**** BEGIN LOGGING AT Mon Feb 21 02:59:57 2011 Feb 21 03:29:16 03Mike Westerhof  07org.openembedded.dev * r26f05a7682 10openembedded.git/recipes/busybox/busybox-1.18.3/slugos/defconfig: SlugOS: busybox - add modutils support in preparation to switch. Feb 21 03:29:22 03Mike Westerhof  07org.openembedded.dev * r27332febf3 10openembedded.git/recipes/update-modules/update-modules_1.0.bb: Feb 21 03:29:22 SlugOS: update-modules - correct SlugOS-specific RDEPENDS override Feb 21 03:29:22 to bring it into alignment with the newer RDEPENDS_${PN} technique. Feb 21 03:46:10 03Mike Westerhof  07org.openembedded.dev * r1f2ef66b15 10openembedded.git/recipes/udev/udev_118.bb: Feb 21 03:46:10 SlugOS: udev - add SlugOS-specific override to drop the dependency on depmod Feb 21 03:46:10 (SlugOS will declare its own dependency so that it can optionally use busybox) Feb 21 03:46:40 Really should upgrade udev; we're WAY behind on that... but that'll have to wait for another day. Feb 21 04:21:06 Tartarus: yes I maintain micro Feb 21 04:35:40 khem, micro.conf shows one Martin Lund as the maintainer. Importing micro/micro-uclibc appears to be his one and only commit. Feb 21 04:36:51 * grg wanders off into la-la land again Feb 21 04:41:48 * mwester checks his atlas for "La-la" Feb 21 04:46:21 la-la land is my state of mind when i look at opkg code Feb 21 04:46:36 ow! it hurts! Feb 21 05:35:13 03Martin Jansa  07master * r2ced18eac5 10openembedded.git/recipes/ (3 files in 3 dirs): Feb 21 05:35:13 aceofpenguins, mokomaze, numptyphysics: bump PR to rebuild with right libpng in RDEPENDS after a44300e54ce49be9ffd34eecabdfaae4faf9c143 Feb 21 05:35:13 Signed-off-by: Martin Jansa Feb 21 06:27:23 gm Feb 21 06:41:27 khem is there by any chance something you can say about http://article.gmane.org/gmane.comp.handhelds.openembedded/42926 ? Seems gcc 4.1.2 libtool adds some dependency_libs that newer gcc's don't Feb 21 07:39:09 good morning Feb 21 07:54:29 mckoan: morning Feb 21 07:55:00 JaMa: hi Feb 21 07:55:05 mckoan: thanks for ACKing patches, can you use proper format for ACK so patchwork will pick it automatically? Feb 21 07:55:57 JaMa: what is proper format for ACK ? Feb 21 07:56:54 Acked-by: real-name Feb 21 07:57:04 like everybody else does :) Feb 21 07:57:28 JaMa: which patch are you referring to? Feb 21 07:58:21 don't remember, just noticed it on 2 patches from last 5 days Feb 21 07:58:23 JaMa: what is your email? Feb 21 07:58:36 those weren't my patches.. Feb 21 07:58:43 JaMa: I acked 2 patches from Tom Rini, properly acked Feb 21 07:59:04 JaMa: commit 87a92185e87fa62f4595db6e125ef100368772a7 for example Feb 21 07:59:47 then maybe I overlooked it (seen only "ACK") while reading >800 e-mails after returning from skiing Feb 21 07:59:53 then sorry Feb 21 08:00:00 JaMa: yw :-) Feb 21 08:07:42 mckoan: http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-February/030112.html http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-February/030113.html Feb 21 08:09:24 mckoan: or 667 and 669 in patchwork Feb 21 08:18:32 JaMa: ok, apologize Feb 21 08:46:09 JaMa: how was the skiing Feb 21 08:49:06 khem: great Feb 21 08:50:46 where did u go to Feb 21 08:53:02 Austria, Saalbach Hinterglemm Feb 21 08:53:23 cool Feb 21 08:54:40 * JaMa apologizes for pushing glib-2.0_2.28.0 too soon Feb 21 08:55:24 03Koen Kooi  07org.openembedded.dev * r78be426591 10openembedded.git/recipes/avahi/avahi.inc: Feb 21 08:55:24 avahi: add libcap2 to DEPENDS Feb 21 08:55:24 avahi will pick it up already if present: http://git.angstrom-distribution.org/cgi-bin/cgit.cgi/testlab/commit/?id=6384666b6ae4781c1d16818b4f0d02f5f9af7906 Feb 21 08:55:24 Signed-off-by: Koen Kooi Feb 21 08:56:20 the binutils issue u reported I wondered if it worked with binutils 2.20 Feb 21 08:57:51 that in webkit-efl? Feb 21 08:58:06 yes Feb 21 08:58:58 want to make sure if its a regression Feb 21 08:59:06 or not Feb 21 08:59:15 I guess so (but webkit-efl had ARM mode forced for some time so I haven't tried it with 2.21, but there was similar issue with 2.21 and latest boost on armv4t with thumb enabled) Feb 21 08:59:43 khem: resulting in http://git.openembedded.org/cgit.cgi/openembedded/commit/?id=b1a15f7dc12a3badf7aa3e3438a63ff5bc1be958 Feb 21 09:00:10 and all worked well with 2.20? Feb 21 09:00:55 boost was building fine with 2.20 IIRC Feb 21 09:02:05 hmm Feb 21 10:23:23 hi rp Feb 21 10:26:31 hi woglinde, rp, all Feb 21 10:26:41 he pb Feb 21 10:27:29 pb 2 years ago, as we where the one day in cherry histon you could have told us how near we where at the arm hq Feb 21 10:28:09 oh, heh. yeah, that place we had lunch is only a few hundred metres from arm. Feb 21 10:29:02 next time I will know it Feb 21 10:29:04 *g* Feb 21 10:29:46 hi all Feb 21 10:30:22 Hi everybody Feb 21 10:30:39 I'm currently trying to define my own distro with OE Feb 21 10:30:56 To see if just the pure base works, I execute bitbake task-base in my build dir Feb 21 10:31:08 hi bluelightning Feb 21 10:31:22 It prints some errors while parsing recipes regarding TARGET_LDFLAGS Feb 21 10:31:24 lacks look at the other distros Feb 21 10:31:32 and what the do in conf Feb 21 10:31:46 What more info concerning this shall I provide (config file on pastebin, perhaps)? Feb 21 10:32:05 lackS: the text of the errors would be a good place to start Feb 21 10:32:12 I followed your user manual.... Feb 21 10:32:49 Error messages look like: ERROR: EOL while scanning string literal (TARGET_LDFLAGS, line 1) while parsing [...]