf/checksums.ini recipes/bluez/bluez4_4.31.bb): bluez4: bump to 4.32 Mar 26 10:43:56 03Jan Lübbe  07jan/master * rb845f79ef6 10openembedded.git/recipes/base-files/base-files/ (om-gta01/fstab om-gta02/fstab): Mar 26 10:43:59 base-files/om-gta02: if /media/card/bind-home exists, bind-mount it to /home/root Mar 26 10:44:01 This will make live easier for those who reflash their system often. Mar 26 10:44:13 03Jan Lübbe  07jan/master * rf814b1e2c7 10openembedded.git/ (conf/checksums.ini recipes/xtrace/xtrace_0.10.1.bb): xtrace: add xtrace, a X11 tracer Mar 26 10:44:16 03Jan Lübbe  07jan/master * rfc8a3b9bc4 10openembedded.git/recipes/bluez/bluez4_4.32.bb: bluez4: stage the include files Mar 26 10:44:21 03Jan Lübbe  07jan/master * r5d20bb7801 10openembedded.git/conf/distro/openmoko.conf: openmoko: switch to bluez4 Mar 26 10:44:24 03Jan Lübbe  07jan/master * re4a3efd808 10openembedded.git/ (4 files in 4 dirs): openobex: add version 1.5 Mar 26 10:44:27 03Jan Lübbe  07jan/master * rd405cebf41 10openembedded.git/ (conf/checksums.ini recipes/bluez/obexd_0.9.bb): obexd: new receipe Mar 26 10:44:30 03Jan Lübbe  07jan/master * r781c26e6cc 10openembedded.git/ (conf/checksums.ini recipes/tinymail/libtinymail_0.0.9.bb): libtinymail: add 0.0.9 Mar 26 10:44:37 03Jan Lübbe  07jan/master * ra7b970c47a 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs: bump openmoko-alsa-scenarios in include gsmbluetooth.state Mar 26 10:44:40 03Angus Ainslie  07jan/master * r663328b0bd 10openembedded.git/recipes/efl1/ecore-native_svn.bb: ecore-x needs to be enabled for edje-sdk to compile Mar 26 10:44:43 03Stefan Schmidt  07jan/master * r7ee7501e42 10openembedded.git/ (conf/checksums.ini recipes/connman/connman_0.13.bb): connman: Add version 0.13 Mar 26 10:44:48 03Angus Ainslie  07jan/master * r588e6f1724 10openembedded.git/recipes/xorg-lib/libxext-native_1.0.4.bb: New libxext-native recipe Mar 26 10:44:51 03Angus Ainslie  07jan/master * r6eefb9570c 10openembedded.git/recipes/efl1/edje-sdk_svn.bb: Add enia library paths Mar 26 10:44:58 03Angus Ainslie  07jan/master * r74a53e093a 10openembedded.git/classes/openmoko2.bbclass: Allow openmoko2 to succeed on unsplitable sections Mar 26 10:45:01 03Jan Lübbe  07jan/master * r8cfc71e186 10openembedded.git/recipes/bluez/bluez4_4.32.bb: bluez4: use update-rc.d Mar 26 10:45:04 03Stefan Schmidt  07jan/master * rce20dec3b3 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions: Bump connman to 0.13 Mar 26 10:45:07 03Angus Ainslie  07jan/master * r83bf34c2c6 10openembedded.git/recipes/efl1/ecore.inc: ship ecore-input for python ecore bindings Mar 26 10:45:12 03Angus Ainslie  07jan/master * r079d74ad0c 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs: Bump EFL version to latest for new python bindings Mar 26 10:45:15 03Jan Lübbe  07jan/master * r51885a8d37 10openembedded.git/recipes/cellhunter/cellhunter_0.4.2.bb: cellhunter: fix RDEPENDS Mar 26 10:45:18 03Angus Ainslie  07jan/master * r0d1497b316 10openembedded.git/ (5 files in 4 dirs): Add orrery to the feed Mar 26 10:45:23 03Angus Ainslie  07jan/master * r30416e7990 10openembedded.git/recipes/xorg-xserver/xserver-xorg-conf/om-gta02/xorg.conf: Add xorg.conf file for om-gta02 Mar 26 10:45:26 03Stefan Schmidt  07jan/master * rdb00b7cd38 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: Mar 26 10:45:29 preferred-om-2008-versions: Use an older connman version that does not fail Mar 26 10:45:35 (2 lines omitted) Mar 26 11:08:54 03Koen Kooi  07org.openembedded.dev * re91b799bcb 10openembedded.git/recipes/dsplink/ (4 files): dsplink: bump PR on recipes for kernel changes Mar 26 11:10:22 hi mickey|cafe Mar 26 11:10:33 hi florian Mar 26 11:51:03 hi Mar 26 11:52:29 I am using the org.openembedded.dev branche and I recently started encountering a strange error when bitbake'ing some packages Mar 26 11:54:00 For example, when building tslib, do_patch fails because of a command error Mar 26 11:54:21 It just outputs the usage options for quilt Mar 26 11:54:50 has anyone encountered a similar problem? Mar 26 12:04:35 mickeyl: ping! Mar 26 12:04:54 yo? Mar 26 12:05:31 python-native fails at run-time. Mar 26 12:05:49 2.5 python in /usr/bin, execute python in staging: Mar 26 12:05:54 "import md5" Mar 26 12:06:02 results in error. Mar 26 12:06:08 sounds like the openssl issue Mar 26 12:06:17 when folks updated from 0.9.7 to 0.9.8 Mar 26 12:06:29 md5 is only tried to import when openssl fails to import Mar 26 12:06:38 and if the python in staging is linked to 0.9.7 Mar 26 12:06:40 it no longer runs now Mar 26 12:06:53 so you need to rebuild it Mar 26 12:07:54 build by two people now, from empty TMPDIR last night. Mar 26 12:08:34 ok, that's not it then Mar 26 12:08:38 no idea offhand then Mar 26 12:08:43 was abroard for 2 weeks Mar 26 12:09:08 one of the tons of changes to core stuff might have broken it Mar 26 12:09:09 :) Mar 26 12:09:16 lucky bugger Mar 26 12:09:21 * mickeyl launches a build Mar 26 12:10:06 mickeyl: hi Mar 26 12:10:11 hey hrw Mar 26 12:10:15 how are things? Mar 26 12:10:35 Ok. I just don't even know where to look. Google comes up with similar problems; general resolution is "uninstall one of the two copies of python and it goes away". I hacked my build by editing the bbclass to stick /usr/bin at the front of PATH before executing the python script that fails. And it only fails with hashlib/md5. Mar 26 12:10:38 I've been doing clean builds alot Mar 26 12:10:55 no