**** BEGIN LOGGING AT Sun Jan 27 02:59:57 2008 Jan 27 06:50:36 !logs Jan 27 06:50:38 Channel logs for #oe are archived at: Jan 27 06:50:39 http://hentges.net/tmp/logs/irc/%23oe Jan 27 06:50:40 Live-logs are available at Jan 27 06:50:41 http://hentges.net/tmp/logs/irc/livelogs/%23oe.livelog Jan 27 06:50:43 See ?? help-logs for usage instructions Jan 27 09:51:27 hi Jan 27 09:51:37 how do I make bitbake generate a feed for the packages i built so far? Jan 27 10:04:22 morning all Jan 27 10:07:40 morning thesing Jan 27 10:07:52 I ordered the book printed by the way Jan 27 10:07:56 should be here in a few weeks Jan 27 10:08:39 wvdschel: ok. Jan 27 10:09:51 wvdschel: if you want to get started in doing something you could look at zaurusd and make it calling "apm -s" when the cancel key is pressed for some time. Jan 27 10:09:52 hi Jan 27 10:13:50 hmm I am compiling glibc and I guess it needs to include my arch specific asm directory that is located in the kernel subdir. is that the right thinking or ams I missing symlinks? Jan 27 10:13:59 thesing, alright Jan 27 10:15:24 keesj: AFAIK you need kernel headers to compile any libc. Jan 27 10:16:12 thesing: yes , I also think so , but apparently it is not done in my setup :( Jan 27 10:16:44 I think people like to compiled against generic kernel headers Jan 27 10:17:06 are you using a generic distro? Jan 27 10:18:02 mamona (based on angstrom) Jan 27 10:18:28 but I am trying to compiled a "x86_64" version Jan 27 10:19:24 perhaps it is a pretty weird idea :p. I have not the feeling the x86_64 is well supported by oe Jan 27 10:19:33 I have no idea how well this arch is supported by oe. Jan 27 10:23:36 someone has recepie by hand for qemu build taking gcc 4.x instead of 3? Jan 27 10:26:24 just let quemu use 4.x is ok? Jan 27 10:26:59 gints: IIRC, qemu doesn't compile well with 4.x; you should probably install gcc-3.3 Jan 27 10:27:21 k.. thx Jan 27 10:27:42 hi, i have a little problem with mtn and the database, i want to update the database, but I allways get a database is locked. The strange thing is, that if i cope the database from my network-folder to the local-folder everything works great Jan 27 10:29:16 Azrael: Using it over nfs? Jan 27 10:29:40 Azrael: Make sure the nfs mount has lock in the options Jan 27 10:29:41 i tried nfs3 and cifs Jan 27 10:30:07 hm Jan 27 10:30:08 actually Jan 27 10:30:10 ok, i will try lock Jan 27 10:30:12 is it a soft or hard nfs mount? Jan 27 10:31:41 its the default Jan 27 10:31:58 nfs v3 or v4? Jan 27 10:32:21 v3 Jan 27 10:34:58 the client-configuration is: remote_Machine:/Daten /Daten nfs nfsvers=3,hard,defaults 0 0 Jan 27 10:38:45 lock is also not working, still the sqlite error: database is locked Jan 27 10:44:56 hm Jan 27 10:45:09 if you're absolutely sure no remote machines will touch it, use nolock in the mount options Jan 27 10:45:14 that'll only do local locks Jan 27 10:46:09 ok, nolock worked ;) Jan 27 10:46:33 thanks a lot Jan 27 10:47:03 np :) Jan 27 10:47:11 I only have 2 computers, an storage server and my local computer, so this should work ;) Jan 27 10:47:48 Ah, cool :) Jan 27 10:48:09 If you're doing a full build of a distro, say, angstrom or openzaurus, you'll need something in the vicinity of 3-10gb, depending on what you build. Jan 27 10:48:51 mhm, thats why i need the server, my computer has only 20Gb, und 2gb are free oO Jan 27 10:49:26 I dont need a full distro, only qtopia gtk and maybe x11, hopefully it will work Jan 27 10:49:39 ah, good luck :) Jan 27 10:49:47 thx :) Jan 27 11:09:25 lol question I just tought about.. when my HD is full .. what film to delete now :) Jan 27 11:09:46 2 G is used already.. .how much more :) Jan 27 11:10:20 gints: tmp/ is where all the output goes Jan 27 11:11:08 gints: if you are not debuging build, add INHERIT += "rm_work" to your local.conf Jan 27 11:11:25 k.. I do Jan 27 11:12:29 yes ..btw nooby question where is supposed to be local.conf :) Jan 27 11:12:39 I configure by hand first :) Jan 27 11:13:23 build/conf/local.conf Jan 27 11:13:35 oh ok tx Jan 27 11:21:43 03xjqian 07org.oe.dev * r094e9bfd... 10/ (1 packages/opie-taskbar/opie-taskbar-images.inc): opie-taskbar-images.inc: fix -dbg listed in PACKAGES multiple times (QA) Jan 27 11:26:15 03rwhitby 07org.oe.dev * rd2c5e48f... 10/ (5 files in 4 dirs): apex: Added automatic fatslug detection. Jan 27 12:14:04 * * OE Bug 3703 has been REOPENED by Jan 27 12:14:06 * * libwnck-2.13.5-r0-do_install Jan 27 12:14:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3703 Jan 27 12:14:16 * * OE Bug 3703 has been RESOLVED (WORKSFORME) by Jan 27 12:14:18 * * libwnck-2.13.5-r0-do_install Jan 27 12:14:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3703 Jan 27 12:14:28 * * OE Bug 3694 has been RESOLVED (WORKSFORME) by Jan 27 12:14:30 * * libsdl-mixer-1.2.6-r1-do_install Jan 27 12:14:32 * * http://bugs.openembedded.org/show_bug.cgi?id=3694 Jan 27 12:24:32 xjqian: Your latest commit should not have been necessary Jan 27 12:25:02 =+ is appended and thus $PN-dbg should already be at the