**** BEGIN LOGGING AT Mon Feb 20 03:00:00 2012 Feb 20 07:48:25 I'm trying to add our own glib into the oe build system, but we get an error when trying to compile dbus-glib: *** You must have either have gettext support in your C library, or use the GNU gettext library. Is this gettext support supposed to be in glibc, or in eglibc? Feb 20 08:00:45 sorry, not glibc, glib-2.0 Feb 20 08:15:32 had anyone success with openjdk-6 on i586, i get allways segaults with compile stage2 with qemu .. Feb 20 08:34:04 I'm trying to add our own glib-2.0 into the oe build system, but we get an error when trying to compile dbus-glib: *** You must have either have gettext support in your C library, or use the GNU gettext library. Is this gettext support supposed to be in glib-2.0,or in eglibc? Feb 20 08:43:40 good morning Feb 20 08:47:43 is there a chance to get WebM enabled in firefox ? Feb 20 08:58:34 morning all Feb 20 08:58:41 morning. Feb 20 08:58:53 morning Feb 20 09:08:07 is there a way to make bitbake silent? right now it outputs a LOT of informations which I truly don't care to much about. I would like the log files, but not need to spam my terminal:P Feb 20 09:16:03 mertsas: console redirection ? bitbake world > /dev/mylog 2>&1 ... Feb 20 09:16:37 s/dev/tmp/ Feb 20 09:16:43 pwgen: yeah, true. I just don't thing to long sometimes:P Feb 20 09:17:36 or start it inside a screen (:-)) Feb 20 09:18:21 pwgen, yeah. That too. For some reason I was crawing a --silent switch, but no need for that really:P Feb 20 09:23:36 is there any package to print gps corordinates from nmei strings Feb 20 09:51:11 hi bluelightning, all Feb 20 10:36:35 i got my changes into the kernel config, u-boot and the i2c speed, thanks to the help from everybody here :) thanks! Feb 20 11:54:24 hello , whats the best to develop digital signage gui ? i will need video player etc... Clutter or DirectFB on C++ ? which one is best for me? Feb 20 12:08:09 Hi there Feb 20 12:11:39 if license file is placedc way inside the sources then how I will provide the path in LIC_FILE_CHECKSUM Feb 20 12:11:51 my file is placed at work/ppce500mc-mgc-linux/fm-lib/fm-lib-01.03.06-r5.1/fm-lib-01.03.06/share/doc/fm-lib-01.03.06 Feb 20 12:12:15 and I ma trying to give it has LIC_FILES_CHKSUM = "file://share/doc/fm-lib-01.03.06/COPYING;md5sum=d58aa0dc8e20d600d3dd06c1abde5085" Feb 20 12:12:53 but I ma getting fm-lib: md5 checksum is not specified for file://share/doc/fm-lib-01.03.06/COPYING;md5sum=d58aa0dc8e20d600d3dd06c1abde5085 Feb 20 12:14:22 what's in WORKDIR variable? Feb 20 12:14:56 Noor: ah and for LIC_FILES_CHKSUM the param is md5 not md5sum Feb 20 12:15:39 JaMa: Yeah noticed that while writing the message and updated the recipe .... it working now :D Feb 20 12:16:02 JaMa: Thanks for noticing that Feb 20 12:21:14 ericben: hi, are you around? Feb 20 12:29:14 i have another question regarding DEFAULT_PREFERENCE ..... I know it controls the versioning selection within single recipe ... does it also control the Provider selection Feb 20 12:30:10 I mean to say of 2 recipes are provideing same thing and in one recipe DEFAULT_PREFERENCE is set to -1 should the recipe that has =ve DEFAULT_PREFERENCE be preferred upon the other one Feb 20 12:30:22 +ve Feb 20 12:30:52 or it only controls the different version of same recipe? Feb 20 12:39:35 can anybody remove my above confusion .... thanks Feb 20 12:41:59 Noor: I'm not sure, but I think you should not rely upon any such mechanisms to select between alternative providers... Feb 20 12:42:13 Noor: can you not just define PREFERRED_PROVIDER_ ? Feb 20 12:42:57 I know it was used in meta-dri.inc as way to denominate it as virtual/libx11 provider, but not sure if it still works Feb 20 12:43:31 and in most cases P_P will be defined anyway, because without it you will get warning about multiple providers Feb 20 12:43:52 this only happens adding layers, usually Feb 20 12:46:46 bluelightning: yes I managed to do that in the end but I was in the impression that by setting DEFAULT_PREFERENCE to -1 I dont