**** BEGIN LOGGING AT Tue Sep 18 02:59:59 2007 Sep 18 04:44:09 Hmm.. Navit? :) Sep 18 05:58:07 alphaone: It's definitely cool :D Sep 18 06:31:20 0,4 http://st-pitch.miniville.fr/  Sep 18 06:31:23 URL http://st-pitch.miniville.fr/ Sep 18 06:38:50 Bloody spam bots. Sep 18 07:57:44 03koen 07org.oe.dev * r97c62854... 10/ (1 packages/openmoko2/openmoko-browser2_svn.bb): openmoko-browser2: add svn version Sep 18 07:57:48 03koen 07org.oe.dev * rdba16850... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs: add openmoko-browser2 Sep 18 08:00:07 morning Sep 18 08:07:21 good morning all Sep 18 08:08:51 yo koen Sep 18 08:22:45 http://labs.vivi.eng.br/blog/?p=31 Sep 18 08:23:13 http://trolltech.com/company/newsroom/announcements/press.2007-09-17.9260755578 Sep 18 08:24:26 heh Sep 18 08:24:43 camono has the browser title: "Trolltech Announces Qtopia Phone Edition Integration with Neo 1973 Phone -- Troll..." Sep 18 08:26:11 ljp: congrats on getting the phone and sxe stuff gpl'ed Sep 18 08:27:00 thanks Sep 18 08:28:19 now we just need a rotary dialer Sep 18 08:36:28 ljp: nice :) Sep 18 09:23:41 morning all Sep 18 09:23:56 yo RP Sep 18 09:27:40 hey RP Sep 18 09:29:34 lightbulb moment: with pH5's changes to stamp logic packaged-staging would become a lot easier Sep 18 09:33:22 koen: In theory yes. If they didn't break all the existing functionality :) Sep 18 09:34:42 did someone experience issue with microphone on intel ac97 (eg: PXA270 boards) and alsa 1.0.14 ? Sep 18 09:35:07 I don't use the 'remove stamps to rebuild' functionality, so that can go :) Sep 18 10:26:38 Random thoughs for the day. Why does linux-libc-headers install into both cross and staging. Why does gcc-cross-initial run an install task at all? Sep 18 10:29:34 cbrake_away: did you manage to get the microphone working on CM-X270 ? Sep 18 10:36:45 sakoman, pong Sep 18 10:57:29 koen, any idea whose autobuild created the gumstix-verdex and omap5912osk errors? Sep 18 10:58:21 mine Sep 18 10:58:38 they look bogus, and I just started a new autobuild from scratch Sep 18 10:58:42 good, mine run created kernels last night :) Sep 18 10:58:44 yeah Sep 18 10:58:51 koen: do you want to host me tonight? Sep 18 10:58:53 I ran one with the smae machines last night Sep 18 10:59:08 has zecke gone walkabout? Sep 18 11:00:08 Crofton|home: zecke is the new OE basement troll Sep 18 11:00:36 does he eat much? Sep 18 11:02:13 XorA: haha Sep 18 11:02:21 only small children Sep 18 11:02:31 that don't go to bed on time Sep 18 11:03:01 three years of ranting, complaining and talking and Qtopia is fully GPL :} Sep 18 11:03:21 until the next version Sep 18 11:03:30 ade|desk: I don't think so Sep 18 11:03:52 zecke: hmmm, that is going to be difficult with work, let me make a few calls Sep 18 11:04:25 koen: don't worry, I can go to Amsterdam as well, and we meet another day Sep 18 11:04:26 zecke, what will you do for an encore? Sep 18 11:07:37 Crofton|home: encore for what? Sep 18 11:08:06 getting Qtopia open sourced Sep 18 11:08:25 Crofton|home: doing the Qt4 dance Sep 18 11:35:32 I really do create a 4M kernel for the dv evm Sep 18 11:37:12 zecke, ljp: Is there a policy in place for future qtopia licences? Sep 18 11:38:43 03xora 07org.oe.dev * rc60ea02b... 10/ (1 packages/gnome/epiphany_2.19.6.bb): gnome/epiphany_2.19.6.bb : add iso-codes to DEPENDS Sep 18 11:38:44 RP: what kind of policy? (and I don't work there yet) Sep 18 11:39:33 zecke: I mean would a theoretical qtopia5 be gpl or would this happen all over again? Sep 18 11:40:12 03xora 07org.oe.dev * rc60ea02b... 10/ (1 packages/gnome/epiphany_2.19.6.bb): gnome/epiphany_2.19.6.bb : add iso-codes to DEPENDS Sep 18 11:41:00 Crofton|home: