**** BEGIN LOGGING AT Thu Feb 01 02:59:57 2007 Feb 01 04:57:19 is bitbake-1.6 (SVN rev 753) broken? Feb 01 04:57:25 it just errors for me Feb 01 04:59:26 error and my local.conf is http://pastebin.ca/335237 Feb 01 05:05:16 bitbake-1.6 (SVN rev 753) error and my local.conf is http://pastebin.ca/335237 Please help Feb 01 05:05:32 new setup, I'm sure I didn't something stupid, but I don't know what Feb 01 05:50:57 Zero_Chaos: perhaps you need to upgrade python or install a package Feb 01 05:52:22 JustinP: python is version 2.2... :-( too old you think? it's not my local.conf? Feb 01 05:52:40 <_law_> i have build a angstrom bootstrap image does anybody know where i can find the modules.conf (modprobe.conf) ? Feb 01 05:52:46 <_law_> i cant find it in /etc Feb 01 05:52:51 Zero_Chaos: you probably need 2.4 Feb 01 05:53:11 fsck Feb 01 05:53:16 JustinP: thanks :-) Feb 01 07:16:33 * Zero_Chaos hugs JustinP Feb 01 07:17:02 JustinP: you were right, running Python 2.5 now, works fine. Although Psycho warns that Python 2.5 isn't stable... Feb 01 07:17:06 I'll find out I guess Feb 01 07:17:14 JustinP: thanks again Feb 01 08:41:08 morning Feb 01 09:18:09 ibot, seen schurig Feb 01 09:18:44 schurig was last seen on IRC in channel #oe, 5d 21h 36m 18s ago, saying: 'mickeyl: we're all coughing like mad, but otherwise good'. Feb 01 09:33:53 ~lart new flight taxes :-( Feb 01 09:33:54 * ibot DoSes new flight taxes :-( Feb 01 09:45:32 hi all. anybody tried to build ncurses with unicode support? Feb 01 10:10:46 ~lart people which use GPG keys which are not signed by anyone Feb 01 10:10:46 * ibot makes a balloon animal out of people which use GPG keys which are not signed by anyone Feb 01 10:11:44 hrw: OI Feb 01 10:15:33 someone here use(d) google desktop search? Feb 01 10:16:48 mickey|zzZZzz: A bit late but nothing changed with collections that I know of Feb 01 10:16:52 morning all Feb 01 10:17:16 hi hrw, RP Feb 01 10:17:23 hi pH5 RP Feb 01 10:22:54 ~lart konqueror Feb 01 10:22:55 * ibot offers konqueror some herring Feb 01 10:37:15 morning Feb 01 10:38:20 does anyone know what the approximate size difference of the xserver-xorg and xserver-kdrive packages is when installed on the target machine? Feb 01 10:47:47 cu Feb 01 10:50:25 hi, all! Feb 01 10:51:13 do anybody remembers how to make stock wm97xx driver work? any device which use it? Feb 01 10:53:19 slapin_nb: look at http://handhelds.org/moin/moin.cgi/HandheldHardwareXref Feb 01 10:55:00 pH5, most of these use own versions, I'm interested in stock driver usage... Feb 01 11:02:19 slapin_nb: opensource.wolfsonmicro.com/~lg/asoc/ Feb 01 11:02:34 slapin_nb: wm97xx is my job Feb 01 11:07:43 XorA, I'm mostly interested in usage guidance for driver (how to connect it to platform). Feb 01 11:08:51 slapin_nb: check the machine files, tosa.c is a good one Feb 01 11:09:00 sound/soc/pxa/tosa.c Feb 01 11:09:55 RP: I have tested ASoC 0.13rc3 on spitz and it works, so you might want to advance the version in your 2.6.20rcX in poky Feb 01 11:15:03 XorA, I'm also interested in touchscreen connection, any samples? or should work as is after sound initialization? Feb 01 11:15:20 slapin_nb: tosa uses it as touchscreen Feb 01 11:15:54 slapin_nb: http://opensource.wolfsonmicro.com/~lg/wm97xx/ Feb 01 11:15:59 slapin_nb: is the touch drivers Feb 01 11:17:39 XorA, ok, which kernel contains samples for usage? Feb 01 11:17:48 slapin_nb: linux-rp Feb 01 11:18:30 Thanks! Feb 01 11:18:40 actually, thanks a lot! Feb 01 11:18:57 slapin_nb: which handheld you trying to use? Feb 01 11:29:03 greetings ! Feb 01 11:38:14 is it possible to tell OE to compile the kernel with a specific compiler version, and not with the one that is used for everything else? Feb 01 11:38:43 Jin^eLD: yes, check the openzaurus-* kernels Feb 01 11:38:48 