/openembedded/recipes/uclibc/uclibc-initial_git.bb Feb 21 10:32:59 hm we have something to distros in manual? Feb 21 10:33:30 lackS: closing " missing ? Feb 21 10:33:42 btw gm everyone Feb 21 10:33:46 woglide: http://docs.openembedded.org/usermanual/usermanual.html#commonuse_new_distro Feb 21 10:33:55 gm eFfem Feb 21 10:34:07 eFfem: I'll check Feb 21 10:36:12 eFfem_work: A missing closing " is not the problem Feb 21 10:36:30 I'm using vim with syntax highlighting, which makes it easy to determine that Feb 21 10:37:08 lackS: ah ok, guessed that based in the scannign string literal, guess without .conf file it is hard to say anything about it Feb 21 10:38:06 wait a sec Feb 21 10:39:11 My local.conf file just contains BBFILES (seems to be correct, since Bitbake scans about 7,500 files), DISTRO and MACHINE Feb 21 10:39:41 if other distro's build the problem is in your distro conf file Feb 21 10:39:48 I'm using some hardware which is not included into OE yet (Toradex Colibri PXA320), and I'm writing the machine file by myself, too. Feb 21 10:40:10 I'll check that -- I'll be back in some minutes. Feb 21 10:40:52 any suggestions for a distro/machine combination that should build relatively quick? Feb 21 10:41:05 lacks funny think is toradex used oe for some years and then dropped support for it Feb 21 10:41:31 I already asked their support and they unfortunately weren't that much of help Feb 21 10:42:02 Might be they come back for *nix support, since this is obviously a bit topic on embedded hardware Feb 21 10:42:12 But for now, they support WinCE only -- officially and inofficially Feb 21 10:42:27 s/bit/sig Feb 21 10:42:34 s/sig/big Feb 21 10:43:02 Monday morning. Learning to type again. ;-/ Feb 21 10:43:04 but anyway I would buy the tegra2 modul instead of ancients pxa320 Feb 21 10:43:36 woglinde, what about driver support? Feb 21 10:43:53 which driver? Feb 21 10:44:07 Tegra2 is kind of bleeding edge, and I wonder whether it is supported that well already by Linux Feb 21 10:44:29 ttp://android.git.kernel.org/?p=kernel/tegra.git;a=summary Feb 21 10:45:17 hm, 99EUR Feb 21 10:45:58 ynezz ;) Feb 21 10:46:06 woglinde, fine :-) Has a successful distro build with OE for that hardware already been reported? Feb 21 10:46:37 lacks I am owning an ac100 toshiba netbook Feb 21 10:46:52 on the weekend we got the kbd and touchbad driver working on .36 Feb 21 10:46:56 touchpad Feb 21 10:48:11 Sounds good, but I unfortunately won't have the time to code drivers for the tegra Feb 21 10:48:52 hi,android is highly non-standard Feb 21 10:49:12 the quantity of work needed depend on your target kernel Feb 21 10:49:31 on how much standard it is and on how much replacement standard stuff exist Feb 21 10:49:45 qualcomm chips has the alsa problem Feb 21 10:49:52 *have Feb 21 10:50:01 but they ran ubuntu on ac100 Feb 21 10:50:08 so maybe they already fixed everything Feb 21 10:50:26 alsa problem? Feb 21 10:50:59 only on qualcomm SOC Feb 21 10:51:12 k Feb 21 10:51:28 Thanks for the info Feb 21 10:51:34 basically android for qualcomm chips lack alsa but have another custom system instead, that use dsp directly Feb 21 10:51:45 basically customs ioctl Feb 21 10:51:48 btw, by problem seems to be distro-file related (surprise, surprise)... Feb 21 10:51:54 there are some alsa drivers but they don't work well Feb 21 10:52:05 never the less there is support for htcdream in openembedded Feb 21 10:52:13 but lots and lots of work were needed Feb 21 10:52:14 k Feb 21 10:52:26 even a different kenrel configuration were an issue Feb 21 10:52:36 Android will be another topic I will be facing during summer Feb 21 10:52:42 for instance without debugfs compiled in the kernel....no wifi Feb 21 10:53:03 so I had to look what option made wifi magically work Feb 21 10:53:04 etc... Feb 21 10:53:26 at the end 2 things are still problematic: Feb 21 10:53:38 *alsa (mplayer works fine, but not linphonec) Feb 21 10:54:07 Sounds like a lot of work Feb 21 10:54:08 *wifi PSM (I didn't succeed with .irq = MSM_GPIO_TO_INT(29) because it froze the phone) Feb 21 10:54:17 for my phone, yet Feb 21 10:54:24 for your device, I can't tell Feb 21 10:54:26 it depends Feb 21 10:54:30 GNUtoo what you mean is /conf/machine/ac100.conf for? Feb 21 10:54:42 ? Feb 21 10:55:07 -> < GNUtoo|laptop> but they ran ubuntu on ac100 Feb 21 10:55:32 yes I was just telling that there was ubuntu for ac100 Feb 21 10:55:45 and ubuntu expect standard kenrel<->userland interfaces Feb 21 10:56:17 ? Feb 21 10:57:28 ubuntu on ac100 lacks lot of things Feb 21 10:57:46 but even with fbdev xserver machine is fast Feb 21 10:58:14 hrw ? Feb 21 10:58:29 hrw ah right you are in cambridge with ogra Feb 21 10:58:42 there isnt lacking so much Feb 21 10:59:20 as I since weekend kbd and touchpad nearly works on .36 Feb 21 10:59:24 +said Feb 21 10:59:26 ;) Feb 21 10:59:42 alsa works now if I remember well Feb 21 10:59:55 only missing is than the xorg driver Feb 21 10:59:59 for newer kernel Feb 21 11:15:50 * ajb_oe wonders why packages insist on overiding libc Feb 21 11:16:07 ajb_oe: what do you mean? Feb 21 11:16:52 pb_: lmsensors-dev wants to install usr/include/linux/i2c-dev.h overiding libc Feb 21 11:17:12 should I just move them to usr/include/lm-sensors/linux/i2c-dev.h? Feb 21 11:17:36 oh, that's a bit weird. I don't think lm-sensors should be shipping that file at all really. Feb 21 11:17:49 indeed Feb 21 11:18:05 I doubt it would break anything, they look broadly the same Feb 21 11:18:07 I guess it might be there as some hangover from old versions of libc which didn't include i2c-dev.h, though it has been in there for years if not decades. Feb 21 11:18:18 I would just remove it from lmsensors-dev altogether Feb 21 11:18:29 so just rm the file in do_install()? Feb 21 11:18:45 yeah Feb 21 11:19:07 * ajb_oe notes the ml mentioned a similer issue with i2c-tools a while back Feb 21 11:31:15 re Feb 21 12:04:45 hi, I'll send a mail to the ml but I've this problem: Feb 21 12:05:08 /usr/lib64/libgnutls.so.26: undefined reference to `gcry_cipher_setkey@GCRYPT_1.2' Feb 21 12:05:15 the solution is the following: Feb 21 12:05:53 + deps.append("libgcrypt" + native) Feb 21 12:05:57 in curl-common.inc Feb 21 12:06:02 I'll explain the background Feb 21 12:06:12 basically libc-headers fail to compile Feb 21 12:06:16 *configure Feb 21 12:06:34 oops sorry Feb 21 12:06:39 It's all confused Feb 21 12:06:47 I'll write a mail with the proper patch Feb 21 12:07:28 lol Feb 21 12:07:54 the problem is that curl-native looks for features of the system it can use Feb 21 12:08:00 yes Feb 21 12:08:07 and it finds libcrypto-dev Feb 21 12:08:12 and enables ldap support Feb 21 12:08:19 so the simple patch is Feb 21 12:08:28 but is libgcrypt a feature? Feb 21 12:08:29 disable ldpa-support or libcrypto via cmdline Feb 21 12:08:33 dont know Feb 21 12:08:35 should be Feb 21 12:08:38 otherwise they suck Feb 21 12:08:46 I' Feb 21 12:08:46 and we have to fix the configure.ac Feb 21 12:08:49 because here's the more in depth error Feb 21 12:09:08 btw. that was mentioned now twice on the ml Feb 21 12:09:16 and I cannt do all the dirty jobs Feb 21 12:09:45 Oops. I'm trying to learn how to use syslog-ng here. What decides how large the logfiles can get? Can't find any configfile in /etc. Feb 21 12:09:46 ok Feb 21 12:09:53 curl-native: do_configure -> checking for send... no Feb 21 12:10:47 anr78 you can use rsyslog Feb 21 12:10:56 thats what the most distros uses now Feb 21 12:11:44 I'm http://www.pastie.org/1589070 Feb 21 12:11:50 so I'll look in the mailing list Feb 21 12:11:56 I read it Feb 21 12:12:09 but right now I'm a bit backward in the messages Feb 21 12:12:43 woglinde: ok. I'm using Angstrom with task-proper-tools added, and syslog-ng seems to be the default. just add the rsyslog recipe and all will be well? Feb 21 12:13:01 hm Feb 21 12:13:09 maybee you can ovveride it Feb 21 12:13:15 or use syslog-ng at all Feb 21 12:14:09 #define HAVE_LDAP_SSL 1 Feb 21 12:14:17 thats the root of evil Feb 21 12:14:18 *g* Feb 21 12:14:40 but anyway I dont think we need ldap in curl-native Feb 21 12:18:44 I'll try adding rsyslog and see if it magically takes over for syslog-ng Feb 21 12:19:11 no Feb 21 12:22:32 woglinde: was that no for me, or ldap/curl-native related? Feb 21 12:23:20 it was for you+# Feb 21 12:23:28 it will not magicly overwritten Feb 21 12:23:40 you will end up with syslog-ng and rsylog installed Feb 21 12:24:32 woglinde, what was the tread's name? maybe it's an old thread? Feb 21 12:24:52 gnutoo holger wrote it the last days Feb 21 12:24:59 and some wrote it before Feb 21 12:25:11 there is no real discussion thread for it Feb 21 12:25:11 ok Feb 21 12:25:23 if you dont have ldap-dev and openssl-dev installed Feb 21 12:25:29 you wont see the error Feb 21 12:25:50 I think we have this problem with serval other autotools -native recipes too Feb 21 12:38:49 any idea for | /home/gnutoo/embedded/oe/oetmps/eee701/sysroots/i686-angstrom-linux/usr/include/asm/sigcontext.h:28: error: expected specifier-qualifier-list before '__u64' in libpcap Feb 21 12:39:36 __u64 xstate_bv; is the line Feb 21 12:40:20 hm seems there is not typedef for __u64 Feb 21 12:40:42 hm and I dont think asm/sigcontext.h should be included directly Feb 21 12:40:54 damn pcap devlopers Feb 21 12:41:03 or thex fixed it Feb 21 12:41:10 and we have an to old versiob Feb 21 12:41:28 check whats the debian package has as patches Feb 21 12:41:39 * woglinde mutters Feb 21 12:41:47 they should all have builders Feb 21 12:41:54 autobuilders Feb 21 12:42:02 they dont have to test oe Feb 21 12:50:32 hmmm Feb 21 12:50:35 how should I fix Feb 21 12:50:41 I include the right headers? Feb 21 12:51:03 #include should normally have that Feb 21 12:51:11 no? Feb 21 12:51:14 the __u64 Feb 21 12:51:21 * GNUtoo|laptop looks Feb 21 12:52:21 http://lxr.free-electrons.com/source/include/asm-generic/int-l64.h?a=arm#L29 Feb 21 12:52:25 but that's kernel stuff Feb 21 12:54:41 http://lxr.free-electrons.com/source/include/linux/types.h?a=arm Feb 21 12:55:08 typedef __u64 u_int64_t; Feb 21 12:55:27 I'll look if types.h is present Feb 21 12:57:48 yeah thats why thats the problem Feb 21 12:58:05 but I think libpcap cannt do it right without kernel-def Feb 21 12:58:07 's Feb 21 12:58:15 thats the whole problem Feb 21 13:03:39 is it expected that ASSUME_PROVIDED is interpreted as a list of regex substrings? E.g. when it contains "make-native", then "automake-native" won't be built Feb 21 13:03:55 no? Feb 21 13:04:21 gnutoo@gnutoo-laptop:~/embedded/oe/oetmps/eee701/sysroots$ ls => i686-angstrom-linux x86_64-linux Feb 21 13:04:42 should I make it depend on kernel headers? Feb 21 13:05:23 gnutoo would be an idea yes Feb 21 13:05:27 ok Feb 21 13:05:29 just to be sure Feb 21 13:05:45 kernel headers are staged in the device specific part of staging/sysroot ? Feb 21 13:05:46 good you spot it *g* Feb 21 13:05:53 yes Feb 21 13:05:56 ok Feb 21 13:05:58 thanks a lot Feb 21 13:12:53 I've added that: Feb 21 13:12:54 DEPENDS = "linux-libc-headers Feb 21 13:13:01 and still the same stuff happens Feb 21 13:13:52 hms Feb 21 13:13:53 grep Feb 21 13:14:04 why it dont includes the type-defs Feb 21 13:26:49 ok Feb 21 13:37:21 linux/types.h in sysroots/i686-angstrom-linux/usr/include/ contains some typedef Feb 21 13:37:32 typedef __u64 __bitwise __le64; Feb 21 13:38:14 kernel headers are staged in the device specific part of staging/sysroot ? Feb 21 13:38:18 yes Feb 21 13:38:19 hmmm Feb 21 13:38:23 they are still not there Feb 21 13:38:37 but are in i686-angstrom-linux Feb 21 13:38:40 ah Feb 21 13:38:41 yeah Feb 21 13:38:43 the machine ie eee701 Feb 21 13:39:00 seems we have still failures when targer == host Feb 21 13:39:10 ah ok Feb 21 13:39:20 host: Feb 21 13:39:28 oops Feb 21 13:39:34 host: x86_64-linux Feb 