first position Jan 27 12:25:29 Does opie-taskbar-images build for you? What machines? Jan 27 12:26:21 Laibsch: I know, but QA barks. And I don't see why you need $PN-dbg in =+? Jan 27 12:26:56 QA should not bark Jan 27 12:27:13 Essentially, your commit does not change anything, really Jan 27 12:27:26 If it barked, it probably still barks Jan 27 12:27:35 Or you have some other problem Jan 27 12:28:03 Laibsch: It did. opie-taskbar-images-dbg listed multiple times in PACKAGE Jan 27 12:28:15 BTW, I don't see the need for $PN-dbg in =+, either, but it was you who made the commit Jan 27 12:28:24 Laibsch: QA even asked you to submit a bug :) Jan 27 12:28:31 Not me Jan 27 12:28:55 Having it listed multiple times in PACKAGE should now list it even one more time Jan 27 12:29:18 xjqian: My bad Jan 27 12:29:32 I read your commit with "-R"-glasses on ;-) Jan 27 12:29:39 The commit is sane Jan 27 12:29:53 no problem Jan 27 12:30:57 Laibsch: I do have a question because I have to patch opie-taskbar-images_cvs.bb to build for opie-1.2.4-pre Jan 27 12:31:33 re Jan 27 12:32:06 Laibsch: it is that do_configure() calls base_do_configure, which does nothing and is the correct behavior in 1.2.3 Jan 27 12:32:55 Laibsch: however in 1.2.4-pre or do_configure() calls qmake_base_do_configure, which brings in all kinds of crap is wrong Jan 27 12:33:34 Laibsch: you know why the discrepency or anything changed recently? Jan 27 12:56:30 no idea, sorry Jan 27 12:56:37 I am firmly in 1.2.3-land Jan 27 12:56:54 bluelightning and polyonymous are the ones to talk to, I think Jan 27 13:00:48 Laibsch: did you find the error in your build environment? Jan 27 13:00:58 No, not yet Jan 27 13:01:22 But I can build images fine from a clean Sonkei environment on another (and fast) computer Jan 27 13:01:39 So, it looks like the local changes I have have brought this problem about Jan 27 13:01:53 It should be relatively easy to bring things back to normal. Jan 27 13:03:24 xjqian: I looked through the log for opie-taskbar-images. Jan 27 13:03:36 Because the commit you made will cause other problems. Jan 27 13:04:05 And I found that the =+ creeped in while I was experimenting with that file, but it was never meant to be really committed. Jan 27 13:04:22 I just overlooked it when doing the "inherit opie" line Jan 27 13:04:28 03pfalcon 07org.oe.dev * rcce3c198... 10/ (1 packages/linux/linux-handhelds-2.6-2.6.21/h5000/defconfig): Jan 27 13:04:28 linux-handhelds-2.6 2.6.21: Enable USB WiFi support for h5000. Jan 27 13:04:28 * Possibly, incomplete. Jan 27 13:04:28 * Based on defconfig from Koen Kooi. Jan 27 13:04:28 * defconfigman r432 Jan 27 13:04:37 03pfalcon 07org.oe.dev * r9814a7ac... 10/ (24 files in 24 dirs): Jan 27 13:04:37 linux-handhelds-2.6 2.6.21: Update to latest defconfigman. Includes: Jan 27 13:04:37 * magician flash boot hacks Jan 27 13:04:38 * mtdblock support moves to modules, as non-essential for flash boot. Jan 27 13:04:40 * Few netcards PCMCIA drivers enabled for hx4700. Jan 27 13:09:44 Laibsch: what do you mean? Jan 27 13:10:05 Just a second Jan 27 13:10:16 I will commit what is the right solution in a minute Jan 27 13:11:50 The right thing was not to remove $PN-dbg, but replace =+ with = Jan 27 13:12:57 Laibsch, xjqian I haven't tried -pre for a few months now, I think bluelightning is the one to talk to. Jan 27 13:13:16 in fact I haven't built 1.2.3 either for quite a while :) Jan 27 13:14:05 polyonymous: thanks, I will try to talk to bluelightning. Jan 27 13:14:48 btw, except for that small annoyance, 1.2.4-pre builds fine with todays SRCDATE Jan 27 13:24:22 Laibsch: collie hardware does support sdio. Jan 27 13:24:39 ?? Jan 27 13:24:56 Did you assume until now that it did not? Jan 27 13:25:07 What benefit does it bring? Jan 27 13:25:19 I read it somewhere that some important pins were not connected. Jan 27 13:25:38 well you could use sdio wifi or gps cards. Jan 27 13:26:10 and use your a microdrive in cf. Jan 27 13:26:46 03Laibsch 07org.oe.dev * rf82b7629... 10/ (1 packages/opie-taskbar/opie-taskbar-images.inc): Jan 27 13:26:46 opie-taskbar-images: fix the PACKAGES line Jan 27 13:26:46 * while doing the "inherit opie" thing in 8725fc715529f6dba921a0a92e01e4e68124bfdb Jan 27 13:26:46 I also committed a broken PACKAGES line I had been testing out locally which was Jan 27 13:26:46 turned out to be unsuitable for the public repo. Jan 27 13:27:12 thesing: Nice Jan 27 13:27:26 I think I don't have an SDIO device, though Jan 27 13:27:36 Is the SD wifi card SDIO? Jan 27 13:28:04 yes. Jan 27 13:28:23 I remember some people had problems because wifi blocked the CF slot which they wanted to use for something else (GPS, possibly) Jan 27 13:28:37 afaik everything that goes into sd-slot and is no flash card is sdio. Jan 27 13:29:25 they should connect gps with normal uart. Jan 27 14:00:07 03xjqian 07org.oe.dev * r08302dde... 10/ (17 files in 6 dirs): Jan 27 14:00:07 gnuplot: update 4.2.2, add 4.3.0+cvs Jan 27 14:00:07 * restructure patch directories Jan 27 14:00:07 * update angstrom-preferred