**** BEGIN LOGGING AT Sat Aug 15 02:59:59 2009 Aug 15 06:16:50 khem: For some reason, SDL is not defined. OE has always built fine until two days ago. Aug 15 06:17:25 john3909: mine is uptodate and builds fine Aug 15 06:17:40 john3909: whats your host config Aug 15 06:17:47 khem: I'm using Ubuntu 9.10 Aug 15 06:18:04 khem: let me get that Aug 15 06:18:22 do you have sdl installed Aug 15 06:18:37 I have it building ok on ubuntu 9.10 Aug 15 06:19:01 what is your distro and machine conf Aug 15 06:20:03 khem: here is my local.conf - http://pastebin.com/m3e1b37d2 Aug 15 06:21:58 beagleboard and angstrom.2008 Aug 15 06:22:49 khem: yes Aug 15 06:24:14 I have different machine and distro but its arm Aug 15 06:24:19 and works fine Aug 15 06:24:33 khem: which package do I install for sdl? Aug 15 06:25:28 I dont have dsl installed too Aug 15 06:25:28 khem: I have libsdl1.2debian installed Aug 15 06:25:35 remove it Aug 15 06:26:19 khem: why did this become a problem only now? Aug 15 06:26:41 well I have same installed Aug 15 06:27:08 khem: libsdl1.2debian? Aug 15 06:27:25 yes Aug 15 06:28:09 khem: In you log file, does SDL get expanded? Aug 15 06:28:28 s/you/your/ Aug 15 06:29:27 sdl does not appear at all Aug 15 06:29:54 khem: then I am at a loss Aug 15 06:30:40 khem: after I made that change, bitbake beagleboard-demo-image works fine Aug 15 06:31:31 khem: are you building gnome-games? Aug 15 06:32:40 khem: some of these files reference sdl.h, which means you have to include the folder usr/include/SDL Aug 15 06:32:45 I have this PREFERRED_PROVIDERS += " virtual/libsdl:libsdl-qpe" in my local.conf Aug 15 06:34:07 khem: where did that come from? Aug 15 06:34:39 khem: I'm using the local.config defined by Koen Aug 15 06:35:37 khem: I can try and add this and see if this makes a difference Aug 15 06:42:44 khem: rebuilding, but cache has to rebuild, which is so slow Aug 15 06:52:36 khem: The only file that references SDL.h is games-sound.c and this only happens if HAVE_GSTREAMER is not defined. Aug 15 06:53:59 khem: compile fails Aug 15 06:54:37 khem: same error as before Aug 15 06:55:06 khem: Thanks for all your help. I'm going to bed as I have to be up early tomorrow. Chears. Aug 15 09:19:22 * rkirti cries. Aug 15 09:19:40 Latest oe tree breaks everything for me :-/ glibc failing Aug 15 09:42:28 hi Aug 15 09:43:18 what should I change to make console-image to autologin without asking user.. ? Aug 15 10:00:19 morning florian Aug 15 10:00:51 good morning Aug 15 10:04:17 rkirti: your tree seems to work as an overlay pretty well... but glib is still wrong Aug 15 10:04:51 florian: huh ? I thought I changed to maemo5.0alpha wherever needed ? Aug 15 10:05:10 yes but its not ...1maemo4+0m5 in there Aug 15 10:05:29 its ... 1maemo2... Aug 15 10:06:45 florian: these versioning changes seem to be more frequent than I guess. I think I will try and check if I can add my DLDIR to a webserver here so we get temporarily get around the source issue Aug 15 10:07:36 florian: I have the opposite issue here - the higher level stuff built well. and then I do a git pull and glibc breaks :-/ Aug 15 10:07:59 hi,does xserver-xorg support fbdev? Aug 15 10:08:17 because I bitbaken xserver-xorg and don't find the fbdev module Aug 15 10:08:22 but there is a kdrive one Aug 15 10:08:41 rkirti: that's a good idea... worst case we can use LinuxToGo for this purpose. Aug 15 10:16:25 found it Aug 15 10:25:01 rkirti: do you have the glib archive at hand... maybe you can try to dcc it so that i can continue building Aug 15 10:25:11 * florian can't find this package any more Aug 