21 13:39:46 target: i686-angstrom-linux Feb 21 13:42:01 03Dongxiao Xu  07master * r1a32c01490 10bitbake.git/lib/bb/ (build.py cache.py runqueue.py siggen.py): (log message trimmed) Feb 21 13:42:01 Introduce stamp-extra-info task flag into stamp filenames Feb 21 13:42:01 For certain tasks, we need additional information in build stamp file Feb 21 13:42:01 other than the task name and file name. stamp-extra-info is introduced as Feb 21 13:42:01 a task flag which is appended to the stamp file name. Feb 21 13:42:01 [Code simplifcations/tweaks from Richard] Feb 21 13:42:02 (From Poky rev: 80cf3e405c923ed95dd09936739e816dbdd92299) Feb 21 13:42:07 03Richard Purdie  07master * rf08da7b435 10bitbake.git/lib/bb/utils.py: Feb 21 13:42:07 bitbake/utils.py: Teach unlockfile about shared mode lockfiles Feb 21 13:42:07 (From Poky rev: fd88588df029604689db9b0e30c55aad68392a5d) Feb 21 13:42:07 Signed-off-by: Richard Purdie Feb 21 13:42:07 what should I do then? Feb 21 13:42:08 03Richard Purdie  07master * rebea0799d8 10bitbake.git/lib/bb/utils.py: Feb 21 13:42:08 bitbake/utils.py: Add option of holding shared lockfiles Feb 21 13:42:08 (From Poky rev: 7857834691868b7f48f732ee78d8770f5473ff68) Feb 21 13:42:08 Signed-off-by: Richard Purdie Feb 21 13:42:09 03Richard Purdie  07master * r3c6caab8bc 10bitbake.git/lib/bb/build.py: Feb 21 13:42:09 bitbake/build.py: Rename message field to avoid DeprecationWarning: BaseException.message warning Feb 21 13:42:10 (From Poky rev: 0388f0758f84c7a8ec975d9be7e384cb65d9af4c) Feb 21 13:42:10 Signed-off-by: Richard Purdie Feb 21 13:42:11 03Richard Purdie  07master * r935b98818f 10bitbake.git/lib/bb/utils.py: Feb 21 13:42:16 bugreport to the mailing list? Feb 21 13:42:17 rais it to the ml Feb 21 13:42:22 same key existing in two sides. This commit provides better output by figuring out differences Feb 21 13:42:22 from addition, removal or hash change. Feb 21 13:42:22 ok Feb 21 13:42:23 (From Poky rev: bcb18738361ee2394bc266911e9e75d5bf1c10cd) Feb 21 13:42:24 03Richard Purdie  07master * r43188e8453 10bitbake.git/lib/bb/process.py: Feb 21 13:42:24 process.py: Avoid deprecation warning Feb 21 13:42:24 bitbake/lib/bb/process.py:15: DeprecationWarning: BaseException.message has been deprecated as of Python 2.6 Feb 21 13:42:26 (From Poky rev: 6cb8fd6def4912e4aa76330649ba42a9ed2694fd) Feb 21 13:42:26 Signed-off-by: Richard Purdie Feb 21 13:42:26 03Richard Purdie  07master * r2cc4a1a867 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:27 yeah hope khem fix it quikcly for you Feb 21 13:42:27 bitbake/fetch2: Fix message about md5 checksums when sha256 is incorrect Feb 21 13:42:27 (From Poky rev: 4a226460d5e158e8adbf74a68717cb9a923e8009) Feb 21 13:42:28 Signed-off-by: Richard Purdie Feb 21 13:42:29 03Richard Purdie  07master * r6f13c23f20 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:29 bitbake/fetch2: Pass over malformatted (empty) mirror url lines Feb 21 13:42:29 (From Poky rev: 20ab1cb0ca10b0c7724f2154fe5b54c939b8a887) Feb 21 13:42:30 Signed-off-by: Richard Purdie Feb 21 13:42:30 03Richard Purdie  07master * r3b72de69f1 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:32 bitbake/fetch2: Correctly handle git mirror tarball fetching Feb 21 13:42:32 (From Poky rev: 7f30131faaa5e3fdb1191c9da95c5683cefafeb5) Feb 21 13:42:32 Signed-off-by: Richard Purdie Feb 21 13:42:33 btw I didn't find the first thread we were talking about? Feb 21 13:42:35 s/?// Feb 21 13:42:50 Signed-off-by: Richard Purdie Feb 21 13:42:50 03Richard Purdie  07master * ra564d47060 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:50 bitbake/fetch2: Add some debug output so its clear when PREMIRRORS, upstream and MIRRORS are being used Feb 21 13:42:50 (From Poky rev: a72d49478e57b05b99cb1b3beec62a9e79a24e7a) Feb 21 13:42:50 Signed-off-by: Richard Purdie Feb 21 13:42:54 03Mark Hatle  07master * rab05499d1c 10bitbake.git/lib/bb/ (fetch2/__init__.py utils.py): (log message trimmed) Feb 21 13:42:54 fetch2: unpack revision Feb 21 13:42:54 Revise the unpack function to have a way to disable the unpack. This is Feb 21 13:42:54 based on the work from "Andreas Oberritter ", see Feb 21 13:42:54 http://cgit.openembedded.net/cgit.cgi/openembedded/commit/?id=2bdfe8519eda8067845019a699acdf19a21ba380 Feb 21 13:42:54 that has nothing to do with the ldap stuff Feb 21 13:42:54 In addition, the to_boolean function comes from the work of Feb 21 13:42:56 "Chris Larson ", see Feb 21 13:42:57 03Richard Purdie  07master * r6348657c5b 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:57 bitbake/fetch2: Revert part of the unpack change until the issues with it are resolved Feb 21 13:42:57 (From Poky rev: b9a97e07b48b20b00feff5b8c7196c3a3b7762b6) Feb 21 13:42:57 Signed-off-by: Richard Purdie Feb 21 13:42:59 03Richard Purdie  07master * rff3ac7274f 10bitbake.git/lib/bb/fetch2/__init__.py: Feb 21 13:42:59 bitbake/fecth2: Ensure BB_FETCH_PREMIRRORONLY being set as false is handled correctly Feb 21 13:42:59 (From Poky rev: 3ef2d39dbf8a573acfbf633c1d5d7cb3cee18788) Feb 21 13:42:59 Signed-off-by: Richard Purdie Feb 21 13:43:02 03Richard Purdie  07master * r850e194f8c 10bitbake.git/lib/bb/ (data.py siggen.py): (log message trimmed) Feb 21 13:43:02 bitbake/siggen.py: Fix whitelisted variable handling Feb 21 13:43:23 woglinde, curl-native stuff: I didn't find the thread, should I look more or do you remember some keywords Feb 21 13:43:44 *hrms* Feb 21 13:43:51 you made me a bit angry Feb 21 13:43:55 ok Feb 21 13:43:59 I'll look better Feb 21 13:44:00 then Feb 21 13:44:03 just disable the fucking ldap Feb 21 13:44:06 and thats it Feb 21 13:44:16 yes but that's a feature Feb 21 13:44:30 ? Feb 21 13:44:35 disbale it Feb 21 13:44:39 we dont need it Feb 21 13:44:46 ahhh Feb 21 13:44:51 #define