thanks Feb 01 11:39:12 Jin^eLD: but that is limited to a pre-made toolchain Feb 01 11:39:35 thats OK, thats exactly the case i have Feb 01 11:40:04 the board I have comes with a prebuilt compiler for the kernel, and a prebuilt toolchain.. I want to exchange the latter Feb 01 11:40:20 but the kernel is too old, and they use a very old gcc to compile it Feb 01 11:40:51 Jin^eLD: exactly the same as 2.4 zaurus kernels then Feb 01 11:41:08 so this kind of stuff? KERNEL_CCSUFFIX= Feb 01 11:41:18 I'll try that, thanks Feb 01 12:01:17 What's the best way to have a gpe-image build in with some more packages? Feb 01 12:01:27 XorA: tried it similar to openzaurus-unstable.conf, they just did an ASSUME_PROVIDED, so I did ASSUME_PROVIDED+=" virtual/mipsel-linux-gcc-2.95" Feb 01 12:01:31 but I get an error ERROR: Nothing provides dependency virtual/mipsel-linux-gcc-2.96 Feb 01 12:01:36 what am I doing wrong? Feb 01 12:02:14 s/95/96/ Feb 01 12:02:26 oh crap I am really stupid Feb 01 12:02:28 sorry :) Feb 01 12:02:56 is ok, I spent 2 hours a couple of days ago working out that get != set Feb 01 12:03:06 :) Feb 01 12:09:13 morning Feb 01 12:09:53 hi Feb 01 12:11:25 XorA: uhm... that did not solve the problem Feb 01 12:11:33 ASSUME_PROVIDED+="virtual/mipsel-linux-gcc-2.96" Feb 01 12:11:35 ERROR: Nothing provides dependency virtual/mipsel-linux-gcc-2.96 Feb 01 12:11:49 unless I am blind the lines match... Feb 01 12:11:52 what else could that be? Feb 01 12:13:01 Jin^eLD: looks the same to me Feb 01 12:13:31 any ideas why it is still not working? Feb 01 12:13:46 what could I have missed or how can I debug this issue? Feb 01 12:13:46 nope Feb 01 12:14:15 Jin^eLD: my bitbake knowledge is lacking there Feb 01 12:14:59 mine too.. well, I'll try around and see how far I get Feb 01 12:15:03 thanks Feb 01 12:20:03 hi all Feb 01 12:20:09 who was the guy working on the DVB-USB stick? Feb 01 12:21:12 hey likewise Feb 01 12:21:17 Jin^eLD: hi Feb 01 12:28:05 what is the correct way to specify a specific revision for the svn fetcher? Feb 01 12:28:23 in my bb file I have SRCDATE="4635" Feb 01 12:28:34 and then rev=${SRCDATE} on the end of the url Feb 01 12:28:57 but if I do that, then none of my patches apply because they are deemed 'outdated' Feb 01 12:30:21 I have looked at the code in patch.bbclass Feb 01 12:30:44 it tries to compare maxdate against srcdate Feb 01 12:30:54 and also against mindate Feb 01 12:31:29 what I don't get is how mindate is 0, maxdate is 20711226... Feb 01 12:31:44 my srcdate sits nicely between these two, yet it still fails Feb 01 12:32:35 if I add a print statement just before the error to check the values them I get: mindate=0, maxdate=20711226, srcdate;2448 Feb 01 12:32:39 looks right Feb 01 12:32:57 yet another print statement shows that: Feb 01 12:32:59 20711226 is less than 2448 Feb 01 12:33:02 !! etf! Feb 01 12:33:21 if (maxdate < srcdate): Feb 01 12:33:21 print "%s is less than %s" %(maxdate, srcdate) Feb 01 12:36:12 are those strings, or integers? Feb 01 12:36:29 if they are strings, 20... will sort less than 24.. irrespective of length Feb 01 12:36:41 ok... looks like they are strings Feb 01 12:36:49 (except mindate, which is an int) Feb 01 12:36:59 try "if int(maxdate) < int(srcdate):" Feb 01 12:37:49 though, that said, I'm not quite sure how you're ending up with a srcdate of 2448 anyway. that's clearly not a legitimate date string. Feb 01 12:38:01 it's a revision number Feb 01 12:38:38 I need that specifiic revision of the src Feb 01 12:38:56 you'd probably do better to put that in ${PV} directly rather than SRCDATE. Feb 01 12:39:10 there might be other places in the code that expect SRCDATE to be an actual date. Feb 01 12:40:10 ok... and how should I name this bb file? Feb 01 12:40:20 my-package_svn2448? Feb 01 