Jan 27 14:30:04 Laibsch: could you commit http://pastebin.ca/874811 ? With this one can build images that start for collie. Jan 27 14:47:04 * * OE Bug 3645 has been RESOLVED (FIXED) by thommycheck(AT)gmx.de Jan 27 14:47:07 * * Collie: interface for mail and charging leds Jan 27 14:47:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3645 Jan 27 14:53:55 I am still kinda stuck with building the glibc. for x86_64 , I had to point the with-headers to the staging/kernel directory and now I have http://paste-it.net/6033 multiple definition of `__libc_multiple_libcs' Jan 27 14:55:06 any thints would be greatlty apreciated :p Jan 27 15:05:46 cold anyone tell me why TARGET_OS would be set to "linux-gnueabi" for collie builds? It should be "linux" but I can't figure out why it is wrong for me Jan 27 15:08:55 its ok here. maybe something in your local.conf? Jan 27 15:10:52 I wish, local.conf looks innocent enough Jan 27 15:11:55 my python-foo is not strong enough to reverse-engineer the setting of TARGET_OS Jan 27 15:16:06 CoreDump: well with fgrep you should be able to do it. Jan 27 15:17:28 trust me, you have to speak python to understand how it is set Jan 27 15:36:53 well, here is one for the python experts. TARGET_OS gets set correctly when I have 'MACHINE = "collie"' in local.conf but it gets set wrong when I use 'MACHINE = "${@bb.data.getVar('OE_MACHINE', d, 1)}"' while OE_MACHINE="collie". What the hell? Jan 27 15:37:54 'MACHINE = "collie"' is reported in both cases! Jan 27 15:38:46 has probably somezhing to do with ARM_ABI ?= "${@['','oabi'][bb.data.getVar('MACHINE',d) in ['collie','h3600', 'h3800', 'simpad', 'htcwallaby']]}" Jan 27 16:02:47 <_law_> hi, any ideas why my kdrive xserver doesnt build? http://paste.uni.cc/18204 Jan 27 16:04:04 * * OE Bug 3708 has been marked as DUPLICATE of bug 3544 by thommycheck(AT)gmx.de Jan 27 16:04:06 * * Touchscreen non responsive on 2.6.24-rc8 Jan 27 16:04:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3708 Jan 27 16:04:25 * CoreDump dunnos Jan 27 16:17:00 why does angstrom build fortran?? Jan 27 16:17:41 gcc-cross-4.1.1 patches for fortran, then dies complaining of MPFR not supporting it Jan 27 16:18:23 are you workign with a recent checkout? Jan 27 16:18:38 there are some packages that need FORTRAN :) Jan 27 16:18:44 so it builds the toolchain Jan 27 16:18:57 it was checked out yesterday, I will resync Jan 27 16:19:05 hmmm Jan 27 16:19:12 I thought a fix was already in Jan 27 16:19:19 Crofton: get out of here, who uses fortran on a handheld? Jan 27 16:19:26 LOL Jan 27 16:19:32 not all targets are handhelds Jan 27 16:19:42 and someone built octave :) Jan 27 16:19:55 * T0mW says nothing about ratfor Jan 27 16:21:31 Crofton: I figured that since we were going to the trouble to build a distro for the Zipit Z2 that I would use the Angstrom distro as it seems that everyones focus is on that? Jan 27 16:22:46 um, it is gcc-cross-4.2.2 not 4.1.1 Jan 27 16:23:09 same error: Jan 27 16:23:24 checking for correct version of mpfr.h... buggy version of MPFR detected Jan 27 16:23:34 configure: error: GMP 4.1 and MPFR 2.2.1 or newer versions required by fortran Jan 27 16:24:54 GPM is 4.2 Jan 27 16:25:31 mpfr is 2.3.0 Jan 27 16:27:25 http://76.74.156.109/show_bug.cgi?id=3363 Jan 27 16:27:43 why would I want to use Angstrom anyway? Is there something magical about it? Jan 27 16:29:26 Crofton: thanks, I'll try the ugly hack Jan 27 16:30:14 Crofton: but, should I use angstrom? I would like to make a nice handheld and am afraid if I use another distro selection that all my work would be for nothing. Jan 27 16:30:48 seems that whenever I attempt to build a gpe / opie / X that I run into compiliation deaths Jan 27 16:31:57 03pfalcon 07org.oe.angstrom-2007.12-stable * rcd2d41ea... 10/ (1 classes/java-library.bbclass classes/java.bbclass): Jan 27 16:31:57 Add missing files from 9005ec7475f13738f08b14c023c457704df3f4af. Jan 27 16:31:57 * mtn sucks at cherry-picking. Jan 27 16:32:01 03pfalcon 07org.oe.angstrom-2007.12-stable * r527ebc53... 10/ (1 delete.txt): delete.txt: Remove dup file, following .dev. Jan 27 16:36:41 Angstrom is a good place to start Jan 27 16:37:09 eventually, we can add feeds to the angstrom website Jan 27 16:37:19 should I drop back to the angstrom-2007-legacy ? Jan 27 16:37:27 http://www.angstrom-distribution.org/ Jan 27 16:37:39 that might be easier atm Jan 27 16:37:56 I have applied the ugly hack Jan 27 16:38:05 but I do bleeding edge stuff Jan 27 16:38:07 Crofton: where did you put that hack? Jan 27 16:39:49 I tried it in org.openembedded.dev/packages/gcc/gcc-cross_4.2.2.bb Jan 27 16:40:54 I can't find where at them moent Jan 27 16:40:55 hello, there is gnash-0.8.0 in your repository(bb of 0.8.1 avaliable here:http://cvs.savannah.gnu.org/viewvc/*checkout*/gnash/gnash_0.8.1.bb?revision=1.1&root=gnash) but does it work inside a browser like konqueror-embedded ? Jan 27 16:40:55 grrr Jan 27 16:41:03 I think in gcc-cross Jan 27 16:41:39 I tried that, I suspect that it may have to be put into gcc_4.2.2.bb and