**** BEGIN LOGGING AT Mon Nov 24 02:59:57 2008 Nov 24 03:18:22 I have a problem with do_patch not creating a run/log file and I'm unable to tell what command is failing. Nov 24 03:18:43 is it possible to tell bitbake to spit out the commands it executes as it executes them? Nov 24 03:25:47 jkridner: have you tried BBDEBUG="yes"? Nov 24 03:26:02 no, only bitbake -D. I'll try that. Nov 24 03:27:23 wow, that is a lot of noise. hopefully it picks up the 'exec' calls. Nov 24 03:44:55 denix0: that was surprisingly useless. Nov 24 03:45:28 it seems odd to me that getting bitbake to tell you what it is doing isn't easier. Nov 24 03:54:57 jkridner: what does the recipe say about do_patch? Nov 24 03:55:27 nothing. it is all implied by SRC_URI. Nov 24 03:55:54 "inherit native" Nov 24 03:56:22 can you try to apply the patch manually in tmp/work? Nov 24 03:56:31 yes. Nov 24 03:56:36 it applies oky. Nov 24 03:56:39 ok. Nov 24 03:57:59 where is the default do_patch created? Nov 24 04:00:10 that's interesting - BitBake never creates run.do_patch and log.do_patch files... Nov 24 04:00:32 jkridner: patch.bbclass Nov 24 04:00:58 k, that is where I'm reading now, but I wasn't sure if that was the place or not. no simple function called do_patch. Nov 24 04:02:03 are those your own recipe/patch or I should be able to reproduce? Nov 24 04:15:37 well, it is on Mac OSX. Nov 24 04:15:45 with Gentoo-prefix. Nov 24 04:16:12 I'd rather just understand the general process for bitbake debug. Nov 24 04:16:29 I must eliminate the voodoo and take matters into my own hands. Nov 24 04:17:08 If you want, I can try the same recipe on Gentoo/x86 for comparison Nov 24 04:17:26 it works fine there. it is just gnu-config-native. Nov 24 04:17:35 I build it often with Gentoo/x86. Nov 24 04:17:44 that is, Gentoo/x86/Linux. Nov 24 04:18:17 http://rafb.net/p/8uY3jN55.html Nov 24 04:18:38 Unfortunately, the do_patch task does not seem to provide enough logging Nov 24 04:19:00 there should be a way to simply have all exec commands print. Nov 24 04:19:44 not really Nov 24 04:19:50 is 'runcmd' the only function used to shell out? Nov 24 04:19:57 but you can try this: Nov 24 04:20:40 well, I think bitbake should be changed to make something like that possible. it is crazy for every person new to bitbake to live with voodoo and not understand what commands are actually invoked by the tool. Nov 24 04:20:44 open patch.bbclass and uncomment print() calls in runcmd() Nov 24 04:22:07 odd those aren't enabled with a debug level. Nov 24 04:22:31 because they are simply commented out Nov 24 04:22:56 and by the way, that is not bitbake code, that is oe code Nov 24 04:23:05 yes, I can see that. I mean that seems like an odd choice (comment vs. debug level) Nov 24 04:23:13 sure. Nov 24 04:23:40 I wonder why there isn't a common function in bitbake for shell command invocation. Nov 24 04:23:55 the patch.bbclass does not seem to use many BitBake functions... Nov 24 04:24:01 seems like something very fundamental. Nov 24 04:24:10 there is Nov 24 04:24:33 k, then maybe I need to hack-up patch.bbclass to use them more? Nov 24 04:25:04 do you know what command invocation function out of bitbake should be used? Nov 24 04:25:42 BitBake provides lots of primitives (bb.debug, bb.mkdirhier, bb.copyfile etc.), not sure why patch.bbclass is not using them Nov 24 04:27:40 this output is very helpful. I can see that 'quilt' is the command being executed and that the quiltrc file being pointed to does not exist. Nov 24 04:27:50 not sure if there is an "exec" primitive in BB, I see lots of places calling os.system Nov 24 04:31:53 hurray: knowing is half the battle: http://tinderbox.openembedded.net/builds/59633/ Nov 24 04:35:59 so, what was the fix? Nov 