**** BEGIN LOGGING AT Fri Oct 05 02:59:57 2007 Oct 05 07:20:04 morning. Oct 05 07:20:46 is there a postprocess command or something similar to avoid a package gets added to a specific machine? for example, i got a board with no apm support so i dont want the apm package installed on that specific machine Oct 05 07:23:23 ok, MACHINE_FEATURES should do it hopefully but im interested in another way also Oct 05 07:51:40 dbus and avahi don't start in angstrom-console-image, is it expected or a known bug? Oct 05 07:57:48 hmm when having apm neither in DISTRO_FEATURES nor in MACHINE_FEATURES why does it still get installed on bootstrap-image? Oct 05 08:33:37 moin Oct 05 08:33:38 nik0n: After removing it, you'll probably need to rebuild task-base Oct 05 08:33:42 hi zecke Oct 05 08:33:44 koen|away: are you already traveling? Oct 05 08:33:45 morning all Oct 05 08:55:01 good morning Oct 05 09:45:38 moin Oct 05 10:37:14 gm Oct 05 10:37:20 I am in Berlin Oct 05 10:37:24 Hotel looks good Oct 05 10:37:34 time to go find the train station Oct 05 10:41:04 Crofton: hey! Oct 05 10:41:09 Crofton: come to the office? Oct 05 10:41:20 Crofton: I'm going to go with you to the train station... Oct 05 10:41:26 I am going to try meeting Koen etal at the station Oct 05 10:41:37 I think I am on the other side of the subway Oct 05 10:41:42 from the office Oct 05 10:41:54 Crofton: well, let us go together? Oct 05 10:42:06 ;) Oct 05 10:42:20 how about we meet at the V subway station Oct 05 10:42:29 * Crofton looks for the map Oct 05 10:43:06 er Zinnowitzer? Oct 05 10:43:07 Crofton: Zinnowitzer Strasse? Oct 05 10:43:22 yeah Oct 05 10:43:28 I get my letters mixed up Oct 05 10:43:35 I can be there in five minutes Oct 05 10:43:40 Especially after 24 hours with not much sleep Oct 05 10:43:41 ok Oct 05 10:43:51 which entrance? Oct 05 10:44:00 I know about the one in the middle of the road Oct 05 10:44:10 Crofton: well, we will go by bus to the station Oct 05 10:44:19 hmmm Oct 05 10:44:22 Crofton: so let us meet at the Subways Oct 05 10:44:26 you are wise in the ways of Berlin Oct 05 10:44:29 ok Oct 05 10:44:37 since I can find it easily Oct 05 10:44:54 I will be at the entrance in the middle of the road? Oct 05 10:44:55 "Subways" as in the store Oct 05 10:45:05 er ok Oct 05 10:45:08 where is that? Oct 05 10:45:13 and I will need three more minutes, will buy you something to drink Oct 05 10:45:29 thanks Oct 05 10:45:30 Crofton: "at the entrance in the middle of the road" when you don't cross the road Oct 05 10:45:37 ok Oct 05 10:45:44 bbl Oct 05 10:46:03 ok Oct 05 10:46:06 see you in a bit Oct 05 10:46:10 l8r Oct 05 10:46:47 gm Oct 05 11:33:49 good morning Oct 05 11:38:08 Hi. Can someone help me resolving a compile problem? Oct 05 11:38:54 I want to bitbake task-base for my target processor: at91rm9200. Oct 05 11:39:36 I COULD bitbake helloworld-image, but bitbake fails when I try to build task-base.... Oct 05 11:41:22 build fails with libtool-cross-1.5.10-r6.... Oct 05 11:46:52 kleckse: Have you symlinks in your paths? Oct 05 11:47:20 i don't think so.. Oct 05 11:47:26 let me check again Oct 05 11:48:26 no.. there are no symlinks Oct 05 11:48:45 I get the following message: Oct 05 11:48:56 NOTE: Task failed: Command Error: exit status: 1 Output: Oct 05 11:48:56 Usage: quilt [--trace[=verbose]] [--quiltrc=XX] command [-h] ... Oct 05 11:48:56 quilt --version Oct 05 11:50:00 for me it looks like someone called quilt with wrong parameters... but I can't see who is calling quilt ;-) Oct 05 12:01:43 kleckse: just checking : bitbake quilt-native first Oct 05 12:02:39 kleckse: (although quilt-native should have been built automatically) Oct 05 12:03:20 bitbake quilt-native succeeded... Oct 05 12:03:59 It should be using the quilt it built then :/ Oct 05 12:08:41 I just rerun bitbake task-base. same issue. Oct 05 12:20:39 how can i find out who called quilt with what parameters? Oct 05 12:21:44 03likewise 07org.oe.dev * r2409f128... 