rebuild the native package. Jan 27 16:42:01 Oh well, I will give angstrom-2007.1-legacy.conf a try Jan 27 16:42:33 yeah Jan 27 16:42:42 hopefully it does not have the fortran issue Jan 27 16:42:46 if that doesn't work, maybe sprinkle some --disable-fortran statements into the gcc bb file configure Jan 27 16:43:02 I have some work I need to do due the first week in Feb Jan 27 16:43:26 I have some work that was due last week. heh Jan 27 16:43:32 after that I need to 1) Hope they give me more work, and 2) Lower their expectations so I have some hacking time Jan 27 16:43:51 today I need to get a xlinix ml403 booting Jan 27 16:44:07 fortunately, the zipit stuff is related to a project coming up Jan 27 16:45:05 cool Jan 27 16:45:38 I need to do some janitor work on gumtsix, then I can look at the zipit stuff Jan 27 16:45:49 Crofton: maybe I am going about this all wrong? I should assume that *somebody* can build angstrom with OE, then Jan 27 16:46:16 choose one of the machines, like simpad and try to build that against angstrom Jan 27 16:46:42 if it builds, then maybe take a look at my machine definition to see what is missing / added? Jan 27 16:46:57 that is always a good idea Jan 27 16:47:22 I think part of the FORTRAN issue is it finds stuff installed in the build machine Jan 27 16:47:46 I see simpad opie on the angstrom website, but you know, I will probably get borked by something missing in my local.conf :( Jan 27 16:48:09 that is the biggest aggravation about OE, the conf files Jan 27 16:48:15 try legacy also, it is based of the stable angstrom Jan 27 16:48:20 ok Jan 27 16:48:22 cul Jan 27 16:48:44 I need to do some painting Jan 27 16:49:49 Crofton: hobby or honey-do ? Jan 27 16:50:10 my office :) Jan 27 16:50:13 lol Jan 27 16:50:30 I have a closet door and a window to finish Jan 27 16:50:47 basically, I am procastinating by being productive .... Jan 27 16:55:16 Legacy if for 2.4 kernels. You could copy conf/distro/angstrom-for-openmoko.inc to conf/distro/angstrom-2007.1.conf Jan 27 16:55:26 s/if/is/ Jan 27 17:16:39 ho all Jan 27 17:16:44 hi all Jan 27 17:17:03 well i fog a trouble ... i'm movng OE to a DM320 TI machine Jan 27 17:17:17 it's something similar to DaVinci ... Jan 27 17:17:48 i try to compile a static executable hello world program but i got illegal istruction ... Jan 27 17:18:08 OE use as gcc parameters '-march=armv5te -mtune=arm926ej-s' ... Jan 27 17:18:22 machine report to be "Processor : ARM926EJ-Sid(wb) rev 3 (v5l)" Jan 27 17:18:31 and "Features : swp half thumb fastmult edsp java " Jan 27 17:18:41 what's wrong ?!?!? Jan 27 17:19:29 weird Jan 27 17:19:39 I find toolchain from omap5912osk works for DV Jan 27 17:21:46 yes also omap (arm9) are similar ... Jan 27 17:23:00 gremlin[it]: check your floating point settings Jan 27 17:24:56 pb_ TARGET_FPU = "soft" Jan 27 17:26:27 that ought to be safe enough Jan 27 17:26:39 I guess you need to find out what the actual instruction is that fails. Jan 27 17:26:41 in fact ... Jan 27 17:26:59 arg ... Jan 27 17:29:02 if i boot with an image build with OE nothing happen except kernel ... with existent Linux in the machine to run static executable only ... there isn't debugging tool already present ;( Jan 27 17:31:02 gremlin[it]: use CONFIG_DEBUG_USER Jan 27 17:31:47 pb_ mhh how ?! Jan 27 17:32:53 gremlin[it]: turn that option on, rebuild your kernel, boot with the resulting kernel, inspect the output Jan 27 17:33:31 ok ! Jan 27 17:34:17 strict oe question in possible to set two packege type ... ipkg and tar in local.conf ? Jan 27 17:39:09 yes, I think so. Jan 27 17:50:24 pb_ booting with kernel (user_debug=31) and OE filesystem i got this message looping on console Jan 27 17:50:30 Code: e2632000 859fc01c 878a200c ef000000 (e3700a01) Jan 27 17:50:30 [1] init: obsolete system call 00000000. Jan 27 17:50:30 Code: e08fa00a e1a00001 e3a070f8 ef000000 (e1a03000) Jan 27 17:50:30 [1] init: obsolete system call 00000000. Jan 27 17:51:43 what kernel are you using? Jan 27 17:52:32 not a vanilla or arm ... came from the company who build the device ... Jan 27 17:53:25 hm. well, it apparently doesn't like the system call standard you're using. Jan 27 17:53:35 are you trying to run an eabi userspace on an old-abi kernel? Jan 27 17:54:33 pb_ ... can be ... the kernel is a 2.6.15 ... no eabi i suppose ... Jan 27 17:56:07 same fortran error in gcc-cross Jan 27 17:56:11 * XorA|gone runs his first OE compiled application on n810 Jan 27 18:02:17 XorA|gone: congratulations. Jan 27 18:02:36 XorA|gone: what application is it? Jan 27 18:02:46 epdfview Jan 27 18:03:19 didn't they include a pdfviewer ;) Jan 27 18:06:20 thesing: its not very good :-) Jan 27 18:07:08 XorA|gone, what is up with the fortran issue? Jan 27 18:07:20 Crofton: what fortran issue? Jan 27 18:07:22 I thought it had been fixed :( Jan 27 18:07:45 T0mW, was running into the mfpr/fortran problem with gcc-cross? Jan 27 18:07:56 no idea, Ive never run into that issue Jan 27 18:08:16 yeah Jan 27 18:08:20 I did Jan 27 18:08:31 but I think it went away Jan 27 