HAVE_LDAP_SSL 1 Feb 21 13:44:54 thats the root of evil Feb 21 13:44:58 yes Feb 21 13:45:00 that was for me Feb 21 13:45:10 I tought it was for the other person you were talking to Feb 21 13:45:13 so there is some code in configure.ac which checks about ldap Feb 21 13:45:30 but I will look it up myself Feb 21 13:45:39 if you dont make it now Feb 21 13:45:49 * woglinde calmes down Feb 21 13:45:55 ok I've an exam tomorrow Feb 21 13:46:16 so feel free to do it Feb 21 13:46:21 what the hell than you are doing here Feb 21 13:46:25 start learning Feb 21 13:46:28 lol Feb 21 13:46:39 basically I've always a compilation that is running Feb 21 13:46:46 whatever I'm doing Feb 21 13:47:01 and you are wasting the time Feb 21 13:47:23 I'll go back to the revisions Feb 21 13:49:24 hm why is my bitbake parsing slow to 2 mins again Feb 21 13:51:22 ps: I only went here when the food was cooking and right after eating Feb 21 14:05:27 hrms what was the STASH variable now again Feb 21 14:12:56 hrms okay Feb 21 14:36:22 ja bitte wenn du schon dabei bist :) Feb 21 14:36:25 oop swrong window Feb 21 14:36:31 ...as always ;) Feb 21 14:37:38 haha Feb 21 14:38:03 :) Feb 21 14:38:20 bitchx, split screen mode in terminal -> I always forget to switch channels Feb 21 14:49:27 Tartarus, did you ever make any conclusions about system-tools-backends failing? Feb 21 15:00:41 Can you jog my memory a little bit more? Feb 21 15:08:00 hmm Feb 21 15:08:05 let me see Feb 21 15:08:44 ah, you replied to Gary Thomas' email Feb 21 15:08:55 regarding a testing release build fail Feb 21 15:09:01 likely he was building gnome Feb 21 15:09:16 Ah yes Feb 21 15:09:29 I hits me on a fedora machine Feb 21 15:09:30 So, PARALLEL_MAKE made it clearer what the problem was Feb 21 15:09:50 normally, I comment it out Feb 21 15:09:54 and deal with it Feb 21 15:10:03 just curious if you had seen a way forward Feb 21 15:11:56 03Leon Woestenberg  07master * r9d4875fd72 10openembedded.git/ (15 files in 3 dirs): ronetix-pm9g45: Added new machine, with kernel 2.6.30. Feb 21 15:18:18 heh, can't find his log right now Feb 21 15:40:55 no hurry Feb 21 15:41:06 I need to get interested in the problem again Feb 21 15:44:27 [6;5~[6;5~heh Feb 21 16:24:57 03Tom Rini  07org.openembedded.dev * r3195d211fa 10openembedded.git/classes/kernel-arch.bbclass: Feb 21 16:24:57 kernel-arch.bbclass: Add powerpc64 mappings Feb 21 16:24:57 A future TODO is to see if there's anything left that must Feb 21 16:24:57 build with ARCH=ppc in the kernel that can't be upgraded to a Feb 21 16:24:57 newer kernel with ARCH=powerpc. Feb 21 16:24:57 Signed-off-by: Tom Rini Feb 21 16:25:07 03Tom Rini  07org.openembedded.dev * ra81045fc1e 10openembedded.git/recipes/openssl/openssl.inc: Feb 21 16:25:07 openssl.inc: Add linux-powerpc64 to the linux-powerpc case Feb 21 16:25:07 Signed-off-by: Tom Rini Feb 21 16:25:08 03Tom Rini  07org.openembedded.dev * r6574dbdc24 10openembedded.git/classes/insane.bbclass: Feb 21 16:25:08 insane.bbclass: Add powerpc64 entries. Feb 21 16:25:08 Signed-off-by: Tom Rini Feb 21 17:17:24 * mwester attempts to build micro to see if it really does whack up the directory structure for staging... Feb 21 17:18:11 I guess I could just to a SlugOS-specific patch to Squid... but that would leave it broken for everyone else. Feb 21 17:18:15 * mwester shrugs Feb 21 17:21:24 it does Feb 21 17:21:41 but much doesnt build for micro Feb 21 17:22:13 as khem said tho, theres a varable to get it right in both cases already Feb 21 17:29:01 Tartarus, the problem is that I cannot apply that variable. Feb 21 17:29:31 either option? Feb 21 17:29:45 The scripts in question refer to a plethora of locations, including /usr/lib /usr/include but they also just reference /usr (with no subdirs)... Feb 21 17:30:13 Well, given that micro is generally broken, I'd prefer to just add squid to the pile Feb 21 17:30:18 Rather than have squid broken for most Feb 21 17:30:26 Agreed. Feb 21 17:31:28 evening Feb 21 17:32:15 But I think a discussion with khem (who is, I think, the maintainer of micro) would be worthwhile, so I will build micro to understand what it does. If there is some way to make it "better", I'll do that, but I don't think we can let referecnes to /usr remain -- but in the case of micro, I don't know with what to replace a bare reference to /usr Feb 21 17:32:20 So, we'll see... Feb 21 17:32:34 * mwester ducks out to grab some lunch while micro builds Feb 21 17:33:20 Looking at the receipt for generating .ipk files there seems to be a lot of extra fields for the ipgk control file. Nice. But those fields does not seem to be included in the ipk files. Feb 21 17:33:50 Are the extra fields dropped from the ipk format? Feb 21 17:34:12 example? Feb 21 17:34:50 Fiels like Homepage, OE Feb 21 17:35:19 And "Source:" Feb 21 17:40:38 gm Feb 21 17:40:49 hey likewise Feb 21 17:43:45 So, yes, thsoe don't make it to the ipk itself as that's a format that's supposed to have as little as possible in it Feb 21 17:43:52 i forget if they end up residing elsewhere however Feb 21 17:44:13 hey Feb 21 17:45:38 Would have been nice to have some extra info in the ipk about the original source. Feb 21 17:46:11 mwester: my advice for micro is to strip it way the heck down and make it work Feb 21 17:46:36 mwester: I started with minimal and was shocked it was building stuff like gtk+ which I sure didn't expect for something called minimal Feb 21 17:46:58 could be a fantastic starting point for people like me walking up attempting to bringup a board Feb 21 17:47:00 demigod2k: that seems a recent regression, I tracked it down to avahi->python->python-gtk Feb 21 17:47:54 likewise: for some reason it was deciding to do libfso and some smartphone related things too? not really minimal Feb 21 