10/ (1 packages/lighttpd/lighttpd_1.4.18.bb): lighttpd: split plugin modules into separate packages. Oct 05 12:22:42 * Bernardo is away: Ausente por agora. Oct 05 12:24:55 kleckse: even if this might be a hammer-nail issue of mine: strace Oct 05 12:27:20 e.g. strace -ff -o foo -eprocess whatevercommand will give you files foo and foo.PID detailing all calls to process related syscalls (e.g. exec) Oct 05 12:46:02 strace doesn't give me a clue :-( Oct 05 12:58:52 03likewise 07org.oe.dev * raac60b18... 10/ (1 packages/fastcgi/fastcgi_2.4.0.bb): fastcgi: Disable parallel build, it's broken Oct 05 13:02:56 i think, that you could increase verbosity level in bitbake and then just grep for quilt part Oct 05 13:03:25 or find a place in bb scripts and add some debug/note output before it's executing quilt Oct 05 13:08:39 I tried to run bitbake with -D -D. finding the quilt part is easy, as build stops after this error. Oct 05 13:09:24 * nud has no login console on tty0 anymore, despite the sysvinit output goes there Oct 05 13:10:02 My problem is that i can't find the part in .bb scripts where quilt is called. I looked at the libtool-cross-1.5.10.bb file. But I don't see how this will call quilt. Oct 05 13:10:12 (I'm quite new to OE) Oct 05 13:11:38 classes/patch.bbclass Oct 05 13:14:42 there search for 'def _runcmd' Oct 05 13:17:03 add something like bb.note("XXX quilt args" + args) after the line 'def _runcmd(self, args, run = True):' Oct 05 13:17:20 it should now tell you with which args is quilt called Oct 05 13:31:51 i get many error like this: ERROR: exceptions must be strings, classes, or instances, not type while parsing /path/to/org.openembedded.dev/packages//.bb Oct 05 13:32:04 anyone saw this before? im tring to build a prebuild toolchain Oct 05 13:59:58 03henryk 07org.oe.dev * r0e2e8223... 10/ (4 files in 3 dirs): gtk-sharp 2.10: add package Oct 05 14:04:53 hi koen Oct 05 14:05:15 I feel a distance Oct 05 14:10:04 * koen_ has finally arrived Oct 05 14:16:06 koen_: how long have you been travelling? Oct 05 14:22:46 * Bernardo is back. Oct 05 14:37:50 ok. its final. I will in Berlin on Sunday :) Oct 05 14:39:41 steliosk: I look forward to meeting in person! Oct 05 14:43:08 * Bernardo is away: Ausente por agora. Oct 05 14:44:11 how would one go about adding entries to /etc/fstab during the image creation process? Oct 05 14:45:59 jsheldon: customize base-files Oct 05 14:46:03 RP: Same here. Will bring that ouzo bottle i promised :) Oct 05 14:46:46 cbrake: hello Cliff! ok that's what I thought, thanks Oct 05 14:47:49 steliosk: :) Oct 05 14:48:09 how can a package add an user with more than one groups ? Oct 05 14:48:21 steliosk: I did commit one change to cross mind but it was well tested ;-) Oct 05 14:56:44 RP: Yeah i seen it. Haven't found any time to test anything yet :( Oct 05 15:12:11 How can I build an image without a kernel embedded in the rootfs? Oct 05 15:12:40 just don't install the kernel image package Oct 05 15:16:29 good morning! Oct 05 15:16:42 Is there a way to force the use of a particular version of a package within an image recipe? Oct 05 15:17:24 gnuplot 4.2.0 pulls in libstdc++ while version 4.0.0 does not Oct 05 15:17:34 I really need the smaller size Oct 05 15:18:15 But I want to make that decision on a per image basis, not in general for the machine Oct 05 15:19:50 micropal: you can put RDEPENDS_kernel-base = "" in your machine.conf Oct 05 15:20:14 thanks I'll try that =D Oct 05 15:20:44 micropal: it will leave a link to uImage in /boot, but you won't have the kernel itself there Oct 05 15:20:44 sakoman: I did this once by blacklisting the bbfiles in local.conf Oct 05 15:21:00 sakoman: I know it's ugly, but works better than preferred_version Oct 05 15:21:36 nud: thought of that, but then it really isn't associated with the image and it's one more thing to