24 04:37:26 I had an ASSUME_PROVIDED for quilt-native in my local.conf. Nov 24 04:37:36 so, the quiltrc file wasn't getting created. Nov 24 04:39:33 quiltrc is required, then? Nov 24 04:39:55 hmmm... that's wiki material, I would suspect. Nov 24 04:41:42 quiltrc gets created by quilt-native, I believe. every patch seems to use it. Nov 24 04:42:07 I only had the problem because I halfway followed http://lists.linuxtogo.org/pipermail/openembedded-devel/2008-November/006896.html Nov 24 04:42:48 sleep time. thanks and goodnight all. Nov 24 04:43:04 gn Nov 24 09:20:24 Hi Nov 24 09:20:29 i try to build iptables for gta01 (bibake iptables) Nov 24 09:20:36 at task 260 it fails to fetch git_www.denx.de.git.u-boot.git._9912121f7ed804ea58fd62f3f230b5dcfc357d88.tar.gz with 404 Nov 24 09:21:01 @http://www.angstrom-distribution.org/unstable/sources/ Nov 24 09:21:15 it seems to be removed Nov 24 09:21:22 i've found it at http://mirror.hentges.net/SonkeiSourceMirror/raw/ Nov 24 09:21:36 so how can i change the location/uri for this tarball? (i've got the newest version of openembedded). Nov 24 10:00:37 morning all Nov 24 10:01:55 good morning Nov 24 10:04:19 03Jeremy Lainé  07org.openembedded.dev * r659a8173e8 10openembedded.git/packages/u-boot/ (2 files in 2 dirs): u-boot: add support for boc01 LCD display Nov 24 10:06:47 Morning there Nov 24 10:07:37 I have a problem here, I can't understand why, but one of my task doesn't seem to output properly the packages, for my image generation Nov 24 10:17:48 Ainulindale: which task package is this? Nov 24 10:18:22 (I just deleted work and stamps, I'll see) Nov 24 10:18:36 http://shr.bearstech.com/git/?p=shr-overlay.git;a=blob;f=openembedded/packages/tasks/task-shr-minimal-x.bb;h=f777f56084ca71f69f20a9ed54afff36ab8f3382;hb=HEAD <= this one Nov 24 10:19:00 It just won't generate the binary localedata Nov 24 10:19:21 (In the image) Nov 24 10:21:58 pb__: well, removing stamps and work didn't work for my first image Nov 24 10:22:23 Though IMAGES_LINGUAS="en-us" could be the reason why, but I can't see why as I generated that already Nov 24 10:22:41 Ainulindale: are you saying that task-shr-minimal-x.ipk is correctly generated, but that the packages it depends on (glibc-binary-localedata-xx) are not? Nov 24 10:23:16 The packages are generated, but are not included in the images Nov 24 10:23:26 For no apparent reason Nov 24 10:23:28 oh, right. Nov 24 10:23:33 I also noticed that if a package was missing in the list Nov 24 10:23:37 It just didn't include the whole task Nov 24 10:23:42 Without any warning Nov 24 10:23:43 can you check the Depends line in task-shr-minimal-x.ipk? Nov 24 10:24:02 if you're on debian, the easiest way to do that is "dpkg-deb -I tmp/deploy/.../ task-shr-minimal-x.ipk" Nov 24 10:24:12 I'm not on debian :-) Nov 24 10:24:15 (gentoo) Nov 24 10:24:17 doh Nov 24 10:24:27 in that case, you might need to ipkg-unbuild and eyeball CONTROL/control by hand Nov 24 10:24:43 I flashed the image anyway Nov 24 10:24:49 So I'll check on it Nov 24 10:24:54 okay Nov 24 10:25:45 mc is useful for inspecting ipkgs as well... you might cahnge the name to something.a Nov 24 10:27:46 ah, good idea Nov 24 10:27:59 even better would be for somebody to contribute a "dpkg-deb -I" equivalent to ipkg-utils Nov 24 10:28:13 Well, no depends on task-shr-minimal-x Nov 24 10:28:34 Ah, didn't say anything, just wrong command Nov 24 10:28:46 http://pastebin.com/mc97a727 Nov 24 10:29:02 This is totally outdated Nov 24 10:29:20 I locally bumped PR to 2 on the buildhost Nov 24 10:29:21 that's rather odd. that list is completely different to what you had in your .bb file Nov 24 10:29:29 ah, it's an old dependency list? Nov 24 10:29:36 what version of the package does