15 10:25:29 mmm Aug 15 10:26:33 in xf86-video-fbdev there are #include "somefile.h" instead of #include ...how do I fix that? do I patch the files or should I look in the autotools? Aug 15 11:01:55 hi,anyone have cross-compiled xorg-drivers ? Aug 15 11:16:29 Gnutoo: tinderbox? Aug 15 11:16:48 works now but it's *very* hackish Aug 15 11:16:51 I hope it runs Aug 15 11:16:54 I've no tinderbox Aug 15 11:17:08 should I connect my laptop to the oe's build system Aug 15 11:17:09 ? Aug 15 11:18:19 http://tinderbox.openembedded.org Aug 15 11:18:31 you can browse through build reports Aug 15 11:18:48 and the FAQ section in the wiki has information on how to send automatic build reports Aug 15 11:18:58 zecke: ping Aug 15 11:20:09 ok thanks Aug 15 13:08:54 If qt4-x11-free-gles and qt4-x11-free both provide qt4x11 why does qt4-x11-free-gles get picked in the absence of any other preference specification? Aug 15 13:09:12 for "bitbake qt4x11"? Aug 15 13:26:30 Laibsch: whichever one parses first or has higher version wins Aug 15 13:26:52 actually I think it might be parses last, but basically the first one bitbake comes across when it does the run queue Aug 15 13:27:39 koen rejected the proposal I showed you yesterday. Essentially leaving me out in the rain again to fix other people's work. Aug 15 13:27:50 I think the fix was proper to get proper selection. Aug 15 13:28:03 All he said was "find out why X gets picked over Y" Aug 15 13:28:11 So, that does not help me much. Aug 15 13:28:15 bitbake qt4x11 without defining more into is just asking to get rendom selection Aug 15 13:28:20 I'd like to fix this properly for everybody in OE Aug 15 13:28:33 bitbake is essentially random Aug 15 13:28:38 But right now, I'm on the verge of fixing it for minimal and forgetting baout the rest Aug 15 13:29:15 XorA|gone: That's just to make the problem more to the point Aug 15 13:29:23 I ran into this issue when doing "bitbake anki" Aug 15 13:29:56 XorA|gone: BTW, would you be so kind to ACK my OE sources mirror network RFC? Aug 15 13:29:59 it is in the end why I wrote the generic blacklist code Aug 15 13:30:06 ? Aug 15 13:30:09 what's that? Aug 15 13:30:12 tell me more Aug 15 13:30:19 classes/angstrom.bbclass Aug 15 13:30:29 what stops Angstrom using bluez3 Aug 15 13:31:30 hm Aug 15 13:31:38 another angstrom-specific solution Aug 15 13:31:51 why not let everybody have the code and benefit from it? Aug 15 13:32:44 because this is open source and not all of use go into toxic shock just from the name angstrom Aug 15 13:36:03 I hope this doesn't sound wrong Aug 15 13:36:11 But you know that is not a proper reason Aug 15 13:36:36 I have yet to see anybody go into toxic shock Aug 15 13:36:43 what is difficult with another distro doing INHERIT += "angstrom" Aug 15 13:36:48 apart from political reasons Aug 15 13:37:02 I can't answer that on a technical level Aug 15 13:37:06 * XorA|gone really doesnt care about names, just code Aug 15 13:37:41 But it sounds a bit like "what is difficult with you giving me root on your box so I can do X?", assuming X is something you may indeed want the other person to do Aug 15 13:37:55 Another distro wants to keep control, I guess Aug 15 13:38:13 so then they are not generic Aug 15 13:38:20 If angstrom.bbclass is generic enough to be used by other distros, then a different name would indeed help Aug 15 13:38:38 I wrote it, I would have no problems with a RFC to rename it Aug 15 13:38:40 currently it sounds to be applicable only to angstrom and people (me included) will not even look Aug 15 13:39:08 but I wrote it to solve an Angstrom