**** BEGIN LOGGING AT Mon Jun 15 02:59:57 2009 Jun 15 05:57:34 anyone know how I can force the use of bluez-libs_3.36 instead of bluez4? I've tried setting PREFERRED_VERSION_bluez-libs ?= "3.36" and PREFERRED_VERSION_bluez-utils ?= "3.36" in my site.conf but it still tries to build bluez4 when I build a package depending on bluez-libs. Jun 15 06:17:20 erbo: what machine, what distro? Jun 15 06:17:59 try the -g switch to bitbake and inspect the dot files for an answer to your question Jun 15 06:18:12 the bluez situation is still somewhat sucky, I agree Jun 15 06:20:02 Laibsch: I'll try that.. It's a custom ARM-machine and angstrom (using stable/2009 metadata) Jun 15 06:24:05 Anyone knows if there are problems with the dri2proto recipe? It complaints about xorg-macros.m4 beeing too old. Jun 15 06:25:22 b0fh: talk to hrw|gone Jun 15 06:25:27 I had the same problem Jun 15 06:25:43 I disabled the 2.x version and everything is OK again now Jun 15 06:26:04 hrw|gone was unable to reproduce the issue when he tried Jun 15 06:26:09 Hmm.. I'll give it a go, thanks. Jun 15 06:36:28 good morning Jun 15 06:40:40 good morning Jun 15 06:41:23 mckoan: can you please comment on my xorg 7.4 RFC? It relates to kaleios. Jun 15 06:49:02 Laibsch: hi Jun 15 06:50:14 Laibsch: must be a patch that broke KaeilOS, I have to investigate Jun 15 06:53:27 no, it broke nothing, not me Jun 15 06:53:37 It's not yet applied Jun 15 06:53:48 It unbreaks things Jun 15 07:07:22 good morningall Jun 15 07:15:35 Laibsch: sorry, could you explain? what RFC ? :-) Jun 15 07:16:32 have you looked for my name? Jun 15 07:16:48 should be obvious Jun 15 07:17:01 I even mentioned kaleios, but apparently that failed to grab your attention Jun 15 07:24:19 Laibsch: any point, how to disable dri2proto 2? The 'preferred_providers' isn't the proper route I guess?? Jun 15 07:24:44 b0fh: several Jun 15 07:24:53 a) delete bb file in question Jun 15 07:25:11 b) set DEFAULT_PREFERENCE to -1 inside the recipe Jun 15 07:27:15 Ah.. no way to override the version beyond modifying the recipe/bb file than? Jun 15 07:34:37 I'm trying to compile qt 4.5 I have this error: http://www.mibbit.com/pb/OSy4cW . But I have recompiled all qt from scratch. I deleted tmp/work ... Jun 15 07:37:48 Laibsch: you asked me to comment your xorg 7.4 RFC, what is it? Jun 15 07:40:19 I'm sorry, I can't google it for you right now Jun 15 07:40:26 But it should be right there in the mailing list Jun 15 07:40:39 Only a few days old, shouldn't be hard to find Jun 15 07:41:01 It affects kaleios so I thought you better be aware of it Jun 15 07:52:18 Laibsch: in ML, ok thank you Jun 15 08:20:51 03Phil Blundell  07org.openembedded.dev * r87a447e42b 10openembedded.git/recipes/netbase/netbase_4.21.bb: netbase: don't ship update-inetd if not doing online package management Jun 15 09:18:54 03Phil Blundell  07org.openembedded.dev * r96a0d3afc6 10openembedded.git/ (classes/lib_package.bbclass conf/bitbake.conf): Jun 15 09:18:54 Revert "bitbake.conf, lib_package.bbclass: fix packaging of ${base_libdir}/*.so" Jun 15 09:18:54 This reverts commit ba434bd1dc9dc8ed61953a1579ea9e993e81cbb2. Jun 15 09:18:54 Acked-by: Richard Purdie, Leon Woestenberg Jun 15 09:27:14 florian: good morning Jun 15 09:27:55 good monring Jun 15 09:28:05 morning rather Jun 15 09:28:18 gm ;) Jun 15 09:29:25 can anyone help me with compiling x-load for omap3evm with oe? i basically have this situation: http://tinyurl.com/n9jeqn Jun 15 09:31:02 03Phil Blundell  07org.openembedded.dev * r0b33842906 10openembedded.git/recipes/procps/procps.inc: procps: adjust FILES so that libproc-*.so lands in ${PN} Jun 15 09:49:06 florian: so, latest on my qemu gpe-image; if I manually symlink /usr to / then it does boot as far as the gpe-login screen. however, the touchscreen doesn't work and none of the text displays correctly, so it's hard to use :-} Jun 15 09:49:26 I guess the touchscreen thing might be a qemu issue, I'll have to poke at that a bit more. lack of fonts is probably just a missing package in the rootfs. Jun 15 09:49:45 the gpe startup scripts do have an awful lot of absolute path references in them, though, which is a bit of a pain. Jun 15 09:57:47 Anybody can help me qith qt 4.5 ? I'm trying to compile it and I have this error: http://www.mibbit.com/pb/OSy4cW . But I have recompiled qt from scratch. I deleted tmp/work ... Jun 15 09:59:22 recalcati_oe: what recipe? Jun 15 10:01:32 pb_: oh well, sounds like this is still some work we have to do. Jun 15 10:02:12 yeah, I might try to convert all that stuff to use autotools. that way we can arrange for $(bindir) to be correctly substituted at build time. Jun 15 10:27:10 pb_: that's a good idea. There shouldn't be many bits missing any more. Jun 15 10:27:18 * florian shoots redboot Jun 15 10:28:56 first though, I'm going to see if I can get this new micro-image to work on my nslu2 Jun 15 10:30:11 gosh, the default busybox config is so bloated. telnetd?! Jun 15 10:35:06 fingerd? Jun 15 10:35:23 heh, no, that one isn't enabled. Jun 15 10:35:31 the ultra-important fsck.minix is in there though Jun 15 10:36:30 plus the famously-important-for-embedded-systems setconsole and setfont utilities. Jun 15 