opkg think you have installed? Nov 24 10:29:49 ... ah, I see, 1.0-r1.1 Nov 24 10:30:05 so, sounds like your PR bump was ineffective for some reason Nov 24 10:30:10 Yeah, it's old. Nov 24 10:30:14 I bet that's because there's no PV Nov 24 10:30:31 no, lack of PV shouldn't make any difference Nov 24 10:30:37 Hmmm. Nov 24 10:30:45 it will just inherit the default PV of "1.0" from bitbake.conf if you don't specify one. Nov 24 10:30:52 What's weird is that it worked perfectly locally Nov 24 10:30:56 And not on the buildhost Nov 24 10:31:08 Even after deleting work and stamps for all the tasks Nov 24 10:31:13 check in tmp/work to see what version (i.e. what PR) of the package was actually built in this run Nov 24 10:31:37 The good one, it seems Nov 24 10:31:52 okay, now check in tmp/deploy/glibc/ipk to see what version is in your repository Nov 24 10:31:59 is that also the good one? Nov 24 10:32:19 Yep Nov 24 10:32:36 Ah Nov 24 10:32:40 Just saw a thing Nov 24 10:32:47 FOrmerly I had the PACKAGE_ARCH thingy Nov 24 10:32:53 There's old ipk... Nov 24 10:33:18 oh, right Nov 24 10:33:35 yes, I think opkg will prefer an old package for a "better" architecture over a new one for a less good architecture. Nov 24 10:33:54 Damn thing. Nov 24 10:33:57 if you have downgraded PACKAGE_ARCH then you probably do need to wipe the old packages from your repository. Nov 24 10:33:58 Well, removed everything. Nov 24 10:34:07 Isn't there a tool or command to do that properly? Nov 24 10:34:36 not that I know of. I just use "rm" :-} Nov 24 10:34:44 Great tool indeed :-) Nov 24 10:34:52 bbiab, meeting now Nov 24 10:34:54 Well, thanks anyway for stepping me through Nov 24 10:52:45 hi. is it possible to have the whole operating system in a cpio.gz root filesystem? that is always trestored to the same state after a reboot? Nov 24 10:53:16 i can boot the system with the cpio.gz initrd kernel option but it complains that it cant mount root ... Nov 24 10:53:39 i tried root=/dev/ram but this does not work either. i just wanted to mount the cpio.gz file as the real rootfs on my target Nov 24 11:15:17 nik0n: how about using an initramfs Nov 24 11:15:34 and just running the whole OS inside the initramfs Nov 24 11:24:06 rwhitby: im trying to do that Nov 24 11:24:14 i have th einitramfs Nov 24 11:24:24 and appendet the initramfs file as the initrd parameter to the kernel Nov 24 12:10:04 hello Nov 24 13:24:13 hi Nov 24 13:26:28 woglinde hi Nov 24 13:43:34 he zecke Nov 24 13:43:51 ho Nov 24 13:44:47 woglinde: I still got my link issue with gettext Nov 24 13:47:28 zecke I am testing here Nov 24 13:47:42 but I am now at binutils Nov 24 13:47:55 so I needs a while Nov 24 13:47:59 I'm building too... but also webkit... and I have to eat... i can barely see Nov 24 13:48:15 you to learn more from me Nov 24 13:48:18 have Nov 24 13:48:25 eating is essential Nov 24 13:48:34 even before programming Nov 24 13:48:36 I need to charge... Nov 24 13:51:53 morning Nov 24 13:53:16 hi hrw Nov 24 13:53:22 o fsck.. Nov 24 13:53:31 I had OOM Killer on home machine Nov 24 13:53:33 with 4GB ram Nov 24 13:53:36 lol Nov 24 13:53:47 ~curse strigi Nov 24 13:53:48 May you be reincarnated as a Windows XP administrator, strigi ! Nov 24 13:54:10 strigi? Nov 24 13:56:04 woglinde do you have some time to look at a recipie from me? Nov 24 13:57:09 mr_nice hm Nov 24 13:57:14 not really Nov 24 13:57:18 and I am at work Nov 24 13:57:25 woglinde ok Nov 24 13:57:29 where I do not have my ssh key yet Nov 24 13:57:42 woglinde: metadata tracker in kde4 Nov 24 13:57:49 hrw lol Nov 24 13:58:00 I wouldnt use such useless beasts Nov 24 14:02:20 heh.. removing ~/devel/OH/ subdirs gave me precious GBs... Nov 24 15:01:18 kergoth: good morning Nov 24 15:01:33 hi pb