http://svn.o-hand.com/view/poky/trunk/meta/packages/linux/linux-rp-2.6.22%2B2.6.23-rc4/binutils-buildid-arm.patch?rev=2726&view=markup Sep 18 11:41:25 RP: there is no promise (yet) that future versions will be GPL Sep 18 11:42:21 RP: e.g. for Qt/X11, Qt/Mac there is the promise when TT stops releasing GPL versions, the latest Qt version will be BSD licensed (it is a contract between KDE e.V. and Trolltech) Sep 18 11:43:32 zecke: Right. I guess that would be the really interesting announcement... :} Sep 18 11:44:19 koen, that looks scary Sep 18 11:44:45 03xora 07org.oe.dev * rc60ea02b... 10/ (1 packages/gnome/epiphany_2.19.6.bb): gnome/epiphany_2.19.6.bb : add iso-codes to DEPENDS Sep 18 11:44:58 Crofton|home, koen: Its a safe fix discussed on the arm list Sep 18 11:45:46 only for binutils 2.18? Sep 18 11:45:59 I imagine it is safe for lower revs ... Sep 18 11:46:14 Crofton|home: That version introduced some new fields in the binaries Sep 18 11:46:18 RP: right, but I assume that to happen sooner or later Sep 18 11:48:31 I assume we need to add this to other versions? Sep 18 11:49:53 Crofton|home: Other versions? Sep 18 11:50:06 03xora 07org.oe.dev * r4dc53c15... 10/ (5 files in 3 dirs): libglade/libglade_2.6.2.bb : version upgrade needed with newer gtk+/glib combo Sep 18 11:50:11 2.6.21? Sep 18 11:50:36 Crofton|home: Yes, I suspect so although I'm not 100% sure Sep 18 11:51:04 gumstix works of .21, but I do not think I have tried a kernel from a 64 bit machine Sep 18 11:51:08 wait, maybe I ahve Sep 18 11:51:21 work machine is 64 Sep 18 11:52:01 03xora 07org.oe.dev * rc60ea02b... 10/ (1 packages/gnome/epiphany_2.19.6.bb): gnome/epiphany_2.19.6.bb : add iso-codes to DEPENDS Sep 18 11:52:05 03xora 07org.oe.dev * r4dc53c15... 10/ (5 files in 3 dirs): libglade/libglade_2.6.2.bb : version upgrade needed with newer gtk+/glib combo Sep 18 11:52:26 CIA-3: you lurve me :-) Sep 18 12:07:12 hmmm looks like pidgin now has issues with '& # x 3' and the like all over the place ..grr. honestly package QA has gone out the window in debian recently Sep 18 12:12:05 heya Sep 18 12:23:12 ~change 22400 jpy to gbp Sep 18 12:23:18 i got some error trying that Sep 18 12:24:35 ibot you suck, its 97 quid Sep 18 12:41:37 XorA, "org.openembedded.dev/packages/libglade/libglade-2.6.2/./no-deprecation.patch not found." Sep 18 12:42:38 koen: so you have to work? Sep 18 12:44:39 sadly, yes :( Sep 18 12:46:36 don't worry :) Sep 18 12:47:20 tuukkah: odd, I must have hit U in vim just before commiting Sep 18 12:48:48 tuukkah: fixed Sep 18 12:58:13 XorA, thanks! Sep 18 13:01:39 morning Sep 18 13:03:36 03xora 07org.oe.dev * ra76bca79... 10/ (1 packages/libetpan/libetpan_0.52.bb): libetpan/libetpan_0.52.bb : new version Sep 18 13:03:41 03xora 07org.oe.dev * rfc268f95... 10/ (1 packages/sylpheed/claws-mail_3.0.1.bb): sylpheed/claws-mail_3.0.1.bb : new version Sep 18 13:03:46 03xora 07org.oe.dev * radfe5eb6... 10/ (1 packages/libglade/libglade_2.6.2.bb): libglade/libglade_2.6.2.bb : remove no-deprecation.patch Sep 18 13:06:23 Hi there, is it possible to use the bb.data.setVar inside the base.bbclass? something like that: ${@bb.data.setVar("DEBRULES", "${EXTRA_OEMAKE} $@", d)} ????? Sep 18 13:06:43 I was trying that in oe_runmake() Sep 18 13:21:08 hi, all! Sep 18 13:21:23 hi Sep 18 13:21:46 do anybody remembers where in glibc-2.4 there's test for kernel headers version? Sep 18 13:44:30 nud: yes, the mic does work, but I had to modify the kernel to eable the 20db MIC gain in the UCB1400 Sep 18 13:44:59 cbrake: ah ok Sep 18 13:45:19 the mic seems to work with alsa 1.0.13 but not with 1.0.14 Sep 18 13:45:36 nud: however, I strong advise you put an external MIC amp on your custom board with a quiet analog section with a filtered