**** BEGIN LOGGING AT Wed Jan 20 02:59:57 2010 Jan 20 04:57:29 Hi, i can't find libdri.so and libglx.so in xorg directory, so, what should i build to have them appear? Jan 20 05:58:54 can anyone please tell me if I can use makefiles to build in Angstrom instead of bitbake recipes? Jan 20 05:59:06 if I can somehow embed my "rather large" makefile in a bitbake recipe that would work as well. Jan 20 06:05:43 nhg, surely you can invoke your makefile from a bitbake recipe Jan 20 06:06:22 grg: are there any examples/docs you can point me to? Jan 20 06:06:36 I'm new to OE Jan 20 06:07:54 nhg, start here http://docs.openembedded.org/usermanual/html/ Jan 20 06:09:00 you'll probably want something like 'do_compile() { oe_make }' in your recipe Jan 20 06:15:41 thanks...then how would the .bb recipe know where my make files is? I assume I have to add it somewhere in the .bb as well? Jan 20 06:24:05 nhg, i suggest you keep reading the user manual. chapter 8 should help out. Jan 20 06:24:43 other recipe files that are lying around might help out too Jan 20 06:32:14 haven't found any so far with references to make files....I'll keep digging. thanks. Jan 20 08:24:40 good morning Jan 20 08:29:08 hrw|gone: ping, you've pushed it to temp branch, not to dev :) http://lists.linuxtogo.org/pipermail/openembedded-commits/2010-January/041115.html Jan 20 08:34:38 how can I enable x11 font 9x15 ? Jan 20 08:34:53 03Khem Raj  07org.openembedded.dev * r2bcf50bbf1 10openembedded.git/recipes/cherokee/cherokee_0.99.24.bb: (log message trimmed) Jan 20 08:34:53 cherokee_0.99.24.bb: Add missing DEPENDS Jan 20 08:34:53 * 0.99 version uses openssl and there is no option to have gnutls Jan 20 08:34:53 so we configure it to use openssl. Jan 20 08:34:53 * Add DEPENDS which were Jan 20 08:34:55 missing and override the default depends coming from cherokee.inc Jan 20 08:34:59 as it has gnutls in depends and would build this redundantly. Jan 20 09:08:05 03Koen Kooi  07org.openembedded.dev * r680760be3a 10openembedded.git/ (conf/checksums.ini recipes/gnome/libbonobo_2.24.1.bb): libbonobo: update to 2.24.2 (automake 1.11.x fixes) Jan 20 09:08:16 03Koen Kooi  07org.openembedded.dev * rdfc707b840 10openembedded.git/ (conf/checksums.ini recipes/gnome/gnome-keyring_2.28.0.bb): gnome-keyring 2.28: update to 2.28.2 Jan 20 09:08:17 03Koen Kooi  07org.openembedded.dev * ra353f4ee72 10openembedded.git/recipes/rygel/rygel_0.4.8.bb: rygel: add hack to get rid of automake 1.11.x dependency Jan 20 09:08:56 what's difference between libpixman and xorg-lib/pixman, libpixman looks the same as recipe (just older versions) Jan 20 09:12:42 shit Jan 20 09:13:01 morning Jan 20 09:13:04 moin Jan 20 09:13:18 mckoan: xfonts-misc? Jan 20 09:15:20 hrw: thx I'll try Jan 20 09:18:33 hrw: there is any xfonts-misc in recipes Jan 20 09:20:22 so build all fonts and check which one you need? Jan 20 09:20:29 03Petr Štetiar  07org.openembedded.dev * rf6c05a045d 10openembedded.git/recipes/xinput-calibrator/ (3 files in 2 dirs): Jan 20 09:20:29 xinput_calibrator: switch to autotools Jan 20 09:20:29 Signed-off-by: Petr Štetiar Jan 20 09:20:29 Acked-by: Marcin Juszkiewicz Jan 20 09:20:33 it is 10-15 minutes of build machine time Jan 20 09:22:55 hrw: do you mean recipes/xorg-font/xfonts-xorg_6.8.bb ? and how do i get them in the final image? Jan 20 09:23:40 mckoan: you are using OE for over year and still did not know how to alter image contents? Jan 20 09:25:28 hrw: the problem is to alter image contents in a OE community accetable way Jan 20 09:25:50 hrw: BTW I thought that was a 10" answer, thanks anyway Jan 20 09:27:10 mckoan: building all fonts takes 10 minutes. then you need to check which pacakge has needed one and install it on device to check does it work Jan 20 09:27:20 unless you reflash device each time with new image Jan 20 09:48:41 hi mickey|munich Jan 20 09:48:49 morning hrw Jan 20 09:49:02 mickey|munich: remove 'temp' branch from git server please Jan 20 09:50:13 done Jan 20 09:50:26 laibsch/scim was scheduled for removal as well ,right? Jan 20 09:55:10 ~curse libnl Jan 20 09:55:12 May you be reincarnated as a Windows XP administrator, libnl ! Jan 20 09:55:50 no idea Jan 20 10:03:06 hello Jan 20 10:07:23 good morning Jan 20 10:07:45 good morning :) Jan 20 10:08:23 I'm trying to build an image with two extra tasks (to do it a bit modular), is there any documentation on that? Jan 20 10:08:47 because right now I'm not sure what I _need_ to define in the image and task(s) files Jan 20 10:09:26 hmmz, that sorted the qte problems sort of :/ Jan 20 10:13:38 to be more specific: I need gstreamer stuff (that works) and qte stuff (doesn't work) Jan 20 10:14:44 lrg: hi Jan 20 10:15:22 morning hrw Jan 20 10:15:42 when I have RDEPENDS_${PN} = "qt4-embedded", I get an error there aren't any buildable dependencies and: Jan 20 10:15:46 NOTE: Runtime target 'qt4-embedded-gles' is unbuildable, removing... Jan 20 10:15:48 Missing or unbuildable dependency chain was: ['qt4-embedded-gles', 'virtual/egl'] Jan 20 10:15:50 ERROR: Required build target 'proes-image' has no buildable providers. Jan 20 10:15:53 Missing or unbuildable dependency chain was: ['proes-image', 'task-proes-qte', 'qt4-embedded-gles', 'virtual/egl'] Jan 20 10:22:16 lrg: did you made something with your nhk15? Jan 20 10:26:06 hrw: not yet, just been doing general stuff atm Jan 20 10:29:36 ok Jan 20 10:31:53 03Marcin Juszkiewicz  07org.openembedded.dev * rc4408fc35e 10openembedded.git/classes/qt3e.bbclass: Jan 20 10:31:53 qt3e.bbclass: fix paths for staging Jan 20 10:31:53 Without that patch uicmoc3-native fails to build. Jan 20 10:32:04 03Martin Jansa  07org.openembedded.dev * r1b426b8382 