issues and last one was a couple of days ago Ithink Mar 26 12:11:20 mickeyl: fine. got another toy, waiting for other to arrive, going to change car etc Mar 26 12:11:35 hrw: sounds like things are working well for you. glad to hear Mar 26 12:11:59 * mickeyl needs to find some new funding for FSO work this year Mar 26 12:12:01 mickeyl: and my direct boss is +8h from me ;D Mar 26 12:12:08 hehe Mar 26 12:12:09 mickeyl: openmoko ended? Mar 26 12:12:13 yep Mar 26 12:12:20 last 4 days Mar 26 12:12:44 mickeyl: are there any users other then openmoko and some random hackers? Mar 26 12:12:59 hrw: unknown Mar 26 12:13:12 spreading the word a bit this year to find out Mar 26 12:13:33 ~curse images Mar 26 12:13:34 May the fleas of a thousand camels infest your most sensitive regions, images ! Mar 26 12:13:40 my image lack /etc/group Mar 26 12:14:47 does anyone know how to build qt embedded 4.5? without X11? Mar 26 12:16:20 in the bb files of 4.4.3 I can't seem to find any reference to x11, but when dry running it still wants to build X11 as well Mar 26 12:16:54 clearly hal as wel depends on X11, which I find quite odd Mar 26 12:42:51 hi, the documentation branch is deprecated, but what does "Please commit doc updates to the main branch." mean? the .dev branch? Mar 26 12:43:10 hi, I want to fix the deprecation warning in one of my recipes in the Jalimo overlay Mar 26 12:43:37 unfortunately the comment in base.bbclass does not give a hint how it should be replaced Mar 26 12:43:43 so how do I fix: Mar 26 12:43:44 FILESPATH = "${@base_set_filespath([ "${FILE_DIRNAME}/openjdk-6-${PV}" ], d)}" Mar 26 12:44:14 I am using this for a couple of recipes to use the same directory for patches (although their PNs differ) Mar 26 12:45:20 mickeyl: sorry to hear that :-( Mar 26 12:45:51 mpoullet|work: ok, the user manual is now under docs/usermanual Mar 26 12:46:11 mwester: ya, it did come a bit as a surprise Mar 26 12:48:26 Q: If I was to port a robotic project to x86 (my ixp4xx slug just can't cut it anymore), what is the suggested OE distro? I need good package support and a lean & mean distro. Mar 26 12:48:55 I could use a plain desktop distro but I'd have to clean it up to make it RT friendly. Mar 26 12:49:22 * mwester wonders why x86, and why not (for example) a beagleboard, or the sheevaplug? Mar 26 12:50:15 Angstrom runs on the beagleboard, and timtimred just added sheevaplug support to Angstrom, IIRC. Mar 26 12:50:22 I'm open to suggestions as the smaller the better but I'm craving something faster than ARM for future vision projects. Mar 26 12:50:26 thebohemian: use just openjdk-6-${PV} and put recipes in one dir? Mar 26 12:50:40 thebohemian: or put common patches in files/ dir Mar 26 12:50:56 my code is all C based. Mar 26 12:51:08 VoodooZ_: sheevaplug has 1.2GHz arm Mar 26 12:51:32 VoodooZ_: I can't do any comparisons, but considering that the beagleboard can do full HD video, and the sheevaplug is a >1GHZ cpu with 512MB ram... Mar 26 12:51:39 I think I saw a post of that board... It's the wall-wart linux pc right? Mar 26 12:51:46 yep Mar 26 12:52:31 the problem is that these HD capable boards tend to make use of a application specific ASIC or chipset which I can't easily use for my cv stuff. Mar 26 12:52:32 the dev kit is a wall wart; with a "daughterboard" that enables comms with the outside world (jtag, console). Mar 26 12:52:46 yep. it did catch my attention... would be small and low power. Mar 26 12:52:56 Is there HW fp? Mar 26 12:53:31 In any case, Angstrom is probably what you would end up with on the target in terms of distro. The gap between Angstrom and SlugOS has closed to almost nothing recently. Mar 26 12:53:44 the reason I wanted x86 is that there's less dependence on cross-tools and it would be cheaper (I have an old P3-900) Mar 26 12:53:53 ok. Mar 26 12:54:56 I tried to no avail last night to get that toolchain working so I'm a bit frustrated.. I guess I could ask in here.. Mar 26 12:57:02 hrw: no, it unfortunately not that easy :( Mar 26 12:57:23 hrw: but with the new scheme it looks like defining FILESPATHPKG is the way to go Mar 26 12:57:32 what do people use in terms of sheevaplug implementation? the SDK board or? Mar 26 12:57:47 hrw: wrote you a mail regarding buglabs' poky btw Mar 26 12:58:31 thx Mar 26 13:01:43 ummm. the beagleboard's DSP looks good. at 149$ it's reasonable too Mar 26 13:19:02 tftp does not see, to answer the phone Mar 26 13:34:02 aloisiojr: hi! Mar 26 13:35:51 aloisiojr: I am about to commit some fairly big changes to qt4.inc, which will make it closer to your own packaging of Qt Mar 26 13:36:16 aloisiojr: unfortunately I no longer have the copy of your packaging to check the exact differences (pastebin expired) Mar 26 13:43:29 aloisiojr: we got qt-4.5 built with OE toolhain but out of OE tree now Mar 26 13:43:39 now is matter of includeing it in OE Mar 26 13:48:31 hmm, fund Mar 26 13:48:33 fun Mar 26 13:48:37 i shouldn't have updated OE Mar 26 13:48:39 nothing builds anymore :D Mar 26 13:48:50 ;d Mar 26 13:48:57 what did you guys do to it in the last week? Mar 26 13:48:58 *sigh*Ü Mar 26 13:49:07 mickeyl: join our quest for Holy Grail Mar 26 13:49:13 ^W^Wstable branch Mar 26 13:50:02 seems no cross compiler gets built Mar 26 13:50:09 ah well Mar 26 13:50:16 back to the last rev Mar 26 14:04:50 mickeyl: hiya Mar 26 14:05:02 03Jeremy Lainé  07org.openembedded.dev * rd392691750 10openembedded.git/classes/qmake2.bbclass: qmake2.bbclass: remove gratuitous export of QMAKE_RCC Mar 26 14:05:05 03Jeremy Lainé  07org.openembedded.dev * r613166bd06 10openembedded.git/recipes/qt4/ (qt4-embedded_4.4.3.bb qt4-x11-free_4.4.3.bb qt4.inc): qt4: overhaul packaging and staging rules Mar 26 14:05:32 hi jeremy_laine Mar 26 14:05:44 good to see you bringin qt in shape again Mar 26 14:06:06 mickeyl: yup, I am pretty confident things are pretty clean now Mar 26 14:06:29 mickeyl: I'd quite like to kill the uicmoc4-native and qmake2-native packages, as they are not used anymore Mar 26 14:07:07 mickeyl: now, building Qt (whether embedded or X11) involves: 1/ building qt4-tools-native 2/ building your target Qt flavour Mar 26 14:07:38 mickeyl: as a test I am going to try updating to Qt 4.5.0 (with a default_preference = -1 for now) Mar 26 14:10:02 so qt4-tools-native can be used if you have programs that use qmake but not Qt? Mar 26 14:10:22 [which was the reason we split out uicmoc and qmake from Qt in the first place] Mar 26 14:13:18 mickeyl: absolutely, for instance sip-native Mar 26 14:13:46 right, good Mar 26 14:14:47 hrw: when will the first stable branch appear? **** BEGIN LOGGING AT Thu Mar 26 14:18:21 2009 Mar 26 15:16:37 mickey|lunch: while(lunch)sleep(1); ping()!; Mar 26 15:19:10 Perhaps this is the problem with Python-native: openssl problem; 0.9.8 is built, except 0.9.7 for -native (due to DEFAULT_PREFERENCE = -1 in -native recipe). I'll set a preferred version so that both match, and try again with an emtpy TMPDIR. Mar 26 15:34:12 03Mike Westerhof  07org.openembedded.dev * r318dd21d6e 10openembedded.git/conf/distro/include/preferred-slugos-versions.inc: Mar 26 15:34:12 SlugOS: set preferred versions for openssl and openssl-native Mar 26 15:34:12 otherwise we get different versions for the two, due to conflicting Mar 26 15:34:12 DEFAULT_PREFERENCEs being set in the various recipes. Mar 26 15:39:37 I'm not sure that'll fix the python-native problem, but it's certainly not right without it. :) Mar 26 16:20:07 he updated us to j and then it broke Mar 26 16:20:08 which stefan? Mar 26 16:20:08 schmidt? Mar 26 16:20:08 there was an email about u-boot Mar 26 16:20:08 and using the git recipe for the latest u-boot Mar 26 16:20:08 ah, sledz Mar 26 16:20:08 no idea Mar 26 16:20:08 I forgot to disagree with usin gthe git recipe for that version :) Mar 26 16:20:09 Just checked, slugOS built. (I *LOVE* my quad-core, 29 minutes to build SlugOS from scratch.) Now I'll build the meta-toolchain, and see if the python problem goes away. :) Mar 26 16:20:09 I need to add a recent u-boot recipe Mar 26 16:20:09 I may need to get one of these quad things Mar 26 16:20:10 mmh Mar 26 16:20:10 *envy* Mar 26 16:20:11 i want a faster build machine as well Mar 26 16:20:11 need to save money now though Mar 26 16:20:11 Yeah. :-( Mar 26 16:20:11 we need to get ev finances setup Mar 26 16:20:23 * mwester_ sends Crofton|work to "Sales 101" for remedial training on proper phrasing for effective salesmanship :-D Mar 26 16:20:25 03Jeremy Lainé  07org.openembedded.dev * r31a8edfb70 10openembedded.git/recipes/qt4/files/ (3 files): qt4: refresh some patches against qt 4.4.3 Mar 26 16:20:25 I used a smiley Mar 26 16:20:49 We would say "and start enabling effective resource usage by .com people" or some other such marketing shlock. :) Mar 26 16:20:49 03Jeremy Lainé  07org.openembedded.dev * r27c22a41fd 10openembedded.git/recipes/qt4/qt4-tools-native-4.4.3/qt-config.patch: qt4: drop obsolete patch Mar 26 16:21:10 obviously the proper ay to do this involves convincing them the money is used wisely to increase the value of OE to their business Mar 26 16:21:18 Ah, there you go! Mar 26 16:21:30 See, we don't need a salesforce, we can do it ourselves. ;-) Mar 26 16:22:26 OE developer summit sponsored by "Your Name Here" Mar 26 16:22:47 For my project I use directfb-1.0.0 (not 1.2.7). That's why I have PREFERRED_VERSION_directfb = "1.0.0" in my machine config file. I'd like to apply project-specific patch for directfb. For this, I copy directfb_1.0.0.bb, directfb.inc and files/ directory into my overlay directory, modify bitbake for patch applying, update PR variable. But somehow OE build system doesn't compile mine "1.0.0-r7" package but generic "1.0.0-r6" one. WTF? Mar 26 16:23:24 Sell adverstising? "This git commit brought to you by the fine people and products of " Mar 26 16:23:40 mwester_: :-) Mar 26 16:23:41 Oh wait. sourceforge already rather does that. Mar 26 16:24:09 If I build directfb with 'bitbake directfb-1.0.0-r7' then OE compiles the proper recipe version but not for generic building through deps Mar 26 16:33:21 hi mickeyl Mar 26 16:33:27 hey pb_ Mar 26 16:33:31 how are things in GB? Mar 26 16:34:09 pretty bad in general, from what I hear, but things are okay in my little part of the country :-) Mar 26 16:37:04 hehe, good for you. Mar 26 16:39:25 tomorrow is my wife's last day at work, so I am making a list of things for her to do around the house during her weeks of leisure :-} Mar 26 16:39:26 hi mickeyl Mar 26 16:39:26 hi florian Mar 26 16:39:26 mickeyl: back home? Mar 26 16:39:27 pb_: hi! Mar 26 16:39:34 what's the trick for changing the path in which patches/files are picked up? Mar 26 16:39:34 pb_: hehe "leisure" Mar 26 16:39:37 I thought opkg ued that trick but not anymore it seems Mar 26 16:39:38 hi florian, yes back in FRA for three days. Monday departing to Bern