power supply. Sep 18 13:46:26 yeah the quality is bad Sep 18 13:46:48 nud: I'm still running 1.0.13 Sep 18 13:46:51 but there will be analog stuff and echo cancellation Sep 18 13:47:56 nud: running a mic level signal over the cm-x270 connector does not work very well -- line level is much better, but maybe we are just lousy analog engineers ... Sep 18 13:48:14 nud: let me find my ucb1400 patch ... Sep 18 13:49:39 nud: this may or may not be useful: http://www.pastebin.ca/701890 Sep 18 13:52:17 03xora 07org.oe.dev * r3ff683ed... 10/ (5 files in 2 dirs): links/links*_2.0+2.1pre30.bb : new version Sep 18 13:53:20 what is that patch for ? Sep 18 13:53:57 nud: alows your to set the +20db MIC gain switch in alsamixer Sep 18 13:54:15 nud: but, if you have external amplification, it is probably not needed Sep 18 13:54:33 cbrake: why don't you put it directly in OE ? ;-) Sep 18 13:54:46 eck, pastebin.ca is slow as hell these days Sep 18 13:55:02 nud: its on my list ... -- just like your mkfs.jffs2 changes ... Sep 18 13:55:39 nud: I like to test things before pushing changes, but my current HEAD builds are failing due to a ATK build problem Sep 18 13:56:03 nud: probably another AMD64 problem :-\ -- not sure why I bother with AMD64, but I guess it makes OE better Sep 18 13:56:07 wow, I stopped way earlier in the stack ;-) Sep 18 13:56:38 nud: well, as of a few days ago, everything got built with a console-image build Sep 18 13:56:56 nud: there was a fix discussed, but not sure if it has been implemented yet. Sep 18 13:57:11 *everything = gtk and friends Sep 18 13:58:21 cbrake: yeah I remember seeing my machine compiling X and other stuff Sep 18 13:58:25 * cbrake kicks off another mtn HEAD build ... Sep 18 13:58:39 nud: you are probably running i686 on HOST? Sep 18 13:58:53 nud: (32 bit) Sep 18 13:59:06 yes Sep 18 14:08:02 hi Sep 18 14:34:33 I am doing a test to try out mdev with busybox. In doing so, I want to build busybox with an all yes config. so, if I change that in the build directory and want to force a build, I do a bitbake -f busybox but it doesn't go and recompile it. is there a command that will rebuild everything without doing a clean and redownloading a new config? Sep 18 14:35:42 christopher: bitbake busybox -c compile -f Sep 18 14:35:53 christopher: then bitbake busybox Sep 18 14:36:27 RP: thanks. so the -c doesn't do a clean where it will redownload ? Sep 18 14:36:44 christopher: bitbake --help will tell you what -c does Sep 18 14:37:00 koen: btw, the -f option works internally by removing the stamp so you do rely on the deleted stamp behaviour ;-) Sep 18 14:38:10 RP: yeah, I looked before asking but it wasn't clear even in the -c of what commands were available. I'll give that a shot Sep 18 14:38:38 christopher: bitbake busybox -c listtasks will tell you whats available Sep 18 14:39:21 RP: cool, thanks Sep 18 14:55:33 http://osnews.com/story.php/18631/Trolltech-Announces-GPL-Version-of-Qtopia-Phone-Edition/ Sep 18 14:56:19 Crofton: your late the the party, everyone over ---> is already running it on Neo :-) Sep 18 14:56:53 * Crofton is always late Sep 18 14:57:06 the interview comments are interesting in this article Sep 18 14:57:12 Crofton: its not your fault, USA is just the wrong side of the world Sep 18 14:58:06 what we need to do is drive all the phone groups to OE so they share work :) Sep 18 14:58:41 Crofton: ljp is already here for Qtopia :-) Sep 18 14:58:56 yeah Sep 18 15:00:10 I guess I may have to pick up a QT book one day Sep 18 15:00:21 * Crofton sucks at UI though Sep 18 15:02:19 * steliosk looks where the phone groups will park if we drive them here Sep 18 15:02:39 obviously we need a parking garage Sep 18 15:02:52 Crofton : hehe Sep 