10openembedded.git/classes/sanity.bbclass: (log message trimmed) Jan 20 10:32:04 sanity.bbclass: handle lack of permission to read /proc/sys/vm/mmap_min_addr Jan 20 10:32:04 * With 2.6.33-rc2-00252-ge9e5521 on my host I noticed that Jan 20 10:32:04 "cat /proc/sys/vm/mmap_min_addr" returns now Jan 20 10:32:06 "cat: /proc/sys/vm/mmap_min_addr: Operation not permitted" Jan 20 10:32:08 Its probably becuse of Jan 20 10:32:10 http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=0e1a6ef2dea88101b056b6d9984f3325c5efced3 Jan 20 10:33:33 I got tired of keeping those two in own repo Jan 20 10:34:19 hrw: thanks :) Jan 20 10:35:54 JaMa|Off: this is simpler version of first your one Jan 20 10:37:51 hrw: seen that on ML.. I don't need that warning too.. (and I have 0 in that sysfs node already..) Jan 20 11:03:05 * mckoan is trying to use patchwork (http://wiki.openembedded.net/index.php/Patchwork) Jan 20 11:04:38 anubody use patchwork system with 2 PC? one where to browse http://patchwork.openembedded.org and one where develop OE connected with SSH Jan 20 11:05:25 I wonder if is possible to wget a mbox from http://patchwork.openembedded.org directly into the development machine Jan 20 11:05:54 mckoan: I think there's a script for that in contrib/patchwork/ Jan 20 11:06:09 pb_: ahhhhhh I remember now, sorry Jan 20 11:06:59 contrib/patchwork/pw-am.sh Jan 20 11:07:05 yep Jan 20 11:07:25 you give it IDs of patches and it fetch/git-am each Jan 20 11:07:47 hrw: exactly what I needed Jan 20 11:08:07 not only you :) Jan 20 11:08:19 Koen wrote first ver, I improved it for multipatches Jan 20 11:08:24 I'll try with patch 1401 Jan 20 11:09:43 looks not so smooth Jan 20 11:11:08 http://pastebin.com/m22eb6ed is it normal to get these messages? Jan 20 11:11:49 shouldn't I see any file added to OE if I do git status ? Jan 20 11:12:10 no, I think it commits the changes for you. try git log. Jan 20 11:12:24 mckoan: it does 'git am' Jan 20 11:13:13 mckoan: the trailing whitespace errors mean that the patch adds trailing whitespace to some lies. Jan 20 11:15:47 pb_: hmmm interesting :-D Jan 20 11:16:14 the patch will still be applied, it's just complaining. Jan 20 11:31:58 Hi ..... do you know how Xorg detects mouse and keyboard?.... I have an image where the keyboard in the console works fine, but Xorg does not detect it. Jan 20 11:33:29 by hal Jan 20 11:35:06 hrw: ok ... will check hal presence. Jan 20 11:40:39 hrw: hal-device does not list any mice .... do you think I need something ? The kernels detects the mouse in the log, so it is not a kernel-driver that is missing. Jan 20 11:58:56 03Martin Jansa  07org.openembedded.dev * r33d74d678b 10openembedded.git/ (2 files in 2 dirs): Jan 20 11:58:56 linux-openmoko-2.6.32: bump SRCREV for suspend/resume fix, enable MMC_UNSAFE_RESUME Jan 20 11:58:56 Signed-off-by: Martin Jansa Jan 20 12:00:57 ah Jan 20 12:03:18 03Martin Jansa  07org.openembedded.dev * r4750e2e563 10openembedded.git/conf/checksums.ini: Jan 20 12:03:18 checksums.ini: add checksums for upcomming xorg upgrades Jan 20 12:03:18 Signed-off-by: Martin Jansa Jan 20 12:10:24 03Martin Jansa  07org.openembedded.dev * rc0002bd2f7 10openembedded.git/ (2 files in 2 dirs): (log message trimmed) Jan 20 12:10:24 stopwatch: new recipe for QT4 based stopwatch Jan 20 12:10:24 * Author: Christof Musik Jan 20 12:10:24 * can be used as stop watch Jan 20 12:10:24 * has countdown capability (e.g. for next holiday) Jan 20 12:10:26 * survives suspend/resume Jan 20 12:10:28 Signed-off-by: Christian Rüb Jan 20 12:39:15 RP: Hi, good work on bitbake! BTW that patch I was referring to, wasn't about wildchar in preferred_version, but this one [git: remove repodir usage] http://www.mail-archive.com/bitbake-dev@lists.berlios.de/msg00667.html, without it I cannot fetch with git and SRCPV Jan 20 12:44:38 strange freenode behaviour today Jan 20 12:54:21 what's the idea with MACHINE_KERNEL_PR and how does it apply to modules built out-of-tree? Jan 20 13:04:46 03Johan Hovold  07org.openembedded.dev * r0eaeda31b9 10openembedded.git/ (6 files in 3 dirs): (log message trimmed) Jan 20 13:04:46 wxwidgets: add 2.9.0 and clean up recipes Jan 20 13:04:46 - Add wxwidget-2.9.0 include with derived rather than hardcoded lib and Jan 20 13:04:46 config script names (depends on debug and charset configure flags) Jan 20 13:04:48 - Use regular expression for LEAD_SONAME rather than hardcoded name. Jan 20 13:04:50 - Fix autoconf issue where configure could not find staging root. Jan 20 13:04:52 - Use INC_PR. Jan 20 13:08:56 my first patchwork + git-am push, I hope all is ok Jan 20 13:09:50 ;) Jan 20 13:10:05 mckoan: marked patch in patchwork as applied and archived it? Jan 20 13:25:20 03Marco Cavallini  07org.openembedded.dev * r78c0a87869 10openembedded.git/conf/distro/include/kaeilos-2009-preferred-versions.inc: kaeilos-2009-preferred-versions.inc: added PREFERRED_VERSION_wxwidgets 2.8.10 Jan 20 13:38:34 hrw: marked patch in patchwork as applied Jan 20 13:40:09 hrw: does exist a patchwork page for applied patches? (or for patches with status != New) Jan 20 13:40:53 mckoan: there are options for filtering Jan 20 13:42:53 hrw: looks like most of patches have status Accepted but only 2 are Applied, is it normal? Jan 20 13:43:34 hrw: aha Applied+ Archived Jan 20 14:00:43 mckoan: can you ack 1499? Jan 20 14:07:17 hrw: I could but I have to test it before, because is stable Jan 20 14:07:55 hrw: and I don't know on which machine test it Jan 20 14:09:23 ok Jan 20 14:10:08 hrw: this is one of the biggest limit of stable Jan 20 14:10:34 prepare better rules for next one Jan 20 14:11:00 hrw, need a more diverse group of people Jan 20 14:11:12 sure Jan 20 14:11:17 it is good that we are all looking and talking about what goes in Jan 20 14:11:38 I got Koen's ack now Jan 20 14:11:52 basically, we are trusting that robert is the java guy :) Jan 20 14:12:34 I added wxwidgets-2.9.0.bb but I want to build with 2.8.10, so I added KAEILOS_WX_VERSION ?