**** BEGIN LOGGING AT Thu Sep 20 02:59:57 2007 Sep 20 05:07:12 i'm trying to create a package that needs on postgresql's headers/libs. i can get postgresql into my image fine, and i see a postgresql-dev dir, but for some reason i can't see the postgresql headers even if i add DEPENDS = "postgresql-dev", and also "bitbake postgresql-dev" gives me "No providers or build target postgresql-dev" Sep 20 05:08:59 i also have a dependency on the pg_config program resulting from building postgresql. it dumps information about the postgresql build, and is typically included as part of the postgresql-dev package in distros Sep 20 05:09:16 can anyone please help? Sep 20 05:12:07 well, i've posted about this issue here, so if anybody has ideas, i'd love to hear them. thanks in advance! Sep 20 05:12:08 http://lists.linuxtogo.org/pipermail/openembedded-users/2007-September/000269.html Sep 20 06:51:26 i want to use directfb as the backend but x11. but after i modified DEPENDS in pango.inc & cairo.inc, the depends for xrender and x11 package is not removed. Sep 20 06:51:28 why it act as this, and how to correct the recipe? Sep 20 07:00:53 hi folks, i made an image for gumstix-connex, but when i try to boot it on the target i get an error like: fax-foo Sep 20 07:01:21 Bad magic number i mean. but when i load the kernel from the memory by flashing it and do loadm it works Sep 20 07:44:09 how can I do to have libstdc++ installed in staging when building C++ programs ? Sep 20 07:45:13 (hello everybody btw) Sep 20 07:54:24 morning all Sep 20 07:57:05 morning Sep 20 07:58:40 which is the package to be built in order to have the right stdc++ lib ? is it gcc or gcc-cross ? Sep 20 07:59:37 nud: gcc-cross Sep 20 08:00:03 ok that corresponds to what I was thinking Sep 20 08:00:26 RP: is the generated libstdc++ supposed to end up in staging ? Sep 20 08:01:28 * nud tries to understand why c++ linking doesn't work anymore Sep 20 08:07:41 nud: yes Sep 20 08:10:20 hm Sep 20 08:10:39 it clearly doesn't, for some reason Sep 20 08:15:21 nud: It probably ends up in cross/arm*/lib Sep 20 08:15:53 I did a find -name "libstd*.so" and there's just nothing Sep 20 08:15:58 in staging Sep 20 08:16:04 I didn't look in cross :$ Sep 20 08:26:31 is it possible with the SRC_URI to copy whole directorys with file:// ?? Sep 20 08:26:59 nik0n: it would be better to tar them Sep 20 08:33:38 XorA worked. thx Sep 20 08:33:43 any ideas for: Sep 20 08:33:43 | NOTE: make -e MAKEFLAGS= zImage CC=ccache mips-linux-uclibc-gcc LD=mips-linux-uclibc-ld Sep 20 08:33:46 | make: *** No rule to make target `zImage'. Stop. Sep 20 08:33:49 | FATAL: oe_runmake failed Sep 20 08:34:08 forgot to set S Sep 20 08:34:39 S = "${WORKDIR}/where-my-kernel-unpacks" Sep 20 08:35:09 no. got that Sep 20 08:35:34 no idea then Sep 20 08:35:53 hmm gonna look at a bitbake -D hopefully it helps Sep 20 08:36:27 look at tmp/work/arm*/packagename/temp/log.do_compile as well Sep 20 08:46:50 RP: it effectively ended up in tmp/cross/arm-angstrom-linux-gnueabi/libstdc++.{a,la,so} (I found it after a rebuild gcc-cross). Now how can I fix things up so that it ends up in staging as it is supposed to ? Sep 20 08:48:04 hmm can't find anything. any1 can have a look at http://pastebin.com/m67ec7cb3 please? Sep 20 08:52:33 hi Sep 20 08:53:00 I posted a fix for Orbit2 on ARM EABI to oe-devel Sep 20 08:53:40 kisg: please put in bugtracker Sep 20 08:53:52 But to get a proper fix into OE the way how the autotools site configuration has to be changed Sep 20 08:53:53 kisg: bugs.openembedded.org that way it can be tracked properly Sep 20 08:54:06 so I wanted to open a discussion about it Sep 20 08:54:17 just like I