18 15:03:34 Crofton: can you also hire those US guys who taser you for asking awkward questions? Sep 18 15:03:37 Crofton : I think we need a space at Second life Sep 18 15:04:07 * Crofton has never tried second lif Sep 18 15:04:23 I don't like second life ... my first one already sucks I don't need another one Sep 18 15:04:29 haha Sep 18 15:04:38 Crofton: you miss little, its like the old 1980's virtual reality, unless you wait about an hour for textures to arrive Sep 18 15:04:56 * steliosk in second life will be a simple user Sep 18 15:05:02 no more code Sep 18 15:06:05 when I have time for a second life, it will be here : http://www.hitechcreations.com/frindex.html Sep 18 15:06:09 shooting users! Sep 18 15:06:21 heheh Sep 18 15:08:26 XorA, I need to check out the youtube videos Sep 18 15:14:53 Crofton: thanks for posting the pointer to the interview! Been a couple of years since I talked with Benoit. Got to get back in touch. Sep 18 15:15:23 Small world Sep 18 15:17:12 Yes it is. Benoit is really a great guy. Takes some amazing amazing photos with his telescope as a hobby Sep 18 15:18:45 * Crofton wonders what hobbies are Sep 18 15:18:53 * Crofton remembers caving Sep 18 15:26:45 Crofton: this is a hobby - http://www.astrosurf.com/benoit/ Sep 18 15:28:21 thanks Sep 18 15:28:24 click on the details of the setup link -- amazing telescope! Sep 18 15:31:14 my father worked on receivers for radio telescopes Sep 18 15:33:38 Crofton: I love that stuff -- don't need any more obsessions though :-) Sep 18 15:40:38 Crofton: got the collections stuff working now. is there a similar technique to have a custom machine conf outside org.openembedded.dev? Sep 18 15:41:10 works the same Sep 18 15:41:28 BBPATH Sep 18 15:41:56 ah, OK I'll try that. Thanks Sep 18 16:00:50 03koen 07org.oe.dev * r3bbc4b2b... 10/ (1 packages/openmoko2/openmoko-browser2_svn.bb): openmoko-browser2: depend on intltool Sep 18 16:00:55 03koen 07org.oe.dev * rfbfe95bb... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs: bump webkit and openmoko-browser2 Sep 18 16:08:46 anyone tried using a dedicated TOOLCHAIN_HOST_TASK? I have added packages to it, but they can't get installed during the sdk_populate task. I suspect ipkg arch's being to blame? Sep 18 16:25:51 mickey|ICE77: host tasks must have the sdk arch Sep 18 16:26:13 RP: ok, what do i need to make that happen? Sep 18 16:26:20 hand-setting the ARCH somewhere? Sep 18 16:26:25 mickey|ICE77: It basically only applies to cross tools. What have you got in mind? Sep 18 16:26:37 well, my idea was to have Sep 18 16:26:47 TOOLCHAIN_HOST_TASK = task-openmoko-sdk Sep 18 16:26:58 task-openoko-sdk RDEPENDing on libmokoui2-dev etc. Sep 18 16:27:04 is that the wrong route? Sep 18 16:27:17 No, but they're not host tools Sep 18 16:27:26 aaah Sep 18 16:27:34 so overriding the TOOLCHAIN_TARGET_TASK? Sep 18 16:27:45 mickey|ICE77: yes Sep 18 16:28:08 host tools are things like gcc-cross, gdb-cross, binutils-cross etc Sep 18 16:28:10 ok, thanks Sep 18 16:28:19 of course. didn't think that through Sep 18 16:29:19 mickey|ICE77: My idea was that distros could override TOOLCHAIN_TARGET_TASK like you describe Sep 18 16:29:39 mickey|ICE77: I let the possibility of overrideing HOST_TASK open too but I doubt anyone will need to change that from the default Sep 18 16:29:46 agreed. Sep 18 16:29:49 s/let/left/ Sep 18 16:33:57 mickey|ICE77: http://svnweb.openmoko.org/?rev=3008&view=rev has some spurious stuff in it Sep 18 16:34:32 mickey|ICE77: svn-commit* sato* Sep 18 16:35:10 right. I'll tell thos Sep 18 16:36:45 mickey|ICE77: http://scap.linuxtogo.org/files/c856ea92ec00812e350c22a04e86b73e.png <- with pH5 22x22 icons Sep 18 16:37:48 pretty. nearly good Sep 18 16:37:59 