Nov 24 15:01:37 hi woglinde Nov 24 15:01:51 pb I found another package which do not compile with thumb Nov 24 15:01:59 uclibc ldso stuff Nov 24 15:13:57 hi ant Nov 24 15:14:54 hi woglinde Nov 24 15:15:32 he he, later I'll bother you about qmake-native Nov 24 15:15:45 FATAL: Unsupported target linux-uclibcgnueabi for oe-g++ qmake spec Nov 24 15:15:52 Package libqpe-opie - 1.2.4 - r0 do_configure Nov 24 15:16:19 yeah Nov 24 15:16:23 should be simple edit Nov 24 15:16:28 in one of the makefiles Nov 24 15:16:29 he..unsure Nov 24 15:16:42 in qmake native there is strange code Nov 24 15:16:52 hm Nov 24 15:16:53 okay Nov 24 15:17:01 never looked into it Nov 24 15:17:38 iirc there is another target for uclibc but not eabi Nov 24 15:17:55 right Nov 24 15:18:10 but it should be trival to add it either Nov 24 15:18:22 only there ? Nov 24 15:20:34 woglinde: for template in linux-oe-g++ linux-uclibc-oe-g++ linux-gnueabi-oe-g++ Nov 24 15:28:37 anybody aware of a good tutorial for initramfs based root filesystems? Nov 24 15:32:58 woglinde, building for atngw100(avr32) I get failed do_patch: NOTE: Task failed: Error: /home/tharvey/oe/atngw100/org.openembedded.dev/packages/uclibc/uclibc-0.9.30/avr32/./ldso_hash_arm.patch not found. - I suspect thathttp://gitweb.openembedded.net/?p=openembedded.git;a=commit;h=c9640f06e3d05300251b8acc7979089b8413e28f is not quite correct? Nov 24 15:34:04 tharvey hm okay Nov 24 15:34:09 yes it isnt right anyway Nov 24 15:34:42 the original requester didnt test if the binutils does really have gnu-hashstyle support Nov 24 15:34:51 tharvey just edit the .bb Nov 24 15:35:01 and remove the avr32 line or comment Nov 24 15:35:08 ya Nov 24 15:35:33 and disable the hash-gnustyle support in the uclibc.config Nov 24 15:35:36 files Nov 24 15:38:21 woglinde, ah... ok wouldn't have caught the change to the config - will you revert that patch then? Nov 24 15:41:31 tharvey yes I will Nov 24 15:41:37 thx Nov 24 15:41:39 I just waited someone raises his hands Nov 24 15:41:43 *g* Nov 24 15:41:50 heh Nov 24 15:49:02 woglinde: ah, what goes wrong with it? Nov 24 15:49:20 it claims it cant find a registr Nov 24 15:49:40 but havent the time to isnpect it for now Nov 24 15:49:59 I am needing a working oe Nov 24 15:50:14 so I fixed it the koen-style Nov 24 15:50:59 args Nov 24 15:51:02 03Jeremy Lainé  07org.openembedded.dev * rc7d514e3de 10openembedded.git/packages/u-boot/ (4 files in 3 dirs): u-boot: add mtdparts support for boc01 Nov 24 15:51:12 we need a newer netbase recpie Nov 24 15:51:21 who works at Mobisense ? Nov 24 16:06:55 i build a rootfs which will be startet as initramfs. during the system bootup i get the error msg that /dev could not be mounted (this is a tmpfs) and tmpfs is build into the kernel. besides that my /tmp tmpfs is mounted from fstab properly. seems like udev wants to mount /dev with a tmpfs and its own device files. anybody experienced similar problems with initramfs? Nov 24 16:19:32 damn tabkey Nov 24 16:27:04 openembedded.org down? Nov 24 16:33:12 tharvey: looks like the linuxtogo server is down Nov 24 16:33:23 tharvey: http://wiki.openembedded.net/index.php/Main_Page seems to be up Nov 24 16:33:59 ah... thx Nov 24 16:34:52 or at least very sloooow Nov 24 16:35:03 does http://openembedded.org redirect there anyway? Nov 24 16:35:10 yes Nov 24 16:35:26 good to know... I'll change my bookmark/notes Nov 24 16:39:46 heh.. removing some builds, old toolchains and 156GB back Nov 24 16:42:24 156GB!?! :-D Nov 24 16:42:55 woglinde: doh Nov 24 16:44:02 args Nov 24 16:44:17 looks like I have to make maven support into oe Nov 24 16:44:50 :) Java gets more and more invasive -- like a weed in your garden. Nov 24 16:45:03 hehe Nov 24 16:45:15 I am getting paid Nov 24 