for OpenExpo Mar 26 16:39:57 pb_: I just had you in mind looking at a small ARM based internet radio at an electronics store :) Mar 26 16:39:58 jeremy_laine: FILESPATH is how file:// urls are handled Mar 26 16:40:10 jeremy_laine: see the FILESPATH lines in bitbake.conf to see how it's set up Mar 26 16:40:11 mickeyl: uh, OpenExpo might be nice... looks like you are pretty busy Mar 26 16:40:23 florian: yes, especially now that Om is no longer funding FSO, I need to spread the word and try finding someone else Mar 26 16:40:40 kergoth: ehm, it's just for a single package Mar 26 16:40:45 yes, i know. Mar 26 16:40:45 kergoth: would that still apply? Mar 26 16:40:52 what bout it? Mar 26 16:40:54 mickeyl: oh... didn't know that. This is really bad news! Mar 26 16:40:57 florian: ah, very good. which one was it? Mar 26 16:41:31 kergoth: my case is the following: various qt4 packages use the same patches, so I put them in "files/" Mar 26 16:41:35 it automatically looks in files Mar 26 16:42:33 mickeyl: heh, yes, she seems to have a punishing schedule of shopping lined up. Mar 26 16:42:33 florian: yes. due to a number of unforseen circumstances, Om had to let go 50% of their personnel. they are sailing on a rough ocean this year. perhaps things will continue next year, but I can't wait for that :) Mar 26 16:44:17 pb_: hehe, wow. poor finances Mar 26 16:44:52 kergoth: this worked fine with a single Qt version, but now I am introducing Qt 4.5.0 and need versions of the patches for 4.5.0, shared across qt4-x11-free and qt4-embedded Mar 26 16:45:45 ah Mar 26 17:07:56 XorA: right now I'm focusing on HTC Touch Pro Mar 26 17:07:57 thats another of almost the same hardware? Mar 26 17:07:58 What about the glofish Harald is working on? Mar 26 17:07:58 hi woglinde Mar 26 17:09:07 XorA: ya, all the qualcomm msm7 based HTCs are somewhat similar Mar 26 17:09:16 and the Touch Pro i like best hardware-wise Mar 26 17:09:22 (glofish) well... Mar 26 17:09:23 hehe Mar 26 17:09:37 I shall have to see how well my kaiser is freed one day :-D Mar 26 17:09:50 (glofish) the modem doesn't work. and until this has not happened I'm not interested. Mar 26 17:10:03 we all know Harald is speedy starting things Mar 26 17:10:07 but slowly completing Mar 26 17:10:15 and it seems noone else is working on that Mar 26 17:10:22 so i don't have high hopes to be honest Mar 26 17:10:43 aloisiojr: my day in office is ending Mar 26 17:10:52 this is more of a git question but when I see messages in oe's git such as: 'Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev' what exactly is that? I thought org.openembedded.dev was the master branch? Mar 26 17:10:58 aloisiojr: I can came back later from home Mar 26 17:11:09 tharvey: this is when people are using git pull without --rebase Mar 26 17:11:21 mickeyl: Mar 26 17:11:22 mckoan: oh nice, i'll be here Mar 26 17:11:46 BTW about qt licenseing issue here is the official answer: We will license Qt under the LGPL for all platforms we currently support. This includes Windows, Linux/X11, Mac OS X, Windows CE, embedded Linux, and the upcoming S60 port. Mar 26 17:12:03 mickeyl: it might be a good idea to try involvign more people. e.g. Nils has one of these and has quite some kernel experience. I just wonder how to motivate him ;) Mar 26 17:12:21 i know the #1 motivation Mar 26 17:12:24 show something that works Mar 26 17:12:40 and with FSO and SHR we can do that soon Mar 26 17:13:02 HTClinux suffers from Android-wannabees Mar 26 17:13:09 so they just want to run Android Mar 26 17:13:17 aloisiojr: does the qt4-x11-free_4.4.3.bb work at the moment ? Mar 26 17:13:20 rather than working on standard interfaces Mar 26 17:13:29 which gives FSO a headache Mar 26 17:13:29 hi florian Mar 26 17:13:51 mickeyl, have to admin git is still pretty alien to me - so your saying because prior commits were not rebased, a general merge has to be done? Mar 26 17:13:54 hi woglinde Mar 26 17:14:21 tharvey: correct. you can either rebase your changes to a tree or merge the changes. in the first case, the merge message will not be seen Mar 26 17:14:27 mickeyl: waht a pain... so starting a proper project for the htc phones and gaining some publicity would be a good start Mar 26 17:14:33 correct Mar 26 17:14:49 when we meet at LT, I'll show you how far I am with Mar 26 17:14:51 it Mar 26 17:15:00 mckoan: i didn't test it yet... i was very busy with work stuffs... Mar 26 17:15:23 mckoan: i taked to jeremy_laine, maybe we should merge with his work too Mar 26 17:15:33 for a long time we have struggled in almost all areas Mar 26 17:15:37 but things are coming along now Mar 26 17:15:39 aloisiojr: I think would be better to verify that all is working with old version before start wirking with 4.5 Mar 26 17:15:39 FSO as middleware Mar 26 17:15:44 Elementary as UI-toolkit Mar 26 17:15:49 kernel 2.6 Mar 26 17:15:56 things looking better than ever imo Mar 26 17:16:14 aloisiojr: I'm facing to too many broken build issues with OE at the moment Mar 26 17:16:22 mickeyl, so when people fail to rebase their commit messages get obscured or just lost in another branch? Mar 26 17:16:29 mickeyl: hrm... maybe we maange to do somethgin good before lt Mar 26 17:16:35 * florian is optimistic Mar 26 17:16:57 aloisiojr: I can't rely on OE recipes more than one day before get broken Mar 26 17:17:00 tharvey: no, the merge is in _addition_ to the commits Mar 26 17:17:29 mckoan: i see.. yep we