10:48:20 * pb_ stabs upslug2 Jun 15 10:48:24 "zImage-nslu2be.bin: file too large or too small: fatal error [std::exception]" Jun 15 10:48:52 I wonder what that means. Jun 15 10:49:42 std::exception? binutils uses c++ now? Jun 15 10:50:19 gold does, yeah. but upslug2 isn't part of binutils anyway :-} Jun 15 10:50:39 ah right, but gold is not default right? Jun 15 10:50:53 no, you have to configure --enable-gold I think Jun 15 10:51:24 ah, apparently the kernel is limited to 1MB. Jun 15 10:51:27 pb_: yeah, I tried with webkit... gold does not know about close()... so I ran into a ulimit Jun 15 10:51:39 shame that oe is building me a 1.2MB zimage. Jun 15 10:51:42 * pb_ stabs the kernel Jun 15 11:06:17 florian, ping Jun 15 11:11:29 03Leon Woestenberg  07org.openembedded.dev * re007568f88 10openembedded.git/conf/checksums.ini: linux-2.6.30: Added to checksums.ini Jun 15 11:53:56 sudh Jun 15 11:57:27 pvanhoof: pong Jun 15 11:57:42 florian, can you remove the tinymail-commits mailing list from linuxtogo? Jun 15 11:57:50 It's not used, and being spammed Jun 15 11:59:09 pvanhoof: not easily... but I have enabled the list administrators to remove their lists some time ago. Jun 15 11:59:28 florian, I have no password for this Jun 15 11:59:58 pvanhoof: you should have got the initial one when you created the project there. Jun 15 12:00:09 But I didn't create the project :) Jun 15 12:00:22 At some point I started receiving those spams from tinymail-commits Jun 15 12:00:47 who did create the project? Jun 15 12:00:51 no idea Jun 15 12:01:37 pvanhoof: http://projects.linuxtogo.org/projects/tinymail Jun 15 12:04:11 florian, "This Project Has Turned Off The Mailing Lists" Jun 15 12:04:22 But I can't login to the mailman interface, and it still sends spam Jun 15 12:04:50 I can't really edit much on that sourceforce thing Jun 15 12:05:05 yes you need to use the mailman interface. Jun 15 12:05:09 morning Jun 15 12:05:13 hi hrw Jun 15 12:05:27 florian, mailman interface's pwd != that page's pwd Jun 15 12:05:29 turning of the lists doesn't change the fact they exist. Jun 15 12:06:20 pvanhoof: yes right... I guess it is possible to do it with administrative privileges on the server with the mailman command line interface Jun 15 12:06:31 I don't have that Jun 15 12:06:59 I can do this, but toady I don't have a time lot left to find out how it works. Jun 15 12:07:05 ok Jun 15 12:07:23 well I don't use the list, so I'd even be fine if it were removed completely Jun 15 12:07:31 ok Jun 15 12:07:46 (just disable the virtual in postfix) Jun 15 12:30:22 dth_ntb: you have mail Jun 15 12:44:55 hello Jun 15 12:45:46 currently bitbake opie-image for machine h1940 pulls the 2.6.17 kernel. How do I change the opie-image to use my new 2.6.22 kernel for the h1940 machine? Jun 15 12:50:27 soz back Jun 15 13:02:22 which recipe indicates the opie-image to compile+install linux-2.6.17 for the machine h1940? how could I change the kernel version desired in this case? Jun 15 13:07:17 hrm, this isn't working too well. Jun 15 13:07:28 does anybody have the nslu2 5ki11z? Jun 15 13:09:33 pb_: I had one Jun 15 13:10:50 do you know where it gets its kernel commandline from? Jun 15 13:10:58 I guess it's configured in redboot somewhere but I can't find out where. Jun 15 13:11:00 redboot Jun 15 13:11:07 use 'fconfig' in redboot Jun 15 13:11:25 command 'exec' runs kernel with given cmdline Jun 15 13:16:56 ah right, thanks Jun 15 13:23:00 ah, not so good Jun 15 13:23:06 RedBoot> fconfig Jun 15 13:23:06 ** Error: Illegal command: "fconfig" Jun 15 13:23:06 RedBoot> Jun 15 13:23:07 doh Jun 15 13:25:13 ;( Jun 15 13:25:34 oh well. I guess I need to get a serial port to see what's really going on. Jun 15 13:29:05 pb_: 'fco' maybe? Jun 15 13:29:26 * florian fighting redboot too Jun 15 13:29:53 florian: there's nothing about "fco" in the help, but I guess it might be undocumented. I'll try it presently. Jun 15 13:30:04 help does say that it understands "fis", any idea what that does? Jun 15 13:30:04 http://www.run.montefiore.ulg.ac.be/~martin/enp-compiling.html helped for me last time Jun 15 13:30:14 fis lists flash partitions Jun 15 13:30:28 if they are done Jun 15 13:30:45 on edb9301 which I have here 'fis' returns many screens of crap Jun 15 13:33:25 well.. fis is a little bit more. its the tool to work with redboot partitions. Jun 15 13:34:24 okay, that doesn't sound like what I need Jun 15 13:39:18 * florian is shocked again what companies call a linux bsp Jun 15 13:43:08 hi all! how do you free the shared memory ? ipcs is not found on my system is there a different port for this? Jun 15 13:50:53 'morning Jun 15 13:51:10 florian: want to extend? Jun 15 13:54:38 03Leon Woestenberg  07org.openembedded.dev * r573fd638e7 10openembedded.git/ (3 files in 3 dirs): (log message trimmed) Jun 15 13:54:38 u-boot: Added 2009.06. Machine mpc8315e-rdb uses it now. Jun 15 13:54:38 Newest u-boot release 2009.06. Jun 15 13:54:38 Also, as the mpc8315e-rdb boards come with a buggy u-boot Jun 15 13:54:38 (possible invalid access due to missing guard, when Jun 15 13:54:40 speculatively executing libz code), its u-boot version Jun 15 13:54:44 is now set to 2009.06 which contains the