12:40:32 or just my-package_2448 Feb 01 12:40:45 my-package-2448_svn+2448? Feb 01 12:43:57 that doesn't really matter. Feb 01 12:44:27 if you have no plan to ever make numbered releases, so the svn revision will always be the definitive numbering, my-package_2448 would be fine. Feb 01 12:44:58 XorA: it seems it wanted the ASSUME_PROVIDED stuff to be on local.conf, I had it in the distro conf file. openzaurus-unstable also has it in the distro, so I wonder why it failed for me Feb 01 12:45:00 if you think you might one day want to have a version 1.0, you might be better off with my-package_0.0+svn2448 Feb 01 12:45:01 right... but I do intend on using stable releases at some point Feb 01 12:45:06 right Feb 01 12:45:54 ok, so: Feb 01 12:45:54 SRCDATE = "2448" Feb 01 12:45:54 PV = "0.0+svn${SRCDATE}" Feb 01 12:46:27 re the patcher problem: Feb 01 12:46:29 srcdate = int(srcdate) Feb 01 12:46:29 mindate = int(mindate) Feb 01 12:46:30 maxdate = int(maxdate) Feb 01 12:46:33 fixes it Feb 01 12:47:43 ok, cool Feb 01 12:48:01 I guess those int()s will also help to avoid a problem in the year 10000 so they sound like good news anyway Feb 01 12:48:19 I'll post a patch on the bugtracker Feb 01 12:48:30 great Feb 01 12:48:32 pb__: man your support dedication goes beyond the grave :-) Feb 01 12:48:50 XorA: mm? Feb 01 12:49:07 pb__: planning to still be working on OE in 10000 Feb 01 12:55:43 oh, heh Feb 01 12:55:44 patch in the bugtracker Feb 01 12:59:06 hey mickeyl Feb 01 12:59:10 moirning Feb 01 12:59:11 morning, as well Feb 01 13:01:02 hi mickey Feb 01 13:01:07 cheers pb_ Feb 01 13:15:33 ~seen likewise Feb 01 13:16:04 likewise was last seen on IRC in channel #oe, 54m 38s ago, saying: 'Jin^eLD: hi'. Feb 01 14:18:02 hi Feb 01 14:21:12 I have installed debian and OE because on OsX i had problems to compile packages. Now ... i have an error in debian while doing any "bitbake something" Feb 01 14:21:21 http://pastebin.ca/335678 Feb 01 14:22:02 alvises: check your BBPATH env variable Feb 01 14:22:13 alvises: you should see some number of packages parsed. Feb 01 14:23:00 alvises: NOTE: Parsing finished. 3948 cached, 0 parsed, 184 skipped, 0 masked. Feb 01 14:23:25 cbrake, /home/alvise/develop/oe/build:/home/alvise/develop/oe/org.openembedded.dev Feb 01 14:25:06 cbrake, seems to be correct :| Feb 01 14:25:16 alvises: looks right to me Feb 01 14:25:28 alvises: have you tried bitbake -DDDD? Feb 01 14:29:47 cbrake, well... redone the export and now works but i have a worst error : http://pastebin.ca/335690 Feb 01 14:36:05 alvises: guessing you don't have local.conf set up correctly. Trying something simple like MACHINE="x86", DISTRO="generic" Feb 01 14:36:24 alvises: or MACHINE="compulab-pxa270" Feb 01 14:36:48 cbrake, thx i try Feb 01 14:39:39 cbrake, same tipe of error Feb 01 14:41:47 cbrake, same type of error Feb 01 14:41:53 i-> y ;) Feb 01 14:42:48 cbrake, tried with DISTRO="generic" and now is working why ? Feb 01 14:42:49 alvises: well do something like introduce errors in local.conf to make sure it is getting parsed. Feb 01 14:43:03 alvises: there is probably some type of setup issue. Feb 01 14:43:24 alvises: what version of Python are you running? Feb 01 14:43:28 cbrake: mmm .. .if i want openzaurus-3.5.4.2 distro ? Feb 01 14:43:34 cbrake, check Feb 01 14:43:50 cbrake, python2.4 Feb 01 14:44:47 alvises: I don't know about OZ. All I build is dev and angstrom from dev.oe.org. Feb 01 14:45:51 cbrake, and for enable full X11 what i have to do on conf ? Feb 01 14:47:07 nick nzg Feb 01 14:47:26 mornin Feb 01 14:47:44 alvises: to build for OZ: Feb 01 14:47:59 alvises: 1) checkout anc use the org.openembedded.oz354x branch Feb 01 14:48:11 alvises: 2) use DISTRO="openzaurus-3.5.4.2" Feb 01 14:48:23 alvises: 3) set your MACHINE correctly (spitz?) Feb 01 14:48:58 JustinP, thanks Feb 01 14:49:17 alvises: 4) for full X11 do this in your local.conf: BBMASK = "diet-x11*.