= "2.8.10" and PREFERRED_VERSION_wxwidgets = "${KAEILOS_WX_VERSION}" but in root I always get 2.9.0, where am I wrong? Jan 20 14:12:35 the problem is that I am more and more java infected Jan 20 14:12:54 hrw: java is worse than H1N1 Jan 20 14:13:45 hrw: lethal for already ill and a few was infected (LOL) Jan 20 14:15:12 hrw: if Koen acked you, does this mean that he tested it in a couple of hours? Jan 20 14:15:42 most likely he looked at it and figured it won't break anything hrw won't immediately fix :0 Jan 20 14:15:47 :) Jan 20 14:16:23 and there is nothing obviously wrong Jan 20 14:16:28 Crofton_|work: I inderstand he's a wizard, but this is unsafe Jan 20 14:16:56 or he was affected by that already and used such fix internally Jan 20 14:17:17 hrw: I'd prefer the latter :-D Jan 20 14:17:31 mckoan, the other option is deadlock Jan 20 14:17:39 because we do not have a lot of java users Jan 20 14:17:48 Crofton_|work: I understand Jan 20 14:17:57 the goal is stable be usable Jan 20 14:18:09 by now several of us have looked it the patch in questions Jan 20 14:18:34 with this fix I should finally have stable/2009 based buildbot builds for BugLabs working Jan 20 14:18:49 time to heat up their server room Jan 20 14:19:49 hrw: I will be less scrupulous in the future ;-) Jan 20 14:20:31 any hint about PREFERRED_VERSION_wxwidgets ? Jan 20 14:20:48 hrw: is buglabs buildbot reporting to tinderbox? (or could be?) Jan 20 14:23:22 is someone looking at Ulf's patches? Jan 20 14:23:38 JaMa|Off: I can make it Jan 20 14:23:46 mckoan: no hint sorry Jan 20 14:24:26 hrw: so must be a problem hard to solve :-( Jan 20 14:25:24 hrw: thanks.. I'm interested if you will be able to build http://tinderbox.openembedded.net/packages/phoneme-advanced-foundation/ Jan 20 14:25:52 hrw: and as you're java infected then I guess you have it in some image/feed already Jan 20 14:26:00 mckoan: is it actually building the wrong version of wxwidgets, or is it just pulling in the wrong one from your deploy directory? Jan 20 14:26:12 JaMa|Off: we use phoneme-advanced-personal Jan 20 14:26:29 hrw: here it hangs indefinitely, not sure if because of low mem or that deadlock.. Jan 20 14:27:14 pb_: I thought each bb recipe was responsible foor its deploy, I'll investigate, thx Jan 20 14:28:20 JaMa|Off: never built foundation Jan 20 14:28:31 hrw: I hope you are not offended because I didn't ack, I simply followed stable rules, I will be less scrupulous in the future ;-) Jan 20 14:28:59 hrw: hmm nvm.. this patch is there long enough.. so last time I tried to build foundation it was already there.. Jan 20 14:32:21 fine with it I am Jan 20 14:32:53 mckoan, different people have different comfort levels with what they will ack Jan 20 14:32:56 this is a good thing Jan 20 14:33:09 Crofton_|work: no prob :-) Jan 20 14:33:31 it is good to be cautious Jan 20 14:35:33 Crofton_|work: in a community you have to balance technical stuff with social one, if too cautious = ugly this could generate social problems :-) Jan 20 14:40:34 can someone please point me to some examples/documentation on how a make file could be used within a .bb (bitbake recipe)? Jan 20 14:41:06 nhg: do_compile() { oe_runmake } Jan 20 14:41:39 in fact, if your makefile is conventionally named, you don't even need that because it is the default. Jan 20 14:41:49 see base_do_compile() in base.bbclass Jan 20 14:42:13 pb_: thanks....then how does the recipe know where my make file is located etc.? Jan 20 14:42:22 nhg: it expects it to be in ${S} Jan 20 14:42:39 if it's somewhere else, you can point to it with -f as normal Jan 20 14:43:05 EXTRA_OEMAKE = "-f subdir/Makefile", that kind of thing Jan 20 14:44:07 ok...I couldn't find much on the oe documentation on that Jan 20 14:45:25 wonder if the makefiles should conform to some standard for this to work...or any makefile will work in a recipe Jan 20 14:45:32 any makefile will be fine Jan 20 14:46:13 well, assuming it doesn't have things like hard-coded compiler names Jan 20 14:46:30 nhg: it just runs make.. if your installed make works with the makefile, it'll work fine. and makefiles are pretty portable, the only thing that could even theoretically choke would be a makefile designed for the old bsd make rather than gnu make, and those are few and far between Jan 20 14:46:33 hey pb_ Jan 20 14:46:35 did you find it somewhere in http://docs.openembedded.org/usermanual/usermanual.html? Jan 20 14:46:37 * kergoth yawns Jan 20 14:46:38 hi kergoth Jan 20 15:33:51 guh Jan 20 15:34:03 I still have the problem with "hwclock: can't open '/dev/misc/rtc' no such file or directory". I know how to create it manually when the system is running but it doesnt sounds like the best idea.. Is there any other way? Why isn't it created "by default" Jan 20 15:34:30 we're running into an issue where we need to not build a given package, but only if its dependency is newer than version XYZ.. in OE tahts not a rpoblem because people would just not explicitly build it.. but we do regular world builds.. Jan 20 15:34:35 jovox: sounds like a bug in your udev/mdev setup Jan 20 15:34:59 * kergoth thinks about how painful itd be to try to add version specific depends/conflicts to oe Jan 20 15:35:01 s/oe/bitbake/ Jan 20 15:35:02 bugs-- Jan 20 15:35:34 kergoth: fairly painful, I think, though it would be very worthwhile Jan 20 15:36:26 iirc, bitbake has a somewhat ingrained assumption that it will only build a single