fix. Jun 15 13:57:05 anybody used ipcs? Jun 15 14:05:32 i just now built a console-image for omap3 evm board. i try to mount the file system via nfs. it hangs after the message Configuring ppp. any idea what this could be due to? Jun 15 15:28:45 i built a file system for omap 3 evm board. the bb file i used is console-image. but when i boot the kernel hangs at hwclock: can't open '/dev/misc/rtc': No such file or directory Jun 15 15:29:00 i try changing inittab to run level 3. but i get same result Jun 15 15:29:56 What were you expecting to happen next? Jun 15 15:30:03 that is, what is the next initscript that ought to run after hwclock? Jun 15 15:31:06 are these run alphabetically? Jun 15 15:31:12 if so it is modutils.sh Jun 15 15:31:28 no, they're run in the numeric order that they appear in rcX.d Jun 15 15:32:04 or alphabetically in the event of a tie, so it runs S01aaa, then S01bbb, then S02aaa, etc. Jun 15 15:42:11 pb_: hi Jun 15 15:42:18 about the patches you committed lately Jun 15 15:42:46 Can you push them upstream or add a line "upstream: OE-only" at the top if they are not applicable? Jun 15 15:43:24 Laibsch: which patches in particular are we talking about? Jun 15 15:44:08 all of them ;-) Jun 15 15:44:29 heh Jun 15 15:44:46 okay, I'll have a look. is there any easy way you can think of for me to get a list of the patches in question? Jun 15 15:44:50 I just saw one of your commits the other, it had a patch and I thought it would be nice if we generally were documenting upstream status Jun 15 15:45:05 yeah, that's not a bad idea Jun 15 15:45:13 git can easily list your commits only Jun 15 15:45:21 I have done this with qgit Jun 15 15:45:30 which also shows you which files are changed Jun 15 15:45:38 ah right, that'd be handy Jun 15 15:45:43 gitk probably has something similar Jun 15 15:46:00 but I know with qgit you have everything you'll need right in front of you Jun 15 15:52:28 pb_: actually, cgit.openembedded.net has the same information Jun 15 15:58:41 is there a way on bitbake recipe to specify a different strip level for a patch? Jun 15 15:59:05 patch=? doesn't seem to make any difference, patch is always executed with -p1 Jun 15 16:03:48 tmartins_: pnum=0 .. stupid, i know, patch's value could be used. someone should think about improving that :) Jun 15 16:04:24 kergoth_: thanx! I apparently misinterpreted the bitbake source Jun 15 16:04:31 kergoth_: Hi, any chance to look at the tslib compilation failure, yet? Jun 15 16:05:31 ugh, not yet, been really busy with work and personal life, thanks for the reminder Jun 15 16:06:03 please, help me here. I have a patch url like this: file://xalan-c_gcc_4_4.patch;patch=1;pnum=2" Jun 15 16:06:25 but executed patch parameters are: -p1 --backup --prefix=.pc/xalan-c_gcc_4_4.patch/ -r /tmp/quilt_push.tndbIb -E -f -i patches/xalan-c_gcc_4_4.patch Jun 15 16:06:41 again, '-p1'... Jun 15 16:06:58 g'day kergoth Jun 15 16:07:48 Laibsch: how do I ask cgit to show me more history? Jun 15 16:07:58 http://cgit.openembedded.net/cgit.cgi/openembedded/log/?qt=author&q=pb%40 just shows me three commits Jun 15 16:08:14 oh, I see the problem, I gave it the wrong username Jun 15 16:08:22 http://cgit.openembedded.net/cgit.cgi/openembedded/log/?qt=author&q=philb works better. Jun 15 16:09:00 gah, nevermind, apparently, a -c clean solved it! thanks again Jun 15 16:09:37 Laibsch: the irda-utils and xcb patches are oe-only, not something that'd be appropriate upstream. Jun 15 16:09:42 the binutils patches are already upstream Jun 15 16:10:11 those are the only ones I noticed from a quick review of the logs. Jun 15 16:10:40 if you can add the information there, that would be wonderful Jun 15 16:10:49 no need to be complete Jun 15 16:11:16 would be nice if more people included that information so we get a critical mass and others follow Jun 15 16:11:40 right Jun 15 16:11:47 I think there are over 4000 patches in OE where it is unclear whether upstream would benefit from having them Jun 15 16:12:15 with "upstream: " at the top we at least now the status Jun 15 16:18:39 03Phil Blundell  07org.openembedded.dev * r75579638d2 10openembedded.git/recipes/irda-utils/irda-utils/sbindir.patch: irda-utils: note that this patch isn't wanted upstream Jun 15 16:18:40 03Phil Blundell  07org.openembedded.dev * r83446b2fb8 10openembedded.git/recipes/xcb/libxcb/configure.patch: xcb: note that this patch isn't wanted upstream Jun 15 16:18:41 03Phil Blundell  07org.openembedded.dev * r4e9311cee0 10openembedded.git/recipes/binutils/binutils-2.19.51/uclibc-segfault.patch: binutils: note that uclibc-segfault.patch is already upstream Jun 15 16:18:41 03Phil Blundell  07org.openembedded.dev * rc744c7e48e 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@new.openembedded.org:openembedded into org.openembedded.dev Jun 15 16:18:44 there we go Jun 15 16:41:52 !oebug 3678 Jun 15 16:42:02 03Phil Blundell  07org.openembedded.dev * r0fc051cfc8 10openembedded.git/recipes/busybox/busybox-1.13.2/micro/defconfig: busybox: add less bloated defconfig for micro Jun 15 16:42:08 * * Bug 3678, Status: HTTP transfer failed :( Jun 15 16:42:34 funny how much extraneous stuff there still is installed in these filesystems. Jun 