*" Feb 01 14:49:38 alvises: then you'll have to edit each of the files that give you a dependency error to depend on x11 instead of diet-x11 Feb 01 14:49:49 I don't know why so many packages are set to depend on diet-x11.... Feb 01 14:50:10 JustinP, mm.. ok ..thanks. Feb 01 14:51:14 alvises: here's a list: packages/xmonobut/xmonobut_0.4.1.bb, packages/meta/meta-gpe.bb, packages/gpe-soundserver/gpe-soundserver_0.4-1.bb Feb 01 14:51:29 just change the diet-x11 in those 3 to x11 Feb 01 14:51:59 does anyone know why drm-kernel inherits from module-base.bbcalss and not from module.bbcalss? Feb 01 14:52:04 JustinP, thx Feb 01 14:54:31 JustinP, compiling :D Feb 01 14:54:42 JustinP, bitbake gpe-image ... Feb 01 14:55:09 JustinP, kernel version remain 2.6.16 isn't it ? Feb 01 15:12:42 alvises: I believe it's 2.6.17 Feb 01 15:40:09 good morning all! Feb 01 15:40:35 koen|gprs, hi ! Feb 01 15:41:11 hey koen Feb 01 15:41:31 morning Feb 01 15:51:35 03koen 07org.oe.dev * rda7c8d01... 10/ (4 files in 3 dirs): cairo 1.3.12: update surface-cache patch to r3 Feb 01 16:13:28 Building an ep93xx arch and I'm getting cc1: error: unrecognized command line option "-mapcs-32" Feb 01 16:13:51 Anyone know what I might be missing? Feb 01 16:14:27 I'm currently assuming this must be purely a toochain problem and can't be related to my kernel version....right... Feb 01 16:23:27 right Feb 01 16:23:31 sounds like you have a bad version of gcc Feb 01 16:26:12 hehe is there a 'good' version of gcc Feb 01 16:27:01 ade|desk: the next version ;) Feb 01 16:27:10 natezg: I recently build a 2.4 kernel (don't ask) on several ARM/Debian systems and encounterd the same error. Feb 01 16:29:26 JustinP, how i look the kernel version before compiling ? Feb 01 16:29:54 JustinP, there isn't the kernel in packages dir Feb 01 17:06:40 cbrake|pb: so does anyone know what version of gcc supports this arguement? Or an easy way to find that out Feb 01 17:09:53 natezg: http://gcc.gnu.org/onlinedocs/gcc-4.1.1/gcc/ARM-Options.html#ARM-Options Feb 01 17:10:25 cbrake:thx Feb 01 17:11:01 I think I just found the problem BTW, looks like I have another arm-linux-gcc in my path already Feb 01 17:11:03 Reading specs from /usr/local/arm/3.4/bin-ccache/ Feb 01 17:11:09 that doesn't seem right Feb 01 17:13:17 I wonder why it sees that, my compile script is clearly prepending to thepath, shoudn't it see the oe tools "first" I dunno Feb 01 17:56:23 hi Feb 01 17:56:30 any idea why i can't checkout? Feb 01 17:56:35 cvs -d :pserver:anoncvs@cvs.handhelds.org:/cvs login kbdd Feb 01 17:56:36  Feb 01 17:56:41 cvs login: authorization failed: server cvs.handhelds.org rejected access to /cvs for user anoncvs Feb 01 18:12:02 alvises: it's packages/linux Feb 01 18:12:10 alvises: bitbake --dry-run virtual/kernel Feb 01 18:22:24 hey guys Feb 01 18:22:31 and girls... Feb 01 18:22:41 where is the OE crowd going to stay for Fosdem? Feb 01 18:27:25 philippe, see the topic Feb 01 18:30:11 Crofton: thanks... Slipped by at the end of my screen/// Feb 01 18:30:30 hey philippe Feb 01 18:37:02 hey koen|gprs__ :-) Feb 01 18:37:49 koen|gprs__: tried gpe-mini-browser 0.20 yet? Feb 01 18:38:14 not yet Feb 01 18:38:22 I'm ~200km away from my devices Feb 01 18:38:34 * koen|gprs__ is near eindhoven this week Feb 01 18:39:07 koen|gprs: aha/ That's why there is the gprs :-) Feb 01 18:47:30 does oe support uImage stuff from uboot? Feb 01 18:48:10 I just found out that I have to use mkimage or something to convert the produced vmlinux.bin to something uboot can handle Feb 01 19:01:05 any idea why i can't ceckout kbdd? Feb 01 19:02:31 sry works now Feb 01 19:21:57 daurnimator: well, there is a problem that I did