remember when building that image :-) Oct 05 15:23:23 nud: for now I just made a copy of gnuplot in my collection and put a DEFAULT_PREFERENCE = "-1" at the top of the 4.2.0 recipe Oct 05 15:23:53 my OE karma is not that high yet ;-) Oct 05 15:24:25 morning Oct 05 16:01:55 I have a package that I'm trying to build (gnash 0.8.1) Oct 05 16:02:04 it's creating an empty include subdirectory. Oct 05 16:02:38 bitbake thinks that means that it should try to include the files which don't exist, and it exits with an error. Oct 05 16:02:58 Thus, I believe that the correct thing to do is convince gnash not to create the empty subdirectory, rather than try to change bitbake. Oct 05 16:03:01 Correct? Oct 05 16:17:44 nelson: yes, we typically patch sources to make them work in OE, unless this is a bitbake problem that should be fixed -- not quite sure exactly from your description Oct 05 16:18:25 it depends on the "cooperation" between configure and bitbake. Oct 05 16:18:48 arguably configure shouldn't be creating an empty subdirectory. Oct 05 16:18:55 nelson: can you pastebin your build log? Oct 05 16:19:27 but if it does and that's not changable, maybe bitbake needs to say "hmmm, empty subdir, I'll skip this." Oct 05 16:20:16 nelson: seems like its worth looking at. Oct 05 16:21:10 cbrake: well, yes, I'm currently trying to figure out how to torture configure into not creating the empty directory. Seems that it's stuffing in shell code which says "hmmm... I'll copy an empty list of files, but first I'll make the subdirectory into which they (won't) go" Oct 05 16:21:44 nelson: heh, getting configure to do what you want can be painful ... Oct 05 16:22:46 I know, and ... bitbake is written in a sane language, which makes the work-around a tempting path even if it increases entropy. Oct 05 16:39:19 ~ping zecke Oct 05 16:39:20 pong zecke Oct 05 17:02:00 * Bernardo is back. Oct 05 17:11:15 * Bernardo is away: Ausente por agora. Oct 05 17:30:48 * Bernardo is back. Oct 05 17:51:58 * Bernardo is away: Ausente por agora. Oct 05 17:59:06 I see a conflict in design between bitbake and auto$WHATEVER. Oct 05 17:59:29 bitbake assumes that if a subdir (in my case /usr/include) exists, that it must contain files. Oct 05 18:00:36 the standard snippet of code written by auto$WHATEVER into the Makefile makes the subdirectory even if $(include_HEADERS) is empty Oct 05 18:02:07 bitbake doesn't do anything with directories. perhaps you mean OE. Oct 05 18:03:39 scruggs: finally solved my gnuplot/libstdc++ issue. Turns out that gnuplot 4.2.0 for some reason uses g++ for its final link phase, wheras previous versions just used gcc Oct 05 18:03:56 So I just reverted to version 4.0.0 Oct 05 18:04:25 Need to check more closely, but I think that brings me to parity with buildroot Oct 05 18:10:20 sakoman: thats great news! Oct 05 18:22:12 * Bernardo is back. Oct 05 18:39:26 scruggs: did your experiment with reverting to an earlier version of dropbear work? Oct 05 19:00:40 scruggs: just tried reverting to dropbear v0.47. Success! No more zero length files with scp Oct 05 19:05:30 * jbs is away: Ausente por agora. Oct 05 19:21:45 re Oct 05 19:23:55 hi florian Oct 05 19:33:24 florian kannst du mal schauen was mit linuxtogo los ist? Oct 05 19:33:29 args Oct 05 19:35:39 woglinde: öh ok ;) Oct 05 19:35:59 florian the webserver hangs Oct 05 19:36:53 woglinde: ah, looks like the upgrade did not restart it Oct 05 19:37:24 hi khe1 /khem Oct 05 19:38:34 florian hehe Oct 05 19:39:30 likewise: its my IRC client giving me all sort of names on reconnection Oct 05 19:57:02 * flo_lap tries to restore the ltg theme from the backup Oct 05 20:04:56 hi Crofton Oct 05 20:05:23 gm Oct 05 20:05:30 almost gn for me Oct 05 20:05:41 just got back from dinner in Berlin Oct 05 20:05:47 no real sleep since I left the US Oct 05 20:09:24 Crofton: good morning! Oct 05 20:10:20 Crofton: finally reached parity