must keep all 4.x working Mar 26 17:17:48 mickeyl, ok thx Mar 26 17:17:53 mickeyl: mobile-linux at LTG is an active project. so this might be a good base. but it lacks some public infrastructure htclinux has - but a website, list and forum we should be able to create. Mar 26 17:18:10 now I launched a x11 build that hwr told me worked and discovered that here doesn't build Mar 26 17:18:32 so I'm ath the same point I was 8 working hours ago Mar 26 17:18:59 florian: actually, I'm pondering whether all these foo_linux project need to have own infrastructure Mar 26 17:19:06 mckoan: did you take a look in what jeremy_laine commit? Mar 26 17:19:12 i'd rather see HTCLinux, Openmoko, OpenEZX working more closely together Mar 26 17:21:54 mickeyl: right... that was the intention people hat developping at hh.org. mobile-linux might be a good successor but is bat at public relations. Mar 26 17:21:54 mckoan: he was wondering that something that i'm doing in the new bbclass shouldn't happen Mar 26 17:21:54 03Robert Schuster  07org.openembedded.dev * raf5fcda661 10openembedded.git/recipes/cacao/ (cacao_0.99.3.bb cacao_0.99.4.bb): Mar 26 17:21:54 cacao 0.99.3: Added stackbase fix patch. Mar 26 17:21:54 cacao 0.99.4: Dito. Mar 26 17:21:54 as normal recipes stageing native binaries Mar 26 17:21:54 hm the opengl es 2.0 sdk pc emu is about 80 mb archive Mar 26 17:21:54 he created qt4-native to be used instead of qmake-native and uicmoc-native packages Mar 26 17:23:11 mckoan: and actually he has a point... should we leave target recipes stage native binaries? Mar 26 17:25:01 aloisiojr: actually I don't know at the moment, I'm focused to getting a stable qt system Mar 26 17:25:09 see you later Mar 26 17:25:15 mckoan: see ya Mar 26 17:41:29 03Jeremy Lainé  07org.openembedded.dev * r49592361f8 10openembedded.git/ (8 files in 3 dirs): qt4: add qt4-x11-free version 4.5.0 Mar 26 17:41:30 03Jeremy Lainé  07org.openembedded.dev * r951b573a5b 10openembedded.git/: Merge branch 'org.openembedded.dev' of ssh://git@git.openembedded.net/openembedded into org.openembedded.dev Mar 26 17:42:04 aloisiojr, mckoan|away : ok I have commited an initial version of Qt 4.5.0, with DEFAULT_PREFERENCE = "-1" Mar 26 17:42:26 I am currently working on qt4-embedded 4.5.0, should be very quick Mar 26 17:46:50 jeremy_laine: was qte discontinued ? Mar 26 17:47:57 aliosiojr yes Mar 26 17:48:10 some parts will be merged with qt-embedded Mar 26 17:50:51 aloisiojr: what do you mean by "qte" ? Mar 26 17:51:21 qt-extended formerly know as qtopia Mar 26 17:51:33 if you mean Qt Extended, yes as woglinde said: http://www.qtsoftware.com/about/news/qt-software-discontinues-qt-extended Mar 26 17:52:24 aloisiojr: btw, which flavour of Qt do you use? X11 or Embedded? Mar 26 17:53:04 damn, I've been pounding my poor CPU over the past 48 hours, building and rebuilding Qt in every possible way Mar 26 17:53:14 03Jan Lübbe  07fso/milestone5.5 * r447862e734 10openembedded.git/packages/efl1/ecore-native_svn.bb: ecore-native: add libxext-native dependancy Mar 26 17:53:31 jeremy_laine: x11 Mar 26 17:53:33 jeremy_laine hehe I bought my a phenom2 4x Mar 26 17:53:36 me Mar 26 17:53:51 make -j4 is fun Mar 26 17:54:00 woglinde: actually to be fair, it's our company's build server, a quad core too Mar 26 17:54:14 hehe okay Mar 26 17:54:26 I am trying the opengl es stuff Mar 26 17:54:31 woglinde: it has a nice SCSI disk as well, it helps Mar 26 17:54:43 I am wonder if we have a chance to integrate it in oe somehow Mar 26 17:55:30 hm let me se the omap gl suff recipes from koen Mar 26 17:59:26 ah Mar 26 17:59:54 jeremy_laine would be nice if you would enable opengl es stuff inside qt for omap3 Mar 26 18:00:17 woglinde: ehm, what does that entail? Mar 26 18:00:54 woglinde: I don't know why opengl is currently disabled in Qt builds, I just picked up the existing flags Mar 26 18:01:10 jeremy faster redenring Mar 26 18:01:31 woglinde: yes, that I can imagine, but basically what deps does it add to QtGui ? Mar 26 18:01:48 woglinde: basically, get the thing working and send a patch or commit it :) Mar 26 18:01:58 libgles-omap3 Mar 26 18:02:31 and enable opengl Mar 26 18:03:57 ah Mar 26 18:04:08 hm didnt look upp at your recipes Mar 26 18:04:23 but I think we should disable the qt-designer compile Mar 26 18:04:30 woglinde: should be easy, look at qt4-x11-free_4.5.0.bb Mar 26 18:04:51 woglinde: remove "-no-opengl" from QT_CONFIG_FLAGS and add whatever deps are needed Mar 26 18:08:47 okay lunch now Mar 26 18:08:48 till later Mar 26 18:08:54 args Mar 26 18:08:56 diner Mar 26 18:08:59 woglinde: lunch ? Mar 26 18:09:29 ah Mar 26 18:25:13 bitbake -c rebuild -b Mar 26 18:25:21 should it clean and build ? Mar 26 18:25:30 rebuild is busted at the moment Mar 26 18:25:35 ah Mar 26 18:25:57 thats why I was wondering it was working few days back Mar 26 18:26:27 what broke it I wonder Mar 26 18:26:58 ah so it wasnt me going insane last night, its actually busted :-D Mar 26 18:27:13 I had written it off to tiredness Mar 26 18:27:34 khem: a change to bitbake. bb.build.exec_task no longer obeys deps Mar 26 18:28:01 khem: so a task can execute clean, and execute build, but it won't obey build's deps, and won't build other packages that any of those tasks need.. Mar 26 18:28:24 is it in cooker Mar 26 18:28:59 the dependency bits were removed from bb.build, bbecause it was duplicating code in the cooker. it was a cleanup commit that RP