18:08:38 I guess its one of these issues that the people who hit it have to fix it Jan 27 18:08:40 I'm trying to install gfortran on the system and see if that fixes it Jan 27 18:08:46 oh well back to paiting Jan 27 18:08:57 T0mW, I suspect that "fixes" it :( Jan 27 18:11:29 XorA|gone: http://76.74.156.109/show_bug.cgi?id=3363 Jan 27 18:12:31 T0mW: means nothing to me Jan 27 18:12:51 XorA|gone: just that I'm not the only one that has hit this one Jan 27 18:47:47 hello, i don't manage to find where 2008.1-test-20080127 's opie-image use the 2.3.12 version of qte : Jan 27 18:47:53 i've searched in opie-image Jan 27 18:48:18 i've searched if there was something included in angstrom-2008.1-test-20080127 Jan 27 18:48:23 (the include) Jan 27 18:48:39 oops s/2008.1-test-20080127/angstrom-2008.1-test-20080127/ Jan 27 18:59:24 03coredump2 07org.oe.dev * r2b629c70... 10/ (3 files in 2 dirs): altboot: Try unb0rking collies keyboard once again... Jan 27 19:13:32 what would be a better place to ask those newbie questions about compiling glibc ? Jan 27 19:14:42 keesj: If you don't ask them here, I'll be the only one asking newbie questions Jan 27 19:14:49 don't leave me alone ;) Jan 27 19:16:37 my opie build stops at gconf-dbus unpack :( ... Jan 27 19:16:57 wvdschel: heh Jan 27 19:18:37 does building-angstrom wiki page have instructions for latest OE repositary or for stable one? Jan 27 19:21:52 gints: no error message? Jan 27 19:21:58 no load? Jan 27 19:23:06 starox|gone: /oe/downloads/trunk_developer.imendio.com_.svn.gconf-dbus_606_.tar.gz: Cannot open: No such file or directory Jan 27 19:23:40 Crofton: Angstrom builds now (gcc-cross) if I installed the gmp + mpfr libs onto the development computer. thanks Jan 27 19:24:07 strange that the ones that OE builds won't work Jan 27 19:24:16 try downloading again.. bitbake -f -c fetch gconf-dbus Jan 27 19:27:37 keesj: http://www.pastebin.ca/875281 Jan 27 19:28:29 it's funy that fetch goes very fast Jan 27 19:30:06 perhaps it does nothing :p Jan 27 19:30:54 seems like Jan 27 19:30:59 and does the tar.gz file exist and is ok? Jan 27 19:31:06 nop :( Jan 27 19:32:48 and no disk full ? Jan 27 19:32:57 it was but now it is not Jan 27 19:33:00 Hello folks ... I've been fighting with OE to get to work with latin1 (well, ISO-8859-1) characters. Does someone know about any distribution in OE that uses it? Jan 27 19:33:07 zecke: hey Jan 27 19:33:12 zecke: hi man Jan 27 19:33:33 yo zecke Jan 27 19:33:40 hi flo_lap Jan 27 19:33:42 re Jan 27 19:34:00 hi pb_ Jan 27 19:34:16 flo_lap: hey, did you find a spare cycle for the e.V. Jan 27 19:34:43 otavio: when you want to upgrade packages, you have to add the checksum to checksums.ini as well Jan 27 19:34:52 erm Jan 27 19:34:53 ERROR: QA Issue: libmudflap.la failed sanity test (installed) in path /opt/openembedded/build/tmp/staging/arm-angstrom-linux/lib Jan 27 19:34:59 zecke: you mean for cmake? Jan 27 19:35:07 this happens every time I build an image Jan 27 19:35:11 zecke: ok, my fault. Want me to send a new patch with it? Jan 27 19:35:24 zecke: not yet, I guress I'll do that tomorrow at work Jan 27 19:35:26 otavio: it is already upgraded Jan 27 19:35:38 zecke: great. Thanks a lot Jan 27 19:35:57 otavio: I'm about to look into Finding Qt4 now Jan 27 19:36:07 zecke: awesome Jan 27 19:37:25 zecke: how do I find out from monotone the history of a file that has been deleted? Jan 27 19:37:35 "mtn log" doesn't seem to work on files that are no longer there. Jan 27 19:37:40 does anyone know what is causing this? Jan 27 19:38:04 wvdschel: causing what, exactly? Jan 27 19:38:13 oh, right, up there Jan 27 19:38:13 wvdschel: it is a non fatal error... why do you care? Jan 27 19:38:20 zecke, hello i'm trying to use qte-2.3.10 instead of 2.3.12 for angstrom-2008.1-test-20080127 but i don't find where it look for the version...i've looked inside images,angstrom and its include Jan 27 19:38:21 then what is it? Jan 27 19:38:38 pb_: http://www.openembedded.org/bonsai/do_query/?default=file and search for your file Jan 27 19:38:38 besides a non-fatal error Jan 27 19:38:55 wvdschel: check www.openembedded.org and what our QA errors mean? Jan 27 19:39:08 GNUtoo-laptop: spare with that old stuff Jan 27 19:39:12 GNUtoo-laptop: spare me eveb Jan 27 19:39:18 even... *gosh* Jan 27 19:40:00 zecke: great, thanks Jan 27 19:40:12 heh, seems it was deleted by koen, hrw, mickeyl and florian Jan 27 19:40:21 obviously some kind of undead vampire file that wouldn't stay down Jan 27 19:40:24 wvdschel: http://www.openembedded.org/more-quality-assurance-tests Jan 27 19:40:27 hehe! Jan 27 19:40:32 pb_: one deleted it, the other merged :) Jan 27 19:40:43 ah yes Jan 27 19:41:02 koen seems to be the original culprit Jan 27 19:41:43 "Xorg: drop stale recipes X11R7* in PV. Those aren't needed anymore after the Xorg cleanup from poky was merged" Jan 27 19:42:03 well, yes, except that apparently the new files don't include everything from the old ones. Jan 27 19:42:24 hehe Jan 27 19:43:16 zecke: do you think my monotone key would