version of a package in any one run, which of course breaks down if you admit versioned dependencies. Jan 20 15:36:28 yeah.. thats what i was thinking.. would have to add a fair bit of extra logic to the cooker/runqueue Jan 20 15:36:48 true Jan 20 15:37:02 not really sure what semantics you could meaningfully give to conflicts at the bitbake level. Jan 20 15:39:29 pb_, what do you mean with udev/mdev "setup"? Configuration or bug in udev itself? Jan 20 15:39:42 configuration, probably, but it could be either Jan 20 15:39:48 would almost have to implement private staging, or locked staging that gets packages uninstalled/installed based on version specific dependencies, to pull it off, since its not normally safe to build multiple versions of something, unless it has no staging files Jan 20 15:40:00 yeah Jan 20 15:40:09 private staging seems like it must be the way to go for all kinds of reasons Jan 20 15:40:29 I think we previously talked about making packaged staging mandatory as a first step in that direction. Jan 20 15:40:40 * kergoth nods Jan 20 15:41:15 * broonie doesn't see why versioned dependencies are a problem there - it'd just say "I can't satisfy this" if the version can't be satisfied Jan 20 15:42:43 broonie: i suppose thats one route, just error out if there are conflicting version requirements.. but it could make world builds impossible Jan 20 15:43:34 yeah, I think that would suck. Jan 20 15:46:42 Presumably if these versioned dependencies already exist then world builds are already impossible. Jan 20 15:47:57 s/already/actually/ Jan 20 15:48:04 otherwise the dependencies are spurious anyway Jan 20 15:48:06 world builds work, just some packages may fail to build Jan 20 15:48:09 thats actually why i brought this up Jan 20 15:48:24 we have this happening at work, we do regular world builds, and we have a package that can't be built with version > XYZ of one of its deps Jan 20 15:48:27 Eh? Surely that means that world builds only partially work? Jan 20 15:48:29 so we're basically hosed, just now Jan 20 15:49:13 with the deps you could at least do something like have world builds drop the packages that can't possibly be built, rather than fail during build Jan 20 15:50:13 that's true. as long as the distro still sets the version preferences so bitbake knows which of the alternative scenarios to choose from, since it can't pick one at random Jan 20 15:50:29 * kergoth is thinking the case where A DEPENDS on C < 1.0 and B DEPENDS on C > 1.0 Jan 20 15:51:16 otherwise it would likely have to drop all packages involved in the conflict until the user selects one Jan 20 15:52:44 it's too bad a recipe can't access bitbake's preferences information, and raise SkipPackage if it sees it wants to build a given version of something Jan 20 15:52:50 would be a workaround, at least Jan 20 16:01:40 damnit.. i know i remember patching this little hack into bitbake for some reason, but i don't remember why.. Jan 20 16:02:42 I am trying to understand how OpenEmbedded works... is the point to create an image by creating a config file and then run bitbake for all packages you want installed on the image? Jan 20 16:02:56 kergoth: a=['cos','tam','asaa'] - and now I want to get 'cos tam asaa' string from that list. how? Jan 20 16:03:11 ' '.join(a) Jan 20 16:03:15 thx Jan 20 16:03:17 np Jan 20 16:03:29 * kergoth uses that ALL THE TIME Jan 20 16:03:30 :) Jan 20 16:03:46 someone got a few (hopefully) mins, to help me creating a libxml-perl module recipe ? Jan 20 16:04:02 (got some problems with Makefile.PL and the 'make test' Jan 20 16:06:21 yoww.. my private staging prototype is .. not the prettiest code I've written Jan 20 16:06:43 what is the proper way of creating an image that should run on a beagleboard with some packages installed? Jan 20 16:10:19 no cpan.bbclass guru present ? :/ Jan 20 16:11:32 how do I update package list again? Jan 20 16:11:50 * kergoth tends to avoid perl wherever possible Jan 20 16:12:38 kristoffer: bitbake package-index Jan 20 16:13:01 thanks Jan 20 16:13:18 hmm Jan 20 16:14:18 kergoth, also, got a new fancy tar.gz bug. Seems like tar.gz builds just fine when manually specifying files. But when building images, all their unbuilt depends only create ipks. Jan 20 16:14:48 i don't think a rootfs_tar even exists right now. Jan 20 16:15:05 you'd have to set IMAGE_PKGTYPE = "tar", and create a rootfs_tar if necessary Jan 20 16:16:26 Im not sure if you follow me, if image X depends on packages a b c d e, I only get packages a b c d e in ipk format. Jan 20 16:17:25 yes? Jan 20 16:18:54 * kergoth doublechcks image.bbclass Jan 20 16:20:53 hmm Jan 20 16:27:55 kristoffer: rootfs_ipk only pulls in the *ipks* of its dependent things. when you bitbake foo, it builds foo and it runs the tasks in foo's dependencies necessary to build foo. if foo is a regular recipe, that means it runs the populate_staging task on those dependencies. if foo is an image, it runs do_package_write_ipk on those recipes, as well as do_populate_staging and do_deploy. Jan 20 16:28:11 aha Jan 20 16:29:18 thx Jan 20 16:29:24 np Jan 20 16:30:05 if you're curious about how i found that, you can see that in package_ipk, there's also a task that updates the package index for the rootfs, it runs before do_rootfs, and does a recrdeptask on do_package_write of its deps. Jan 20 16:51:19 kergoth: are the issues with packagedstaging and wrong arch solved in your tree? Jan 20 16:51:38 no, we still see warnings about it.. it doesn't break, but we see warnigns about no arch set and using default Jan 20 16:51:51 thus, rebuilding all natives... Jan 20 16:52:36 which wrong arch problem? Jan 20 16:52:40 target stuff being host_arch.ipk