saw on the list for other issues Sep 20 08:54:34 kisg: mail hasnt come through yet Sep 20 08:54:50 kisg: but if its a test that needs its result hard coded then thats probably fine Sep 20 08:55:05 oh: should have checked the archives Sep 20 08:55:27 the problem is that you need different settings for ARM OABI and EABI Sep 20 08:55:39 because to structure alignment is different Sep 20 08:56:22 there is arm-linux and armeb-linux Sep 20 08:56:29 nud: Its fine in cross Sep 20 08:56:44 hm Sep 20 08:57:04 oh except is eb Edndian Big or EABI :-( Sep 20 08:57:40 but I guess in principle there is no problem with an arm-oabi arm-eabi site if that is needed Sep 20 08:57:55 ok Sep 20 08:58:04 kisg: I did see that mail and it does look like something we need to fix with different site files as you suggest Sep 20 08:58:22 then i will post a proper patch to the BT Sep 20 08:58:48 oh was this mail 1-2 days ago, I think I skipped across it Sep 20 08:58:58 thank you guys Sep 20 08:59:13 yes, I did not want to rush things Sep 20 08:59:16 kisg: Its really just a question of how to do it as neatly as possible Sep 20 08:59:36 exactly, that is why I asked it on the list Sep 20 09:03:11 kisg: I agree it needs separate files for oabi vs eabi. Have you any ideas how you're going to detect which we're using and which file to include? Sep 20 09:06:24 RP: I thought I saw in the angstrom*.conf the distinction between "arm" and "arm-oabi" Sep 20 09:07:02 but I am still new to OE, so I will have to dig in deeper Sep 20 09:07:25 kisg: We can't assume DISTRO=amgstrom really Sep 20 09:07:38 actually I just found a variable called ARM_ABI in angstrom conf Sep 20 09:07:40 mmh Sep 20 09:07:43 ok Sep 20 09:07:53 the linker error says undefined reference to `std::basic_istream >::read(char*, int)@GLIBCXX_3.4' Sep 20 09:08:29 RP: so should we like parse the compiler output or something similar? Sep 20 09:10:37 kisg: siteinfo works before we've built the compiler :/ Sep 20 09:11:03 so should I rebuild those libs that fail linking ? Sep 20 09:11:07 kisg: It could be we need to define some new variable like TARGET_ABI to handle this Sep 20 09:11:32 nud: You could try but your build seems a little broken tbh... Sep 20 09:11:45 RP: I was going to suggest something similar: I think the debian folks introduced a new architecture for EABI Sep 20 09:12:22 kisg: I think there is some variable already actually, probably TARGET_OS Sep 20 09:13:20 kisg: Yes, you can use TARGET_OS Sep 20 09:13:39 RP: with linux-gnueabi as the value? Sep 20 09:13:52 kisg: Yes, its linux for oabi Sep 20 09:14:23 so it basically contains the OS part from the autotools target definition? Sep 20 09:14:31 RP: thank you for your help Sep 20 09:14:57 RP: I will try to come up with a proper patch and open an issue for it in the BT Sep 20 09:15:08 kisg: I think the syntax to handle target_os specific files already exists too so its just a question of moving the entries to files with the right names Sep 20 09:15:32 RP: great, thank you Sep 20 09:16:05 kisg: You will need to tweak siteinfo.bbclass but once you look at the file, you should see how to do it Sep 20 09:16:14 sure Sep 20 09:17:20 RP: Actually I also saw the discussion you started about changing the way staging works Sep 20 09:17:31 RP: Could I ask a possibly naive question? Sep 20 09:17:41 kisg: sure Sep 20 09:18:26 good morning all Sep 20 09:19:31 RP: I saw an initiative about packaged staging on the OE wiki. What is the greatest obstacle in using the target packages for the staging area? Sep 20 09:19:31 XorA: Did you say it was the restaurant hotel or garden hotel? I think you said restaurant? Sep 20 09:19:36 RP: RH Sep 20 