16:45:18 for it Nov 24 16:45:25 or droped out Nov 24 16:45:27 hehehaa Nov 24 16:45:38 ;) I figured you must be; I didn't think you would do that voluntarily!! :-D Nov 24 16:46:10 It's good to see paid work add to OE, though. Nov 24 16:46:47 ;) Nov 24 16:47:11 it is good to be paid for OE related work Nov 24 16:47:35 hrw yes but do not know now if building maven suckz Nov 24 16:48:45 oh not Nov 24 16:48:48 ant Nov 24 16:48:49 args Nov 24 16:52:18 pb still there? Nov 24 16:54:03 woglinde pasted "uclibc thumb error" at http://paste.lisp.org/display/70968 Nov 24 16:54:36 woglinde: some stuff for it you maybe will find in jalimo overlay Nov 24 16:54:46 hrw hehe just checked Nov 24 16:54:48 no maven Nov 24 16:54:51 but ant maybee Nov 24 16:55:07 will ask robert on wendsday here at work Nov 24 16:57:50 okay bye for now Nov 24 16:57:54 I am heading home Nov 24 16:59:14 trying to understand diff between PROVIDES and RPROVIDES for creating a virtual package - is it that you would use PROVIDES for something that is needed for other packages during buildtime (like a lib) and otherwise RPROVIDES for runtime only (or both if needed?) Nov 24 16:59:30 tharvey right Nov 24 16:59:54 thx Nov 24 16:59:57 hm Nov 24 17:00:10 but for virtual runtime I do not know exactly Nov 24 17:00:30 mabyee then the depends of the real package is used Nov 24 17:01:02 and shared lib deps are found automatic Nov 24 17:01:13 okay bye Nov 24 17:05:18 broonie: Nov 24 17:05:31 Tab completion? Nov 24 17:05:47 broonie: lol, again :) Nov 24 17:06:06 sometimes the CTRL+F9 for ion3 get lost and I tabe bro return Nov 24 17:06:10 to get to my browse workspace Nov 24 17:33:38 ibot, botmail for woglinde: ah, this is a compiler bug. what version of gcc do you have? Nov 24 18:27:39 03Jeremy Lainé  07org.openembedded.dev * r30926fd252 10openembedded.git/packages/linux/ (13 files in 2 dirs): linux-2.6.26: add patches for boc01 hardware Nov 24 19:16:29 03Michael 'Mickey' Lauer  07org.openembedded.dev * rdd2d70e621 10openembedded.git/packages/tasks/ (task-cli-tools.bb task-x11.bb): Nov 24 19:16:29 task-cli-tools: ship s3c24xx always, it's supersmall Nov 24 19:16:29 task-x11: the xserver subpackage needs to be architecture-specific Nov 24 19:21:27 pb_? Nov 24 19:33:11 good day all! Nov 24 19:52:25 Ainulindale: you rang? Nov 24 19:54:48 pb_: heh :-) Nov 24 19:54:49 Well Nov 24 19:54:53 I had time to look at this Nov 24 19:54:57 And I still don't understand Nov 24 19:55:00 (my outdated task) Nov 24 19:55:10 I cleaned it, cleaned work and stamps Nov 24 19:55:13 Regenerated it Nov 24 19:55:16 Installed it Nov 24 19:55:19 I have the good version Nov 24 19:55:21 But not the good depends Nov 24 19:55:27 (Which are still the one of the old task) Nov 24 19:55:39 oh dear. Nov 24 19:55:48 you mean, opkg info prints the good Version but the bad Depends? Nov 24 19:55:52 Yep Nov 24 19:56:28 pb_: I did a basic rm tmp/work/*/task* tmp/stamps/*/task* tmp/deploy/glibc/ipk/*/task* Nov 24 19:56:40 (rm -rf of course) Nov 24 19:56:51 hm, very strange. what does "bitbake -e -b whatever.bb |grep ^RDEPENDS" say? Nov 24 19:57:34 Something very log :-) Nov 24 19:57:43 heh Nov 24 19:57:46 can you put it in a pastebin? Nov 24 19:57:52 But seems to be good. Nov 24 19:58:09 okay Nov 24 19:58:19 http://pastebin.com/m6aa1d335 Nov 24 19:58:29 have you tried unbuilding the .ipk in your tmp/deploy/glibc/ipk directory and inspecting its control file? Nov 24 19:58:31 (I'm also looking for advices for this locale thing) Nov 24 19:58:38 Nope, not yet Nov 24 19:58:51 Some pointers as to how to unpack it? Nov 24 19:59:12 there should be a program (script) called ipkg-unbuild in tmp/staging/HOST_SYS/bin Nov 24 19:59:54 run it on the .ipk in question and it will unpack the