observe on some embedded systems Feb 01 19:22:00 oops wrong window Feb 01 19:22:01 and wrong nick :) Feb 01 19:42:30 ah, FYI it is set by the kernel Feb 01 19:42:53 mapcs-32 is getting set by the arch makefile in the Cirrus kernel Feb 01 19:43:24 * koen|gprs wonders who in its right mind would use the cirrus kernel Feb 01 19:43:26 it's Feb 01 19:43:30 a) a ugly hack Feb 01 19:43:33 b) ancient Feb 01 19:43:40 Someone who's using a board not supported by the mainstream kernel Feb 01 19:44:05 write a mail to lennert and have support in ~10 minutes Feb 01 19:44:28 Surely that's an exaggeration Feb 01 19:44:59 There are lots of ep93xx configurations not supported by the mainstream kernel Feb 01 19:45:09 I read the posts, I understand why, but there it is Feb 01 19:45:24 Extra work not to use the existing kernel Feb 01 19:45:55 so you say "I'm not going to ask, but stay with a non-compiling kernel" Feb 01 19:45:58 great attitude Feb 01 19:46:04 hello all Feb 01 19:46:12 didn't say that Feb 01 19:46:17 hey likewise Feb 01 19:46:20 hi likewise Feb 01 19:46:39 Jin^eLD, koen|gprs : hey back Feb 01 19:47:01 koen:I'll ask, but I doubt I'm the first to do so Feb 01 19:47:09 likewise: were you the one asking about the wis8100 board? Feb 01 19:47:12 natezg: Does you need your kernel to be upstream? Feb 01 19:47:17 Jin^eLD: yes Feb 01 19:47:35 likewise:don't really care if it works Feb 01 19:47:41 natezg: sorry, that sentence got mangled while editing Feb 01 19:47:49 natezg: which target board? Feb 01 19:48:05 likewise:edb9302A clone Feb 01 19:48:53 natezg: I think OE has a very complete and recent kernel for that, but I could be mistaken Feb 01 19:49:07 likewise: I am having a hard time bringing it into OE... old kernel, based on a release candidate, ancient toolchain, their divers can not be built as modules, lots of stuff like that Feb 01 19:49:08 likewise: pretty recent Feb 01 19:49:29 likewise: I think it would be easy to update it to 2.6.20rc :) Feb 01 19:49:44 likewise: and maybe include the discontigmem patch Feb 01 19:49:45 likewise:I tried OE first,it didn't work Feb 01 19:50:22 The port is probably relatively simple, but I'm trying to take the path of least resistance here Feb 01 19:50:39 unfortunately the resistance is only clear looking backward Feb 01 19:56:13 koen|gprs: I'm looking at upgrading linphone in OE (we have 1.6.0 working in console mode for nslu2 and yeaphone handsets). Feb 01 19:56:29 rwhitby: go for it Feb 01 19:56:35 There is an existing linphone package, but no maintainer listed. Feb 01 19:56:41 rwhitby: although I still think telepathy should be used Feb 01 19:56:56 I can build the console version, but have no way to test the console version. Feb 01 19:57:27 http://swik.net/telepathy Feb 01 19:58:06 actually, there's a question there: if a package needs to be significantly different (at build time and at run time) between console and GPE, should I just make two .bbs using a common .inc, or use a distro or machine override in one .bb ? Feb 01 19:58:40 koen|gprs: I'm just packaging this - someone else has done the work to get it to work, so changing it to another package is not an option for the limited time I have available for this task. Feb 01 19:58:44 use a .inc and have the big one RRPEPLACES the small one Feb 01 19:59:22 any convention on the naming of the two packages? Feb 01 19:59:29 linphone-console, linphone-gpe ? Feb 01 19:59:37 linphone and linphone-console? Feb 01 19:59:43 ok Feb 01 19:59:47 * koen|gprs thinks the linphone-gpe for should go away Feb 01 20:00:01 s/for/fork/ Feb 01 20:01:05 BTW, this new .bb just uses a small patch on the standard linphone sources. Feb 01 20:01:21 * koen|gprs should start writing a presentation for bossaconference Feb 01 20:01:27 any idea what linphone-hh is? Feb 01 