**** BEGIN LOGGING AT Mon Jan 30 02:59:59 2012 Jan 30 07:19:11 gm Jan 30 08:02:34 good morning Jan 30 08:25:49 kergoth, elc? Jan 30 08:30:21 Russ, did I see that you might be going to ELC? Jan 30 08:30:27 yup Jan 30 08:30:33 k Jan 30 08:46:18 oh, people going to ELC :) Jan 30 08:46:38 * dm8tbr is not yet sure if he'll make it :/ Jan 30 10:16:13 morning all Jan 30 10:16:32 good morning Jan 30 10:17:00 mornin Jan 30 10:17:21 hi Jin^eLD, hi pb_ Jan 30 10:18:05 hi florian Jan 30 10:19:11 morning Jan 30 10:20:20 hi ka6sox Jan 30 10:21:14 hi all Jan 30 10:21:27 florian: no fosdem this year? Jan 30 10:21:43 mckoan: I will be there... Jan 30 10:22:19 florian: good to know, I didn't see you in the fodem2012 page in OE website Jan 30 10:22:53 mckoan: I didn't manage to add myself yet... maybe I should do it right now Jan 30 10:23:05 hi guys Jan 30 10:23:15 florian: see you there then ;-) Jan 30 10:23:52 oh, someone added me... very good Jan 30 10:26:04 no FOSDEM for me this year :-( Jan 30 10:26:19 I had hoped to go but it doesn't look like it's going to work. next year though! Jan 30 10:26:54 pb_: too bad :-( Jan 30 10:27:12 pb_: Who is guilty - company or family? :-) Jan 30 10:27:21 probably too early for someone to push something up so I can be sure that github replication is working Jan 30 10:30:26 florian: both, really. my company sent me to California the week before last and I don't think my family will put up with me going away again so soon. Jan 30 10:31:03 also I am feeling a bit impoverished at the moment because tomorrow is tax day. Jan 30 10:31:26 what a well known effect :-) Jan 30 10:34:59 git from oe's master will be slow this morning, might want to use github. Jan 30 10:35:10 I need to move 1TB of data around. Jan 30 10:36:57 yeah, bit of a shame but never mind. next year I will plan things better. Jan 30 12:29:01 denix: Hello Jan 30 12:29:10 denix: you sent me a pull request, right? Jan 30 12:29:29 denix: I just commented on it. Please take a look and fix it when you can. Jan 30 13:34:04 how can I prevent the fetch step from decompressing a .gz file? Jan 30 13:40:53 likewise_: just add ;unpack=0 to the SRC_URI entry Jan 30 13:46:37 bluelightning: thanks, much appreciated! Jan 30 13:46:41 np Jan 30 13:52:57 otavio, done, I've updated the pull request Jan 30 13:54:46 otavio, btw you're the one doing git://github.com/OSSystems/meta-mozilla.git right? Jan 30 13:55:02 I've a problem with firefox, it exits with illegal instruction Jan 30 13:55:04 denisATeukrea: yes Jan 30 13:55:08 gdb gives something like that: Jan 30 13:55:13 denisATeukrea: cool Jan 30 13:55:28 denisATeukrea: please post an issue with all information you have please. Jan 30 13:55:36 denisATeukrea: so we track it there Jan 30 13:55:43 ok Jan 30 13:55:50 denisATeukrea: btw, you have a fix for the issue already? Jan 30 13:55:56 0x2c050044 <+540>: ; instruction: 0xffffd3d4 Jan 30 13:56:09 is that normal in gdb? Jan 30 13:56:17 or is it the cause of the illegal instruction? Jan 30 13:56:59 also there is an issue with the toolchain and the imx 35 kernel Jan 30 13:57:04 I'll look at it later tough Jan 30 13:57:07 not now Jan 30 13:57:12 (for the toolchain issue) Jan 30 13:57:19 basically it doesn't boot : Jan 30 13:57:51 if I remember well the init is right tough Jan 30 13:58:00 since it printed uncompressing Linux ... done Jan 30 13:58:03 or something like that Jan 30 13:58:16 I'll check right now Jan 30 13:58:35 machine quick start board Jan 30 