17:48:37 I was looking more for a starting point with busybox and that's about it Feb 21 17:49:23 for that matter python seems like a bit much for something minimal Feb 21 17:50:52 * mwester examines the micro sysroot, and wonders why? Feb 21 17:51:23 I mean, really, what does it cost to create the /usr directory level? Feb 21 17:51:29 A few inodes? Feb 21 17:52:28 demigod2k: I know, I am all for a truely minimal image. It seems to bloat every month. Feb 21 17:52:46 But on a constructive note, might we not fix the staging problem much easier by adding a symlink from usr to the sysroot? That way sysroot/usr/include and sysroot/include could resolve to the same path. Feb 21 17:53:00 likewise: it's something I'll be on once I manage to get over the learning curve, assuming we keep with this path at work Feb 21 17:53:37 What am I missing with regard to this collapsed directory structure? Feb 21 17:53:38 likewise: I think the *.dot file is pretty helpful in figuring that out. I was unable to run graphviz (too big?) but I assume whole trees could be pruned once a person can see it all laid out Feb 21 17:54:18 demigod2k: there is a very useful tool here: https://github.com/scottellis/oe-deptools/blob/master/oey.py Feb 21 17:54:51 demigod2k: been doing this since 2005, when we selected OE :-) Feb 21 17:55:19 ya we've made everything with buildroot all along; started on the first product with a 1ghz + DDR2 socket Feb 21 17:55:25 mwester, but then run time is still going to have issues. But that might help building Feb 21 17:55:38 mwester, IMHO, we need to have a conversation about micro and what's valid and what's not.. Feb 21 17:55:54 Yeah, I suppose that's a valid concern -- one would rather find the issues at buidl time than at run time. Feb 21 17:56:01 demigod2k, Did I ask you about bitbake -u depexp -g target ? Feb 21 17:56:05 finally its time for us to step up and run the full versions of various software but it's been a steep approach. Feb 21 17:56:08 If not, with 1.12.0 I know it's in a release :) Feb 21 17:56:19 Tartarus: Nope or if you did I missed it. unfortunately I'm at home today -- we're in a snowpocalypse today Feb 21 17:56:47 12+ inches that I had to snowblow this morning. truck got stuck, pulled right back in and called off for today :) Feb 21 17:57:27 * mwester thinks demigod2k must live in Minnesota Feb 21 17:57:44 nope michigan. actually my workplace is open, but my neighborhood got buried badly Feb 21 17:58:05 Ah, ok. Just a bit of ice here in Chicago. Feb 21 17:58:18 likewise: is that tool linked anywhere on the wiki? looks interesting Feb 21 17:58:22 there is tech work in the midwest? Feb 21 17:58:36 demigod2k: no I just found it today when I saw GTK dragged into minimal.... Feb 21 17:58:46 Crofton|work, not nearly enough... which is why I practically live in airports and hotels. Feb 21 17:59:02 But you travel a lot too, don't you? Feb 21 17:59:23 Crofton|work: in general tough to say. we're a staff of about 15 and supply automotive service tools. assembly by a contract manufacturer, all software and hardware design in house Feb 21 18:00:03 there are three of you here now, cbrake, mwester and you :) Feb 21 18:00:28 cbrake is somewhere in the midwest, too? Didn't know that. Feb 21 18:00:37 I'm skeptical since I see people doing "embedded electronics" and half the time it's a board with a PIC that a college student could do Feb 21 18:00:56 but there are some decently high tech companies around if I weren't working where I am Feb 21 18:02:52 it's also nice because a comparable salary goes a lot further in michigan than it does in california Feb 21 18:03:05 do we have a firewall distribution in oe, some sort , ipcop like or so ? Feb 21 18:03:40 * mwester noted on his trip to Detroit last week that one can buy a house in Detroit REALLY cheap anymore! Feb 21 18:03:54 is tzdata_2011a.bb out of data already? The build keeps failing to fetch the tarball Feb 21 18:04:24 ajb_oe: same problem here. get it from http://strijp.sidebranch.nl/ Feb 21 18:04:50 ajb_oe: I just put it there for a colleague :-) Feb 21 18:05:21 * ajb_oe grabs Feb 21 18:06:06 mwester: all except the high price of car insurance it's a pretty cool place too Feb 21 18:07:27 OE HEAD constantly fails on me on Ubuntu 10.04 LTS. Seems all host compiler related when building -native packages. Anyone else running into this? Feb 21 18:08:55 Tartarus: How does 2011.03 relate to master? Is it cherry-picked? Feb 21 18:10:05 Tartarus: regaring your comment... what would I want to pass for -u -g for example Feb 21 18:11:27 03Tom Rini  07org.openembedded.dev * rd74ffffd27 10openembedded.git/classes/insane.bbclass: Feb 21 18:11:27 insane.bbclass: Fix a thinko in powerpc64 support Feb 21 18:11:27 Signed-off-by: Tom Rini Feb 21 18:15:18 Tartarus: + elif re.match('(p|owerpc)64$', a): return 'powerpc' Feb 21 18:15:49 Tartarus: (p|power)pc64? Feb 21 18:15:56 mwester: I'm in Akron/Clev OH area Feb 21 18:20:15 I guess to match ppc64? Feb 21 18:28:23 likewise: do I need to do anything special other than putting those files in the download dir, bitbake keeps trying to re-download them from the broken mirrors Feb 21 18:29:28 You need the matching .md5 file Feb 21 18:32:32 mwester: ahh yes, missed that - thanks Feb 21 18:46:16 re Feb 21 18:46:30 wb Feb 21 18:50:49 * mwester reworks the squid patch to accomodate micro, and feels too queasy to complete the hackery. Feb 21 19:15:30 Ok, much easier way -- kinda punts, but it will work for the use-cases in squid at least. Feb 21 19:26:02 hi, any hints about this: Feb 21 19:26:06 http://pastebin.com/8hd2qeN3 ? Feb 21 19:26:27 I mean the "fatal: Not a valid object name 1" message, where does it come from? Feb 21 19:26:29 yeah Feb 21 19:26:38 SRCREV isnt donde right Feb 