sizewise with buildroot Oct 05 20:14:06 excellent Oct 05 20:16:16 ah email deleted Oct 05 20:17:34 Crofton: enjoy some German beer with dinner? Oct 05 20:17:46 I did :) Oct 05 20:17:58 of course we went to a Mexican place Oct 05 20:18:12 Good for sleep -- not that you are likely to have much trouble getting to sleep :-) Oct 05 20:18:19 heh Oct 05 20:18:28 having trouble staying awake now Oct 05 20:18:30 Crofton: I hate that more then anything Oct 05 20:18:32 ok Oct 05 20:18:34 It's the staying asleep that's always the issue for me Oct 05 20:18:40 yeah Oct 05 20:18:45 When I travel, I want the local food Oct 05 20:18:58 this is local Mexican food :) Oct 05 20:19:08 Lot's of coming wide awake at 2am on my first night Oct 05 20:19:14 The locals chose Oct 05 20:19:27 Worst Mexican food I ever had was in Oslo Oct 05 20:19:28 heh Oct 05 20:19:35 This is not bad Oct 05 20:20:18 Only time I've ever done that was eat at a Kentucky Fried Chicken in Singapore Oct 05 20:20:23 badbadbadbad Oct 05 20:20:28 heh Oct 05 20:20:33 bad everywhere! Oct 05 20:20:50 I saw an old school Kentucky fried chicken on the way on from the airport Oct 05 20:21:02 What time will you get started tomorrow? Oct 05 20:21:11 10 AM Oct 05 20:21:24 Fight the good fight on task-base :-) Oct 05 20:21:30 yeah Oct 05 20:21:41 we need to get that broken up somehow Oct 05 20:21:51 I need to get my head back togehter alos Oct 05 20:21:55 urg Oct 05 20:21:57 tired Oct 05 20:22:02 Go to sleep! Oct 05 20:22:09 ok I am going to "sleep" Oct 05 20:22:10 gn all Oct 05 20:22:14 GN Oct 05 20:22:17 Crofton: good night Oct 05 20:22:28 hi pH5 Oct 05 20:26:37 hi flo_lap Oct 05 20:28:14 would there be any reason for busybox's /bin/stty not to work with baud rates over 115200? Oct 05 22:09:49 when does oedem start tomorrow? Oct 05 22:35:04 hey all Oct 05 22:35:13 how to do change the output dir from tmp to something else? Oct 05 22:38:27 Tartarus: I believe its the TMPDIR variable Oct 05 22:38:43 you can set it in your local.conf or in your site.conf or something, at least if I am not mistaken Oct 05 22:39:04 I was hoping it was something like that, thanks Oct 05 22:40:51 hey Oct 05 22:41:01 hi Oct 05 22:41:49 evening Oct 05 23:08:30 * * OE Bug 2424 has been RESOLVED (WORKSFORME) by carbonfreeze(AT)gmail.com Oct 05 23:08:33 * *  After reboot, FN/Home keys do not work on first gpe login Oct 05 23:08:34 * * http://bugs.openembedded.org/show_bug.cgi?id=2424 Oct 05 23:08:43 * * OE Bug 2766 has been RESOLVED (WORKSFORME) by carbonfreeze(AT)gmail.com Oct 05 23:08:45 * *  No brightness adjustment in gpe-login Oct 05 23:08:47 * * http://bugs.openembedded.org/show_bug.cgi?id=2766 Oct 05 23:08:55 * * OE Bug 2953 has been created by mdevico(AT)qualcomm.com Oct 05 23:08:57 * * gstreamer_0.10.14.bb do_configure fails for x86_64 Oct 05 23:08:59 * * http://bugs.openembedded.org/show_bug.cgi?id=2953 Oct 05 23:09:07 * * OE Bug 2954 has been created by alessandro.iurlano(AT)gmail.com Oct 05 23:09:09 * * New program and recipe: ipkgorphan Oct 05 23:09:11 * * http://bugs.openembedded.org/show_bug.cgi?id=2954 Oct 05 23:09:19 * * OE Bug 2955 has been created by carbonfreeze(AT)gmail.com Oct 05 23:09:21 * * Unable to change terminal as normal user after logging into gpe Oct 05 23:09:23 * * http://bugs.openembedded.org/show_bug.cgi?id=2955 Oct 05 23:09:27 * * Flood control triggered, clearing queue... Oct 05 23:09:31 * * [334] pending messages removed. Oct 05 23:09:41 * * OE Bug 3124 has been created by cwixon(AT)usa.net Oct 05 23:09:42 * * mrxvt crashes and brings down X when forced to close Oct 05 23:09:44 * * http://bugs.openembedded.org/show_bug.cgi?id=3124 Oct 06 01:30:05 * * OE Bug 3125 has been created by autobuild(AT)openembedded.org Oct 06 01:30:06 * * webkit-gtk-0.0+svnr25808-r1-do_configure Oct 06 01:30:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3125 **** ENDING LOGGING AT Sat Oct 06 02:59:56 2007