15 16:42:52 good to see someone cleaning it up Jun 15 16:43:00 yeah, it's a slow process Jun 15 16:43:01 Good to see people still care Jun 15 16:43:10 OSS is slow Jun 15 16:43:14 slow but steady Jun 15 16:43:17 I'm still getting an /etc/opkg directory being created by something but I don't know what Jun 15 16:43:46 while you're at it, care to look at the tinylogin recipes? Jun 15 16:43:59 sure, what needs doing to them? Jun 15 16:44:06 They don't fetch anymore because tinylogin is now rolled into busybox Jun 15 16:44:46 oh, right. gotta find a new SRC_URI for those. Jun 15 16:45:03 I think it's basically not looking for the old source Jun 15 16:45:20 hi all Jun 15 16:45:30 but taking the busybox source and creating the right kind of configuration so that the output is just tinylogin Jun 15 16:45:37 there is a combination disto/image that work on avr32 ngw100 ? Jun 15 16:45:41 That is my understanding from what the website says Jun 15 16:46:42 gremlin[it]: Angstrom console-image works Jun 15 16:47:01 XorA|gone, thanks ! Jun 15 16:47:47 Laibsch: yah, I suspect that doing this will yield rather bigger binaries than the old tinylogin. Jun 15 16:47:49 worth a go though Jun 15 16:48:31 but, right now my mission is to reduce bloat rather than increase it, so I think I will park that for now and look at it some other day. :-} Jun 15 16:49:17 for micro we don't need a separate tinylogin, the busybox internal tools will do Jun 15 16:49:42 Laibsch: ping Jun 15 16:50:20 bbiab Jun 15 16:51:02 doh, image_rootfs.ipk itself is creating this unwanted /etc/opkg directory! Jun 15 16:51:11 mkdir -p ${IMAGE_ROOTFS}${sysconfdir}/opkg/ Jun 15 16:51:18 no wonder I couldn't find it in any package. Jun 15 16:51:59 CoreDump|cf-18: what do you think the issue is with cdbot2? Bugzilla works fine at the new installation, looks like mails are flowing, too. And DNS should have propagated. Jun 15 16:52:04 !oebug 3678 Jun 15 16:52:19 * * Bug 3678, Status: HTTP transfer failed :( Jun 15 16:52:41 pb_: makes sense. I was just wondering if you might want to take a look at it while "you're in the area" Jun 15 16:52:46 ant__: pong Jun 15 16:52:53 Laibsch: about opie / bluez problem (missing pref. providers) see http://fr.pastebin.ca/1461231 Jun 15 16:53:02 I meant to look for you because I have issues with kexecboot Jun 15 16:53:03 second problem is worst... Jun 15 16:53:18 I have nothing to report currently, though, because I restarted compilation from scratch Jun 15 16:53:37 opkg-cl Jun 15 16:53:43 will wait for input... Jun 15 16:53:47 03Phil Blundell  07org.openembedded.dev * r4319baa438 10openembedded.git/classes/rootfs_ipk.bbclass: rootfs_ipk: move "mkdir /etc/opkg" inside ONLINE_PACKAGE_MANAGEMENT conditional Jun 15 16:54:32 for bluez: http://fr.pastebin.ca/1461233 Jun 15 16:55:06 Laibsch: finally, your 'compiler not found' after rebuilding from pstage..confirmed :/ Jun 15 16:55:14 03Denys Dmytriyenko  07org.openembedded.dev * rc71ae35816 10openembedded.git/recipes/alsa/alsa-utils_1.0.18.bb: (log message trimmed) Jun 15 16:55:14 alsa-utils: fix linking against host libc.so/libm.so Jun 15 16:55:14 Caused by expecting ncurses library in /usr/lib. Jun 15 16:55:14 For some people it's just a warning: Jun 15 16:55:14 ld: warning: library search path "/usr/lib" is unsafe for cross-compilation Jun 15 16:55:16 ant__: yes, bluez needs more consideration Jun 15 16:55:18 ld: skipping incompatible /usr/lib/libm.so when searching for -lm Jun 15 16:55:20 And for some people it breaks the build: Jun 15 16:55:41 ant__: I restarted from scratch right now, rather than looking for a fix Jun 15 16:55:57 I'm retrying with gpe-image now first Jun 15 16:56:14 Laibsch: didn't you have a patch to fix the "opkg-cl hangs forever" thing by adding --force-defaults or something? Jun 15 16:56:16 And have locally removed bluez from the gpe-tasks Jun 15 16:56:19 Laibsch: look in your /cross ... former was /cross/armv5te/ after is /cross/bin Jun 15 16:56:26 pb_: yes, I do Jun 15 16:56:34 (after rebuild from pstage) Jun 15 16:56:36 did you check it in, or is it still pending? Jun 15 16:56:53 pb_: I don't have it checked in, since koen objected Jun 15 16:57:17 And it will mask those problems whereas right now, they are "in your face" Jun 15 16:57:55 can you remind me what koen's objection was? Jun 15 16:57:56 I am quite confident that by now I know how to properly fix the particular problem I was having myself Jun 15 16:58:22 IIRC, "fix the underlying problem rather than taping things over" Jun 15 16:58:31 which has some merit to it, I think Jun 15 16:58:47 well, it's true that the underlying problem should be fixed as well. Jun 15 16:58:48 OTOH, do_rootfs should really never ask for input during compilation Jun 15 16:59:02 but, opkg-cl hanging forever is just a bug in its own right, and deserves to be fixed independent of anything else. Jun 15 16:59:12 well, his objection is that you won't see the problem anymore if force-default is committed Jun 15 16:59:26 there is no justification for having it just sit there and spin, no matter how badly broken its input might be. Jun 15 16:59:33 pb_: andrea 31830 99.9 0.1 11788 8900 pts/0 R 07:13 159:22 opkg-cl -f /oe/build/tmp/staging/i686-linux/etc/opkg.conf -o /oe/build/tmp/rootfs/opie-image install task-base-extended task-opie-base task-opie-base-applets task-opie-base-inputmethods task-opie-base-apps task-opie-base-settings task-opie-base-decorations task-opie-base-styles task-opie-base-pim task-opie-extra-settings task-opie-bluetooth task-opie-irda Jun 15 16:59:33 initscripts udev sysvinit sysvinit-pidof tinylogin opkg-nogpg angstrom-version angstrom-feed-configs opkg-collateral Jun 15 16:59:38 before i killed it... Jun 15 16:59:49 I'm sorely lacking CPU cycles to test the many patches I have accumulated lately Jun 15 17:00:03 159 mins stolen..sh@t Jun 15 17:00:08 And I also need to do a few things while I'm here in .jp Jun 15 17:00:27 yeah, I guess it's possibly true that it would be better to have it output a diagnostic rather than just carrying on. Jun 15 17:00:36 ant__: locally apply the patch I sent to the ml Jun 15 17:00:44 maybe opkg just needs a --stop-on-errors flag, or something Jun 15 17:00:44 I'll do , thx Jun 15 17:00:56 or, even better, it could detect when its console is /dev/null and not try to read from it Jun 15 17:01:08 pb_: I wouldn't know how to do that Jun 15 17:01:17 without some research for which I lack the time Jun 15 17:01:22 yeah, fair enough Jun 15 17:01:24 pb_: isn't bluez - mismatch underlying problem ? Jun 15 17:01:31 yes Jun 15 17:01:41 bluez is still rather unfinished Jun 15 17:01:42 ant__: in that particular instance, yes. Jun 15 17:01:52 k Jun 15 17:01:55 it could happen with other packages, of course Jun 15 17:02:17 * ant__ crossing fingers and doing vodoo rituals Jun 15 17:03:50 Laibsch: opie should be set to bluez3 imho Jun 15 17:04:06 iirc it somehow worked sometimes Jun 15 17:04:40 is there any reason bluez3 is inherently better for opie? Jun 15 17:05:01 unsure about compatibility with bluez4 and dbus Jun 15 17:05:03 personally, I think bluez should be kept out of the picture until the maintainer has fixed things properly Jun 15 17:05:29 ant__: it's really a distro decision which bluez version to use. it'd suck a bit to try to use one version for opie and one for everything else, within a single distro. Jun 15 17:05:29 we have to ask Paul 'Bluelightning' Jun 15 17:05:30 none of the Zauri has built-in bluetooth Jun 15 17:05:51 Laibsch: true, though almost all of the ipaqs do Jun 15 17:05:55 I happen to have 2 CF cards Jun 15 17:06:05 my approach to dealing with this is right now "when in doubt, drop it from the image" - locally, that is Jun 15 17:06:29 ant__: well, you have to start somewhere. Jun 15 17:06:43 when the qemuarm image does not even calibrate for me, I have bigger problems Jun 15 17:06:48 he Jun 15 17:07:00 And if bluez keeps giving me headaches, it's dropped Jun 15 17:07:14 until other, more important things have shaped up Jun 15 17:07:59 pb_: you were somehow involved with blueprobe iirc? Jun 15 17:08:06 yeah Jun 15 17:08:15 ok, fine Jun 15 17:08:44 what would be *oday* the best way to sense a Socket CF card (CSR, btuart)? Jun 15 17:08:46 udev? Jun 15 17:08:51 *today* Jun 15 17:09:19 yeah, I guess so Jun 15 17:09:41 03Phil Blundell  07org.openembedded.dev * r9772552877 10openembedded.git/recipes/images/micro-base-image.bb: micro-base-image: new image Jun 15 17:09:41 03Phil Blundell  07org.openembedded.dev * rd1c51b45fc 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@new.openembedded.org:openembedded into org.openembedded.dev Jun 15 17:09:47 or hal, if you have it Jun 15 17:09:55 I ended up editing /sysconfig/bluetooth by hand Jun 15 17:10:39 oh, yes, now there are these demons... Jun 15 17:10:40 "ERROR: no cross compiler built yet?" - Anybody else getting lots of those? Jun 15 17:10:46 I Jun 15 17:11:00 but only after cleaning all but /pstage Jun 15 17:11:06 (as RP suggests) Jun 15 17:11:08 * pb_ go home now Jun 15 17:11:09 later all Jun 15 17:11:37 Laibsch: the compiler resides indeed in a wrong path Jun 15 17:12:07 the 'armv5te' part is skipped during rebuild of /cross Jun 15 17:12:15 in my case Jun 15 17:12:19 ant__: I restarted from scratch Jun 15 17:12:23 and still get those Jun 15 17:12:34 oh boy, bad bad Jun 15 17:14:31 more exactly, I got 'CROSSCOMPILER not found' or the more subtle 'CC cannot create executables' Jun 15 17:21:41 Laibsch: btw I've seen thesing on #kexecboot. I'll ask him to undo the version split in linuxtogo as you suggests Jun 15 17:22:37 yes, I saw he logged off and on a few times a few minutes ago Jun 15 17:24:55 ok, now I have the phrases in the buffer Jun 15 17:25:52 brb Jun 15 17:45:53 ant__: usually armv5l works for me (i have arm_v5t_le toolchain) Jun 15 17:50:05 rhc: this is the toolchain as built from scratch: http://fr.pastebin.ca/1461279 Jun 15 17:50:11 Laibsch: ^^ Jun 15 17:50:13 bbl Jun 15 17:51:36 ant__: that does not speak to me Jun 15 17:54:07 I suggest to add libapache2-mod-python to DEPENDS Jun 15 17:54:38 As well as python-mysqldb|$blah-pgsql|$blah-sqlite Jun 15 17:54:45 sorry, wrong window Jun 15 18:01:14 re Jun 15 18:07:48 florian: wb Jun 15 18:09:07 florian: thanks for your email :-) Jun 15 18:09:42 pb__: yw :) Jun 15 18:11:19 hmmm, no-one ever did answer the very serious questions brought up about the adding new