threw in Mar 26 18:29:57 you can make a custom do_rebuild that works for the most part, obeying task deps, but it'll only act on the package, wont' build that package's deps, so don't rebuild if it hasn't already been built at least once to build its deps.. Mar 26 18:30:54 is someone planning to fix it Mar 26 18:31:31 right now I am using a shell script to call the clean and build tasks as work around Mar 26 18:31:48 i have a do_rebuild that fits the description above, it works, as long as the package's deps are already built Mar 26 18:31:55 better than whats in there now, probably Mar 26 18:31:57 re Mar 26 18:31:58 still not perfect, but *shrug* Mar 26 18:32:00 hi khem Mar 26 18:32:07 and xora and kergoth Mar 26 18:32:34 * kergoth wonders what he did with that commit.. Mar 26 18:32:52 woglinde: how is new machine treating you Mar 26 18:32:56 kergoth: Mar 26 18:33:01 khem hm nice Mar 26 18:33:08 only graphic suckz a little bit Mar 26 18:33:16 because its ati hd3300 Mar 26 18:33:22 khem: its easy to write, it just pokes at the deps stamps, little recursive function that runs exec_task on each task, obeying the deps stamp contents Mar 26 18:33:22 ugh ati Mar 26 18:39:09 khem there is no phenom2 chipset with integrated nvidia yet Mar 26 18:39:26 woglinde: and AMD would want to use ATI Mar 26 18:39:41 03Chris Larson  07kergoth/master * r10b7252b5e 10openembedded.git/classes/ (base.bbclass packaged-staging.bbclass patch.bbclass): Mar 26 18:39:41 Shorten some full paths printed to the user. Mar 26 18:39:41 Adds a base_path_out convenience function, which prepares a full path for Mar 26 18:39:41 display to the user. The initial implementation just makes it relative to Mar 26 18:39:41 ${TOPDIR}. This function is then used for some messages outputted to the Mar 26 18:39:45 user (packaged-staging, patch application, clean, unpack tasks). Mar 26 18:39:47 Signed-off-by: Chris Larson Mar 26 18:40:56 khem hm yes Mar 26 18:41:17 hm for omap3 we can even enable the powervr driver in qt Mar 26 18:41:49 khem, patch for adding kernel-headers package to kernel.bbclass: http://www.pastebin.ca/1373203 - I added a var to specify install location and added to task-base-bare.bb - it works for me (I was doing something stupid yesterday when I said it wasn't) Mar 26 18:45:05 tharvey: why do you install them in usr/src/linux-headers-${KERNEL_VERSION}-${MACHINE} Mar 26 18:45:13 why not in /usr/include Mar 26 18:45:48 just seemed more appropriate... the debian way Mar 26 18:45:55 if anyone's bored, please test your builds with those two commits on the kergoth/master branch and see if you like the build output a bit better. kills some messages, cleans it up a little Mar 26 18:46:46 khem, you mean instead of /usr/include/linux-headers-${KERNEL_VERSION}-${MACHINE}? you definately need to have the version and machine in there Mar 26 18:47:24 tharvey: hmm ok, I see I think you are needing them for module development Mar 26 18:47:50 I was trying to see if we could use same headers for userspace Mar 26 18:49:54 yes, I need them for externally built modules - its possible this is useful for libc stuff but I'm not clear what the diff is between sanitized headers are and if theya re needed there Mar 26 18:50:06 you dont want to use the headers from the kernel as the /usr/include/{linux,asm} dirs, ever. Mar 26 18:50:11 there's a reason linux-libc-headers exists Mar 26 18:51:01 kergoth: he is generating sanitized headers Mar 26 18:51:25 thats what is essentially needed for glibc/userspace Mar 26 18:52:29 I'd trust headers that are manually sanitized over some generator, personally. not to mention we should be consistent with what every other distro is doing.. unless something's changed there that I don't know about, linux-libc-headers is still the way to go, and you still want the actual headers from the kernel in a separate location for external module builds Mar 26 18:52:40 * kergoth shrugs Mar 26 18:53:34 kergoth: from 2.6.18 onwards kernel provides the sanitized headers with headers_install Mar 26 18:53:42 ah, i stand corrected Mar 26 18:53:48 so we do not need to sanitze them manually Mar 26 18:53:52 anymore Mar 26 18:54:08 well, we were never the ones doing the sanitizing :) but nobody does now, thats good Mar 26 18:54:17 this is what happens when I'm out of it for 3 years.. Mar 26 18:54:20 hehe Mar 26 18:54:44 so essentially linux-libc-headers package is redundant Mar 26 18:54:59 we still need kernel headers but we can use same sources that we will use for kernel Mar 26 18:55:32 still a good idea to distinguish between the sanitized ones and the stock kernel headers, probably want to package both Mar 26 18:55:45 thats ok. Mar 26 18:56:04 okay, i'll shut up and go back to what i was doing now, thanks for the info :) Mar 26 18:57:24 tharvey: do you need santized headers for external modules ? Mar 26 18:58:10 not sure... what exactly is 'sanitized'? Mar 26 18:58:27 pretty sure sanitized, if its at all like it used to be, is sanitizing for -userland- Mar 26 18:58:34 so userland can pull in headers for ioctls and stuff Mar 26 18:58:36 not for module Mar 26 18:59:40 jeremy_laine: i'm taking a look in qt4 recipes, regarding qt4-tools, i have some questions... Mar 26 19:00:07 aloisiojr: go ahead, but I'll be leaving soon Mar 26 19:00:28 jeremy_laine: i think it's very important to build the tools from the same source that you build the qt4 itself Mar 26 19:00:34 well keep