13:58:46 denisATeukrea: you might want to try to rebuild firefox with updated toolchain as IIRC we got some GCC fixes latelly Jan 30 13:58:59 I use angstrom-bleeding Jan 30 13:59:02 I'll try Jan 30 13:59:17 denisATeukrea: but in any case, report it as we can track it in github and others might help Jan 30 13:59:25 report but where? Jan 30 13:59:36 denisATeukrea: koen is using firefox afaik. He might also contribute with information about it Jan 30 13:59:43 ok Jan 30 13:59:46 denisATeukrea: in the repository; in github itself. Jan 30 13:59:52 ok Jan 30 14:00:07 also we should add meta-mozilla to http://www.openembedded.org/wiki/LayerIndex Jan 30 14:00:42 denisATeukrea: https://github.com/OSSystems/meta-mozilla/issues Jan 30 14:00:52 thanks Jan 30 14:00:56 I'll recompile it first Jan 30 14:01:04 denisATeukrea: I don't have wiki write rights; if you can, please add it and meta-chicken as well ;) Jan 30 14:01:15 meta-chicken? Jan 30 14:02:32 what's the goal of meta-chicken? Jan 30 14:02:46 what's Chicken toolchain Jan 30 14:02:50 denisATeukrea: it provides Chicken toolchain support Jan 30 14:03:04 denisATeukrea: it is a scheme based language Jan 30 14:03:11 ah ok Jan 30 14:03:18 denisATeukrea: http://www.call-cc.org/ Jan 30 14:03:20 denisATeukrea: we use it a lot here Jan 30 14:03:34 ok Jan 30 14:04:09 I just added meta-mozilla Jan 30 14:04:10 :) Jan 30 14:04:22 ok Jan 30 14:04:29 can add meta-chicken now as well Jan 30 14:04:34 ok thanks Jan 30 14:04:46 the problem is that I must remember my wiki password Jan 30 14:05:07 and my main laptop is busy and inacessible Jan 30 14:05:35 (it is transfering the content on different hdds) Jan 30 14:06:18 note that I got firefox working in SHR Jan 30 14:06:33 it just fails with angstrom-bleeding Jan 30 14:06:44 no idea if it works with the normal angstrom Jan 30 14:07:28 ok, meta-chicken added also Jan 30 14:07:36 bluelightning: thanks a lot Jan 30 14:07:42 denisATeukrea: interesting Jan 30 14:07:47 http://scap.linuxtogo.org/ has screenshots of it on gta04 Jan 30 14:12:07 ah no it blocked at starting kernel for the qsb kernel Jan 30 14:12:25 which is a very bad sign Jan 30 14:12:52 http://pastie.org/private/3fiu9h08ugxqb1itdztgq Jan 30 14:13:09 compiling with an old codesourcey made it boot Jan 30 14:15:06 also I tried to make work the 2.6.38 and I failed to succeed in time and abandoned but I did quite some work Jan 30 14:30:01 denisATeukrea: firefox javascript jit requires armv7; you can disable it by some about:config variable Jan 30 14:32:01 yes that's where it failed Jan 30 14:32:05 in the JIT Jan 30 14:32:22 but strangely the CPU is an armv7 Jan 30 14:32:28 but thanks for the tip Jan 30 14:32:42 because for SHR I'd like to try it on armv4 Jan 30 14:33:53 denisATeukrea: http://pastie.org/3282468 Jan 30 14:34:27 denisATeukrea: it's not armv7 only, but neon Jan 30 14:34:48 ok but I've neon I think on my board Jan 30 14:34:59 it's the quick start board from freescale Jan 30 14:35:02 imx53 Jan 30 14:35:30 Features : swp half thumb fastmult vfp edsp neon vfpv3 Jan 30 14:35:41 CPU architecture: 7 Jan 30 14:48:13 does somebody else see lockups with recent bitbake? e.g. when git url is not fetchable, bitbake hangs and can not be interrupted with ctrl-c Jan 30 14:53:57 there is apparently a hang bug somewhere, but no one has been able to nail it down sufficiently to be able to reproduce it Jan 30 14:58:09 kergoth: have you seen my message on #yocto today? Jan 30 14:59:05 haven't gotten