still work for checking in? Jan 27 19:43:47 GNUtoo-laptop: generally, there is something called PREFERRED_VERSION (see the bitbake manual) this either gets set to 2.3.12 (e.g. in a file in conf/distro) or 2.3.12 is the highest version and wins anyway Jan 27 19:43:51 pb_: it should Jan 27 19:43:56 great Jan 27 19:44:06 * pb_ tries to remember what machine it is on Jan 27 19:44:12 I guess on my old laptop Jan 27 19:44:28 hehe, I hope you do remember your password as well Jan 27 19:44:43 otherwise we are going to have a pb2... Jan 27 19:45:32 hm Jan 27 19:45:34 I think I do Jan 27 19:45:39 zecke, i know that but i don't know in which file i should change it Jan 27 19:46:03 GNUtoo-laptop: quite easy. does it get set? change it there? Jan 27 19:46:26 can you remind me how to get the key out of my old database and import it into the new one? Jan 27 19:46:35 GNUtoo-laptop: is it not getting set? and you read the angstrom documentation regarding Opie... Jan 27 19:46:37 zecke, ahh...i understand if it's not set it takes the higher version Jan 27 19:46:49 pb_: is it in your db? or is it in .monotone/ Jan 27 19:47:09 pb_: a mtn --db foo.db db migrate should extract the keys to the .monotone dir Jan 27 19:47:55 okay Jan 27 19:47:58 wvdschel: read the link? Jan 27 19:49:13 ah right, I seem to have something in .monotone/keys already Jan 27 19:49:20 not yet, I don't have the time right now Jan 27 19:49:23 I want the @openembedded.org one, right, not pb@handhelds.org? Jan 27 19:49:52 wvdschel: why did you ask then? ;) Jan 27 19:50:08 how do I make bitbake generate an ipkg feed? Jan 27 19:50:25 zecke, because a) I didn't know it would be such a long answer Jan 27 19:50:43 b) and I thought it would be semi-serious Jan 27 19:50:56 wvdschel: the answer is one paragraph :) Jan 27 19:51:06 the first one? Jan 27 19:51:07 wvdschel: bitbake package-index (but technically bitbake does not generate your feed) Jan 27 19:51:44 pb_: @openembedded.org is generally preferred but do what you want Jan 27 19:52:08 zecke, yeah, I know, but I didn't know how to put it Jan 27 19:52:20 wvdschel: :) Jan 27 19:52:34 zecke: your @o.o isn't working, at least it failed last time I tried Jan 27 19:52:50 I should have said "how do I generate the package feed" is suppose, assuming we still do the things we issue the commands for :P Jan 27 19:52:53 otavio: hehe, it is not a mail address :) Jan 27 19:52:58 gah.. where OE keeps logs? want check why gconf-dbus fetch does nothing Jan 27 19:53:13 gints: how are you? where have you been? Jan 27 19:53:34 zecke: hi.. I am fine.. I was used to programming for symbian Jan 27 19:53:36 zecke: OMG! it is used on the commit logs. That is confusing! Jan 27 19:53:38 gints: I can give you mines if you want Jan 27 19:53:52 otavio: on openembedded.org/people it says "NOTE: The monotone key addresses may or may not be valid email addresses. Please do not attempt to use them as such." Jan 27 19:54:04 ahhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh Jan 27 19:54:20 hehe :) Jan 27 19:54:22 keesj: your gzip would be fine :) Jan 27 19:54:32 add -vvv -d to your command , and the logs are located in a diretory called temp Jan 27 19:55:29 gints: http://box.mmapps.net/oe_gconf_dbus.zip (contains a few more files that stated with the same root) Jan 27 19:56:29 gints, symbian and open source? Jan 27 19:56:46 wvdschel: no :( Jan 27 19:56:55 I thought so Jan 27 19:57:09 symbian is pretty hard to program for as an open source developer Jan 27 19:57:12 wvdschel: there is just some opensource projects Jan 27 19:57:18 at least I found it to be Jan 27 19:57:31 wvdschel: symbian is beatch Jan 27 19:57:35 yeah, but very few, and very few that work on UIQ Jan 27 19:57:50 and even fewer that work on the newer UIQs Jan 27 19:58:09 zecke: well ... but it's _confusing_ Jan 27 19:58:11 I have to get back to studying Jan 27 19:58:21 see you all later Jan 27 19:58:32 otavio: I agree and you are probably not the first to mail there Jan 27 19:58:46 otavio: specially when reading the git changelog, where these are mail addresses :} Jan 27 20:00:51 zecke: I just found that opie coming back to Angstrom... so I come to test :) Jan 27 20:02:24 probably I found no better enviroment yet for embedded devices like opie yet Jan 27 20:13:19 gr, this fonts thing is a nuisance Jan 27 20:13:32 it seems that it wasn't just one file that was erroneously deleted, there are a whole load of them Jan 27 20:14:33 * Crofton is in awe of people that understand things like fonts Jan 27 20:14:48 and is also glad he uses machines without displays most of the tim Jan 27 20:16:08 well, I seem to have managed to fish enough bits out of the archive that at least bitbake will do something other than report dependency problems now :-} Jan 27 20:18:14 Crofton: it even gets worse if you have to throw internationalisation in... Jan 27 20:18:57 keesj: many thx.. I can continue now to opie-image :) Jan 27 20:19:09 cool Jan 27 20:20:29 mhh have sense if OE create in tmp/work the dir "arm-oabi-angstrom-linux-gnueabi" ... it's