now the main problem is the MBScaledPixmap bug Sep 18 16:38:01 yeah, the application icons are still 32x32 Sep 18 16:39:46 * koen wonders when the traintickets to berlin will arrive Sep 18 16:45:02 * RP has his OEDEM flight booked now Sep 18 16:45:24 from newcastle? Sep 18 16:46:06 sakoman, great Sep 18 16:46:20 long lunch, we had to cruise the career fair for swag Sep 18 16:46:33 heh Sep 18 16:46:42 we also have a career fair here Sep 18 16:46:47 koen: Newcastle -> Amsterdam -> Berlin, nothing direct anymore :-( Sep 18 16:48:00 the good news is that at amsterdam you won't leave the secured zone, so no extra cehcks Sep 18 16:48:10 chicks? Sep 18 16:48:12 oh, sorry Sep 18 16:48:13 misread Sep 18 16:48:15 *cough* Sep 18 16:48:37 mickey|ICE77: :) Sep 18 16:49:24 bah, my todolist is out of quick-and-easy tasks Sep 18 16:49:26 koen: Last time I went through Amsterdam everything got scanned again (I think that was to Finland) Sep 18 16:50:04 koen: Down to implement packaged staging? :) Sep 18 16:50:24 I actually am formulating a strategy for p-s Sep 18 16:50:48 it involves merging the sysroot and gcc wrapper stuff and the stamp rework Sep 18 16:51:02 please not the gcc wrapper Sep 18 16:51:07 in a branch :/ Sep 18 16:51:19 including a modified bitbake? Sep 18 16:51:44 koen, finish school :) Sep 18 16:51:50 my idea is to have a patch to bitbake included you have to apply by hand Sep 18 16:52:04 absorbing 3rd party projects is a bad idea in general Sep 18 16:52:43 koen: I'd like to see a proposal for how this magic stamps patch for bitbake is going to work Sep 18 17:52:09 03koen 07org.oe.dev * r5684dae5... 10/ (6 files in 2 dirs): bluez: add 3.18, with the cups and alsa split discussed on the mailinglist Sep 18 17:55:00 03koen 07org.oe.dev * r9a65859e... 10/ (1 packages/bluez/bluez-libs_3.19.bb): bluez-libs: update to 3.19 Sep 18 17:55:04 03koen 07org.oe.dev * r4f606468... 10/ (4 files in 2 dirs): bluez-utils*: update to 3.19 Sep 18 18:26:23 koen, ping Sep 18 18:27:16 pong Sep 18 18:27:36 There were some fixes committed to eglibc after a big merge last week Sep 18 18:27:49 I am not getting the latest rev checked out when I build Sep 18 18:28:00 for OE Sep 18 18:29:42 vi conf/distro/includes/sane-srcrevs.inc Sep 18 18:29:44 03koen 07org.oe.dev * rc926e75b... 10/ (1 packages/python/python-pybluez_0.13.bb): python-pybluez: update to 0.13 Sep 18 18:29:54 every svn packages is now pinned Sep 18 18:30:46 time to watch dr house Sep 18 18:30:48 bbl Sep 18 18:30:55 koen enjoy Sep 18 18:37:14 koen, I got disconnected Sep 18 18:49:44 khem: vi conf/distro/includes/sane-srcrevs.inc Sep 18 19:04:55 hi koen, if i'm porting to a arm5te target, is there any reason to NOT use the EABI toolchain? Sep 18 19:13:30 HopsNBarley_: only if you're on kernel <2.6.16 Sep 18 19:14:18 koen, thanks - i'm 2.6.19. Sep 18 19:53:56 pxa270/ucb1400 record seems broken with 2.6.23-rc6 Sep 18 20:06:35 koen, upstream eglibc vsn is 3531 Sep 18 20:07:09 where as the one I get from OE is 3468 Sep 18 20:07:32 koen, How can I fix that Sep 18 20:09:52 vi conf/distro/includes/sane-srcrevs.inc Sep 18 20:10:02 then change the eglibc* entries Sep 18 20:11:36 hmm pxa270/ucb1400 record works now -- strange ... Sep 18 20:28:17 could anyone here tell me how exactly bitbake calculates dependancies between recipees Sep 18 20:28:39 and whether those dependancies are cached anywhere (and if so, how the cache can be flushed) Sep 18 20:29:41 polz: you have build time dependencies DEPENDS, and run time dependencies RDEPENDS Sep 18 20:30:00 polz: bitbake -g may help understand what the build time dependencies are Sep 18 20:30:07 cbrake: I've set up an overlay... Sep 18 20:30:41 and in it, I have a task recipee called