Jan 20 16:52:47 Or target stuff gettting binconfigs ? Jan 20 16:53:00 wrong arch in the staging packages Jan 20 16:53:49 have a nice rest of the day Jan 20 16:54:06 ant_work, what do you mean? Jan 20 16:54:54 even target ipks get the wrong arch (the host one) Jan 20 16:54:58 see udev e.g. Jan 20 16:55:42 Er, I don't follow Jan 20 16:55:50 I see armv6 ipks in the armv6 ipk dir Jan 20 16:56:00 from my build out right now of a buncha pstaging stuff Jan 20 16:56:28 iirc was staging_helper in packages-staging.bbclass Jan 20 16:56:38 I don't have oetree here Jan 20 16:57:02 issue is rebuild from pstage takes longer as expected... Jan 20 17:02:57 Tartarus: i think his concern is with the staging ipks, not the target ipks. the staging ipks have both an arch that opkg knows about, as well as an arch/os embedded in the filename. if the target is embedded in everything, then the native prebuilts wouldn't be shared properly. i don't think we made many changes to it, so I'm not entirely clearly on what the exact problem is right now, but thats the area of his concern, as far as i can t Jan 20 17:05:36 Ah Jan 20 17:05:51 Yeah, there's a lot of overlap Jan 20 17:06:07 and like I was saying with binconfigs, host junk ends up in the target ipks anyhow Jan 20 17:06:41 My kind of thought atm is that the host vendor needs to be encoded anyhow Jan 20 17:07:06 Since we sure aren't making LSB host bins :) Jan 20 17:08:04 Tartarus: the point is opkg thinks all staged-packages are for i686 (my host) and a rebuild from pstage rebuilds all again Jan 20 17:09:25 A rebuild on a not-i686 host? Jan 20 17:09:43 on the same host, if I delete all but /pstage Jan 20 17:09:58 OK, that sounds like a bug somewhere then Jan 20 17:10:08 I'm doing a whole bunch of start from scratch builds Jan 20 17:10:22 my pstage.bbclass isn't that different Jan 20 17:10:27 before that, try to wipe all but that /pstage dir Jan 20 17:10:34 But the big issue I'm hitting is deciding something is invalid Jan 20 17:10:36 when it shouldn't be Jan 20 17:10:43 ant_work, do you redefine .... Jan 20 17:10:56 DEPLOY_PSTAGE_DIR ? Jan 20 17:11:10 I stick that outside of $TMPDIR Jan 20 17:11:23 ok, thios leads toother issues :) Jan 20 17:11:34 with package_stagefile_shell Jan 20 17:11:43 but even Jan 20 17:11:45 What issues? Jan 20 17:11:53 I haven't seen any, but maybe I'm not hitting the right things Jan 20 17:12:03 I suppose some very naughty do_deploys could be a problem Jan 20 17:12:07 but the big ones I thought were fixed Jan 20 17:12:17 if you are lucky enoughyou don't ever need that package_stagefile_shell hack Jan 20 17:12:29 only a few bootloaders Jan 20 17:15:48 Tartarus: found http://blog.denix.org/2008/09/getting-even-more-dangerous-with-pstage.html Jan 20 17:17:03 03Simon 'Morphis' Busch  07org.openembedded.dev * rc87ea56a9e 10openembedded.git/recipes/preboot/ (5 files in 3 dirs): preboot: new recipe; a dual boot solution for the palm pre (WIP) Jan 20 17:18:12 Hmmm Jan 20 17:18:28 Haven't hit a point in my builds where it wants to install the kernel again Jan 20 17:18:32 lemme go poke the pstage file Jan 20 17:21:53 Bah, I see fakeroot needs some special handling Jan 20 17:23:44 bbl Jan 20 17:26:50 Tartarus: i was thinking yesterday about looking at some of the other autopackage tools. they had some scary stuff for binary modification to make it more portable. i think it may have gone in and manipulated the glibc versioned symbols, can't recall.. twisted, but interesting Jan 20 17:26:52 heh Jan 20 17:27:27 That's a bit drastic, yes Jan 20 17:28:07 they were really anal about binary portability.. not that i can blame them, its a mess, even with LSB around Jan 20 17:28:11 heh Jan 20 17:30:16 Yeah, but LSB helps a bit Jan 20 17:30:20 Not perfect I suppose Jan 20 18:25:09 http://hrw.pastebin.ca/1759215 - my recent OE helper in Python Jan 20 18:45:45 bye Jan 20 18:55:01 03Marco Cavallini  07org.openembedded.dev * r91c888e06f 10openembedded.git/conf/distro/include/kaeilos-2009-preferred-versions.inc: kaeilos-2009-preferred-versions.inc: errata in PREFERRED_VERSION_wxwidgets Jan 20 18:58:59 hallo Jan 20 19:07:33 03Stefan Schmidt  07org.openembedded.dev * rfae0d64b58 10openembedded.git/conf/checksums.ini: checksums.ini: Add missing sum for linuxkernel-2.6.24-patch Jan 20 19:09:12 re Jan 20 19:09:27 hey florian Jan 20 19:09:41 hi kergoth Jan 20 19:10:15 http://www.embedded.com/201200638 .. nice little series, decent refresher Jan 20 19:10:19 hrm Jan 20 19:15:57 03Thomas Zimmermann  07org.openembedded.dev * r1a6ebaa5ed 10openembedded.git/ (2 files in 2 dirs): Jan 20 19:15:57 mickeydbus2: add new recipe Jan 20 19:15:57 Signed-off-by: Klaus Kurzmann Jan 20 19:23:21 jo Jan 20 19:35:55 Hmm, trying to build glibc 2.9 for armv4. Getting alot of undefined reference to __divdf3 __muldf3 __ltdf2...and so on coming from unwind-pe.c Jan 20 19:36:19 hm Jan 20 19:36:25 softfloat problems Jan 20 19:46:49 pb_, the missing /dev/misc/rtc problem I was talking about before.. Do you have any suggestions on where I should start? I don't know much about how udev work (if it's udev that's supposed to create it) Jan 20 20:26:54 man, irc is dead today Jan 20 20:27:40 so last century, right? Jan 20 20:28:13 heh Jan 20 20:28:42 jovox: I don't know that much about udev myself. I guess you'd need to check the documentation. Jan 20 20:29:04 I use mdev on all my OE systems nowadays. Jan 20 20:33:08 woglinde, thanks again for the help with gnome-gmaes Jan 20 20:33:09 pb__ ok Jan 20 20:45:59 re Jan 20 20:46:07 crofton no prob Jan 20 20:46:12 hi pb and kergoth Jan 20 20:46:20 hey Jan 20 20:49:43 If I manually create the /dev/rtc with mknod I still get and error: "hwclock: can't open '/dev/rtc': No such device". Could this have anything todo with my kernel configuration? Jan 20 20:49:58 jovox yes Jan 20 20:50:05 make sure rtc is enabled Jan 20 20:50:19 or /dev/rtc has the wrong acess rules Jan 20 20:50:42 wrong access rules.. uhmm udev config? Jan 20 20:53:20 03Stanislav Brabec  07org.openembedded.dev * rc71044be4d 10openembedded.git/recipes/xorg-driver/ (2 files in 2 dirs): Jan 20 20:53:20 xf86-input-tslib: Fix double free crash while on chvt or exit. Jan 20 20:53:20 * For more see recipes/xorg-driver/xf86-input-tslib/double-free-crash.patch Jan 20 20:57:16 woglinde, is there any documentation about configuring the kernel in OE? I have never really understood how it's done. Jan 20 20:58:17 http://www.xora.org.uk/2009/12/10/openembeddedangstrom-kernel-workflow/ Jan 20 20:58:20 might help Jan 20 20:58:42 thanks Jan 20 20:58:45 03Stanislav Brabec  07org.openembedded.dev * ra197e5c95c 10openembedded.git/recipes/xorg-xserver/ (2 files in 2 dirs): Jan 20 20:58:45 xserver-xorg: Worked around server termination after chvt on some fbdev devices. Jan 20 20:58:45 * The patch is just a quick hack that fixes chvt on some fbdev devices. Jan 20 20:58:45 * For more read recipes/xorg-xserver/xserver-xorg/hack-fbdev-ignore-return-mode.patch Jan 20 20:58:50 I'll read through it Jan 20 21:08:58 Crofton_|work, one thing that confuses me is this "bitbake virtual/kernel". and that the kernel recipes are in linuc/linux_version.bb Jan 20 21:10:47 so lets say I want to configure a specific version.. xora does: bitbake linux-elphel -c configure. How is that translated to a specific kernel in oe? Jan 20 21:12:03 PROVIDES. Every recipe, by default, provides its package name, name dash version, name dash version dash revision. It's likely only one recipe provides linux-elphel, so that recipe will be used. Multiple recipes provide virtual/kernel, however Jan 20 21:12:52 Ok I see Jan 20 21:15:24 And from where should I do the configuration? ie, make menuconfig Jan 20 21:19:18 jovox if the default defconfig is not good foor you do Jan 20 21:19:31 bitbake -c menuconfig virtual/kernel Jan 20 21:20:05 reboot, brb Jan 20 21:26:11 eFfeM, thanks. Problem is I don't have any X on the build machine.. Do you know where to configure it to run in terminal? Jan 20 21:49:50 Tartarus: ping Jan 20 21:50:33 hi laibsch Jan 20 21:51:02 hi woglinde Jan 20 21:51:05 hi Laibsch Jan 20 21:51:22 Tartarus: I kept an old log resuming the pstage issue: http://fr.pastebin.ca/1759461 Jan 20 21:51:37 I'll try in a few mins a new rebuild from pstage Jan 20 21:52:11 jo ant Jan 20 21:52:20 Laibsch: have you a mom to look into the e2fsprogs/TLS issue? Jan 20 21:52:25 woglinde: ^^ Jan 20 21:52:30 Laibsch: sorry Jan 20 21:52:55 seems plain uclibc issue Jan 20 21:53:17 hm? Jan 20 21:53:31 some macro you said... Jan 20 21:53:32 ant__: not an issue now? I have managed to build a rootfs image from totally just cached pstage stuff Jan 20 21:53:33 03Khem Raj  07org.openembedded.dev * r8364b075ed 10openembedded.git/recipes/gdb/ (gdb.inc gdbserver.inc): Jan 20 21:53:33 gdb: Add RRECOMMENDS for uclibc-thread-db when using uclibc for system C Jan 20 21:53:33 library. Jan 20 21:53:33 * Fix PR 3273 Jan 20 21:53:33 Signed-off-by: Khem Raj Jan 20 21:53:47 Tartarus: I'll tell you in a few mins ;) Jan 20 21:53:48 That said, instead of noarch, it's hostarch so that warning is gone, but it's not quite right either Jan 20 21:54:35 woglinde: http://tinderbox.openembedded.net/public/logs/task/4452892.txt Jan 20 21:55:09 What is needed to do a bitbake -c menuconfig without running X? IIRC there where some arguments that could be passed to bitbake? Jan 20 21:55:16 woglinde: in configure: checking for thread local storage (TLS) class... __thread Jan 20 21:55:29 minimal-uclibc btw Jan 20 21:55:54 ant hm Jan 20 21:55:59 false tls macro Jan 20 21:56:12 lookinto my pulseaudio patch Jan 20 21:56:37 or better the tls macro is plain wrong Jan 20 21:56:49 but that happend normaly only for c++ stuff hm Jan 20 21:58:24 Tartarus: hm.. Jan 20 21:59:17 I'm looking at the control file in staging-udev-c7x0-angstrom-linux-gnueabi_141-r20_i686-linux.ipk Jan 20 21:59:29 c7x0 ... i686... Jan 20 21:59:34 Yeap Jan 20 21:59:36 Architecture: i686-linux Jan 20 21:59:36 That's what I mean Jan 20 21:59:39 it's not noarch anymore Jan 20 21:59:41 Its i686-linux Jan 20 21:59:45 Which is less wrong Jan 20 21:59:50 ?? Jan 20 21:59:51 Since for example, stuff with binconfig in it Jan 20 21:59:56 c7x0 = armv5te Jan 20 21:59:57 ie staging-freetype-... Jan 20 22:00:00 host = x86 Jan 20 22:00:04 has a host bit in it Jan 20 22:00:21 Not all target stuff is contaminated like this Jan 20 22:00:22 But some is Jan 20 22:00:35 just cross, sdk are ok it seems..let see Jan 20 22:00:41 Well Jan 20 22:00:47 They're all going to be your host arch Jan 20 22:00:54 Which I'm saying is just less wrong Jan 20 22:01:03 D: Jan 20 22:01:16 bitter Jan 20 22:02:01 woglinde: btw there is another issue with uclibc and old headers (libtrace) Jan 20 22:02:38 ltrace Jan 20 22:02:39 http://tinderbox.openembedded.net/public/logs/task/4446605.txt Jan 20 22:02:58 it's just an incomplete/old header Jan 20 22:03:17 strangely another target in the same dir defines the missing things... Jan 20 22:04:00 clearly that's not happening with glibc/eglibc Jan 20 22:04:18 ask khem about it Jan 20 22:04:22 sorry have no time for this Jan 20 22:04:37 I think uclibc needs updates. fix is trivial, though Jan 20 22:05:27 you know, I'm building different combos distro/libc just to check the image sizes... Jan 20 22:05:40 sure Jan 20 22:05:48 atm minimal-uclibc looks good, is just a bit broken... Jan 20 22:06:21 Is it bitbake that launches