have to set PREFERRED PROVIDER Feb 20 12:47:32 Noor: I don't think that's the case Feb 20 12:48:33 bluelightning: then we can safely say that DEFAULT_PREFERENCE only handles the versions within the same recipe? Feb 20 13:01:06 Noor: and withing same layer.. Feb 20 13:01:40 Thanks JaMa Feb 20 13:35:14 I have a kernel which I wish to patch up with a driver. Can I do that in the do_patch step by saying that the driver is present in a git repo? Feb 20 14:17:10 hi otavio Feb 20 14:17:12 hi ericben Feb 20 15:50:18 Noor: LIC_FILES_CHKSUM is always relative to ${S}. Feb 20 15:50:32 * kergoth yawns and stretches Feb 20 16:04:33 hi ppl Feb 20 16:04:42 do we have support for m68k arch? Feb 20 16:04:49 (any device) Feb 20 16:05:56 * kergoth doubts it, but isn't sure Feb 20 16:07:46 grep -r m68k meta/conf gives no result.. Feb 20 16:09:13 haven't heard of anybody doing anything with m68k, no Feb 20 16:09:42 if anyone knows then I guess it would be the freescale h4sx0rs. Feb 20 16:46:33 Hi all Feb 20 17:41:15 anyone have issues with opkg ignoring /etc/opkg.conf? Feb 20 17:41:28 I've been trying to google this and have got nowhere. Feb 20 17:41:51 I can update if I -f /etc/opkg.conf, and it pulls the package file Feb 20 17:41:59 btu that's as far as I get Feb 20 17:42:18 muriani: shouldn't it be in /etc/opkg/opkg.conf? Feb 20 17:42:26 hm Feb 20 17:42:45 it's here in /etc/opkg.conf on my machine Feb 20 17:42:54 I can copy it there and see what happens Feb 20 17:43:44 ah, ok, it updates properly now Feb 20 17:44:17 but list only shows locally installed packages Feb 20 17:49:49 did you run opkg update? Feb 20 17:49:54 prior to listing? Feb 20 17:50:35 yes Feb 20 17:53:24 hm Feb 20 17:54:41 it says it's downloading the packages file, and claims it's putting the updated list of available packages in /var/lib/ipkg/armv5te Feb 20 17:54:45 but the file is empty. Feb 20 17:55:04 well maybe check where it is downloading it from, what if its really empty on the server? Feb 20 17:55:06 oh, hm. It seems the packages file IS empty. Feb 20 17:55:09 wtf Feb 20 17:55:28 This makes using repositories problematic. Feb 20 17:55:59 :) Feb 20 17:56:02 is it your server? Feb 20 17:56:16 Yes, but not just mine Feb 20 17:56:30 the current nightly ipk repo on yocto is the same way Feb 20 17:56:50 http://downloads.yoctoproject.org/autobuilder-images/nightly/CURRENT/ipk/ Feb 20 17:57:04 well, the question was targetting at - did you set it up, is it your build or is it someone elses? Feb 20 17:57:16 on mine, yes, I set it up Feb 20 17:57:16 because you need to bitbake package-index if you want the opkg stuff to be updated correctly Feb 20 17:57:20 ah Feb 20 17:57:22 I didn't know that. Feb 20 17:57:49 I'd had it working before, and didn't do package-index then. Something else must have hit it. Feb 20 17:57:52 ok, I'll try that Feb 20 17:57:53 Thanks Feb 20 18:04:50 i am doing somthing stupid... I am trying to get a custom .bb file for the linux-firmware files Feb 20 18:05:00 and i keep getting a fatal: not a valid object name 1 Feb 20 18:05:04 when i try to run it Feb 20 18:05:08 i know its something i caused Feb 20 18:05:20 if i post my bb file can any one point me in the right direction? Feb 20 18:05:32 git repo in SRC_URI and no SRCREV defined? Feb 20 18:06:04 well i tried defining SRCREV but im not sure quite what to put there Feb 20 18:06:11 i did find that in my google searches of the problem Feb 20 18:06:23 git revision which should be used Feb 20 18:06:39 and if i want what ever the current revision is? Feb 20 18:07:08 then set it to ${AUTOREV} Feb 20 18:07:22 but in recipe there should be some stable tested revision Feb 20 18:07:32 and you can overwrite it to AUTOREV in your local.conf Feb 20 18:08:59 Jin^eLD: bingo, that put me on the right track. Going smoothly now. Feb 20 18:09:08 :> Feb 20 18:10:06 JaMa: i did that... got a new error Feb 20 18:10:09 JaMa: http://slexy.org/view/s21EtETn5p Feb 20 18:10:12 is my current file Feb 20 18:10:49 not AUDOREV Feb 20 18:11:03 oh... damnit. Feb 20 18:11:36 * Thex1le hangs head Feb 20 18:12:14 JaMa: that worked Feb 20 18:12:15 thanks Feb 20 18:12:19 cant belive i was that dumb Feb 20 18:32:10 some days I really hate computers Feb 20 18:33:54 * bluelightning was thinking about some generic code to pick up typos in variable names today Feb 20 18:34:21 we could at least catch repeated letters and missing/extra underscores pretty easily Feb 20 18:35:09 the question would be the overhead. thankfully it'd be config time not recipe time, probably, so maybe not too bad. its the recipe time / anonfunc / recipeparsed processing that causes issues Feb 20 18:35:10 heh Feb 20 18:37:28 * JaMa would prefer "unknown" declaration warnings then typos detection but that would need something like setting valid variables to "empty" value if they are supposed to be empty Feb 20 18:42:41 03Henning Heinold  07org.openembedded.dev * r4835b72edb 10openembedded.git/recipes/ecj/ (3 files): Feb 20 18:42:41 libbecj-bootstrap: use sort for build order of the javafiles and introduce INC_PR, backported from meta-java Feb 20 18:42:41 * on some filesystems find orders the files to build not the Feb 20 18:42:41 right way, using sort seems to fix it, thanks Stefan Ring Feb 20 18:42:41 for the patch Feb 20 18:42:41 * use INC_PR for better revision handling Feb 20 18:42:44 03Henning Heinold  07org.openembedded.dev * rb725ced47f 10openembedded.git/recipes/icedtea/ (2 files in 2 dirs): Feb 20 18:42:44 icedtea6-native: add linux kernel version 3 support Feb 20 18:42:44 * bump PR Feb 20 18:42:45 03Henning Heinold  07org.openembedded.dev * rdce169d6e8 10openembedded.git/recipes/jamvm/ (jamvm-initial/libffi.patch jamvm-initial_1.4.5.bb): Feb 20 18:42:45 jamvm-inital: rework libffi detection Feb 20 18:42:45 * this patch allows us to real enable or disable libffi and Feb 20 18:42:45 it does not let sneak host libs in Feb 20 18:42:45 03Henning Heinold  07org.openembedded.dev * rbac9a73e0e 10openembedded.git/recipes/cacao/ (cacao-native.inc cacao-native_hg.bb): Feb 20 18:42:59 * shark is disabled since 1.11 Feb 20 18:42:59 * rework patches Feb 20 18:42:59 * rework inc files for jamvm Feb 20 18:42:59 * bump PR Feb 20 18:42:59 03Henning Heinold  07org.openembedded.dev * r6d8534ea84 10openembedded.git/recipes/ecj/ (ecj-bootstrap-native.bb files/ecj.in): Feb 20 18:42:59 ecj-bootstrap-native: update ecj script Feb 20 18:43:00 * merge with script from java-initial Feb 20 18:43:00 * bailout after 50 tries Feb 20 18:43:01 03Henning Heinold  07org.openembedded.dev * r4b8579316e 10openembedded.git/recipes/mozilla/ (rhino-native_1.7r2.bb rhino_1.7r2.bb): rhino: add version 1.7r2 Feb 20 18:43:01 03Henning Heinold  07org.openembedded.dev * r8add1e3e9c 10openembedded.git/conf/distro/include/angstrom-jalimo.conf: angstrom-jalimo.conf: prefer icedtea 1.11.1 for openjdk6 Feb 20 18:43:02 03Henning Heinold  07org.openembedded.dev * r4e75705872 10openembedded.git/recipes/openjdk/openjdk-6-common.inc: openjdk-6-common.inc: fix hardcoded path for allowing to build with linux 3.x Feb 20 20:00:58 people commit to .dev still? ;) Feb 20 20:01:45 kergoth: heh Feb 20 20:02:26 kergoth: I asked henning to submit it to 2011.03 branch but he was not interested Feb 20 20:02:45 heh, well, at least it's somewhere. someone can always pull it back if they need it there Feb 20 20:13:07 kergoth thats what I said Feb 20 20:13:22 work was for kaelios Feb 20 20:13:24 mainly Feb 20 20:13:35 or focused Feb 20 20:13:39 ah Feb 20 20:13:43 but so everybody who needs can pickup Feb 20 21:38:17 hi Feb 20 21:38:57 hi rlrosa Feb 20 21:39:12 hi woglinde Feb 20 21:39:37 i almost got everything working :) Feb 20 21:39:57 cool Feb 20 21:40:22 i did bitbake -c devshell virtual/kernel, changed a file, and got a patch from it. the kernel recipe i'm building is linux_3.0.bb. if i add it there, then it'll complain about the patch not applying. Feb 20 21:40:36 did you use quilt? Feb 20 21:40:42 where did you put the patch? Feb 20 21:40:47 at the end? Feb 20 21:40:48 nop, just git Feb 20 21:40:50 yes Feb 20 21:40:55 maybee some other