members procedure Jun 15 18:19:48 * Laibsch missed the question being raised Jun 15 18:20:06 Laibsch: that the procedure is wrong and based on a misreading of the statutes Jun 15 18:20:15 e.V or RW? Jun 15 18:20:22 oh, e.V. Jun 15 18:20:33 in a nutshell, what was the question again? Jun 15 18:21:09 Laibsch: it seems incredible strange that 2 votes get you in, reading both the KDE and the google translation of the eV statutes it looks like to me it require 1 proposer, 2 seconders then a majority vote Jun 15 18:26:23 ah crap my email went directo to Crofton and not the list Jun 15 18:26:47 that may explain things ;-) Jun 15 18:28:05 XorA|gone: my reading of the KDE translation is that you need either a proposer and two seconders, _or_ a majority vote at a GM. But I agree, it is ambiguous in the translation and somebody who reads German should give us a definitive answer. Jun 15 18:28:28 XorA|gone, I see swine fever is in Scotland Jun 15 18:28:38 pb_, agreed Jun 15 18:28:47 they way it reads is ridiculous Jun 15 18:28:56 pb__: if that is the case a crazy person wrote the statutes Jun 15 18:29:03 someone should ask someone in the kde ev Jun 15 18:29:05 pb__: as Crofton|work pointed out it makes zero sense Jun 15 18:29:27 I mean, if it only takes two people, why do you need the alternatives? Jun 15 18:30:30 and also with only two people getting a block vote takes only a couple of days Jun 15 18:30:48 and the people dont even need to be anything to do with OE for a hostile takeover Jun 15 18:31:15 er? you've lost me with that last part. Jun 15 18:31:50 pb__: if two people can add a hundred new members who all vote there way Jun 15 18:31:59 pb__: what is to stop a hostile takeover? Jun 15 18:32:59 oh, I see. well, it wouldn't exactly be a hostile takeover if it was initiated by people who were already members. Jun 15 18:33:12 "hostile" generally means that it's an external agency making the action. Jun 15 18:33:30 but you're right, it would seem there is nothing to prevent what you describe. Jun 15 18:34:19 * XorA|gone suspects quite a few competing commercial entities and rival buildsystems will soon have enough votes to make that happen Jun 15 18:34:52 maybe I just spent too long running societies in Uni where the atmosphere isnt friendly Jun 15 18:35:07 however, even requiring more votes for membership doesn't solve that problem, it just raises the bar very slightly. Jun 15 18:35:38 rasises the bar to technically 50% of the members Jun 15 18:35:44 the members planning the takeover just need to arrange for their zombie friends to appear plausible enough to pass a vote. Jun 15 18:35:46 if they actually bother voting Jun 15 18:36:08 and from being done in one day to probably weeks Jun 15 18:36:29 and, assuming that most of the incumbent members probably won't bother to conduct a huge amount of due diligence, I doubt this is too big a hurdle. Jun 15 18:37:57 anyway, without knowing the actual meaning of the german statutes, this is a sterile discussion. Jun 15 18:38:11 hello Jun 15 18:38:17 there isn't much point in further debate until we have a definitive answer on that. Jun 15 18:42:54 alright Jun 15 18:43:06 I took a quick look on admittance procedures Jun 15 18:43:27 there's quite a few things here, for example there are three different types of members Jun 15 18:43:44 We're talking about the active membership role her Jun 15 18:43:51 e Jun 15 18:43:52 could you guys help me in saving lots of time? I bitbaked opie-image and afterwards did a git pull. now it says that because of the changed DISTRO_PR i will have to rebuild everything. I ignored this by manually changing tmp/distro_pr from .1 to .2 and rebuilding. it broke only in the end: http://pastebin.mandriva.com/13104 Jun 15 18:44:16 Basically, the way XorA|gone is reading the statute is correct Jun 15 18:44:24 one member proposes you Jun 15 18:44:47 you get two more supporters before a vote ist cast Jun 15 18:45:23 you need to get 50% or more in that vote Jun 15 18:45:30 me Jun 15 18:45:35 simple majority Jun 15 18:45:40 * florian doesn't read it like this Jun 15 18:45:46 no? Jun 15 18:46:09 what is your interpretation, you should be better informed, I just skimmed for the relevant part Jun 15 18:46:16 maybe I misunderstood Jun 15 18:46:37 i don't seen an indication that two of the three quoted ways need to be fulfilled. Jun 15 18:47:22 florian: it doesnt read like a list to me Jun 15 18:47:44 florian: We agree that 4.2 contains the relevant part? Jun 15 18:47:47 Die aktive Mitgliedschaft wird auf Vorschlag eines aktiven Mitglieds mit Zustimmung Jun 15 18:47:48 zweier anderer aktiver Mitglieder durch Beschluss der Mitgliederversammlung oder Jun 15 18:47:48 der aktiven Mitglieder durch Internetwahl mit einfacher Mehrheit erworben. Jun 15 18:47:51 apart from the fact that both interpretations lead to a situation that doesn't make much of sence Jun 15 18:48:09 makes a lot of sense to me Jun 15 18:48:15 first step: you get a proponent Jun 15 18:48:40 second step: you need to others to join in proposing you for membership Jun 15 18:49:26 third step (only if 1 and 2 are fulfilled): a vote is cast (offline or online), majority vote decides Jun 15 18:49:47 