gnome-terminal and thereby should be configured? Jan 20 22:06:35 minimal-eglibc is smaller than angstrom-glibc (I have to check better but I think the pacjkages installed are almost the same) Jan 20 22:13:41 hello Jan 20 22:13:53 hi khem Jan 20 22:13:58 hi woglinde Jan 20 22:14:51 khem ant has some questions Jan 20 22:17:48 Do you guys always configure your kernels in X? Jan 20 22:18:02 * JaMa|Off never Jan 20 22:18:52 JaMa|Off, so what did you have to do to make it not starting gnome-terminal? Jan 20 22:18:56 jovox: see TERMCMD TERMCMDRUN in conf/bitbake.conf Jan 20 22:19:19 jovox: I use screen istead of gnome-terminal which is not installed here at all Jan 20 22:20:06 hi gnutoo Jan 20 22:20:36 woglinde: k Jan 20 22:20:42 ant__: whats up Jan 20 22:20:51 hi khem Jan 20 22:20:57 oh the ltrace/uclibc Jan 20 22:21:07 just discussing about failure in minimal-uclibc Jan 20 22:21:11 that will require a header update in uclibc Jan 20 22:21:14 woglinde, hi Jan 20 22:21:19 it seems so Jan 20 22:21:26 ant__: what image are you building Jan 20 22:21:34 ant__: you could select nplt Jan 20 22:21:36 woglinde, can I commit one more abiword fix? Jan 20 22:21:40 for arm/uclibc Jan 20 22:21:46 its in good shape Jan 20 22:21:54 I build console - opie- and x11-image Jan 20 22:22:00 and all your TLS worries will be o ver Jan 20 22:22:03 iirc it fails already in console-image Jan 20 22:22:35 ant__: I think nptl will get you further I am sure Jan 20 22:22:48 just set the DEFAULT_PREFERENCE Jan 20 22:23:11 I have build opie-image and native-sdk-image recently with uclibc/nptl/arm Jan 20 22:23:36 khem: you make me remember when I first rebuilt my Gentoo around glibc-nptl 5-6 yrs ago... Jan 20 22:23:42 JaMa|Off, ah ok, thank you Jan 20 22:23:45 nobody cares about LT anymore so you will see more and more packages broken Jan 20 22:24:03 ant__: well then here you go again Jan 20 22:24:03 also gnumeric has a doltlibtool problem....should we wait? Jan 20 22:24:07 this time with uclibc Jan 20 22:24:12 GNUtoo|oeee sure Jan 20 22:24:17 ok Jan 20 22:24:18 I am only bugfixer Jan 20 22:24:23 I'll comit tomorrow Jan 20 22:24:34 about gnumeric I don't know what to doo Jan 20 22:24:40 if we should fix or wait Jan 20 22:25:02 GNUtoo|oeee my point was remove dolt stuff from the configure.in/ac Jan 20 22:25:35 I know but after the discussion + the upstream sumission is it still what I should do? Jan 20 22:26:27 hm Jan 20 22:26:30 we are oe Jan 20 22:26:36 we make it build Jan 20 22:27:05 ok Jan 20 22:27:16 so I'll fix it too Jan 20 22:27:23 sure Jan 20 22:27:39 I've tons of fix to do Jan 20 22:27:43 otherwise the lovecroftpiddy guys come up Jan 20 22:27:54 while he cannt build demo-image Jan 20 22:27:54 what's that? Jan 20 22:27:58 ah ok Jan 20 22:31:57 btw what image should I build for having illume Jan 20 22:32:06 illume-image seem for phone and doesn't work Jan 20 22:32:23 it should be replaced by the shr images Jan 20 22:32:43 demo or x11 image Jan 20 22:32:44 I would like a generic image Jan 20 22:32:47 ok Jan 20 22:32:52 x11 install illume? Jan 20 22:32:57 I'll look Jan 20 22:41:19 JaMa|Off, I changed it to screen and did bitbake -c menuconfig virtual/kernel. But it seem to hang. I run it with -DDD to see what happends and it stops at "screen -D -m -t 'linux Kernel Configuration' make menuconfig" Jan 20 22:42:38 mmm Jan 20 22:42:52 hmm it opened in a new screen Jan 20 22:42:59 the deps names don't say they depends on illume Jan 20 22:43:05 I'll look in bitbake -i Jan 20 22:44:11 gnutoo hm oh Jan 20 22:44:16 than it was e17 only Jan 20 22:45:56 ? Jan 20 22:46:04 ok I'll look Jan 20 22:46:38 there is e-image Jan 20 22:47:02 but what I wonder is if it askes the profile to choose at the beginning Jan 20 22:47:15 if so I should use this one in the future Jan 20 22:47:58 strange there are gpe things inside Jan 20 22:55:59 I'm running out of time here and thinking of hiring an experienced oe developer for a couple of hours.. Jan 20 23:10:52 arg, wtf.. this recipe parses fine in oe and freaks under mvl6 post-oe-sync, yet i can't find a root cause Jan 20 23:10:59 it seems like it should be failing under oe too Jan 20 23:13:21 man this is weird Jan 20 23:14:03 this konq recipe was setting exec_prefix = ${palmtopdir} but palmtopdir is ${libdir}/opie, and libdir is ${exec_prefix}/lib .. recursive at expansion time, yet doesn't fail to parse when using oe alone Jan 20 23:14:05 * kergoth scratches head Jan 20 23:20:42 Tartarus: unfortunately rebuild from packaged staging is still failing miserably for my target... Jan 20 23:22:57 03Chris Larson  07org.openembedded.dev * r2f6df5f071 10openembedded.git/recipes/rpm/rpm-native_4.4.2.3.bb: Jan 20 23:22:57 rpm-native: kill unnecessary varprefix/localstatedir manipulation Jan 20 23:22:57 Signed-off-by: Chris Larson Jan 20 23:28:32 03Chris Larson  07org.openembedded.dev * r0b76c230a4 10openembedded.git/recipes/konqueror/konqueror-embedded_20060404.bb: Jan 20 23:28:32 konq-e 20060404: don't override the exec_prefix var Jan 20 23:28:32 Signed-off-by: Chris Larson Jan 20 23:32:14 god i hate these "No Python in STAGING_INCDIR" messages Jan 20 23:32:21 exceptions from var expansions piss me off Jan 20 23:32:40 hm Jan 20 23:32:45 kergoth: Yes, I wish we had a nicer way of doing that. KERNEL_VERSION was similar Jan 20 23:32:55 hi rp Jan 20 23:33:01 hi woglinde Jan 20 23:33:24 kergoth: I fixed the bitbake IPC and event issues and am much happier with master now Jan 20 23:33:26 there has to be a better method than this. Jan 20 23:33:39 saw your reply about the progress, good to hear, nicely done Jan 20 23:33:42 kergoth: agreed. The python and person stuff is insane :/ Jan 20 23:33:47 * kergoth will take a look