in mine I build the kernel-headers via the kernel's 'make headers_install' which should be appropriate for userland right? Mar 26 19:00:59 03Jeremy Lainé  07org.openembedded.dev * rd3224e2c27 10openembedded.git/recipes/qt4/ (3 files in 2 dirs): qt4-embedded: add version 4.5.0 Mar 26 19:01:01 currently the source to qt-tools is qt-embedded.. Mar 26 19:01:04 aloisiojr: true, but I don't want to build the native tools *both* for qt4-x11 and qt4-embedded, it needs to be shared Mar 26 19:01:11 aloisiojr: yes, that's not a problem Mar 26 19:02:12 jeremy_laine: it's not a problem if both tools (x11 and embedded) were the same Mar 26 19:02:17 jeremy_laine: can we ensure that? Mar 26 19:02:19 tharvey: yes Mar 26 19:02:23 aloisiojr: yes, I build both versions Mar 26 19:02:57 aloisiojr: the tools are the same, the config files for qmake are not Mar 26 19:04:46 jeremy_laine: i mean the code of the tools Mar 26 19:05:21 do we have the same moc/qmake/uic ? Mar 26 19:05:34 i need to check this deeply... Mar 26 19:05:40 aloisiojr: yes, absolutely sure Mar 26 19:06:21 aloisiojr: if you want proof it works: the qmake2-native and uicmoc4-native have always been built from qt/embedded source and used both for qt-x11 and qt-embedded in OE Mar 26 19:06:42 tharvey: kernel's make install_headers uses unifdef program to remove certain parts of kernel headers to make it usable for userspace Mar 26 19:06:56 tharvey: the difference is do you need a copy of kernel headers ditto Mar 26 19:07:11 or do you need the one's generated with headers_install Mar 26 19:07:22 jeremy_laine: yep... and this was one of the reasons that i decided to change the build structure a little Mar 26 19:08:07 jeremy_laine: but if you ensure this, ok, np Mar 26 19:08:40 tharvey: I sense that you might need kernel headers as it is Mar 26 19:08:54 aloisiojr: I have just done a diff on the "tools" directories: Mar 26 19:08:56 diff -x qvfb -urN qt-embedded-linux-opensource-src-4.5.0/tools/ qt-x11-opensource-src-4.5.0/tools Mar 26 19:09:21 => no difference, the only difference is "qvfb" which is embedded-specific Mar 26 19:11:00 jeremy_laine: right, so we just need to pay attention on some possible modifications in one of these trees, in this case we must stop to share the tools Mar 26 19:12:10 aloisiojr: I really, really doubt that will ever happen but OK Mar 26 19:12:32 Bah Mar 26 19:12:37 what's the right syntax for BBMASK? Mar 26 19:13:14 jeremy_laine: as you are about to leave, tomorrow we discuss about merge my qt4.bbclass with yours qt4.inc Mar 26 19:13:34 jeremy_laine: did you take a look in the file that day? Mar 26 19:13:50 Tartarus: its a regexp IIRC Mar 26 19:14:04 hmm Mar 26 19:14:16 BBMASK = "" Mar 26 19:15:28 btw, i got the a9 stuff and possibly a bit more backported Mar 26 19:15:31 it's like 20 patches Mar 26 19:15:31 khem, so thats what is called 'sanitized headers', the unidef that 'make install_headers' does? Yes, if thats not suitable for building external modules then its not what I need - perhaps there needs to be two packages created Mar 26 19:15:34 aloisiojr: I did at first, but as I said, the pastebin for your file has expired Mar 26 19:15:34 and i haven't yet run make check Mar 26 19:15:39 tharvey: I just looked on my ubuntu box. I think the heades in /usr/src/linux-headers-2.6.28-11/ Mar 26 19:15:46 they do not seem to be sanitized Mar 26 19:15:48 aloisiojr: so I haven't had a chance to pick out the differences Mar 26 19:15:57 jeremy_laine: ow shit :( Mar 26 19:16:08 Tartarus: cool Mar 26 19:16:08 jeremy_laine: we do this tomorrow Mar 26 19:16:19 aloisiojr: OK, got to go now, give Qt 4.5.0 from OE a spin and see how it works Mar 26 19:16:20 Tartarus: do you have a9 hardware to run on Mar 26 19:16:34 So, BBMASK = ".*local/packages/gcc/gcc.*.bb" should work, yes? Mar 26 19:16:38 khem, not yet Mar 26 19:16:48 jeremy_laine: ok Mar 26 19:16:50 i'm gonna make sure a8 stuff is still good for now Mar 26 19:16:52 bye! Mar 26 19:16:56 then a9 when we get the hw Mar 26 19:17:05 yeah Mar 26 19:17:13 khem, so you think the headers from say ubuntu are just cp -r of kernel's include? Mar 26 19:17:24 tharvey: seems like cp -al Mar 26 19:17:26 :) Mar 26 19:17:45 well now that the headers have moved into arch its a bit more Mar 26 19:20:13 tharvey: the headers I was talking about that we call linux-libc-headers in OE are bundled into linux-libc-dev on debian Mar 26 19:20:39 the real kernel headers seem to export all headers and kbuild Mar 26 19:21:52 * khem needs food Mar 26 19:23:09 khem, ya - I honestly haven't been clear what the issues were with the current linux-libc-headers that OE uses and why you were interested in what I was doing to replace them - if they are 'sanitized' in the same way that 'make install_headers' does then they should fit your need but perhaps not mine (testing now to see if I can build modules aginast them) Mar 26 19:23:30 tharvey: I dont think you can Mar 26 19:23:48 tharvey: you need raw kernel headers Mar 26 19:24:28 tharvey: make install_headers is not meant for out of source modules I think Mar 26 19:24:33 its meant for userspace for sure Mar 26 19:25:09 for out of source modules you need something like what debian linux-headers package provides Mar 26 19:27:03 tharvey: I just verified the headers in usr/include are not same as in /usr/src/linux-headers-2.6.28-11 Mar 26 19:27:46 tharvey: so you need to copy stuff around. Its like a mini linux