caught up on all the scrollbacks yet. Jan 30 14:59:20 kergoth: fwiw: after reverting switch to futures d104f29871c04a5a36600a35b2568b49e5b21ca0 I haven't seen bitbake hanging in last couple of days Jan 30 14:59:23 I can reproduce it here 100%, but have no idea how to debug multithread python programs remotely. Jan 30 14:59:40 bitbake hangs in _process_worker() Jan 30 15:00:04 i have no idea what _process_worker even is Jan 30 15:00:26 it is in the backported futures class Jan 30 15:01:31 the python 2.7 multithread code looks very questionable... e.g. queues.py has Jan 30 15:01:32 if not self._poll(block and (deadline-time.time()) or 0.0): Jan 30 15:02:40 and? Jan 30 15:04:22 at any rate, the switch to futures was to fix a known hang bug that occurred at random when encountering a parse failure Jan 30 15:04:30 now it seems to be random as well, but not regarding parse failures Jan 30 15:05:33 * JaMa haven't had hangs before the switch, but ok Jan 30 15:05:55 then you didn't run into parse failures, or you were lucky Jan 30 15:05:59 i had reports about it every other day Jan 30 15:24:20 ensc|w, otavio it still fails Jan 30 15:24:34 I'll gdb it again Jan 30 15:25:30 *gdb+gdbserver Jan 30 15:25:39 because even with 1G of ram on target it OOMs Jan 30 15:32:12 * kergoth digs in and tries to find the hang via inspection Jan 30 15:32:18 kergoth: hang disappears when I handle ParseError exceptions in parse_next() (cooker.py) like other ones (executing self.shutdown(clean=False)) Jan 30 15:32:43 ah, interesting Jan 30 15:33:20 that doesn't seem unreasonable Jan 30 15:33:22 * kergoth looks Jan 30 15:38:08 kergoth: btw, it would be nice, when the bb.fatal() for ParseError exceptions mentions the name of recipe Jan 30 15:39:51 ensc|w: they almost always do. the string stored in the exception generally holds fn as well. Jan 30 15:39:54 e.g. Jan 30 15:39:56 134: raise ParseError("%s:%d: unparsed line: '%s'" % (fn, lineno, s)); Jan 30 15:40:15 the bb.fatal() just grabs its string form, which should include its first argument Jan 30 15:40:46 also: 81: raise ParseError("%s is not a BitBake file" % fn) Jan 30 15:42:44 kergoth: recipe name does not appeasr here: http://pastie.org/3282833 Jan 30 15:43:55 thats an expansion error, not a parse error Jan 30 15:43:57 afaict Jan 30 15:45:07 ah ok... I just looked at the first exception which was handled by this branch Jan 30 15:48:09 should definitely show the recipe or file for that case, though, indeed Jan 30 15:49:11 denisATeukrea: humm ... which machine you reproduce the firefox issue? Jan 30 15:51:10 ensc|w: mind sending a patch to add the missing shutdown() to the bitbake list? I doubt I'll have time to look into this at all today Jan 30 15:51:59 otavio, I replayes on github: imx53qsb with angstrom-bleeding Jan 30 15:53:15 kergoth: ok; but it can take some time Jan 30 16:16:33 hi Jan 30 16:17:44 khem: Hi , JaMa told me to ask my question to you. I am building shr-image from a fresh builhost using JaMa's chroot and I am having trouble while building eglibc. here is the issue Jan 30 16:17:45 http://pastebin.com/raw.php?i=bY89seDf Jan 30 16:18:11 khem: and here is my shlib.lds http://pastebin.com/EU2TuseE Jan 30 16:18:34 do you have a clue ? Jan 30 16:19:48 I looked BTW, and I don't see anything obvious Jan 30 16:19:52 sorry Jan 30 17:02:32 good morning Jan 30 17:03:05 GarthPS: paste complete build log of eglibc that snippet tells me nothinh Jan 30 17:03:21 khem: ok. Jan 30 17:03:37 khem: one moment I rebuild it Jan 30 17:07:15 khem: here it ishttp://pastebin.com/GzYG8uqB Jan 30 17:07:21 http://pastebin.com/GzYG8uqB Jan 30 17:09:50 GarthPS: what awk do you have on your host Jan 30 17:10:07 GarthPS: make sure its not mawk but GNU awk (gawk) Jan 30 17:10:24 khem: 4.0.0 Jan 30 17:10:48 khem: it is GNU Awk 4.0.0 Jan 30 17:15:35 GarthPS: what distro is on your build host Jan 30 17:16:01 khem: kubuntu 11.10 x64 but I am also using a chroot Jan 30 17:25:32 any reason oe.path.remove would fail? Jan 30 17:27:19 yes, numerous reasons Jan 30 17:29:55 GarthPS: thats strange do you get this error all the time ? Jan 30 17:30:31 khem: yes all the time I try . even if I clean before Jan 30 17:31:22 khem: what's strang is that shr-chroot is used by many people and GarthPS is only one (at least first one) to see it Jan 30 17:32:51 JaMa: I did a complet from scratch buildhost using this shr-chroot 3-4 days ago, perhaps no one did this too Jan 30 17:33:06 yet Jan 30 17:33:20 GarthPS: I have kubuntu 64bit 11.10 I never saw this problem Jan 30 17:35:10 GarthPS: anyway clean eglibc and rebuild it we have to see whats happening when shlib.lds is generated Jan 30 17:35:27 the logs you posted doesnt do it so hard to tell Jan 30 17:35:37 khem: ok Jan 30 17:36:03 pb_: PREMIRRORS should be hit before url in SRC_URI right ? Jan 30 17:36:53 yes, thats why it's "PRE" :) Jan 30 17:37:04 thats what documentation says as well :) Jan 30 17:51:38 anyone have an idea why a python script installed on the image has these two lilnes at the top Jan 30 17:51:40 root@usrp-e1xx:~# head /usr/bin/gnuradio-companion Jan 30 17:51:41 #!/home/balister/src/git/oe-core/build/tmp-angstrom_2010_x-eglibc/sysroots/x86_64-linux/usr/bin/python2.7 Jan 30 17:51:41 #!/usr/bin/env python Jan 30 17:52:00 khem: my log file is about 9.2 Mo.. I send you by mail ? Jan 30 17:52:23 hmm my mail will crap on me Jan 30 17:57:10 megaupload? Jan 30 18:00:13 Crofton: thats an error Jan 30 18:00:32 03Ulf Samuelsson  07master * r9389f986ac 10openembedded.git/recipes/python/ (python-2.6.6/07-linux3-regen-fix.patch python_2.6.6.bb): Jan 30 18:00:32 python-2.6.6: Fix python-2.6.6 build failure on linux-3.x Jan 30 18:00:32 python-2.6.6 build will crash when host is running linux kernel 3.x. Jan 30 18:00:32 Fix backported from openembedded-core. Jan 30 18:00:32 meta/recipes-devtools/python/python/07-linux3-regen-fix.patch Jan 30 18:00:32 Signed-Off-By: Ulf Samuelsson Jan 30 18:00:42 03Ulf Samuelsson  07master * r1df2db51aa 10openembedded.git/recipes/freetype/freetype_2.4.3.bb: Jan 30 18:00:42 freetype_2.4.3.bb; Change order of configuration. Jan 30 18:00:42 Build of freetype-native is broken on Ubuntu 11.10 x64 - possible others. Jan 30 18:00:42 Problem fixed in openembedded-core 2.4.4 recipe Jan 30 18:00:42 Backporting the change will allow the build to complete. Jan 30 18:00:43 openembedded-core-contrib/meta/recipes-graphics/freetype Jan 30 18:00:43 Signed-Off-By: Ulf Samuelsson Jan 30 18:00:43 03Ulf Samuelsson  07master * r250219ae88 10openembedded.git/ (21 files in 4 dirs): Jan 30 18:00:44 openssl-1.0.0d: Update to openssl-1.0.0e Jan 30 18:00:44 Backported from openembedded-core Jan 30 18:00:45 meta-openembedded/meta-oe/recipes-connectivity/openssl Jan 30 18:00:45 Signed-Off-By: Ulf Samuelsson Jan 30 18:03:40 moin Jan 30 18:03:59 khem: http://dl.free.fr/q1G3euZ7T Jan 30 18:08:08 khem yes :) Jan 30 18:08:17 I am looking for clues on fixcing it :) Jan 30 18:19:50 GarthPS, hi