problem Aug 15 13:39:15 let me take a look at what's inside Aug 15 13:39:23 I see Aug 15 13:39:27 I dont even know if other people have that problem Aug 15 13:39:39 yeah not exactly the largest chunk of python :-) Aug 15 13:40:20 Hehe Aug 15 13:40:30 But that only warns about problems, right? Aug 15 13:40:39 It does not help actually solving the issue Aug 15 13:40:49 so, we're still not much farther Aug 15 13:40:58 the compile will currently fail as well Aug 15 13:41:06 it stops the blacklisted package appearing Aug 15 13:41:13 so, that will be what you get as a warning ;-) Aug 15 13:41:29 Can it be used with overrides? Aug 15 13:42:00 it should follow all the standard overide rules Aug 15 13:42:14 Actually, I was thinking OE needs some kind of mechanism to say combination of, for example, PN/PV/DISTRO/MACHINE does not work Aug 15 13:42:26 a bit more generic than COMPATIBLE_MACHINE Aug 15 13:42:40 that tuple may even include something like gcc used Aug 15 13:49:01 in this case, I would want the blocked tuple to be (all distros)/(non-omap-machines)/qt4-x11-free-gles Aug 15 13:49:53 XorA|gone: That is VERY interesting code. I think it's the essential hint to get me going in the right direction of fixing "bitbake world" without removing bb files. Aug 15 13:50:22 Is the gcc used in a run stored in some variable? Aug 15 13:50:23 beware it gives a crazy python dump if you use it in combo with bitbake -b Aug 15 13:50:43 Laibsch: no, that would violate causality Aug 15 13:50:46 OK, so not yet ready to be turned by default, he? Aug 15 13:51:04 Laibsch: any call to skippackage with bitbake -b does that Aug 15 13:51:15 Laibsch: as you rip out the recipe under bitbakes nose Aug 15 13:51:21 If we had some generic blacklist Aug 15 13:51:47 what would variables be to test for incompatibility? Aug 15 13:51:49 PN Aug 15 13:51:51 PV Aug 15 13:51:53 MACHINE Aug 15 13:51:55 DISTRO Aug 15 13:52:28 anything else? Aug 15 13:52:35 generically you wouldnt include DISTRO as its distro.conf that should set blacklist Aug 15 13:52:35 Am I missing something obvious? Aug 15 13:54:41 What about two PNs conflicting? I'm of course aware of CONFLICTS in the bb. But that does not support versions. Aug 15 13:54:48 or so I think Aug 15 13:55:27 this is about the point steam comes out of my ears Aug 15 13:56:31 ? Aug 15 13:57:00 I'm just thinking of what things may conflict Aug 15 13:57:00 brain overheats Aug 15 13:57:17 Get some cool beer ;-) Aug 15 13:57:25 as lubricant Aug 15 13:57:42 cool G&T :-D Aug 15 14:01:02 G&T? Aug 15 14:01:10 Gin and Tonic Aug 15 14:01:36 yeah, the Brits like that I heard Aug 15 14:01:50 It helps getting old ;-) Aug 15 15:02:54 man I wish -dbg packages werent so huge :-) Aug 15 15:07:51 pulseaudio stopped building recently. The error here: http://pastebin.com/d584b6af0 is understandable, but can someone please tell me why such a thing should happen ? Aug 15 15:10:50 rkirti: new and bizarre :-) Aug 15 15:28:00 hi Aug 15 15:28:22 there are sombeody? Aug 15 15:28:53 nope :-D Aug 15 15:29:48 hahah Aug 15 15:29:57 I've a (simple) question Aug 15 15:30:07 I have my image recipe Aug 15 15:30:30 with some "thigs" inside tha compile (in my case drivers for xserver) Aug 15 15:30:33 XorA|gone: :-/ nooo Aug 15 15:31:01 when I comment some of these insode and rerun bitbake continue to include the driver!!! Aug 15 15:31:17 How I can to rerun the package list in the image? Aug 15 15:34:24 rkirti, I agree with XorA|gone Aug 15 15:35:27 * rkirti is facing vague build problems with some core components after the last git pull :-/ Aug 15 15:35:51 no? Aug 15 