task-openmoko-polz Sep 18 20:30:45 which builds fine Sep 18 20:31:27 and then I have a recipee for an image, polz-openmoko-devel-image, which DEPENDS on task-openmoko-polz Sep 18 20:31:36 and when I try to build that, I get: Sep 18 20:31:49 | ERROR: Cannot satisfy the following dependencies for task-openmoko-polz: Sep 18 20:31:50 | libmp3lame0 (>= 3.96.1) libsdl-1.2-0 (>= 1.2.9) libsdl-1.2-0 (>= 1.2.9) libgl1 (>= 6.5.2) libsdl-1.2-0 (>= 1.2.9) libsdl-1.2-0 (>= 1.2.9) Sep 18 20:32:01 and I don't even know where it's getting these libs from Sep 18 20:33:15 I suspect it cached the dependancies from when I tried building my image for x86 Sep 18 20:33:35 or perhaps from when I tried adding mplayer to the image... Sep 18 20:33:53 polz: did you wipe /tmp before you switched from x86? Sep 18 20:33:57 nope. Sep 18 20:34:20 if I wipe it, will that mean a full rebuild ? Sep 18 20:34:26 polz: yes :-( Sep 18 20:34:45 I'd really much rather try to find out where these libs are coming from Sep 18 20:35:13 on my computer, a rebuild takes more than 2 days :( Sep 18 20:35:17 polz: run bitbake -DDDD and that may give you some info Sep 18 20:35:39 polz: I suspect the above are run time dependencies as you did not see them when building Sep 18 20:36:49 perhaps... but I doubt that Sep 18 20:37:01 I'll try removing all the packages from my task. Sep 18 20:37:03 polz: you could expand the ipk with ar and take a look, but who knows how far down the dependency tree the above are occuring Sep 18 20:37:30 polz: you have tried bitbake -g and then looked at the *.dot files? Sep 18 20:37:37 nope. Sep 18 20:41:55 cbrake: even if my task doesn't depend on anything, bitbake still thinks the image should depend on those libs Sep 18 20:43:24 polz: can you pastebin the output of bitbake -DDDD ? Sep 18 20:43:32 just a sec Sep 18 20:44:37 make that a minute... each run of bitbake seems to take almost that long Sep 18 20:50:00 cbrake: bitbake -DDDD produces about 900K of output. Sep 18 20:50:24 =:-\ Sep 18 20:50:31 polz: make -DDD then? Sep 18 20:51:11 btw, what happens if you delete the sqlite database bitbake uses as it's cache ? Sep 18 20:51:32 polz: no idea Sep 18 20:52:18 -DDD seems to be the same Sep 18 20:52:23 I'll try -DD Sep 18 20:53:27 that saved a few K, but it's still well over 900K Sep 18 20:53:35 hello anyone clued in on HOWTO add a package to the rootfs such as "mii-tools" which has a bb file Sep 18 20:54:09 that got it down to 400K Sep 18 20:54:21 cbrake: do you still want to see the output ? :P Sep 18 20:54:55 polz: sure Sep 18 20:55:36 cbrake: I'll get it available on the net Sep 18 20:56:40 s/get/make/ Sep 18 20:58:04 grrr... each run of bitbake takes ages Sep 18 21:02:23 koen: There's something wrong with that bluez-util update: http://bugs.openembedded.org/show_bug.cgi?id=3030 Sep 18 21:03:25 cbrake: http://ecg.fri.uni-lj.si/~polz/bitlog.txt Sep 18 21:03:52 it should be up there in less than a minute (scping now) Sep 18 21:05:33 cbrake: OK, it's there, if you can, please try making sense of it. Sep 18 21:05:41 and I'll see if bitbake -g works here Sep 18 21:07:16 polz: the failure is a run time (ipk) dependency failure Sep 18 21:07:47 polz: I would guess that xtscal has a RDEPENDS on libsdl and friends ... Sep 18 21:08:34 cbrake: I've tried rebuilding xtscal... didn't help Sep 18 21:09:27 polz: ok, the dependencies are probably for all packages then -- not necessarily xtscal Sep 18 21:09:49 I've just created the .dot files ? Sep 18 21:10:24 polz: you can look at them with a text editor and search for dependency clues Sep 18 21:11:57 polz: the other thing you can do is modify the do_rootfs script in build/tmp/work/...