20:02:19 (linphone-hh has as description "gpe edition" but also has --disable-gnome_ui) Feb 01 20:04:32 rwhitby: a bitrotted fork Feb 01 20:05:00 ok, so I can ignore it then. Feb 01 20:05:36 looks like the latest version existing in OE (1.3.9) is console-only (and just called "linphone") Feb 01 20:05:50 s/1.3.9/1.3.99.8/ Feb 01 20:06:26 koen|gprs: do you know anyone who can test the gpe edition of linphone? does anyone care? should I just add a console-only version ? Feb 01 20:07:13 rwhitby: anecdote: linphone-gpe had a bug that cause a immediate segfault for over a year before anyone noticed Feb 01 20:07:46 I bet you could replace it with a goatse picture and no-one would notice Feb 01 20:08:34 perhaps I will just add the console version (as linphone, since the existing one is console only) and just leave the gpe bits commented out in the .bb file. Feb 01 20:09:36 (especially since you think the gpe one should be replaced by something else) Feb 01 20:10:23 linphone, ekiga, kphone and friends are all old-style apps Feb 01 20:10:44 each using its own selconftained voip stack without integration with anything else Feb 01 20:10:53 looks like preferred-gpe-versions points to 0.12.1 Feb 01 20:11:00 0.12.2 Feb 01 20:11:18 * koen|gprs mentions telepathy and farsight again for the logs Feb 01 20:11:34 koen|gprs: what will openmoko use? Feb 01 20:11:51 I hope farsight and telepathy Feb 01 20:12:06 since you can use those via dbus Feb 01 20:12:21 task-gpe has linphone-hh only (and it's commented out) Feb 01 20:12:28 likewise, the power supplies arrived Feb 01 20:16:10 koen|gprs: OK, I'm going to add linphone 1.6.0 as console only (with preferred_version = "-1"), then send an RFC to oe-devel to ask for objections in making it the preferred latest version. Feb 01 20:16:35 rwhitby: sounds ok Feb 01 20:17:42 then if someone objects wanting a gpe version, they can test the support in 1.6.0 and I will happily move console-only out to a separate package for them. Feb 01 20:50:42 03rwhitby 07org.oe.dev * rb5d537ef... 10/ (1 packages/linphone/linphone_1.6.0.bb): linphone: Added 1.6.0 (console-only) version. Feb 01 20:50:48 03rwhitby 07org.oe.dev * r3113cfe3... 10/ (1 packages/yeaphone packages/yeaphone/yeaphone_0.1.bb): yeaphone: New package. Support for common USB handsets as a client for linphone. Feb 01 20:50:54 03rwhitby 07org.oe.dev * r848e0c2b... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: Added yeaphone. Feb 01 20:50:59 03rwhitby 07org.oe.dev * rb14608bd... 10/ (1 conf/distro/include/slugos.inc): slugos.inc: Set preferred linphone version for yeaphone Feb 01 20:51:05 03rwhitby 07org.oe.dev * r43ba3756... 10/ (1 packages/linphone/linphone_1.6.0.bb): linphone: Uncomment the default preference on 1.6.0 Feb 01 20:52:38 while generating glibc packages I have an issue Feb 01 20:52:39 cannot open locale definition file `/usr/share/i18n/locales/byn_ER': No such file or directory Feb 01 20:52:42 NOTE: Task failed: localedef returned an error. Feb 01 20:52:46 How can I exclude that locale? Feb 01 21:00:44 Zero_Chaos: BROKEN_LOCALES I think. Check the glibc-package.bbclass Feb 01 21:03:35 thanks Feb 01 21:03:42 JustinP: thanks again Feb 01 21:08:44 Crofton|home: got that, thanks again. Let me know the grand total I need to bring to FOSDEM. Feb 01 21:11:30 will do Feb 01 21:16:24 JustinP: I can't find that in glibc-package.bbclass Feb 01 21:18:36 JustinP: n/m I think I got it Feb 01 21:23:40 Is Angstrom a successor to OZ as in OZ is going away, or it Angstrom a fork? Feb 01 21:26:15 Zero_Chaos: I believe Angstrom is intended to replace OZ Feb 01 21:26:35 I see. Feb 01 21:38:29 I asked Linuxdevices to put OpenEmbedded in a separate category from handhelds.org in the 2007 survey, and they did :-) Feb 01 21:38:39 http://linuxdevices.com/news/NS2797442785.html