21 19:27:07 Tartarus: ah, so I have to specify the revision? Feb 21 19:27:09 need a tag or hash not just branch in uri Feb 21 19:27:22 Tartarus: I can't just say 'pull the HEAD of this branch'? Feb 21 19:27:36 no Feb 21 19:27:45 not for comitting anyhow Feb 21 19:28:08 * Tartarus is feeding the baby, 1 handed typing... Feb 21 19:28:14 Tartarus: :) Feb 21 19:28:25 filip: http://gitorious.digitalstrom.org/dss-oe/dss-oe/blobs/master/targets/common/recipes/dss/dss_git.bb Feb 21 19:29:08 Jin^eLD: would it be accepted as a commit to OE? Feb 21 19:29:21 the above recipe always builds master head, however only if new commits are available Feb 21 19:29:28 filip: no idea, I'm not a core dev :) Feb 21 19:29:43 but I think I "stole" this idea from some other recipe that was in OE Feb 21 19:29:52 yeah, i dont think so Feb 21 19:29:58 yeah, linux-jlime-ben-nanonote_2.6.36.bb has this Feb 21 19:30:03 Tartarus: is this a bad thing? Feb 21 19:30:12 Tartarus: even though there ARE recipes with AUTOREV? Feb 21 19:30:16 I think there were some docs on AUTOREV Feb 21 19:30:17 yeah, no pushing autorev Feb 21 19:30:23 okay, I'll behave Feb 21 19:30:26 defaults need to be set Feb 21 19:34:13 (stuff in oe that is autorev is a bug that should get fixed, btw) Feb 21 19:39:04 cbrake, ping? Feb 21 19:39:58 ka6sox: hey Tom Feb 21 19:40:39 hiya, can you mod the backupscript to not pull oe_stats for now? its offline and won't be back for a bit till its fixed(longer term thing) Feb 21 19:41:44 pulling 16million records that dont' change every day is kinda silly. Feb 21 19:42:59 ka6sox: thinking what oe_stats is ... Feb 21 19:43:31 ka6sox: ahh, the build results Feb 21 19:44:41 ka6sox: so that is the oe_tinderbox database Feb 21 19:44:48 ka6sox: done Feb 21 19:45:57 ka6sox: wow, even bugzilla is 700MB for sql dump!!! Feb 21 19:49:13 yup Feb 21 19:49:33 thats what I mean...I back those up onsite & offsite daily Feb 21 19:49:55 ka6sox: ok, so if you're backing them up, I'll not bother Feb 21 19:50:19 ka6sox: so just to confirm, I'll not back up bugzilla either Feb 21 19:51:36 cbrake, anything that uses the common mysql server doesnt' ahve to be backed up...so mediawiki,tinderbox, bugzilla etc are covered. Feb 21 19:53:28 bbl....popsicle time. Feb 21 20:01:14 heh.. I see nobody cares about build stats server Feb 21 20:02:46 Jay7, or you've captured the requirements of the biggest crashers of the old build stats server :) Feb 21 20:03:06 if I change a recipe does anything need to be done to reread it? (so it doesn't go off any caching issue) Feb 21 20:03:22 Tartarus: yeah.. :) Feb 21 20:04:00 duffolonious: bump PR Feb 21 20:04:04 I thought the requirements as described were very much a super-set of what I wish for such a service. Feb 21 20:04:18 In other words, i would be happy with a smaller set. ;) Feb 21 20:04:36 mwester: our req is client and reports on server mostly :) Feb 21 20:04:57 i.e. working oe_stats :) Feb 21 20:05:24 but huge builders like Tartarus needs more :) Feb 21 20:06:31 well.. linkedin profile is updated.. what to break else?.. Feb 21 20:06:43 ah, yes.. kexecboot :) Feb 21 20:19:24 ericben: around ? Feb 21 20:26:23 Jin^eLD: ahh, I see - thanks. Feb 21 20:26:55 'PR = "${INC_PR}.1"' - how does the INC_PR work in this case? Feb 21 20:29:11 INC_PR comes from an included file Feb 21 20:29:19 look for include or require statements in the recipe Feb 21 20:29:38 the INC_PR variable will be defined there Feb 21 20:29:50 so if you change the included file you should bump inc_pr otherwise pr Feb 21 20:31:01 Jin^eLD: I see, so in this case I did it the wrong way, thanks again Feb 21 20:31:42 no problem :) Feb 21 21:15:57 -32 celsius.. Feb 21 21:16:06 a bit cold Feb 21 21:16:20 -6C and blizzard conditions here.. ;P Feb 21 21:33:45 hiho Feb 21 22:44:58 That's strange... Shouldn't 'git clone git.openembedded.org/openembedded' get you a load of files? Feb 21 22:47:44 It counts a lot of files ('remote: Counting objects: 420344, done') and compresses them ('remote: Compressing objects: 100% (120259/120259), done') but then nothing more happens :S Feb 21 22:51:50 And I didn't get any files either :s Feb 21 23:17:26 http://www.linux.com/news/embedded-mobile/mobile-linux/408379:the-mobile-linux-roundup-an-overview-for-developers Feb 21 23:25:37 03Mike Westerhof  07org.openembedded.dev * r04e52f9cf7 10openembedded.git/recipes/meta/slugos-packages.bb: Feb 21 23:25:37 SlugOS: slugos-packages - update to keep the package feeds building Feb 21 23:25:37 (demote squid until patch accepted; promote a few that are fixed; Feb 21 23:25:37 and clarify comments regarding a few that need yet to be fixed.) Feb 21 23:26:07 I am having trouble cloning the oe repo. 'git clone http://repo.or.cz/r/openembedded.git' shouldn't take 2 seconds and not produce any files, right? Feb 21 23:28:20 Nope Feb 21 23:29:17 Check the wiki for the correct url to use; the channel topic is out-of-date. Feb 21 23:32:27 It IS from the wiki ^^ Feb 21 23:32:38 From the "getting started" page Feb 21 23:34:58 Well, I just pushed a commit to something there, so clearly the primary is up and running. Feb 21 23:35:19 I'd use that instead of the repo.or.cz mirror. Feb 21 23:35:56 Well, the thing is 'git clone git://git.openembedded.org/openembedded' does not seem to work for me either Feb 21 23:36:51 This is the output (no files created): http://pastebin.com/B3XVw2LB Feb 21 23:38:33 Try checking out a branch Feb 21 23:38:52 Should be a link on the wiki to the git phrasebook - i've found that helpful. Feb 21 23:41:03 Can I check out a branch without having cloned first? Feb 21 23:41:03 mwester: not only that but a walkthrough would be nice. not everybody (including me) knows how to work it Feb 21 23:42:59 Should be in the wiki -- I thought