09:19:59 XorA: Thats what I thought, thanks. Someone is making me doubt that, thats all ;-) Sep 20 09:20:09 RP: kill the doubters Sep 20 09:25:51 kisg: The problem is we have more in staging than is covered by the -dev packages and sometimes the files in staging are modified compared to a target device Sep 20 09:26:21 kisg: Also, -dev packages don't cover native staging Sep 20 09:27:08 RP: I see Sep 20 09:27:54 TARGET_OS hs all the info you need for eabi/oabi Sep 20 09:28:23 the only thing we can't handle is oabi/vf Sep 20 09:28:28 p Sep 20 09:28:31 koen|away: We've established that now ;-). Sep 20 09:28:56 just making sure : Sep 20 09:29:00 ) Sep 20 09:29:14 koen: thanks Sep 20 09:29:20 * koen|away stabs hildom-im Sep 20 09:29:50 kisg: jo napot! Sep 20 09:30:10 koen: jo napot neked is! :) Sep 20 09:31:21 zecke: im at schiphol airport now Sep 20 09:32:02 koen|away: aeh? Sep 20 09:32:13 I'm in Eindhoven now :) Sep 20 09:32:17 koen|away: what are you doing? Sep 20 09:32:21 'working' Sep 20 09:34:49 could somene mention fakeroot in #3034? Sep 20 09:35:18 the n800 lacks my bugzilla credentials Sep 20 09:35:59 zecke: om-browser2 looks nice n qvga Sep 20 09:36:50 zecke: did chris' fix for fingercroll fix the bug you saw in your browser? Sep 20 09:41:13 hey mickeyl Sep 20 09:53:08 03xora 07org.oe.dev * r4e3175da... 10/ (3 files in 3 dirs): Sep 20 09:53:08 mplayer/mplayer_svn.bb : update the svn build again and also add ssvb's Sep 20 09:53:08 turbo speed IDCT routine from https://garage.maemo.org/projects/mplayer/ Sep 20 10:07:27 RP: after rebuilding the c++ dependencies, it now links fine Sep 20 10:09:34 /home/dp/zaurus/build-ang/tmp/staging/arm-angstrom-linux-gnueabi/lib/libpoppler- Sep 20 10:09:34 glib.so: undefined reference to `vtable for __cxxabiv1::__si_class_type_info@CXX Sep 20 10:09:34 ABI_1.3' Sep 20 10:09:36 WTF Sep 20 10:11:31 XorA: my friend! Sep 20 10:11:50 I've been struggling with the same issue since yesterday Sep 20 10:11:50 nud was that what happened to you? Sep 20 10:11:58 XorA: same error messages Sep 20 10:12:07 try rebuilding the library Sep 20 10:12:29 nud: poppler in this case? Sep 20 10:12:32 yep Sep 20 10:12:42 in my case it was pwlib and opal but yet, now it builds Sep 20 10:12:54 I guess you're compiling evince or something Sep 20 10:12:58 nud: yes Sep 20 10:13:21 I guess there has been some ABI change in OE's libc++ yesterday Sep 20 10:13:36 but C++ packages weren't rebuilt Sep 20 10:13:54 * XorA glares at someone Sep 20 10:14:29 what's annoying is that we are likely to have unpredictable breaks if there are C++ packages that were not rebuilt... Sep 20 10:15:11 XorA: actually I'm glad it happened to someone else :p Sep 20 10:34:55 how do you setup an ipkg feed server ? I'd like to avoid generating the image and flashing my card each time I update the soft I'm writing Sep 20 10:38:21 actually the best thing would be to have ssh://myhost/stuff/build/tmp/deploy/ipkg as the source Sep 20 10:38:49 I didn't find any howto on how to do that easily, just one about how to set up FEED_URI Sep 20 10:42:14 nud: cheers that fixed it Sep 20 10:42:31 sorry if i misunderstand, but i've simply shared deploy/glibc/ipk with apache and put the subdirs as arc/gz in /etc/ipkg/build.conf Sep 20 10:42:46 XorA: :-) Sep 20 10:44:37 koen|gprs: package evince-2.20.0: completed Sep 20 10:48:28 great Sep 20 10:49:55 hmmm Sep 20 10:50:27 the math actually checks out in this railgun articlr Sep 20 10:50:57 even if their drawing is highly misleading Sep 20 10:52:32 * koen|gprs should've briught his laptop Sep 20 10:53:02 koen : you said that last time... Sep 20 10:53:21 indeed Sep 20 10:54:06 koen what version rom do you