Feb 01 21:38:58 finally Feb 01 21:54:00 JustinP: any idea how to use GLIBC_BROKEN_LOCALES? I added it to my local conf :=byn_ER ... but that didn't work Feb 01 21:59:15 Zero_Chaos: just edit the one you found in the glibc stuff. := most definately will nto work here Feb 01 21:59:29 Zero_Chaos: you can worry about doing it "right" later ;-) Feb 01 22:01:02 JustinP: I didn't really find a definition of it in the glibc stuff, just it was used Feb 01 22:02:05 Zero_Chaos: glibc_2.3.5+cvs20050627.bb:GLIBC_BROKEN_LOCALES = "sid_ET tr_TR mn_MN te_IN" Feb 01 22:02:18 Zero_Chaos: sorry, it wasn't set in the bbclass Feb 01 22:02:31 (this is .oz354x, but I'm sure it can be set for the .dev) Feb 01 22:05:38 I'm in oz354x, but I'm adding Feb 01 22:05:56 but yeah, I should be good now I think. Feb 01 22:06:19 I can just add GLIBC_BROKEN_LOCALES += "byn_ER UTF-8" right? Feb 01 22:06:27 to my conf/local.conf Feb 01 22:06:44 maybe Feb 01 22:06:48 I think so Feb 01 22:06:48 heh Feb 01 22:06:51 but I'm not sure Feb 01 22:06:56 it might happen the other way around.... Feb 01 22:07:00 but += not := right? Feb 01 22:07:06 yes += Feb 01 22:07:17 but our local.conf += may happen before the glibc = Feb 01 22:07:25 and it would be overwritten Feb 01 22:07:29 feel free to try it, though Feb 01 22:08:05 I thought conf/local.conf overrode everything.... Feb 01 22:09:01 * JustinP isn't sure Feb 01 22:09:07 let me know if it works ;-) Feb 01 22:09:29 it seems a bit weird to be setting GLIBC_BROKEN_LOCALES in a .conf file at all. Feb 01 22:09:50 for a given version of glibc, any particular locale is either broken or it isn't. it shouldn't depend on your environment or other variable factors. Feb 01 22:09:50 this is true Feb 01 22:09:58 which is why I told him to just edit the bb Feb 01 22:10:57 pb_: true, but I'm building from a system with a very old glibc and some locales do not exist, so they fail to build in OE Feb 01 22:11:12 it shouldn't matter what locales you have in your host system. Feb 01 22:11:34 pb_: localedef checks the local system Feb 01 22:11:36 the locale data is meant to be taken from the glibc you're building, not the one you happen to have installed. Feb 01 22:11:45 if it's reading data from your host environment, that's just a bug. Feb 01 22:11:56 well, it's a bug then Feb 01 22:12:10 in that case, it'd be better to fix the bug than turn off the locales in question Feb 01 22:12:33 pb_: by all means. is it a bitbake bug or a Zero_Chaos bug? Feb 01 22:12:42 Zero_Chaos: or, if you just don't know how to do that post a bug and drop the locale locally with GLIBC_BROKEN_LOCALES so you can work around it Feb 01 22:12:58 Zero_Chaos: OE bug I think (with the glibc recipe) Feb 01 22:13:02 it's probably a bug in glibc's .bb file Feb 01 22:13:25 or, more accurately, in glibc-package.bbclass Feb 01 22:13:29 Okay, I'll file a bug, but I am a bit over my head on the glibc bitbake stuff here. Feb 01 22:14:12 localedef_opts = "--force --old-style --no-archive --prefix=%s --inputfile=%s/i18n/locales/%s --charmap=%s %s" % (treedir, datadir, locale, encoding, name) Feb 01 22:14:24 at a guess, that "datadir" probably ought to be os.path.join(treedir, datadir) Feb 01 22:14:29 or something along those lines, anyway Feb 01 22:14:36 what is %s? Feb 01 22:14:52 string substitution Feb 01 22:15:04 Zero_Chaos: see printf for history Feb 01 22:15:45 inputfile=%s that %s turns into /usr/share Feb 01 22:18:18 Zero_Chaos: replace datadir with os.path.join(treedir, datadir) and see if that fixes it Feb 01 22:18:54 as if rebuilding glibc is fast, but yeah, I'll give that a shot Feb 01 22:20:07 I know, I know Feb 01 22:20:17 we appreciate the testing, though Feb 01 22:20:32 if it works do an mtn diff of that change and attach it as a patch to the bug