Jan 30 18:20:41 GarthPS, would a pastebin work? I can't see the content(I use gnash) Jan 30 18:24:19 GarthPS: and sed which version of sed do you use Jan 30 18:24:54 hi, is anyone by any chance aware of webui toolkits (like Wt, but I'd like to see if there happens to be alternatives)? Jan 30 18:27:12 GarthPS: jenkins did after switching to eglibc-2.15.. Jan 30 18:27:32 GarthPS: and I did clean_toolchain too Jan 30 18:27:37 2-3 days ago Jan 30 18:27:37 eFfeM: here we use awful with chicken Jan 30 18:28:00 otavio: don't know that one, will google it now Jan 30 18:28:30 eFfeM: not really a toolkit for ui but more for web development Jan 30 18:29:20 well, I have an app, very simple form, they asked me to come up with a web interface for it. Jan 30 18:29:42 khem: GNU sed version 4.2.1 Jan 30 18:29:59 just some number/text/selector box things, nothing special, only thing is that the page needs to be created on the fly Jan 30 18:30:15 and dynamic updates might be nice (but I know this is somewhat tricky) Jan 30 18:30:34 otavio: you mean this: http://wiki.call-cc.org/eggref/4/awful Jan 30 18:31:32 also considering ruby but not sure about the footprint of the basic thing Jan 30 18:35:00 eFfeM: you might look at dynamicmagic Jan 30 18:35:25 searching Jan 30 18:37:05 otavio: do you have a link? only stuff I find is for audio processing Jan 30 18:40:29 eFfeM: https://github.com/nex3/dynamicmatic Jan 30 18:40:38 otavio: thanks! Jan 30 18:40:49 eFfeM: np Jan 30 18:41:57 khem: It seems we have a issue with oe-core toolchain. I was talking with Koen about this issue and he said that Angstrom-Bleeding is using oe-core toolchain. This bug only happens when using it. https://github.com/OSSystems/meta-mozilla/issues/1 Jan 30 18:41:58 GNUtoo-netbook: no to big Jan 30 18:42:04 ok Jan 30 18:42:21 khem: GNU sed version 4.2.1 Jan 30 18:42:41 GarthPS: your problem is that linker script is generated wrongly Jan 30 18:42:52 GarthPS: now why that happens I dont know Jan 30 18:43:06 eFfeM: enyo from webos looks promising Jan 30 18:43:43 ynezz: will investigate, thanks! Jan 30 18:44:22 http://www.enyojs.com/samples/playground Jan 30 18:44:48 I've meant this one http://enyojs.com/samples/stylematters/ Jan 30 18:45:52 khem yeah :/ I will redoo a complet buildhost from scratch then . again. thx anyway! Jan 30 18:47:00 otavio: ping Jan 30 18:47:06 denix: pong Jan 30 18:47:17 otavio: which pull request? Jan 30 18:47:25 denix: ?? Jan 30 18:48:16 otavio: you mentioned I sent you a pull request - did I misdirected something? I don't remember sending you a pull request... Jan 30 18:48:40 otavio: angstrom-bleeding uses gold to link Jan 30 18:48:52 in addition to using gcc from meta-oe Jan 30 18:49:19 it works with oe-core toolchain but fails with meta-oe is that the summary ? Jan 30 18:49:24 denix: it was Denis Carikli Jan 30 18:49:47 khem: it seems to work with angstrom and fail with angstrom-blending Jan 30 18:49:55 otavio: ah, not me. ok then :) Jan 30 18:49:56 bleeding Jan 30 18:50:01 denix: sorry Jan 30 18:50:03 np Jan 30 18:50:41 pb_: i'm getting reports of this failing - the best thing I can determine is because of NFS Jan 30 18:50:48 pb_: i can't reproduce though Jan 30 18:51:08 otavio: he's GNUtoo, btw Jan 30 18:51:23 hi Jan 30 18:51:29 GNUtoo: :) Jan 30 18:52:04 GNUtoo: you are using angstrom-bleeding ? Jan 30 18:52:09 yes I tried Jan 30 18:52:17 to get more performances at runtime on target Jan 30 18:52:28 I tried for work Jan 30 18:52:44 