use for bootldr on hx4700 ? Sep 20 10:54:26 XorA: do you link with gcc or g++? Sep 20 10:55:02 zecke: gcc Sep 20 10:55:17 zecke: but thats all inside libtool magic, not my choice Sep 20 10:55:33 XorA: add -lsupc++ and see what is happening Sep 20 10:55:45 zecke: rebuilding poppler worked Sep 20 10:55:55 zecke: so the test case is now destroyed Sep 20 10:56:14 ouch Sep 20 10:58:39 ~lart the horrendous hack that is C++ Sep 20 10:58:40 * ibot pries the horrendous hack that is C++'s back open with a screwdriver and flashes a new bootldr to the horrendous hack that is C++ Sep 20 11:01:23 03xora 07org.oe.dev * r451f5e51... 10/ (4 files in 3 dirs): mplayer/mplayer_svn.bb : convert to SRCREV Sep 20 11:01:28 03xora 07org.oe.dev * r4c020d03... 10/ (1 packages/mplayer/files/makefile-nostrip-svn.patch): mplayer/files/makefile-nostrip-svn.patch : forget to push latest version Sep 20 11:03:33 ~logs Sep 20 11:03:34 [logs] apt/ibot/infobot/jbot/purl all log daily to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz Sep 20 11:31:51 koen, I just tried to reproduce your bug in OE without your patch Sep 20 11:31:55 koen, your dbus bug Sep 20 11:32:06 koen, I don't get a segfault Sep 20 11:33:24 great Sep 20 11:33:38 thanks for testing Sep 20 11:33:49 koen|gprs: Can I drop the patch now? Sep 20 11:34:18 yes please :) Sep 20 11:36:14 koen|gprs, thos: pushed :) Sep 20 12:50:11 morning Sep 20 13:04:21 how can I forbid a particular version of a package ? I'd like to prefer alsa 1.0.13 due to a bug on 1.0.14 but PREFERRED_VERSION_alsa-lib = "1.0.13" doesn't work, nor does PREFERRED_VERSION_libasound = "1.0.13". Sep 20 13:05:01 excluding alsa.*1\.0\.13 seems to work but it's not very clean Sep 20 13:33:00 *bored* Sep 20 13:33:46 cbrake: ping Sep 20 13:33:51 hi koen|gprs Sep 20 13:33:54 nud: hello Sep 20 13:34:37 cbrake: do you do driver development ? Sep 20 13:34:50 nud: yes Sep 20 13:35:14 me too Sep 20 13:35:17 for osx/linux Sep 20 13:36:47 03rpurdie 07org.oe.dev * rf0c70338... 10/ (5 files in 3 dirs): dbus: Drop fix-segfault patch. Thei issue it fixed can no longer be reproduced (tested by thos) Sep 20 13:36:53 03daniel 07org.oe.dev * r4c89fe83... 10/ (1 packages/openmoko2/libmokoui2_svn.bb): libmokoui2: Touch gtk-doc.make to fix issues with gtk-doc support (Thomas Wood) Sep 20 14:27:11 Looks like epiphany DEPENDS on libstartup-notification -- any objections to adding that? Sep 20 14:33:50 hi koen|gprs Sep 20 14:34:14 neat, midori and epiphany both build :-) Sep 20 14:34:15 * mwester sees yet another full openmoko build that ran without fetch errors of any sort - hurray for SRCREV! Now he just has to figure out what to do with several boxes of incense, and a bucket of chicken entrails now that they are unnecessary... Sep 20 14:35:54 help, i'm trying to create a package that needs on postgresql's headers/libs. i can get postgresql into my image fine, and i see a postgresql-dev dir, but for some reason i can't see the postgresql headers even if i add DEPENDS = "postgresql-dev", and also "bitbake postgresql-dev" gives me "No providers or build target postgresql-dev" Sep 20 14:36:59 i guess that makes sense because postgresql-dev is a package and not a recipe but then how do i make those headers visible to the package i'm building? the package is a python package, so the compilation is run via a setup.py Sep 20 14:37:37 the setup.py also runs the pg_config program resulting from building postgresql, which dumps information about the postgresql build Sep 20 14:38:10 mwester, can you send them to me? I'm doing bringup on a new board (-; Sep 20 14:39:14 i've spent the past 24 hours trying to solve this, it's been an enormous time (and