**** BEGIN LOGGING AT Wed Jul 12 02:59:57 2006 Jul 12 04:04:13 does anybody use something other than xxdiff for 3-way merges in monotone? Jul 12 04:04:25 kdiff3 here Jul 12 04:05:26 thanks. I have had no luck getting xxdiff to compile on FC 4..:( Jul 12 04:13:27 03mwester 07org.oe.dev * ra9953622... 10/packages/nslu2-binary-only/ (3 files in 2 dirs): unslung-rootfs: rc.sysinit - empty win_share file on boot Jul 12 04:15:45 NA|Zzz: Any hints on how to set it up? Jul 12 04:16:25 hvontres: apt-get install kdiff3 worked for me.. Jul 12 04:18:14 NA|Zzz: No, I mean is there anything you need to do to get monotone to use it? Jul 12 04:23:27 hvontres: Hmm, not as far as I know, no Jul 12 06:34:59 morning koen Jul 12 06:40:20 good morning all Jul 12 06:42:25 * zwelch waves to koen Jul 12 07:01:59 morning all Jul 12 07:03:47 <_law_> morning all Jul 12 07:15:51 hey Dirk & _law_ Jul 12 07:20:46 CoreDump|afk: http://openzaurus.org/wordpress/kernel-26-on-poodle-status-page/ is a bit outdated :) Jul 12 07:20:53 ping CoreDump|afk Jul 12 07:24:42 morning Jul 12 07:24:48 hey XorA Jul 12 07:24:59 what can I break today :-) Jul 12 07:25:26 XorA: Got Poodle? Jul 12 07:27:06 hvontres: no Jul 12 07:39:26 morning all Jul 12 07:41:24 hey RP Jul 12 07:41:46 hey RP Jul 12 07:49:47 morning Jul 12 07:50:07 hey tkp Jul 12 07:50:07 do any of you guys know much about the oe python package? Jul 12 07:50:19 tkp: that'd be mickey|zzZZzz :) Jul 12 07:50:36 ah ok... he wrote it? Jul 12 07:50:42 yes Jul 12 07:50:50 ook cool Jul 12 07:50:56 that's the PR = "ml" Jul 12 07:51:31 ah yes...: Michael 'Mickey' Lauer Jul 12 07:52:02 I'm just a little confused about what it actually provides Jul 12 07:52:11 it does: require python-${PV}-manifest.inc Jul 12 07:52:26 and that manifest file sugest it provides a whole bunch of stuff Jul 12 07:52:38 including python-logging Jul 12 07:53:02 yet after installing it, I don't seem to actually hae python logging Jul 12 07:53:30 perhaps I need to specify which of those I want to actually include Jul 12 07:53:51 tkp: Did you actually install python-logging? Jul 12 07:53:52 but bitbake python-logging is unrecognized Jul 12 07:54:13 ERROR: Nothing provides dependency python-logging Jul 12 07:54:33 tkp: When you build python it'll generate a python-logging package. Jul 12 07:54:49 hmm. I built python already Jul 12 07:55:02 tkp: But when you install you need to explicity request python-logging Jul 12 07:55:14 oh ok Jul 12 07:55:30 so I can't bitbake python-loggong standalone Jul 12 07:55:40 but I can include it in RDEPENDS for my target? Jul 12 07:55:51 yes Jul 12 07:55:54 ahhh.. Jul 12 07:55:55 cool Jul 12 07:55:58 s/yes/you should/ Jul 12 07:56:03 :) Jul 12 07:56:04 tkp: No, because all the python things are related, they are just broken up so you don't have to build them all. Jul 12 07:56:24 RP: Morinig :) Jul 12 07:56:26 tkp: Yeah, look at my ridff-backup .bb for an example - it RDEPENDS on several python modules. Jul 12 07:56:40 v8jlene: ok, thanks Jul 12 07:57:30 tkp: "build them all"? I actually meant "install them all"... Jul 12 07:57:44 yeah, thought so! Jul 12 07:58:17 v8jlene: where is this ridff-backup.bb ? Jul 12 07:58:32 tkp: packages/rdiff-backup/rdiff-backup.inc Jul 12 07:58:52 ah ok. not ridff-backup.bb then :) Jul 12 07:59:03 morning Jul 12 07:59:46 hey hrw|work Jul 12 07:59:52 tkp: Yeah, just checking if you were awake ;) Jul 12 07:59:57 ;) Jul 12 08:00:32 hrw|work: http://ewi546.ewi.utwente.nl/tmp/gvim_7.0.bb Jul 12 08:01:13 koen: nice Jul 12 08:01:13 so in specifying something like python-netclient, you automatically get the core too? Jul 12 08:01:42 hrw|work: Depends: gtk+ (>= 2.8.9), libatk-1.0-0 (>= 1.10.1), pango (>= 1.13.2), libcairo2 (>= 1.2.0), libglib-2.0-0 (>= 2.12.0), libxrender1 (>= 0.9.1), libpng12-0 (>= 1.2.8), libc6 (>= 2.4), libfreetype6 (>= 2.2.1), libz1 (>= 1.2.3), libfontconfig1 (>= 2.3.95), libxt6 (>= 1.0.2), ncurses (>= 5.4), libx11-6 (>= X11R7.1-1.0.1), libsm6 (>= X11R7.1-1.0.1), libice6 (>= X11R7.1-1.0.1) Jul 12 08:02:05 hrw|work: but I have no idea how to handle it with RREPLACES/RCONFLICTS, etc Jul 12 08:03:27 tkp: Yep. You'll also get all the things that python-netclient rdepends on if you rdepend on it. Jul 12 08:03:39 cool Jul 12 08:03:43 http://openzaurus.org/wordpress/2006/07/12/sound-support-for-poodle26/ Jul 12 08:05:24 zwelch: ping Jul 12 08:06:03 hrw|work: when do you think you or CoreDump might be able to build an opie image to test? Jul 12 08:06:38 kergoth: pong Jul 12 08:07:26 hvontres: maybe today Jul 12 08:07:29 zwelch: just curious as to whether you've made any progress wrt mingw32 & such Jul 12 08:07:41 hrw|work:Woot :) Jul 12 08:07:44 kergoth: ummmm... what are you looking for? :) Jul 12 08:08:01 * zwelch just rebuilt minisplat for mingw32 today to prove it still works Jul 12 08:08:06 well, I am off to bed now :) Jul 12 08:08:14 hvontres: 'night Jul 12 08:08:29 and i'm just about to commit bb files for the last package to my tree Jul 12 08:08:47 well, all but the suite... and it'll be a bear, since it wraps all the others Jul 12 08:09:35 kergoth: honestly, i'd want some handholding to do it: someone experienced (like you) to answer all my stupid questions as they come up Jul 12 08:10:28 not me yet, i'm rusty :) getting back into oe after nearly a year absence Jul 12 08:10:30 but okay, makes sense Jul 12 08:10:32 :) Jul 12 08:10:41 and, to be blunt, i'd love to see some cash for the effort; i _am_ supposed to be making money by selling myself to do these things ;) Jul 12 08:10:50 koen: you know what timezone mickey|zzZZzz is in?/ Jul 12 08:11:02 tkp: 10:30 AM Jul 12 08:11:02 however, i'm likely to whack at it on my own anyway, as is The Way Jul 12 08:11:27 10:30 am.. little late to be zzZZzz !! Jul 12 08:12:02 honestly, the fact that i'm here and contributing is because i'm getting paid (a tiny bit) to accomplish the minisplat port Jul 12 08:12:34 * zwelch ceases his confession and goes back to figuring out what's wrong with faad2 Jul 12 08:12:40 * koen wonders Jul 12 08:12:52 should be include kergoth's bitbake vim stuff in our vim package? Jul 12 08:13:14 re faad2: it's a known bug w/ gcc 4.1.1: http://www.audiocoding.com/modules/newbb/viewtopic.php?topic_id=457&forum=2 Jul 12 08:13:48 koen: if i pastebin a patch for that, can we skip the bugzilla step? :) Jul 12 08:13:55 03koen 07org.oe.dev * rda0a622f... 10/packages/vim/gvim_7.0.bb: vim: add a gtk2 gui version of 7.0, needs work Jul 12 08:14:24 zwelch: sure Jul 12 08:14:46 zwelch: bugzilla is purely their to to keep stuff from getting lost Jul 12 08:14:50 yeah Jul 12 08:15:02 1.08, where's /win 29 Jul 12 08:15:06 bah Jul 12 08:17:26 kergoth: hey Jul 12 08:17:33 hey Jul 12 08:17:44 kergoth: I don't get what your saying. It must be too hot :) Jul 12 08:17:49 hrm, i forgot to set up that dns change for the oz feeds, didnt i Jul 12 08:18:16 hehe. still adjusting to the new work schedule. 