Laibsch: that is exactly how I read it, thanks for being clearer Jun 15 18:50:58 I never read it like this.. if people would like to see it interpreted in this way we need to get some voting system in place *now* Jun 15 18:51:11 XorA|gone: I think your interpretation is in line with what is written Jun 15 18:51:21 I note in the German there isnt the badly placed commas confusing matters Jun 15 18:51:23 florian: again, how did you see this? Jun 15 18:51:53 florian: I think we always needed a voting sytem, Im annoyed I sent my email to the wrong place Jun 15 18:51:56 Laibsch: Like I posted to the list some days ago. Three options to join and not two. Jun 15 18:52:47 "The active membership is acquired on proposal of an active member with agreement of two of other active members by resolution of the meeting of the members or the active members by Internet choice with simple majority." Jun 15 18:53:40 that to me reads exactly like Laibsch translated Jun 15 18:54:06 three options to join? Jun 15 18:54:09 "or by sacrifice of a chicken" Jun 15 18:54:11 I kind of see only one Jun 15 18:54:29 or is prpplague's the second option I am overlooking? ;-) Jun 15 18:54:46 for an active member that is the route to take Jun 15 18:55:04 florian: can you find out what KDE e.V. are doing? Jun 15 18:55:04 one can join not only as active member, there are two further member types Jun 15 18:55:56 That's nasty.. we should make sure to make this section more clear and simplify the process. I don't see a reason for a vote for each member. Jun 15 18:56:21 well, maybe the right thing is for the founding members to have a vote amongst themselves to clarify the articles, and then take whatever action is needed to admit new members as a result. Jun 15 18:56:47 obviously there's a fundamental choice to be made about what you're trying to achieve with the articles, in terms of how easy or hard it should be for new members to join. Jun 15 18:57:25 yes right, that's the way to go Jun 15 18:57:38 if it's going to require a vote of all the active members, in practice that probably means that new members are going to have to wait a few months to join, since the general meeting only occurs once per year and I imagine that having internet votes every few days would get annoying as well. Jun 15 18:57:47 but anyway, any volunteers to take a look at a voting system? Jun 15 18:58:05 pb_: I don't see any major issues Jun 15 18:58:16 I think kde uses one not run by them Jun 15 18:58:18 it's kind of like what we have when adding people for RW Jun 15 18:58:35 Laibsch: no, there's no major issue, it's just a decision that needs to be taken. Jun 15 18:58:37 Laibsch: The fact that we have to discuss this now indicates a major issue. Jun 15 18:58:52 note that adding people for RW does currently only require about two people in favour, not a majority of all developers Jun 15 18:59:11 you cast a vote, for RW it has to be unanymous (no objections). In both cases not everybody has to vote Jun 15 19:00:07 pb_: adding new members would not require a majority of all devs,either Jun 15 19:00:15 The third step is a vote Jun 15 19:00:18 yah, that's another ambiguity in the statutes, at least the english ones. Jun 15 19:00:29 pb__: It should be fine to have a comparably short period for voting and the majority usually defines as majority of votes instead of members. Jun 15 19:00:32 in all practicality, the vote would be called for Jun 15 19:00:40 I would interpret a "simple majority of the active members" to mean that you have to get 50% in favour, not just 50% of those who bother to vote Jun 15 19:00:41 It would include a deadline Jun 15 19:00:56 all votes cast by then are counted Jun 15 19:01:06 right Jun 15 19:01:16 at least, that's how written resolutions are dealt with in uk company law. it might be different in germany I guess. Jun 15 19:01:28 the new member is in if (number of votes in favour/number of votes cast) >= 0.5 Jun 15 19:01:36 http://ev.kde.org/reports/2007.php item 11, it shows how KDE do it Jun 15 19:01:50 pb__: it isn't... Jun 15 19:03:01 personally, I don't see any ambiguity in the statutes, they are perfectly clear Jun 15 19:03:42 Laibsch: could you explain this to the members list? Jun 15 19:04:08 certainly not now Jun 15 19:04:24 I'm close to dying because of sleep deprivation ;-) Jun 15 19:04:41 Laibsch: Im sure we can give you 24 hours :-D Jun 15 19:04:51 * Crofton|work is glad that English is not the only language where two people read the same document and come to different conclusions Jun 15 19:05:14 re-write the statues in Z Jun 15 19:05:45 Crofton|work: German is great to write ambuguous texts :-) Jun 15 19:05:54 heh Jun 15 19:05:57 so is English Jun 15 19:06:06 You guys should learn Japanese Jun 15 19:06:12 listen to people discuss the US constitution :) Jun 15 19:06:14 heh Jun 15 19:06:15 It's meant to be ambiguous Jun 15 19:06:38 I find German anything but clear-cut, like a marching tune Jun 15 19:06:45 tada humpta tata Jun 15 19:06:45 I major share of German lawyers are occupied with ambiguous laws ;) Jun 15 19:06:48 ;-) Jun 15 19:07:00 those aren't ambiguous Jun 15 19:07:05 they are full of loopholes Jun 15 19:07:22 and people trying