oabi or eabi ?!?!? Jan 27 20:20:54 eabi, probably Jan 27 20:21:33 ufff ... Jan 27 20:32:28 * pb_ waits patiently for the kernel to download Jan 27 20:32:35 well, maybe not quite so patientl Jan 27 20:32:38 y Jan 27 21:41:29 pb_: hehe, at least you are not waiting on monotone :} Jan 27 21:43:17 someone remind me what causes EOL errors when runnign bitbake on a new install? Jan 27 21:43:42 Crofton: bitbake tries to expand a variable but it was not defined Jan 27 21:45:28 KERNEL_VERSION maybe Jan 27 21:45:30 hmmm Jan 27 21:47:21 ah TMPDIR = 'foo" Jan 27 21:50:04 hi lrg Jan 27 21:50:24 hey flo_lap Jan 27 21:55:09 'morning' all Jan 27 21:55:23 RP: morning :) Jan 27 22:06:19 zecke: heh, right Jan 27 22:06:25 pb_: :) Jan 27 22:06:28 * pb_ waiting on glibc compile now Jan 27 22:06:35 pb_: I operate a git mirror as well Jan 27 22:06:36 hi lrg Jan 27 22:07:45 ah yes, I think I remember that from before Jan 27 22:07:59 well, for now I seem to be doing ok with monotone Jan 27 22:09:15 oh, drat, I forgot to set PARALLEL_MAKE Jan 27 22:09:20 no wonder it's going so slowly Jan 27 22:09:48 I like to leave || make off so I can surf while building :) Jan 27 22:10:48 heh Jan 27 22:11:26 ah, do_install now Jan 27 22:11:43 anyone seen a schedule for the embedded rrom at fosdem yet? Jan 27 22:11:47 brief visit from lrg there Jan 27 22:13:17 Crofton: coming soon. Working on it ;-) Jan 27 22:13:45 ah Jan 27 22:14:04 I was looking on the embedded-kernel-track.org site Jan 27 22:14:17 Hopefully it is as good as last year :) Jan 27 22:14:26 pb_: Which files got lost with the xorg update? Jan 27 22:15:01 Crofton: embedded-kernel-track.org site is deprecated in favour of the fosdem website. Jan 27 22:15:20 ok Jan 27 22:15:22 Crofton: and looking at the submissions it will be good this year too. Jan 27 22:15:28 you should make a note there :) Jan 27 22:16:06 RP: font-misc-misc and font-cursor-misc, plus a pile of things that they seem to depend on to build. Jan 27 22:16:50 font-alias, font-util and something else that I forget Jan 27 22:19:40 pb_: :(. Sounds like they got lost during the conversion but that shouldn't have happened... Jan 27 22:23:49 ah yes, xorg-font-common.inc has font-alias and font-util-native in DEPENDS, neither of which exist anymore in the current packages tree. Jan 27 22:24:12 and font-util-native depends on bdftopcf-native, which also seems to have disappeared Jan 27 22:24:23 Do they have modern xorg equivalents? Jan 27 22:25:32 I'm not quite sure what you mean. Jan 27 22:25:57 All those packages are still in the Xorg upstream tree, as far as I know. They're certainly still in X11R7.3. Jan 27 22:26:09 Yes, thats what I mean... Jan 27 22:26:34 I mean they're not just in the old style xorg cvs/release system Jan 27 22:26:40 oh, right Jan 27 22:27:02 no, the old (now deleted) .bb files were using the X11R7.0 tarballs Jan 27 22:27:20 or X11R7.1 in some cases, I think Jan 27 22:27:21 I sounds like they were removed by mistake... Jan 27 22:29:07 yeah, it does sound that way Jan 27 22:29:19 assuming that they still compile, I guess I'll put them back. Jan 27 22:29:36 hello bitbake opie-image fails at opie-taskmar-images http://pastebin.com/m4b5314ef Jan 27 22:30:10 Strangely my commit mailbox doesn't have the deletion mail in it... Jan 27 22:30:37 I was assuming these files were pre X11R7.0 but that isn't the case... Jan 27 22:31:12 http://www.openembedded.org/bonsai/view/rev/31683/ Jan 27 22:31:39 at a first glance, none of the files deleted there seem to be pre-X11R7.0 Jan 27 22:32:38 no. The worrying thing about that commit is that poky doesn't use xorg-font/*... Jan 27 22:32:51 ah Jan 27 22:32:58 !seen psokolovsky* Jan 27 22:32:58 psokolovsky (n=psokolov@nat-users.is.ua) was last seen quitting from #oe 15 days, 6 hours, 12 minutes ago stating (Read error: 110 (Connection timed out)). Jan 27 22:33:08 I guess that would explain why nobody has noticed before. Jan 27 22:33:09 !seen pfalcon* Jan 27 22:33:10 No matches were found. Jan 27 22:33:17 :( Jan 27 22:33:31 pb_: There are users of OE besides poky though... Jan 27 22:34:38 mm. I guess none of them use xorg-font either. :-} Jan 27 22:34:54 evo's search function is useless, I do have this commit mail... Jan 27 22:34:58 heh Jan 27 22:35:52 good night Jan 27 22:36:00 flo_lap: good night Jan 27 22:36:07 * pb_ bedtime too Jan 27 22:36:08 night all Jan 27 22:36:11 flo_lap: 'night Jan 27 22:37:02 pb_: I'll go through that commit and see what else might have got lost... Jan 27 22:40:55 RP: thanks Jan 27 22:41:58 ah, closet doors in the office again Jan 27 22:43:54 'night pb_ Jan 27 22:44:35 hmm, some guy claiming cheap omap stuff in coming weeks in a blog ... Jan 27 22:49:20 pb_: Ah, I understand what koen did, it was a case of removing everything with the XORG7.