src tree what you need Mar 26 19:29:18 khem, hmmm... perhaps I need to look at say, ubuntu's src RPM for linux-headers and see what they do - I'm a bit surpirsed that the kernel kbuild system doesn't offer building the right thing for external modules Mar 26 19:29:48 sounds like your saying I need to do more than cp -al - which makes sense, you don't need all the arch tree's just the one for your arch Mar 26 19:32:07 tharvey: yep what you are currently doing is not sufficient Mar 26 19:32:47 ubuntu does not have rpms, if might try fedora ones Mar 26 19:32:50 they have rpms Mar 26 19:35:29 true... I'm not any better with rpms than debs heh Mar 26 19:35:38 re Mar 26 19:36:15 I think we need three recipes in the kernel. one for kernel proper, one for linux-libc-header and another one for linux-kernel-headers Mar 26 19:38:41 which one would be for external modules? linux-kernel-headers or kernel proper? Mar 26 19:39:51 aloisiojr: hi Mar 26 19:39:58 linux-kernel-headers Mar 26 19:40:24 have you seen jeremy_laine msg? Mar 26 19:40:30 so what is kernel proper then? Mar 26 19:40:40 what we do now Mar 26 19:40:54 building kernel and modules and packaging Mar 26 19:41:07 ah... yes the kernel image/mods w/o headers Mar 26 19:41:18 or may be it can be done the way you are doing Mar 26 19:41:43 that seems better once you figure the right bits Mar 26 19:42:31 are you sure all 'make headers_install' does isn't just strip out the additional architectures? Mar 26 19:42:50 yes Mar 26 19:42:55 it does more than that Mar 26 19:51:47 anyone good with debian packaging that can point me to the source debian uses for building linux-headers-*.deb? Mar 26 19:56:58 mckoan: hi Mar 26 19:57:38 aloisiojr: I would try building a qemu system with qt Mar 26 19:57:56 mckoan: i was taking a look in the current qt4 build structure.. Mar 26 19:59:10 mckoan: qemu system with qt ? Mar 26 19:59:47 yes Mar 26 19:59:56 just to test it Mar 26 20:00:39 mckoan: oh i got it Mar 26 20:01:12 do you mean you already did it? Mar 26 20:01:47 mckoan: no.. i mean that I understood Mar 26 20:02:06 i was talking to jeremy some minutes ago, about qt4-x11 and qt4-embedded use the same qt4-tools-native Mar 26 20:02:09 aloisiojr: ack ;-) Mar 26 20:02:33 I am using angstorm and would like to install gcc and header files on my beagle. I managed to build and install gcc but am now missing the standard header and headers for libs Mar 26 20:02:42 is there a awy to get this? Mar 26 20:02:57 mckoan: qt4-tools-native get the qt4-embedded source to build qmake/moc/uic Mar 26 20:03:22 mckoan: he ensured that the code for tools/ are the same for both flavors of qt Mar 26 20:04:41 ok Mar 26 20:07:35 mckoan: i want to merge my qt4.bbclass (http://pastebin.com/m102b09af) with current recipes/qt4/qt4.inc Mar 26 20:07:46 mckoan: back in 5min Mar 26 20:10:23 tharvey: probably all generated from a single kernel deb , try apt-get source linux-image-$(uname -r) followed by looking in the debian subdirectory Mar 26 20:14:42 keesj, thx - looking at that now Mar 26 20:18:19 03Tick Chen  07org.openembedded.dev * r26a8720fde 10openembedded.git/recipes/mokoko/mokoko_svn.bb: mokoko: fix the broken bb file SRCPV -> SRCREV Mar 26 20:24:29 aloisiojr: I'll try tomorrow morning, obrigado :-) Mar 26 20:25:02 keesj, that proved uninteresting - kernel source and modules but what I really am after is the recipe the deb uses - what do you mean by 'debian directory' perhaps there's something I'm missing Mar 26 20:27:00 tharvey: let me have a look :p , Mar 26 20:30:04 perhaps I missed something... I was mislead when I did a dpkg -L linux-image-$(uname -r) Mar 26 20:32:36 I did apt-get source .. , now I have a linux-xxx directory where I ran the command , and if you look into that directory there is a control file and the whole debian stuff in there Mar 26 20:33:33 so this is the source for all those packages and you do something like dpkg-buildpackage it will build the packages Mar 26 20:36:13 ya, I see I missed what apt-get source did... looking at control now Mar 26 20:53:32 hi all, i got this error in final step of build final image "genext2fs: group descr 10. free blocks count == 0 (corrupted fs?)" ... any clue ? Mar 26 21:03:02 hello all Mar 26 21:03:14 i have problem when building meta-toolchain Mar 26 21:04:04 i must make symling from curl-sdk-7.18.2-r0 to curl-7.19.0-r1 Mar 26 21:04:17 to build correctly, can you fix this ? Mar 26 21:09:43 hi Mar 26 21:10:24 i am looking for resources of fastest boot (from bios to minimal desktop) , can you help ? Mar 26 21:10:42 check the page on the elinux.org wiki Mar 26 21:11:32 ok Mar 26 21:11:55 i had it in my history ;) Mar 26 21:12:11 heh :) Mar 26 21:12:29 I'm sure there's other resources, presentations, etc, but thats the main one i know of off the top of my head Mar 26 21:13:55 ok thanx , i am on moblin now but i am sure i can find something better Mar 26 21:23:08 is there a way to tell bitbake to re-run the 'install' task without doing a clean? bitbake -cinstall doesn't run as it thinks it already has Mar 26 21:25:52 bitbake -f -c install Mar 26 21:25:54 -f == force Mar 26 21:25:58 bitbake --help is your friend Mar 26 21:29:18 hi, all Mar 26 21:30:28 does anyone here know of a patch for fontconfig which adresses the problem of an infinite loop on fc-cache for arm-platforms? Mar 26 21:32:06 kergoth, thx Mar 26 21:32:15 np **** ENDING LOGGING AT Fri Mar 27 02:59:57 2009