15:36:00 nobody? Aug 15 15:37:00 Crofton|work, XorA|gone : here is another one : http://pastebin.com/d41d1fa85 , what package should i update to give it the newer xorg-macros it needs ? Aug 15 15:37:43 * XorA|gone doesnt see these issues with Angstrom Aug 15 15:38:16 but I am using Angstrom Aug 15 15:38:35 odd, Im just doing builds and have seen that sort of thing Aug 15 15:38:41 did you mix stabe/dev ? Aug 15 15:39:17 XorA|gone: no, I only work on .dev Aug 15 15:44:12 hello woglinde Aug 15 15:44:36 jo rkirti Aug 15 15:54:51 03Graeme Gregory  07org.openembedded.dev * r55afc69df3 10openembedded.git/recipes/transmission/ (files/config files/init transmission_1.73.bb): Aug 15 15:54:51 transmission_1.73.bb : makes this LSB compliant and stop messing with /home Aug 15 15:54:51 added a /etc/default/transmission-daemon to contain paths and values used Aug 15 15:54:51 by the daemon. Moved the homedir and config to /var/lib/transmission-daemon Aug 15 15:55:01 03Graeme Gregory  07org.openembedded.dev * rb4147042b4 10openembedded.git/recipes/transmission/transmission_1.73.bb: Aug 15 15:55:01 transmission_1.73.bb : fix a couple of typos to do with localstatedir and Aug 15 15:55:01 make /etc/default/transmission-daemon a conffile Aug 15 16:00:17 finally been meaning to do that for ages Aug 15 16:00:27 hm? Aug 15 16:08:17 ---------------- Aug 15 16:08:27 I've a (simple) question about OE: I have my image recipe with some "things" inside that compile (in my case drivers for xserver). when I comment some of these insode and rerun bitbake continue to include the driver!!! --- How I can to rerun the package for remake the list in the image? Aug 15 16:08:46 ???? Aug 15 16:08:55 there is no automatic recompilation Aug 15 16:09:02 only when PR is bumped Aug 15 16:10:00 woglinde: I'm new in OE: how I can do that Aug 15 16:10:25 ah Aug 15 16:10:29 the image Aug 15 16:10:41 hm yes the driver maybee have other dependcies Aug 15 16:10:47 inside the recipes Aug 15 16:11:00 so they get compiled, but dont end up in the result image Aug 15 16:11:16 most of the recipes have a DEPENDS line Aug 15 16:11:33 where is defined which packages need to build first Aug 15 16:13:04 yes, I comment this drive (in my case xf86-video-via) Aug 15 16:13:48 xf86-video-via dont compile in i586-generic Aug 15 16:14:01 drm.h:567: error: expected specifier-qualifier-list before 'uint32_t' Aug 15 16:14:25 because of this I want to exclude from my xorg server Aug 15 16:15:04 hm Aug 15 16:16:58 woglinde: this is my image : http://pastebin.com/m2b537d62 Aug 15 16:17:28 again image only tells what ends up in the image Aug 15 16:17:36 not what is build Aug 15 16:18:43 if I chage the PR variable try to remake all? Aug 15 16:19:11 ???? Aug 15 16:19:22 not I misunderstood you first Aug 15 16:19:39 you have to change the xorg recipes to not build via driver Aug 15 16:19:51 in my image I have PR variable (PR = "r0" ) Aug 15 16:20:30 if I change to r1 and rerun bitbake try to remake the image but, in this case dont add my commented driver Aug 15 16:20:41 then bitbake check this variable for this Aug 15 16:21:46 the first time I run bitbake with a PR new, bitbake recheck all dependences and store this somewere asociated to the PR value Aug 15 16:21:57 my question is WHERE store this info Aug 15 16:24:29 no Aug 15 16:24:38 PR is only usefull for updates Aug 15 16:24:49 if already build a recipe Aug 15 16:24:55 and the PR get bumped Aug 15 16:25:04 the recipe will be rebuild Aug 15 16:25:10 ok Aug 15 16:25:18 and then on the device you can easily update with opkg Aug 15 16:25:44 but how I can to