this weekend probably Jan 20 23:33:53 s/person/perl/ :/ Jan 20 23:34:11 kergoth: I've made bitbake master the default in poky - we'll see what that shows up :) Jan 20 23:34:17 nice. Jan 20 23:34:27 i think my boss would shoot me if i tried something like that :) Jan 20 23:34:57 kergoth: I can imagine :) Jan 20 23:35:21 Poky master has development and we just made a stable release so I don't feel too bad about it in my case Jan 20 23:35:29 * kergoth needs to set aside more time to play with OE proper builds rather than just mvl6 stuff Jan 20 23:35:32 cool Jan 20 23:35:55 My plan is to look at this parser branch now, merge that, then call it 1.10 Jan 20 23:36:35 kergoth rp made all packges BBCLASSEXTEND in poky Jan 20 23:36:42 we need to start it in oe too Jan 20 23:37:21 woglinde: I just made BBCLASSEXTEND much nicer in bitbake master too with a few bugfixes Jan 20 23:38:09 Made it part of the parsing and cache code rather than a sneaked in addon Jan 20 23:38:42 fewer files many better parse speed Jan 20 23:38:47 args means Jan 20 23:39:29 woglinde: Next, I'm wondering about multi threaded parsing Jan 20 23:39:51 rp yes Jan 20 23:40:33 i just finished the sync of the oe features/bugfixes into mvl6, so I'm hoping we can start converting to new style staging and BBEXTENDCLASS too Jan 20 23:41:49 RP: i'll go ahead and rebase the conversion of bb.msg to use the logging module on top of current master and push it on a branch for review. one of those things thats been floating around for a while, don't think I ever got around to pushing it Jan 20 23:42:10 will be fun making the thing apply, I'm sure Jan 20 23:42:34 * kergoth heads home Jan 21 00:00:49 how is Asus' reputation for Linux support of it's netbooks? Jan 21 00:10:58 03Graeme Gregory  07org.openembedded.dev * r2154e8eb1a 10openembedded.git/recipes/gnome/at-spi_1.28.0.bb: Jan 21 00:10:58 at-spi_1.28.0.bb : fix compile on systems without host installen gnome devel Jan 21 00:10:58 The configure script was generation paths in /usr/* when looking for Jan 21 00:10:58 orbit-2-idl and bonobo idl files so feed our staging paths in instead. Jan 21 00:11:01 03Graeme Gregory  07org.openembedded.dev * rd99a0541bf 10openembedded.git/conf/checksums.ini: checksums.ini : alternative URLs for Thunar and xfce4-dev-tools Jan 21 00:11:04 03Graeme Gregory  07org.openembedded.dev * rca0ab6d91b 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git.openembedded.org/openembedded into org.openembedded.dev Jan 21 00:11:15 Laibsch: excellent Jan 21 00:11:34 great Jan 21 00:11:45 I'm considering one of the new N450 models Jan 21 00:11:54 the battery runtime is just lovely Jan 21 00:11:58 I have had the 702 and have got the 901 Jan 21 00:12:06 Hello Jan 21 00:12:42 We're having a ugly and strange issue with sqlite3 package Jan 21 00:13:14 When running bitbake with -DDD it shows the right version as choosen as provider, but it doesn't build it Jan 21 00:15:52 ANY help? :-) Jan 21 00:16:22 without the log Jan 21 00:16:25 no help Jan 21 00:16:25 I and mario-goulart are looking at this issue for more then 1 hour without any idea Jan 21 00:16:45 woglinde: we have the log, mario-goulart can you paste the log file somewhere? Jan 21 00:17:17 woglinde: what worries me is that it doesn't mark the tasks of it Jan 21 00:18:21 woglinde: parenteses.org/mario/misc/log.bz2 Jan 21 00:18:36 (118k compressed, 3.6M uncompressed) Jan 21 00:19:01 Just a min. Jan 21 00:19:12 The upload is not finished yet. Jan 21 00:19:23 Laibsch: just be careful about the GPU they use. I don't know which one is but mine, for example, is horrible regarting linux support Jan 21 00:19:29 Laibsch: mine is GMA500 Jan 21 00:19:36 Laibsch: Poulsbo base Jan 21 00:19:39 Laibsch: Poulsbo based Jan 21 00:20:26 otavio: I thought the N450 has the GPU on chip? Jan 21 00:20:26 otavio iegd Jan 21 00:20:49 Laibsch: maybe; and depending on which GPU it can be a problem Jan 21 00:21:08 Laibsch: mine works in fb only or with closed (and horrible to use) drivers Jan 21 00:21:28 woglinde: iegd? Jan 21 00:21:57 otavio, woglinde: the upload is done: http://parenteses.org/mario/misc/log.bz2 Jan 21 00:22:06 mario-goulart: thx Jan 21 00:23:02 m Jan 21 00:23:06 hm Jan 21 00:23:14 cant see an error too Jan 21 00:23:27 what is bitbake doing in the end? Jan 21 00:25:53 woglinde: bitbake doesn't even try to beuid sqlite3 when we invoke it passing the image as argument. When we pass sqlite3 as argument, it builds it. Jan 21 00:26:28 aha Jan 21 00:26:32 re raster Jan 21 00:26:41 bitbake -g Jan 21 00:26:51 and look into deps graph Jan 21 00:26:59 maybee it isnt in there Jan 21 00:27:04 so it will never got build Jan 21 00:27:09 using the image as argument? Jan 21 00:27:24 yes Jan 21 00:27:44 woglinde: yo wogz. wassap? Jan 21 00:27:50 woglinde: I'd bet you got it. Jan 21 00:28:52 woglinde: that's it. Thanks a lot. Jan 21 00:29:19 woglinde: we were assuming firefox-3.5.5 was depending on sqlite3. In fact it is not. Jan 21 00:29:39 uh Jan 21 00:29:52 yeah thats clear an dep-error Jan 21 00:30:02 poor koen didnt notice it Jan 21 00:30:04 hm Jan 21 00:31:15 :-) Jan 21 00:32:19 okay good nite Jan 21 00:41:50 03Mario Domenech Goulart  07org.openembedded.dev * r0f3c7c020d 10openembedded.git/recipes/mozilla/ (4 files): Jan 21 00:41:50 firefox: added sqlite3 as dependency for firefox >= 3.5 Jan 21 00:41:50 Signed-off-by: Mario Domenech Goulart Jan 21 00:41:52 03Otavio Salvador  07org.openembedded.dev * rbc8ce0794c 10openembedded.git/recipes/parted/parted_1.8.8.bb: Jan 21 00:41:55 parted: depends on util-linux-ng to provide libuuid Jan 21 00:41:57 Signed-off-by: Otavio Salvador Jan 21 00:53:26 'nite **** ENDING LOGGING AT Thu Jan 21 02:59:56 2010