basically we must have a fast html5 browser Jan 30 18:53:00 so I had to try angstrom-bleeding to get more performances Jan 30 18:53:06 but it didn't work well with firefox Jan 30 18:53:13 GNUtoo: ok. Jan 30 18:53:14 rogram received signal SIGILL, Illegal instruction. Jan 30 18:53:14 0x2c050022 in markFunctionEpilogueDone (this=0x2fc00358) at vm/Stack-inl.h:445 Jan 30 18:53:25 yes Jan 30 18:53:29 this tells me it was executing 0x2c050022 which should not happen in arm mode Jan 30 18:53:40 ah? Jan 30 18:53:41 is your rootfs compiled in thumb mode Jan 30 18:53:45 what's 0x2c050022 Jan 30 18:53:51 I suppose so Jan 30 18:53:56 but I've no idea Jan 30 18:54:02 it's standard angstrom-bleeding Jan 30 18:54:46 so it should contain some thumb Jan 30 18:55:13 no it should not Jan 30 18:55:17 this tells me it was executing 0x2c050022 which should not happen in arm mode //me doens't understand Jan 30 18:55:21 ok Jan 30 18:55:35 no nowadays thumb is desactivated? Jan 30 18:55:51 so I am thinking you are somewhere running into this unaligned load optimization Jan 30 18:55:57 ahhh ok Jan 30 18:56:29 GNUtoo: thumb is there the default is arm mode Jan 30 18:56:37 ok Jan 30 18:57:09 GNUtoo: if you can debug this issue. so find out more on whats going on I can assist you Jan 30 18:57:20 ok Jan 30 18:57:38 enable kernel debugging Jan 30 18:57:51 and see if it reports address range where it gets the error Jan 30 18:57:58 right Jan 30 18:58:01 and what instruction opcodes were executred Jan 30 18:58:55 khem, btw is there a huge difference between a kernel compiled with unalignement + patched for not failing in the init and one compiled without Jan 30 18:59:35 khem, I'll look but not right now, maybe next monday(I've fosdem in between etc...) Jan 30 18:59:45 no worries Jan 30 18:59:52 the disasm there looks ok Jan 30 18:59:55 ok thanks Jan 30 18:59:56 ok Jan 30 18:59:58 nothing eye popping Jan 30 19:00:13 what's the thing at the end with an undefined instruction? Jan 30 19:00:22 I think you need to step thru that function and see why pc gets corrupt value Jan 30 19:00:26 because I looked at the rest and it seemed ok for me too Jan 30 19:00:35 ok Jan 30 19:00:50 0x2c050040 <+536>: ldrsbteq r4, [pc], -r8 Jan 30 19:00:51 0x2c050044 <+540>: ; instruction: 0xffffd3d4 Jan 30 19:01:00 that looks like data pool to me Jan 30 19:01:18 what's sbt? Jan 30 19:01:36 signed byte transposed Jan 30 19:01:39 ok Jan 30 19:02:02 gcc seldom can generate those instrs Jan 30 19:02:07 ok Jan 30 19:02:19 sometimes disassemblers dont know where data area starts Jan 30 19:02:27 ahhh ok Jan 30 19:08:10 does anyone know how to disable all RECOMMENDS? Jan 30 19:08:23 or if there is some preferred method Jan 30 19:08:40 from what I can tell it's rootfs/package manager specific.. Jan 30 19:08:59 I don't believe I put in a switch to disable it in RPM, but it should be able to be removed by adding a filtering mechanism Jan 30 19:09:55 if you look at package_rpm.bbclass you can probably comment out hte processing recommended packages section Jan 30 19:14:33 gosh hehe Jan 30 19:31:46 fray: thanks… was thinking about just adding an option in meta/ classes/base.bbclass or something to avoid build them as well? Jan 30 19:33:23 it's really a backend thing.. Jan 30 19:33:36 since a lot of the items are built for many other reasons Jan 30 19:41:19 fray: but when trying to build a *small* image all these deps add up Jan 30 20:53:53 kergoth, will you be at ELC? Jan 30 20:57:19 