force a rebuild of the recipe without change PR? Aug 15 16:25:58 bitbake -c clean recipe Aug 15 16:26:05 bitbake recipe Aug 15 16:26:50 ok... Aug 15 16:27:14 but if my recipe is my image, this clean don't clean all the already compiled packaged? Aug 15 16:27:20 (are a lot!!) Aug 15 16:27:28 nope Aug 15 16:27:32 only the image Aug 15 16:27:37 ahhhhhhhhhhhhh Aug 15 16:27:39 TXS!!!! Aug 15 16:27:46 no prob Aug 15 16:27:56 but better you read the doku Aug 15 16:28:14 I don't want to wait another 3 day compiling ! :-) Aug 15 16:28:16 you had have known about bitbake -c commands Aug 15 16:28:34 buy a phenom2 board Aug 15 16:28:36 Yes, I'm reading the docs, but this was not clear to me Aug 15 16:28:45 its cost not much Aug 15 16:29:12 woglinde: well.. where are you from? Aug 15 16:31:30 woglinde: here in Argentina 1U$S=$3.85 and a computer now have a new tax of 30% Aug 15 16:32:00 woglinde: (to this add a 21% of IVA tax!) Aug 15 16:32:20 iva = tax to added value Aug 15 16:32:47 03Henning Heinold  07org.openembedded.dev * r1c28b135a2 10openembedded.git/recipes/libgsm/libgsm_1.0.12.bb: Aug 15 16:32:47 libgsm: fix install, don't use hardcoded include path Aug 15 16:32:47 * bump PR Aug 15 16:32:54 okay Aug 15 16:32:58 03Henning Heinold  07org.openembedded.dev * r03170c9374 10openembedded.git/recipes/ffmpeg/ffmpeg_0.5.bb: Aug 15 16:32:58 ffmpeg: remove libx264 presets files Aug 15 16:32:58 * bump PR Aug 15 16:32:59 03Henning Heinold  07org.openembedded.dev * r62c1a02c6e 10openembedded.git/recipes/ffmpeg/ffmpeg_0.5.bb: Aug 15 16:33:00 ffmpeg: fix --prefix for minimal(-uclibc) distribution Aug 15 16:33:00 yes argentina has problems Aug 15 16:33:02 * bump PR Aug 15 16:33:09 woglinde: a lot :-) Aug 15 17:06:24 03Klaus Kurzmann  07shr/import * rca0352f3ff 10openembedded.git/conf/distro/include/ (shr-autorev-unstable.inc shr-autorev.inc): Aug 15 17:06:24 shr-autorev(-unstable): set fixed rev for connman Aug 15 17:06:24 Signed-off-by: Klaus Kurzmann Aug 15 17:27:25 03Koen Kooi  07org.openembedded.dev * r2a1fada625 10openembedded.git/recipes/ffmpeg/ (ffmpeg.inc ffmpeg_0.5.bb): Aug 15 17:27:25 ffmpeg: but presets in their own package Aug 15 17:27:25 * people, adding subpackages is easy, doing rm in do_install* is pretty much always wrong Aug 15 17:48:54 morning kergoth Aug 15 17:48:57 hey Aug 15 17:57:20 * rkirti frowns Aug 15 17:58:10 When I switch from TMPDIR 1 to TMPDIR 2 for a different build and then come back to TMPDIR 1 again to make some tiny changes to my first image...why should bitbake start things again Aug 15 17:58:36 even if the local.conf is exactly the same as before, bits and pieces like binutils-cross get built all over again Aug 15 18:43:01 could OpenEmbedded be used to create a base image for an EDB9302? Aug 15 18:44:32 and why are the terms EDB9302 and EP9302 used interchangeably? Aug 15 19:31:26 re Aug 15 19:39:54 jo booxter Aug 15 19:40:43 jo ;) Aug 15 19:40:48 hi Aug 15 19:40:55 hi Aug 15 20:43:29 03Sebastian Krzyszkowiak  07shr/import * r1dd0704c26 10openembedded.git/recipes/images/shr-image.inc: shr-image: add fix for overwriting DISPLAY variable when using X forwarding Aug 15 20:43:38 03Sebastian Krzyszkowiak  07shr/import * rda6c162df0 10openembedded.git/recipes/images/ (2 files in 2 dirs): shr-image: add elementary configuration to /etc/profile.d/ Aug 15 21:21:02 03Klaus Kurzmann  07shr/import * r8581a8d218 10openembedded.git/recipes/images/shr-image/etc/profile.d/elementary.sh: Aug 15 21:21:02 shr-image: adjust ELM_FINGER_SIZE Aug 15 21:21:02 Signed-off-by: Klaus Kurzmann **** ENDING LOGGING AT Sun Aug 16 02:59:57 2009