patches interfere Feb 20 21:41:19 ok, but that should be the place? (somewhere in linux_3.0.bb) Feb 20 21:41:45 at the end of course Feb 20 21:41:56 of SRC_URI Feb 20 21:42:16 you can check it with -e or -DDD I think Feb 20 21:42:33 when bitbaking virtual/kernel? Feb 20 21:42:48 yes bitbake -e gives you enviroment setting Feb 20 21:43:00 -DDD gives you debug output of bitbake Feb 20 21:44:19 ok i have a lot to look at now. i'll it out Feb 20 21:44:22 thanks woglinde Feb 20 21:44:32 sure Feb 20 22:44:57 any one willing to look at a .bb file and tell me what im doing wrong? Feb 20 22:44:59 http://slexy.org/view/s20l0uLHPA Feb 20 22:45:32 it seems to download the firmware via git ok and the backage builds fine but it doesnt seem to copy the firmware files it downloaded into /lib/firmware in my image Feb 20 22:51:58 uh, you aren't installing anything Feb 20 22:52:13 i figured that was the case... lol Feb 20 22:52:20 do_install is what puts files into the packages, and all you do there is install -d, which is the same as mkdir Feb 20 22:52:25 you make a lovely empty directory and thats it Feb 20 22:52:26 ah... Feb 20 22:53:11 so if i wanted to just copy all the git files into lib firmware? Feb 21 01:16:50 anyone know how to overide file collisions between packages? I need a package i wrote to overwrite some fimrware files Feb 21 01:18:07 are you colliding with packages coming from different recipes ? Feb 21 01:18:25 khem: yes Feb 21 01:18:42 well then you have to make a choice Feb 21 01:18:49 delete it from the other recipe Feb 21 01:19:01 if you think your recipe needs to provide them Feb 21 01:19:18 are these unrelated recipes Feb 21 01:19:49 no, mine provides the entire firmware tree from linux-wireless Feb 21 01:20:07 the various packages provide certain files but older versions Feb 21 01:24:22 hmmm Feb 21 01:25:05 what recipes are you conflicting with Feb 21 01:51:33 khem: some of the kernel firmware packages Feb 21 01:51:55 khem: i just redid my package to only copy the files i KNOW i need. i just had the space and so i figured getting all the firmware couldnt hurt Feb 21 01:52:04 the ones i need shouldnt conflict Feb 21 01:52:44 hey. what is the way to contribute patches to meta-oe? Feb 21 01:54:00 email to the openembedded-devel list with [meta-oe] in the subject line Feb 21 01:55:37 you mean, using git's builtins? Feb 21 01:55:56 what about the github repo - like pull requests there or such Feb 21 01:59:43 khem: thanks for the help Feb 21 02:06:06 CMoH: this document http://www.openembedded.org/wiki/How_to_submit_a_patch_to_OpenEmbedded should help Feb 21 02:06:27 thanks Feb 21 02:07:57 and plus what kergoth suggested i.e. adding [meta-oe] to subject prefix Feb 21 02:09:42 i see many mails have the [oe] prefix as well - what does that mean? Feb 21 02:13:49 CMoH: generally that should mean classic OE Feb 21 02:14:27 hmm... it seems it's been misinterpreted by most Feb 21 02:16:16 yes there is one ml and lot of layers which use it to submit patches Feb 21 02:16:45 however beaware that maintainers may care for patches with proper headers Feb 21 02:16:52 bloody idiot - git removed the carefully placed [meta-oe] prefix i added! :D Feb 21 02:17:17 you can replace it afterwards too Feb 21 02:17:43 after sending the mail? Feb 21 02:18:05 i used git send-email :D Feb 21 02:21:56 no before sending ofcours Feb 21 02:23:45 your goal should be to get the patches should be noticed quickly Feb 21 02:25:55 i don't get it... git send-email's man page says i've done things correctly. i guess editing the subject line is something not done very often Feb 21 02:29:20 one other thing: we have an issue tracker marker within the commit message. how can one use these patches across different organizations? Feb 21 02:32:39 great Feb 21 02:32:49 now i forgot to add a v2 to it Feb 21 02:32:52 aaaarrrgggh Feb 21 02:37:08 well, i hope someone picks it up - thanks for your input khem Feb 21 02:37:18 next time i guess i'll have more experience :) **** ENDING LOGGING AT Tue Feb 21 02:59:58 2012