* string in PV Jan 27 22:50:11 sometimes knowing how to lots of work with simple scripts is not a good thing :) Jan 27 22:50:58 RP: I just compiled 2.6.24 (copied from -dev to -stable), created some patches, and recompiled opie-image for poodle. WPA-WIFI works now out of the box on poodle :-) Jan 27 22:51:07 RP: Ok, if I send you the patches ? Jan 27 22:51:19 helgeD: great :) Jan 27 22:51:32 helgeD: I don't deal with angstrom stable, they need to go to someone else Jan 27 22:52:06 RP: Ok, but the defconfig-poodle patch and the other (small) linux-patches should go to -dev, right? Jan 27 22:52:21 helgeD: If patches are needed, yes. Jan 27 22:52:35 I think there is a plan to copy kernel updates from .dev in the next few weeks Jan 27 22:52:47 RP: yes, are needed. Doesn't build without (image too big and compile errors). Jan 27 22:54:16 helgeD: ok, patches welcome then :) Jan 27 22:54:17 hi to all Jan 27 22:54:32 i need to recompile a kernel for Familiar Distribution Jan 27 22:54:38 i have recompile it Jan 27 22:54:44 and now? what i have to do? Jan 27 22:55:01 if i use ipkg i have more problem, i want to do this manually. everyone can help me? Jan 27 22:57:15 RP, can help me? Jan 27 22:58:04 leno29: This channel is more about automating that process and using openembedded to do the work for you... Jan 27 22:58:45 and you can explaine me in what mode i can automating it? Jan 27 23:02:17 leno29: Familiar no longer uses openemdedded so I'm not sure this is the right place to ask for help with familiar Jan 27 23:15:29 RP: I just sent off my mail with patch for poodle to you. Jan 27 23:18:04 helgeD: ok, thanks. I'll have a look shortly Jan 27 23:27:40 GNUtoo-laptop: I am seeing the same problem and it prevents me from building an opie-image Jan 27 23:28:04 Laibsch, wow we are 2 now...and i also want to build opie-image Jan 27 23:29:45 Well, I think there are more than just the two of us Jan 27 23:29:52 http://bugs.openembedded.org/show_bug.cgi?id=2721 might be interesting, too Jan 27 23:29:59 Laibsch, thanks Jan 27 23:31:08 wvdschel has seen the problem, too, I think Jan 27 23:34:01 xjqian|away: Did you successfully compile opie-taskbar-images? Jan 27 23:34:11 is http://bugs.openembedded.org down? Jan 27 23:34:18 use .net Jan 27 23:34:32 thanks Jan 28 00:00:31 We lots 149 recipes in that commit... Jan 28 00:00:35 lost Jan 28 00:06:22 pb_: I've restored them along with an attempt at fixing the file version numbers. If it doesn't work from here, patches welcome! Jan 28 00:13:42 03pfalcon 07org.oe.dev * r33c44c7c... 10/ (6 files in 2 dirs): (log message trimmed) Jan 28 00:13:42 bluez-utils 3.23 and friends: Set FILESPATH explicitly and correctly. Jan 28 00:13:42 * Having a "files" dir in a multi-package directory is shooting everyone Jan 28 00:13:42 in leg. There should be per-package dirs or else stuff will be taken from Jan 28 00:13:42 god-knows-where. Jan 28 00:13:43 * So, add FILESPATH for all bluez-utils descendant packages. Jan 28 00:13:45 * This fixes build of bluez-cups-backend, which failed in do_patch Jan 28 00:13:49 03pfalcon 07org.oe.dev * ra99dc9b3... 10/ (1 classes/patch.bbclass): Jan 28 00:13:51 patch.bbclass: Show full path of a patch. Jan 28 00:13:53 * This is for consistency with existing non-patch fetch message in base.bbclass. Jan 28 00:13:55 Allows to catch pulling file from wrong override dir quickly. Jan 28 00:13:57 03rpurdie 07org.oe.dev * rf905ec58... 10/ (155 files in 6 dirs): xorg: Add packages accidently dropped in 778aeb6e918d6ec53d4abbfa67d0d3832c8d717e, updated to match current version scheme (thanks to pb_ for spotting) Jan 28 00:21:49 GNUtoo-laptop: I fixed the problem Jan 28 00:21:54 I'll push shortly Jan 28 00:49:12 03Laibsch 07org.oe.dev * re3c06e0e... 10/ (1 packages/opie-taskbar/opie-taskbar-images.inc): opie-taskbar-images: There is nothing to compile and thus nothing to configure. Fixes breakage in do_configure. Jan 28 00:50:46 ~seen pgfeller Jan 28 00:50:49 pgfeller was last seen on IRC in channel #oe, 160d 4h 45m 2s ago, saying: 'polyonymous: thanks again & cu'. Jan 28 01:08:47 Laibsch: opie-taskbar-images_cvs works fine for me. After all, it's just moving those .png files around. Jan 28 01:09:59 xjqian: Which is why it failed after adding "inherit opie" Jan 28 01:10:16 It seems like it thought it needed to do do_configure Jan 28 01:10:20 Fixed Jan 28 01:11:35 Laibsch: great, thanks Jan 28 02:04:18 03Laibsch 07org.oe.dev * r3f46620a... 10/ (4 files in 4 dirs): Jan 28 02:04:18 libvncserver: update to version 0.9 Jan 28 02:04:18 * latest version 0.9.1 fails, needs some looking into Jan 28 02:04:18 * consolidate bb file into packages/vnc directory Jan 28 02:04:18 * reorder fields according to StyleGuide Jan 28 02:18:04 * * OE Bug 2411 has been RESOLVED (FIXED) by Jan 28 02:18:06 * * .bb for libvncserver 0.9.1 Jan 28 02:18:08 * * http://bugs.openembedded.org/show_bug.cgi?id=2411 Jan 28 02:18:16 * * OE Bug 2125 has been RESOLVED (FIXED) by Jan 28 02:18:18 * * libvncserver_0.9.bb Jan 28 02:18:20 * * http://bugs.openembedded.org/show_bug.cgi?id=2125 Jan 28 02:21:04 * * OE Bug 1312 has been RESOLVED (FIXED) by Jan 28 02:21:06 * * Build of libvncserver failed Jan 28 02:21:08 * * http://bugs.openembedded.org/show_bug.cgi?id=1312 **** ENDING LOGGING AT Mon Jan 28 02:59:57 2008