to screw each other by passing through those loop holes Jun 15 19:07:39 and generally people liking to fight in courts :-D Jun 15 19:07:54 Thanks for looking at this Laibsch Jun 15 19:08:16 re Jun 15 19:08:45 I find German anything but *unclear*, like a marching tune Jun 15 19:08:49 that is what I meant Jun 15 19:08:52 very straight Jun 15 19:09:04 like a marching tune: tada humpta tata Jun 15 19:09:15 everybody marching to the same rhythm Jun 15 19:09:25 very clear-cut Jun 15 19:11:53 gm Jun 15 19:17:48 Laibsch hi Jun 15 19:17:57 do you need to talk about databases still? Jun 15 19:18:11 Eventually, yes Jun 15 19:18:16 are you the one to talk to? Jun 15 19:18:33 yep yep Jun 15 19:19:28 hosting more databases at OSU for you guys is fine, just fire requests to my inbox... Jun 15 19:23:06 i'm working with with gumstix&wifistix with oe is there a command like lsusb lspci to see the modules plugged in to board? Jun 15 19:23:13 no way nothing work ... avr32/ngw100 arm/nslug ... bha ... Jun 15 19:23:40 likewise_, do you have any posters we can use for Linuxtag? Jun 15 19:24:01 Crofton: yes, got the design files. Jun 15 19:25:30 I have the one we used at FOSDEM Jun 15 19:25:37 (the printed one) Jun 15 19:25:51 Crofton, florian: the PDF is in the openembedded-dev ml: 3 Feb "FOSDEM poster final" Jun 15 19:26:04 I' Jun 15 19:26:12 I'll not be at LinuxTag myself :( Jun 15 19:26:30 someone planned a wedding at the wrong date, at the wrong location Jun 15 19:26:37 ok Jun 15 19:27:00 I seem to have bricked my overo board pretty good, set bootdelay=0, my NAND image fails to bring up a login prompt over serial, and I changed my boot settings to only boot from NAND, so booting from SD doesn't work... Jun 15 19:27:01 maybe I shall forward that email to mickey|dinner ? Jun 15 19:27:12 Or someone else who might have a chance to print some? Jun 15 19:27:30 I think if I had a u-boot.bin that didn't read its environment from NAND I could recover it Jun 15 19:27:33 florian: Or is the one you scooped from FOSDEM still OK? Jun 15 19:28:11 actually, I should be asking this in #beagle first... Jun 15 19:28:22 likewise_: I think so... will check tomorrow Jun 15 19:28:52 jimsheldon: connect jtag Jun 15 19:28:54 gremlin[it]: whats up? Jun 15 19:29:08 hrw: yeah that might be my last resort Jun 15 19:30:20 XorA|gone, about avr32 in fail on parsing some recipes ... espeak_1.37.bb Jun 15 19:33:19 mhh the error is so early that isn't shown on tinderbox :s Jun 15 19:34:48 XorA|gone, http://pastebin.com/m7b866c29 Jun 15 19:35:35 gremlin[it]: uclibc? Jun 15 19:36:02 XorA|gone, MACHINE = "atngw100" DISTRO = "angstrom-2008.1" nothing more on local.conf Jun 15 19:36:46 gremlin[it]: ANGSTROMLIBC = "uclibc" is also needed Jun 15 19:36:53 hi all! lspcmcia -v lists my wireless module marvell 802.11 but it seems that no driver found for this... do i have to load a new image or is there a easier way?? any help greatly appreciated Jun 15 19:36:55 gremlin[it]: no avr32 support in glibc Jun 15 19:37:03 XorA|gone, ah ... Jun 15 19:41:51 XorA|gone, ok no errors on parsing ... now cross fingers for build ;) Jun 15 19:42:07 gremlin[it]: it worked a couple weeks ago I last did it Jun 15 19:43:55 orges: I think this channel might not be the best place for you to ask. What distribution and machine are you using? Jun 15 19:44:01 XorA|gone: reading through the members-list on the web, I share your concern that the current members were added against the statutes Jun 15 19:44:14 angstrom & gumstix Jun 15 19:44:16 and this may very well become a *very* big problem Jun 15 19:44:33 orges: probably best to try #angstrom in the first instance Jun 15 19:44:40 any disgruntled person can challenge anything OE does in the future Jun 15 19:45:15 Laibsch: thanks for looking at it as a native speaker, please take your time and send an email to the list, I think you are more understandable than me Jun 15 19:45:45 pb__: thanks for respond :) nobody says a word at #gumstix or #angstrom Jun 15 19:46:21 orges: your last question on #angstrom was hours ago and made no sense to me Jun 15 19:46:47 :) ok i'll update it Jun 15 19:49:29 Laibsch, once we resolve the wording, we can fix the problem easily Jun 15 19:51:21 and like XorA|gone says, get some sleep Jun 15 20:07:03 03Klaus Kurzmann  07shr/import * r26661c14ea 10openembedded.git/conf/distro/include/shr-autorev.inc: Jun 15 20:07:03 shr-autorev.inc: bump EFL_SRCREV to the new recommended rev (41040) Jun 15 20:07:03 Signed-off-by: Klaus Kurzmann Jun 15 20:07:04 03Klaus Kurzmann  07shr/import * r4f06d20c57 10openembedded.git/recipes/efl1/ecore.inc: Jun 15 20:07:04 ecore.inc: remove ecore-add-includes patch (upstream) Jun 15 20:07:08 Signed-off-by: Klaus Kurzmann Jun 15 20:07:10 03Klaus Kurzmann  07shr/import * rc1e8c85812 10openembedded.git/recipes/efl1/evas/rtl_cursor.patch: Jun 15 20:07:13 evas: update rtl-cursor patch to apply to the new EFL_SRCREV Jun 15 20:07:15 Signed-off-by: Klaus Kurzmann Jun 15 23:22:55 03Philip Balister  07org.openembedded.dev * r37d2ff3681 10openembedded.git/recipes/gnuradio/ (gnuradio.inc gnuradio_3.2.0.bb): gnuradio : Fixes for usrp support files packaging. **** ENDING LOGGING AT Tue Jun 16 02:59:57 2009