sanity) sink Sep 20 14:42:57 zeeeee: cross compiling is difficult :-( Sep 20 14:43:21 zeeeee: you are correct, there is no postgresql-dev recipe -- that is a package, not a recipe Sep 20 14:43:24 cbrake: i'm not sure this is even a cross compiling issue; i just want to make some compiled files visible Sep 20 14:43:37 *compiled and header files Sep 20 14:43:49 zeeeee: have you verified the headers are in tmp/cross? Sep 20 14:44:25 cbrake: no they're not, they're only in.... Sep 20 14:45:04 ./work/i486-linux/postgresql-8.2.4-r0/install/postgresql-dev/usr/include/ Sep 20 14:45:44 zeeeee: there is likely a problem with the postresql recipe then Sep 20 14:46:41 zeeeee: the headers must make it into the tmp/cross (or maybe it is tmp/staging dir). tmp/work is all temporary Sep 20 14:46:47 cbrake: http://rafb.net/p/856OOo70.html Sep 20 14:47:03 cbrake: yeah i think they need to be in tmp/staging Sep 20 14:47:11 however even copying them there manually did nothing! Sep 20 14:47:47 e.g. i manually copied pg_config into staging/i686/bin/ and the include/ and lib/ files into staging/i686/... Sep 20 14:48:35 i instrumented the setup.py with a debug print of the PATH environment variable and saw .../i686/bin/ Sep 20 14:48:41 so that's where i put it Sep 20 14:48:48 but still no go...i'm at the end of my wits Sep 20 14:50:26 zeeeee: well, I'd guess you have several problems -- the postgresql is build system needs patched for cross compiling, and your package may not be working properly. Sep 20 14:50:45 zeeeee: have you build any of the other python extensions, and looked at the build environment to see how things are being done. Sep 20 14:51:10 zeeeee: the files in tmp/work/....//temp/run.do_xxx are very useful Sep 20 14:51:35 zeeeee: you can run them manually for debugging, and compare them with other packages Sep 20 14:52:34 zeeeee: add a package to OE is difficult, just because the inherit difficulty in cross compiling. This is why the working stuff in OE has so much value. Sep 20 14:59:37 zeeeee: You want the headers in tmp/staging/i484-linux/... Sep 20 15:01:52 * mwester prefers i686-linux ;) Sep 20 15:06:59 RP: ok i just copied the pg_config program to the i486-linux/bin dir (setup.py runs this first before compiling anything/using any headers) but that still doesn't work Sep 20 15:07:22 always the same error: http://rafb.net/p/VmDCkT78.html Sep 20 15:08:47 zeeeee: Which pg_config binary is it actually running? Sep 20 15:09:01 zeeeee: Then you need to find out which libreadline that binary actually wants Sep 20 15:09:09 RP: nothing is, that's the problem - setup.py says it cannot find a pg_config binary Sep 20 15:09:52 zeeeee: cross binaries like that should usually be in tmp/staging/i686-linux/bin/i486-linux Sep 20 15:15:28 RP: here is the relevant code in setup.py: http://rafb.net/p/3CHWgW33.html Sep 20 15:16:06 i added this debug print which showed that directory you pointed out as well as its parent dir (i686/bin): print 'PATH', os.environ['PATH'] Sep 20 15:16:55 RP: what's meant by libreadline? Sep 20 15:18:41 RP: ok, i added readline to my RDEPENDS but that didn't seem to change anything Sep 20 15:18:48 (and to DEPENDS) Sep 20 15:20:40 zeeeee: When it runs pg_config it gets the error "libreadline.so.4: cannot open shared object file: No such file or directory" Sep 20 15:20:53 zeeeee: pg_config is linked against a library it can't find Sep 20 15:22:10 RP: oh shoot, you're right - i missed that, i kept seeing a very similar error message Sep 20 15:22:33 RP: what else would i need aside from adding readline to DEPENDS and RDEPENDS? Sep 20 15:23:25 (so manually copying the file(s) into place does work, it's now a