//temp and see if there are debug options you can add to ipkg-cl to give you more information Sep 18 21:12:09 polz: and then run the do_rootfs script manually Sep 18 21:12:33 polz: this is where OE is brilliant -- many of the operations are run from generated shell scripts which makes debugging somewhat convenient Sep 18 21:12:42 cbrake: there's no trace of libsdl in either of the generated .dot files Sep 18 21:13:15 polz: libsdl may be and ipkg name then and not a recipe name Sep 18 21:13:36 cbrake: yeah, but shouldn't it show up in the dependancy graph ? Sep 18 21:17:35 polz: recipe names and package names are not necessarily the same. One recipe often generates many packages. Sep 18 21:18:08 polz: you could search the dot files for sdl instead of libsdl ... Sep 18 21:18:19 cbrake: tried that. Sep 18 21:18:30 that didn't help, either Sep 18 21:18:44 I really think libsdl must be cached somewhere Sep 18 21:18:57 polz: have you "rebuilt" your task? bitbake -c rebuild Sep 18 21:19:48 I think so, but I'll try that again Sep 18 21:19:52 how can I "rebuild" an image ? Sep 18 21:20:09 because as I've said, the task builds without any trouble Sep 18 21:20:26 CM: there is no closing quote on SRC_URI Sep 18 21:21:03 CM: edit bluez-utils3.inc and change file://hcid.conf \ to file://hcid.conf " and rebuild and tell me if it works Sep 18 21:21:37 polz: bitbake will only build a task if it things it is out of date, so if you have made changes without bumping the PR int he task recipe, nothing is changing Sep 18 21:21:44 polz: bitbake -c rebuild Sep 18 21:22:20 polz: and then try rebuilding your image Sep 18 21:23:14 how do I _rebuild_ the image ? Right now, I'm running bitbake because bitbake -c build doesn't work at all Sep 18 21:24:50 polz: bitbake -c rebuild && bitbake Sep 18 21:25:03 polz: you are rebuilding the task, and building the image Sep 18 21:25:10 XorA: You mean remove the two file things below? Sep 18 21:25:19 OK, I've rebuilt the task and am waiting for the image to build Sep 18 21:25:50 btw, where exactly do .ipk files get built ? During the task build or during the creation of the image ? Sep 18 21:25:51 CM: no, change backslash to quote Sep 18 21:26:03 Ok, I'll try that Sep 18 21:26:10 But it looks weird Sep 18 21:26:40 cbrake: 10x, that did it ! Sep 18 21:26:57 polz: phew ... Sep 18 21:27:05 XorA: Totally didn't work. I'll try to comment out the two lines below Sep 18 21:27:07 CM: looks like koen removed some patches and forget to reclose the quotes Sep 18 21:27:11 this whole build system somehow manages to make me feel stupid every day :/ Sep 18 21:27:19 and I guess it's right. Sep 18 21:27:35 polz: its really very simple -- just hang in there and it will click one of these days :-) Sep 18 21:27:44 CM: we talking the same file bluez/bluez-utils3.inc Sep 18 21:27:59 XorA: I was just wondering the same.. Wait a sec Sep 18 21:28:01 polz: make sure you understand what just happened Sep 18 21:28:11 cbrake: sorry, I don't... Sep 18 21:28:34 I thought bitbake was supposed to detect the changed .bb file and rebuild the task by itself Sep 18 21:28:41 and I don't know why it didn't Sep 18 21:28:43 CM: sorry I would do it myself, but Im in the middle of another package upgrade Sep 18 21:28:51 polz: no, it only rebuilds if the version changes Sep 18 21:28:58 polz: it will reparse the file if the file changes Sep 18 21:29:16 polz: but not rebuild. Any change to a bb files requies an explicit rebuild, or bump the version # Sep 18 21:29:23 XorA: http://rafb.net/p/t5Gn2792.html Sep 18 21:29:48 hmm.. and the version is set by PR, right ? Sep 18 21:29:50 CM: 3.inc in filename Sep 18 21:30:03 polz: yes, that is the last chunk of the ver Sep 18 21:30:18 Ah, ok. Sorry Sep 18 21:30:27 CM: no probs, its damn confusing