Crofton|work: yeah Jan 30 20:57:27 cool Jan 30 20:57:33 wie will meet at last Jan 30 20:57:46 indeed Jan 30 20:57:55 I haven't been to a conference since 2004 Jan 30 20:57:57 sad, really Jan 30 20:58:02 yeah Jan 30 20:58:03 long time Jan 30 20:58:34 it use to be small conference back them Jan 30 20:58:42 s/them/then Jan 30 21:04:54 Crofton|work: your python errors .. which package is that script coming from Jan 30 21:05:32 evening all Jan 30 21:06:03 gnuradio Jan 30 21:06:05 not upstream Jan 30 21:06:25 is there something I am supposed to run to fis up scripts> Jan 30 21:06:26 gm Jan 30 21:06:31 hm, should get the yocto dudes to change their getting started instructions to suggest pulling from github, not oe.org Jan 30 21:08:41 pb_: good idea Jan 30 21:09:45 not that I have any idea which particular yocto dude is responsible for that page of course. :-} Jan 30 21:10:31 on another topic, does anybody happen to have a working mingw32 layer? Jan 30 21:10:42 if you send an email to yocto ml they should notice it Jan 30 21:11:09 Crofton|work: do you have #!/usr/bin/python in your package Jan 30 21:11:25 Crofton|work: you might want to replace that with #!/usr/bin/env python Jan 30 21:11:31 pb_: mingw wow Jan 30 21:11:57 yeah, it's a bit tiresome writing to the yocto list because it doesn't allow non member posting Jan 30 21:13:24 khem, as in the source OF THE PACKAGE? Jan 30 21:13:56 yes Jan 30 21:22:36 Crofton|work: another thing you should look for is #!@PYTHON@ Jan 30 21:22:55 since that will also expand to staging area during build Jan 30 21:23:05 replace that with #!/usr/bin/env python as well Jan 30 21:24:56 yeah, hunting the source now Jan 30 21:26:38 hmm, it uses usr/bin/env already Jan 30 21:26:51 for some reason oe-core adds the otehr it appears Jan 30 21:28:13 I am rebuilding everytihing atm Jan 30 21:28:23 will look in workdir and logs and see what I can figure out Jan 30 21:30:38 sigh, zlib unfetchable Jan 30 21:30:42 * pb_ stabs it Jan 30 22:28:11 has anyone considered, or already worked on, creating public AMI's for amazon EC2 with OE trees prebuilt? Jan 30 22:31:31 it could save new users a lot of time not having to wait for an entire image to build on their own hardware Jan 30 22:35:08 yes I think so Jan 30 22:35:17 look at bunny's talk at CCC Jan 30 23:32:01 Hi guys, any ideas why files in my overlay aren't taking precedence over the defaults or willing to have a go at helping me solve this. Jan 30 23:32:57 I've double-checked the basics like priority in my layers conf and the like and I am able to build recipes in my layer Jan 30 23:33:36 you probably have your order in BBLAYERS wrong Jan 30 23:33:46 bbfile_priority only coverns recipe priority, not config file / classes Jan 30 23:35:10 IOTW my/layer should be first in BBLAYERS? Jan 30 23:37:45 there are a few network tool recipes (e.g. ntpclient, bonjour) that were in classic but aren't in meta-oe. is it just waiting for someone to submit a recipe or is there an intention to limit the recipes in meta-oe? i'm happy to submit the recipes once they're up and running. Jan 30 23:41:31 icanicant: you can submit a patch to add them to meta-oe I think Jan 30 23:42:18 kergoth: reordering my BBLAYERS seems to have done it, thanks. It didn't like PREFERRED_PROVIDER_virtual/kernel = "linux-3.2", though Jan 30 23:43:42 bluelightning: thanks, i guess they're added as when they're needed then. i'll send the patches. cheers! **** ENDING LOGGING AT Tue Jan 31 02:59:58 2012