different problem i'm seeing) Sep 20 15:24:52 RP: postgresql was built with just 'readline' in its DEPENDS Sep 20 16:56:01 re Sep 20 16:56:55 wb Sep 20 16:59:38 *hungry* Sep 20 17:00:02 * zecke_ too Sep 20 17:00:26 * zecke_ is getting a soup + mexican food and hugs his family Sep 20 17:01:15 * koen|gprs is in a train and will probably vist the bk at the end of the line Sep 20 17:01:34 * chouimat|lunch just had a nice pea soup Sep 20 17:03:30 Anyone seen this error with current OE/OM ? Sep 20 17:03:41 ecognized option `-Qy' Sep 20 17:03:41 | make[2]: *** [g++spec.o] Error 1 Sep 20 17:03:51 /build/tmp/cross/bin/arm-linux-as: unr Sep 20 17:03:55 1st line of the error Sep 20 17:04:05 when building gcc-cross-4.1.1-r11 Sep 20 17:05:14 4.1.1?!?!? Sep 20 17:09:17 yeah Sep 20 17:09:26 also saw with 4.1.0 Sep 20 17:09:39 i think i did downgrade to 4.0.2 but couldn't convince it to downgrade glibc to 2.4 Sep 20 17:20:39 hello, i meet a problem about using ipkg-cl. i modified the recipe for gtk(including pango & cairo).it is reported that dependency failed on libx11-6 when do_rootfs. i have checked the status file in rootfs, there is no words on libx11. Sep 20 17:20:44 please help Sep 20 17:54:15 gm Sep 20 17:54:54 are there any .bb's that use a daily snapshot tarball for source? Sep 20 17:55:16 hi likewise Sep 20 17:58:18 HopsNBarley: hi theree Sep 20 17:59:26 any opinions on a boot-from-NAND consumer device? i'm just so used to seeing NOR boxes... Sep 20 17:59:56 for consumer stuff, perfectly fine Sep 20 18:07:45 likewise, thanks. we are having a debate (-; Sep 20 18:24:28 HopsNBarley: NAND in heavy read-write use mode? Sep 20 18:24:53 HopsNBarley: or written only occasionally (with a purpose other than atime :-))? Sep 20 18:27:52 steliosk, we should propose an OE talk for the TI developers conference .... Sep 20 18:30:23 steliosk: board arrived, thanks! Sep 20 18:30:40 * Crofton heads for lunch Sep 20 18:39:15 likewise, it won't get much write traffic. i was concerned about ECC handling at boot time, but the board/chip we are looking at has some of this hardware. Sep 20 18:48:07 hi all Sep 20 18:48:42 i continue have problem with binutils-cross_2.18 ... (i think a week or more) target is arm-oabi ... any clue ? Sep 20 18:51:52 gremlin[it]: whihc host machine? Sep 20 18:52:02 x86_64 Sep 20 18:52:46 gremlin[it]: there is a bug report on this, with workarounds for 2.18, search for binutils in bugs.oe.org Sep 20 18:52:47 i notice thta on the do_stage log but i'm not sure is the only trouble ... "rmdir: /mnt/hd1/oe/ipaq/tmp/cross//mnt/hd1/oe/ipaq/tmp/cross: No such file or directory" Sep 20 18:53:04 notice the repetition of path ... Sep 20 18:53:10 gremlin[it]: one "workaround" is to use binutils 2.17. Sep 20 18:53:15 thnka likewise i take a look Sep 20 18:53:36 gremlin[it]: I noticed a lot of hardcoded paths are still in the metadata. Sep 20 18:58:29 likewise: howdy Sep 20 18:59:09 likewise: what document viewer is used in OE for ARM for viewing pdf's doc files Sep 20 19:00:12 khem_: I've noticed evince, have not tried it yet Sep 20 19:00:20 evince or epdfview Sep 20 19:00:30 epdfview allows stylus drag of pdf so its nicer Sep 20 19:00:31 and for MS words Sep 20 19:00:35 abiword Sep 20 19:00:53 do you know how much size they incur ? Sep 20 19:01:03 look at package browser Sep 20 19:04:46 XorA|gone: and for excel files ? Sep 20 19:05:51 Crofton : Yeah. Why not. Sep 20 19:06:35 koen : nice. hopefully the others will arrive tomorrow Sep 20 19:07:19 khem_: howdy. you removed the eglibc patch. But does eglibc build for you? Sep 20 19:07:39 yes just yesterday did a clean build Sep 20 19:07:41 !oebug 2993 Sep 20 19:07:59 I also