contents into your current directory Nov 24 20:00:23 tar: all/task-shr-minimal-x_1.0-r3.1_all/data.tar.gz: Cannot open: No such file or directory Nov 24 20:00:26 That doesn't look good. Nov 24 20:00:35 oh, drat, that's an ipkg-unbuild bug Nov 24 20:00:41 it can't unpack packages with no files in Nov 24 20:00:45 (I'm lucky :-) ) Nov 24 20:00:52 you'll have to do it by hand in that caes. Nov 24 20:00:53 case Nov 24 20:01:06 the .ipk file is actually an ar archive, you can bust it apart with "ar x whatever.ipk" Nov 24 20:01:19 that should give you a control.tar.gz file which you can unpack as normal Nov 24 20:02:01 (good to know) Nov 24 20:02:20 Yep, wrong one Nov 24 20:02:33 weird Nov 24 20:02:37 No RDEPENDS in there Nov 24 20:02:45 No appropriate depends either Nov 24 20:02:59 And it's the mirror of what I was compiling before Nov 24 20:03:05 there's no such thing as RDEPENDS at the ipk level. the RDEPENDS from the .bb file will be translated to Depends: here. Nov 24 20:03:08 So I thought there was a cache of some sort somewhere Nov 24 20:03:16 pb_: yeah that's why I precised Nov 24 20:04:06 I don't think there's any cache of that stuff, other than the bitbake parse cache which shouldn't have any effect. Nov 24 20:04:27 What bothers me is why it keeps adding old dependencies Nov 24 20:04:33 There's even a dependency on X Nov 24 20:04:40 is the datestamp on the .ipk file correct? you're definitely looking at the right file? Nov 24 20:04:40 Whereas I use task-x11-illume for that Nov 24 20:05:21 Yep Nov 24 20:05:45 odd. I can't think of any way you could get the wrong Depends in the ipk if the RDEPENDS in bitbake's world is right. Nov 24 20:05:49 Listen I'll recompile it as we speak and we'll check Nov 24 20:06:26 Ok, timestamp right. Nov 24 20:07:17 Wrong dependencies. Nov 24 20:07:49 crazy Nov 24 20:07:54 what command did you use to rebuild? Nov 24 20:08:04 bitbake task-shr-minimal-x Nov 24 20:08:10 (with a clean before) Nov 24 20:08:19 try with bitbake -b , that bypasses the parse cache Nov 24 20:08:40 bitbake -c clean first, again Nov 24 20:09:11 (Indeed, indeed) Nov 24 20:10:18 Same again Nov 24 20:10:47 crumbs Nov 24 20:10:53 I wonder where on earth it can be getting that from. Nov 24 20:11:11 Same here :-) Nov 24 20:11:49 I have no clue about that as it builds fine on my home computer Nov 24 20:12:36 (Though it doesn't seem to include my RRECOMMENDS) Nov 24 20:12:43 (In the generated task) Nov 24 20:12:50 (or in the image) Nov 24 20:12:56 hrm. Nov 24 20:14:02 Hmmmm Nov 24 20:14:03 well, the code to do this is pretty straightforward, you can see it in package_ipk.bbclass. Nov 24 20:14:10 There's something in pkgdata Nov 24 20:14:12 ARCH related. Nov 24 20:14:33 I'm afraid it could be the problem Nov 24 20:14:37 ah, what's that? Nov 24 20:14:50 Well Nov 24 20:14:51 RDEPENDS Nov 24 20:14:53 And FILES Nov 24 20:14:56 Guess what? :-) Nov 24 20:15:07 (Thanks find .) Nov 24 20:15:32 aha Nov 24 20:15:58 afk for a few minutes now Nov 24 20:16:04 Thanks for stepping me through again :-) Nov 24 20:17:28 have a nice rest of day Nov 24 20:20:10 gn Nov 24 20:55:38 hi, just wondering if anybody knows what the preferred method of setting DNS servers is for angstrom... since /etc/resolv.conf is a symlink to a tmpfs, I assumed using dns-nameservers in /etc/network/interfaces would work, but it doesn't seem to... any other ideas? Nov 24 21:07:32 germann: that would be a question for #angstrom, I suppose Nov 24 21:14:14 pb_: Didn't work, again :-/ Nov 24 21:17:06 pb_: I asked there first... no response Nov 24 23:11:03 Hello, i'm running gpe-image on device but after calibration dialog can't change anything at NewUserDialog. Can anuone help? **** ENDING LOGGING AT Tue Nov 25 02:59:57 2008