you create :-) Feb 01 22:20:38 JustinP: merely saying I can do it in a minute when it fails again, but you might not hear back from me as I'm going to the bar ;-) Feb 01 22:20:55 :-) Feb 01 22:20:58 JustinP: LoL, a mtn diff? how about a normal diff? Feb 01 22:20:58 :-) Feb 01 22:21:52 Zero_Chaos: just type "mtn diff ." in that dir. monotone will make a diff Feb 01 22:21:59 a normal diff is fine, too (well, unified) Feb 01 22:22:35 I can type mtn diff . I suppose... I'm just not very familiar with monotone so I had no idea what you meant until you said the command Feb 01 22:22:40 * Zero_Chaos kinda hates mtn... Feb 01 22:23:13 why? Feb 01 22:23:24 * JustinP is going to the Monotone Summit next week. :-) Feb 01 22:23:50 JustinP: I just always have trouble with it and stuff. and installing libboost made me angry. Feb 01 22:24:21 JustinP: really I don't know it well, I have little room to talk... it has just been a source of headache when I feel it should "just work" Feb 01 22:25:43 what kind of trouble? Feb 01 22:25:51 * JustinP runs gentoo and it installed perfectly Feb 01 22:27:53 JustinP: I know on gentoo it works great, but this is a... ahem, hacked up (legally) server running RH9 Feb 01 22:28:13 JustinP: I installed boost from source and it installs the libs to a non-standard directory Feb 01 22:28:47 ah Feb 01 22:29:00 I'm currently rebuilding boost for cygwin Feb 01 22:29:07 trying to get the testing bits installed Feb 01 22:29:12 I banged my head over that crap for more than a day. Feb 01 22:29:30 FYI, OE doesn't run on RH9 ;-) Feb 01 22:29:35 but it is almost working Feb 01 22:29:38 if not for the bug Feb 01 22:30:49 you could have tried the static monotone binary they distribute Feb 01 22:31:06 JustinP: 0.28 is the last one they distributed for RH9 Feb 01 22:31:31 everything else is dynamically linked to a newer glibc than I have Feb 01 22:31:56 Zero_Chaos: the monotone static binary on venge.net/monotone? Feb 01 22:32:57 JustinP: this one? http://www.venge.net/monotone/downloads/mtn-0.32-linux-x86.bz2 I didn't understand how to use it... I uncompressed it but then what? it's not a tar Feb 01 22:33:20 read getting started Feb 01 22:33:29 it is the database for montone Feb 01 22:33:40 oh wait Feb 01 22:33:45 LoL Feb 01 22:33:47 that is the mtn binary :) Feb 01 22:33:48 Crofton|home: try to keep up Feb 01 22:33:53 heh Feb 01 22:34:09 Crofton|home: I'm having issues, but I'm not a n00b, this is just a fscked up system. Feb 01 22:34:18 Zero_Chaos: yeah, it's the binary. Just set it executable and run it Feb 01 22:34:37 JustinP: LoL, wish I had known that a week ago, oh well, problem solved anyway ;-) Feb 01 22:35:47 funny as hell to me, this is a 2.8Ghz Xeon with RH9, the OS is so much older than the system.... Feb 01 22:36:57 Why RH9? Feb 01 22:37:07 MV? Feb 01 22:37:12 Crofton|home: that is what is on the server, and I cannot change it. Feb 01 22:37:26 Crofton|home: but the server is fast, so I wanted to use it. Feb 01 22:38:13 heh Feb 01 22:38:58 Crofton|home: I have full level sudo so I'm just hacking the hell out of, installing new python and a bunch of stuff from sources.... it is working... Feb 01 22:39:06 Crofton|home: just a bit more entertaining then usual :-) Feb 01 22:39:10 yeah Feb 01 23:10:59 JustinP: just FYI, it failed. I'm going out but I'll file a very detailed bug tomorrow Feb 01 23:11:08 same error Feb 02 00:27:50 Zero_Chaos: cool Feb 02 01:30:05 has anyone had any problem compiling jpeg for spitz? It compiled ok for collie but for spitz it complains about tags Feb 02 02:26:28 hmmm I am looking for info on embedded ARM development under linux, not putting linux on an ARM processor. Google seems to return tons of info on running linux on ARM, any hints? **** ENDING LOGGING AT Fri Feb 02 02:59:59 2007