it was. I'm barely competent in even git basics; I keep a copy of the git phrasebook about to help me figure out what to do. Feb 21 23:44:04 I think you have to have a clone before you can check out a branch Feb 21 23:44:50 git clone git://git.openembedded.org/openembedded openembedded Feb 21 23:44:54 git checkout -b org.openembedded.dev --track origin/org.openembedded.dev Feb 21 23:45:02 Something like that should do the trick. Feb 21 23:46:14 Yeah, I though so too, but git clone gives me no files :S Feb 21 23:46:32 Right. Feb 21 23:47:08 the clone gives you a copy of the repo -- that copy is (given the syntax I provided) in openembedded/.git (hidden dir) Feb 21 23:47:28 http://pastebin.com/sRbqvPa3 Feb 21 23:47:31 You can't work with the repo data directly; you need to get a copy of the correct versions -- use the checkout command to do so. Feb 21 23:47:39 ls -al shows me nothing Feb 21 23:47:44 Not even the openembedded dir Feb 21 23:49:10 The above 'git clone' command is supposed to at least leave _some_ files in the current directory? As it is now, I don't get anything new Feb 21 23:49:24 I expected to at least get the openembedded dir Feb 21 23:49:36 or a hidden .git dir Feb 21 23:49:44 but not even that :S Feb 21 23:50:03 Your git output looks truncated. Feb 21 23:50:32 It should go from Coutning to Compressing to Receiving. Where's the Receiving line? Feb 21 23:50:49 Well, that's all that is presented to me. I didn't get anything more and thus the problem Feb 21 23:51:54 Ok. The problem is that you have something going on preventing your git from actually pulling down the data. Don't know why you get no error or warnign message, but cutting and pasting the line from your pastebin has my Fedora 12 system downloading data right now. Feb 21 23:52:15 Okay, that's weird Feb 21 23:52:37 Permissions? Disk space too low? Wrong type of filesystem? Feb 21 23:53:25 52GB ought to be enough, new user in his home dir, ext4 fs Feb 21 23:53:47 Wrong git config? :S Feb 21 23:53:56 my wild guess check for a verbosity flag for debug output Feb 21 23:54:22 these same sort of things always drive me nuts Feb 21 23:54:27 I tried 'git clone -v ...' but it did not produce any more output Feb 21 23:55:28 http://pastebin.com/vwHVWGki Feb 21 23:56:25 This is a new system, using an old home folder. But the user running 'git clone' i brand new, so there shouldn't be any configuration errors Feb 21 23:57:06 common distro, or one of those "roll your own"? Feb 21 23:57:35 sabayon linux, gentoo derivative Feb 21 23:57:50 Used it with oe in the past Feb 21 23:58:59 I'll have to look into this more in the morning. It's past midnight here and I have to get up early tomorrow Feb 21 23:59:27 Maybe I'll install Ubuntu in VirtualBox just to get the files Feb 21 23:59:48 Good night, all =) Feb 22 00:20:09 Well this is annoying Feb 22 00:20:17 imgur doesn't like how big the matrix of builds I did is Feb 22 00:27:39 lock him in the basement. Feb 22 00:27:54 kicking imagemagick around now to convert to pmg Feb 22 00:27:56 png Feb 22 00:27:59 and crush, if needed Feb 22 00:28:01 er, sorry, no that's what you do with a disagreeable igor, not imgur. Feb 22 01:03:06 http://dl.dropbox.com/u/3284539/feb18matrix_withsimplifyflags.pdf for my weekend run Feb 22 01:03:22 With some micro's dropped out to get it to finish up Feb 22 01:07:31 I am easily amused at times. Just noticed a hash I need to reference ends in beeef (and yes, I know it's got one extra e) Feb 22 01:10:15 Just think of how much more fun we'd have if we had adopted a 5-bit/digit system instead of Hexadecimal... :D Feb 22 01:11:09 The pdf isn't showing much here. Feb 22 01:11:37 it's large (8mb) Feb 22 01:12:01 but does load here, heh Feb 22 01:12:27 Yes, it is -- 20 pages, which took a long time for acrobat to "prepare for reading" (whatever that means) -- but it's mostly blank. Feb 22 01:13:05 The first page has some text (upper right), the last page has the "Page generated" message, and a few in the middle have a page number -- but that's all that's rendered here. :( Feb 22 01:13:18 I'll try to save it and open it with a different reader. Feb 22 01:14:02 .k Feb 22 01:14:27 evince and chrome's built-in are happy Feb 22 01:14:30 here anyhow Feb 22 01:18:14 mwester, oh, can you ack pinning gvfs for slugos/ Feb 22 01:18:32 Ok, had to reset the color mappings for the pdf viewer for some reason. But it's all visibile now. :) Feb 22 01:19:53 Sure, i can ack that -- I haven't tested it yet, since I've been busy with a few other things that won't build that have messed up my tmpdir. Feb 22 01:20:05 k Feb 22 01:20:21 It's a non-behavior change for slugos Feb 22 01:20:34 You pin glib-2.0 already and gvfs 1.6.6 is default, until the end of the rest of the series I posted Feb 22 01:21:48 Ah, ok. Then that makes it real easy. :) Feb 22 01:24:00 native-sdk-image <--- is that something that each distro should strive to have working for their distro/machine? Feb 22 01:28:10 If you want on target device development, yes, otherwise, no. Feb 22 02:11:28 native-sdk-image built just fine. Now I have to figure out what it was that it built. :D Feb 22 02:12:58 ah, task-sdk-native on console image. :) So both apparently build fine fot eh NSLU2. I'm sure it won't boot, but it builds. Feb 22 02:16:17 Want me to add that to my matrix for *slug* and nslu2* ? Feb 22 02:19:46 I'm not sure if it will help anyone -- if you think so, then do so -- but it's not generally useful (although adding task-sdk-native as an installable meta-package for SlugOS might be useful and make a lot of users happier) Feb 22 02:21:00 BTW, if you have excess capacity, you can add qemuarm and sheevaplug machines for SlugOS -- those build. **** ENDING LOGGING AT Tue Feb 22 02:59:57 2011