in that dir Sep 18 21:30:58 OK, by next week, I might even become familiar enough with this build system to actually try and code something Sep 18 21:31:00 CM: http://rafb.net/p/1wByid56.html Sep 18 21:31:01 sigh. Sep 18 21:31:20 XorA: *sigh* I did look in the wrong file all along, that's why I didn't see any fault in the SRC_URI Sep 18 21:31:38 I suspected it was something easy like that, but I just couldn't see it. Now it built fine.. Sep 18 21:31:43 CM: anyway if you can verify Ill push in a minute Sep 18 21:31:45 XorA: Thanks a lot :-) Sep 18 21:32:05 It built Sep 18 21:32:39 CM: pushing to repo Sep 18 21:33:24 XorA: Closing the bug too or should I? Sep 18 21:33:31 CM: done Sep 18 21:33:50 :) Sep 18 21:33:51 CM: bug close that is Sep 18 21:34:00 I noticed. Thanks again Sep 18 21:52:34 any E experts in the house? Sep 18 21:54:21 03xora 07org.oe.dev * r372eeb2b... 10/ (1 packages/bluez/bluez-utils3.inc): bluez/bluez-utils3.inc : correct quote at end of SRC_URI Sep 18 21:55:53 XorA: I don't like E, I stick to plain old beer.. ;) Sep 18 21:56:09 * CM waves goodnight Sep 18 21:57:33 lol Sep 18 21:57:48 beer rox Sep 18 21:59:50 XorA: aloisiojr is working with bbs of efl Sep 18 22:00:01 XorA: I'm changing the current bb files (and creating others) to make cvs's versions of E Sep 18 22:01:13 http://pastebin.ca/702513 anyone had such issues? Sep 18 22:01:30 aloisiojr: I have hit http://bugs.openembedded.org/show_bug.cgi?id=2992 Sep 18 22:07:27 XorA: i'll take a look Sep 18 22:11:33 XorA: no clues :( Sep 18 22:13:54 aloisiojr: Ill prod mickey|zzZZzz when he is next on Sep 18 22:17:21 XorA: do you have idea how to export a var to the global database: d? I'm trying to set a var inside the oe_runmake to get it later, but it seen the databases are different... Sep 18 22:17:56 vivijim: no, thats voodoo Sep 18 22:18:10 hahahaha Sep 18 22:18:36 * XorA sings Voodoo people hoodoo people Sep 18 22:19:53 XorA: hoodoo == who do? hehehe Sep 18 22:21:39 ... what you do dare do people o/~ :) Sep 18 22:26:44 time to sleep Sep 18 22:31:28 03xora 07org.oe.dev * r994032a8... 10/ (1 packages/poppler/poppler_0.6.bb): poppler/poppler_0.6.bb : new version of poppler Sep 18 23:32:38 gst-plugins just died on me .... Sep 19 00:11:29 two questions... Sep 19 00:12:21 1) how point to bb file that I have a build depends (not a run depends) Sep 19 00:12:50 morning Sep 19 00:13:05 if you answer DEPENDS, I'll say no... because bb make these packages as my run depends too Sep 19 00:13:18 rwhitby: morning... but here is evening :) Sep 19 00:16:09 ~ugt Sep 19 00:16:10 from memory, ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Sep 19 00:17:38 Crofton|home and ibot: ok! i've got it! morning rwhitby! :) Sep 19 00:20:08 no clues for my question??? even you ibot?? :D Sep 19 00:20:36 DEPENDS is builds RDEPENDS is runtime Sep 19 00:20:55 mind you, I get it backwards some times :) Sep 19 00:22:26 Crofton, but when the packages is built, the deb has DEPENDS as runtime depends :( and build-dep is empty... Sep 19 00:22:52 I am not familiar with the deb stuff Sep 19 00:22:58 maybe there is a bug? Sep 19 00:24:25 Crofton, hum ok.... I take a look in the code, tks the help... Sep 19 00:24:44 np Sep 19 02:34:02 anyone have an OE manual that is a complete whole file in some file? one big HTML? PDF? a tarball of all the HTMLs (with the links fixed)? Sep 19 02:34:16 or maybe a document that explains what all it _can_ do for me Sep 19 02:35:00 I was reading the online HTML document ... didn't yet find what I wanted to know when I reached chapter 4 and there was no "next" link to continue with Sep 19 02:35:23 so that document seems to be broken in that way **** ENDING LOGGING AT Wed Sep 19 02:59:57 2007