upped the svn rev Sep 20 19:08:07 now it builds fine Sep 20 19:08:56 khem_: which svn rev did build ok for you? Sep 20 19:13:22 hmmm 3531 likewise Sep 20 19:13:49 do update your copy and try Sep 20 19:14:16 btw. I just did it for arm so if you are usng other arches then there might be problems that I have not seen Sep 20 19:14:22 khem_: gnumeric Sep 20 19:14:31 khem_: x86 actually Sep 20 19:14:41 likewise: it should just build Sep 20 19:14:45 did u try Sep 20 19:14:51 with all updates Sep 20 19:14:56 XorA|gone: thanks Sep 20 19:16:03 khem_: will commence a fresh build Sep 20 19:16:27 ok lemme know if you hit into bumps Sep 20 19:16:51 I just build angstrom-console-image for osk5912 from scratch Sep 20 19:16:52 khem_: there is support to include patches depending on svn revision. with maxrev=3447 for example Sep 20 19:17:07 ah Sep 20 19:17:21 khem_: very new stuff: http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-August/007387.html Sep 20 19:17:33 khem_: haven't tried so myself Sep 20 19:17:44 yeah I was following it on ml Sep 20 19:17:52 havent tried myself Sep 20 19:40:24 03xora 07org.oe.dev * reee55938... 10/ (3 files in 3 dirs): Sep 20 19:40:24 mplayer/files/simple_idct_armv5te.S : reduce MAX_NEG_CROP to same as Sep 20 19:40:24 dsputil.h until SSVB has managed to get his patch series into mplayer. Sep 20 19:40:29 03mickeyl 07org.oe.dev * rdbabf684... 10/ (1 packages/openmoko2/moko-gtk-engine.bb): add moko-gtk-engine, a gtk theme engine to improve painting performance Sep 20 19:40:34 03cbrake 07org.oe.dev * r766789e1... 10/ (1 packages/gnome/epiphany_2.19.6.bb): epiphany: add libstartup-notification to DEPENDS Sep 20 19:40:39 03mickeyl 07org.oe.dev * rea8da631... 10/ (1 conf/distro/include/moko-autorev.inc): moko-autorev.inc: add entry for moko-gtk-engine Sep 20 19:40:44 03xora 07org.oe.dev * r293b1add... 10/ (1 packages/evince/evince_2.20.0.bb): evince/evince_2.20.0.bb : new release with gnome 2.20 Sep 20 20:27:38 ~change 1 usd to cad Sep 20 20:27:43 i got some error trying that Sep 20 20:28:05 Crofton: probably a division by zero error Sep 20 20:28:12 grrr Sep 20 20:28:34 USD being nearly zero? :-) Sep 20 20:28:39 http://www.bloomberg.com/apps/news?pid=20601082&sid=awhVMIymbX0w&refer=canada Sep 20 20:28:55 apparently they were equal for a short time Sep 20 20:40:04 likewise, (_; Sep 20 20:45:04 koen: "I god fed up with the lusers complaining of the problems whilst Sep 20 20:45:07 waving their enlarged penises." Sep 20 20:45:11 koen: http://www.mail-archive.com/ion-general@lists.berlios.de/msg02004.html Sep 20 20:53:32 the correct monotone work flow is (work), commit, pull, merge, sync, right? Sep 20 20:58:04 HopsNBarley: push works too Sep 20 20:59:56 zecke, thanks. Sep 20 21:42:47 http://tree.celinuxforum.org/CelfPubWiki/ELCEurope2007 Sep 20 21:42:52 anyone going there? Sep 20 21:43:53 likewise: I hope my talk gets accepted and then I will be there... Sep 20 21:45:53 zecke: good luck, on what subject? Sep 20 21:46:04 WebKit Sep 20 22:17:55 good nite Sep 20 23:56:15 I currently have problems creating a bugtracker account, so whoever wants to fix this: I have a patch for http://bugs.openembedded.org/show_bug.cgi?id=2813 at http://www.ploetzli.ch/forumtst/fix-libsdl-net.patch Sep 21 01:21:05 How can I get bitbake to tell me where a dependancy comes from? Sep 21 01:44:18 Tartarus use it with -g and it generates .dot files Sep 21 01:44:31 you can look into them with vim and search for your deps Sep 21 01:45:49 (or use dot to render them to ps or something, then zoom and scroll around and try to follow the arrows :) Sep 21 02:17:33 I'll try -g, thanks **** ENDING LOGGING AT Fri Sep 21 02:59:57 2007