8 hour time difference is a bit much, staying up till 6am every night now Jul 12 08:18:21 :) Jul 12 08:20:35 hmm, autoconf 2.60? time to check its changelog Jul 12 08:21:13 any bets on how many projects will break with 2.60? Jul 12 08:21:36 koen: does gvim also work on terminal? Jul 12 08:21:54 * koen <- vim noob Jul 12 08:22:11 XorA: it has the same binary name, so I sure hope so Jul 12 08:22:26 XorA: it does Jul 12 08:23:40 ~lart rmk patch tracker Jul 12 08:23:40 * ibot shoots rmk patch tracker in the head Jul 12 08:23:42 * kergoth hugs vim Jul 12 08:23:48 ibot: botsnack Jul 12 08:23:48 aw, gee, koen Jul 12 08:28:16 hrw|work: cool, so gvim is a total replacement Jul 12 08:30:25 XorA: gvim is vim with graphical mode support. Jul 12 08:30:45 XorA: gvim == vim -g Jul 12 08:35:58 ~lart crackberry Jul 12 08:35:58 * ibot pours gasoline all over crackberry, ignites the fire, and then enjoys some toasty marshmallows with the glorious blaze Jul 12 08:39:17 zecke: sounds like gui for vim is not in oz. Jul 12 08:39:48 cant wait for a good linux based push email system for all those addicts Jul 12 08:41:07 "The real art of conversation is not only to say the right thing at the right place but to leave unsaid the wrong thing at the tempting moment." Jul 12 08:41:37 koen: http://rafb.net/paste/results/DMsq0F36.html Jul 12 08:41:49 fixes me bug Jul 12 08:42:13 although now that i think about it, i guess it didn't deserve the PR bump Jul 12 08:44:55 patch unexpectedly ends in middle of lineHunk #1 FAILED at 1. Jul 12 08:46:04 03lenehan 07org.oe.dev * r7f1dd72d... 10/packages/dosfstools/ (3 files in 2 dirs): Jul 12 08:46:04 dosfstools/dosfstools-native 2.10: Fix for lib headers which don't pull in Jul 12 08:46:04 the types that dosfstools expects such as those provided with Fedora Core 4 Jul 12 08:46:04 and 5. The mkdsofs.c program is using types of the style __u8 which it Jul 12 08:46:04 implictly gets via an include of linux/hdreg.h with some version of libc Jul 12 08:46:04 headers. Explicity include linux/types.h to ensure we get those defines Jul 12 08:46:06 directly. Resolves bug #1155. Jul 12 09:01:25 mickey|zzZZzz: lol. I just tried Qtopia 2.2.0 and their Qt is named libqt.so Jul 12 09:01:59 mickey|zzZZzz: anyway we germans overengineer. Randonly chaging the SO name is what customers demand Jul 12 09:02:21 hahah Jul 12 09:03:57 * zecke thinks Qtopia 2.2.0 only compiles in a time vacuum Jul 12 09:04:06 because on any Linux distri != debian oldstable Jul 12 09:04:15 the gcc will not allow some of the code... Jul 12 09:06:25 ~lart tk Jul 12 09:06:25 * ibot beats tk over the head with a microkernel Jul 12 09:14:17 03koen 07org.oe.dev * rbee08e33... 10/packages/angstrom/task-angstrom.bb: task-angstrom: ship module tools Jul 12 09:20:56 hey woglinde Jul 12 09:21:00 kergoth, up late, or early? Jul 12 09:21:22 late Jul 12 09:21:26 heh Jul 12 09:21:42 I'm back from a week in Italy, so I am up early Jul 12 09:21:44 working for openedhand now. 8 hour time diff. have to be up late to ensure some overlap Jul 12 09:21:48 he koen Jul 12 09:21:49 ah thats cool, how was the trip? Jul 12 09:21:57 very cool Jul 12 09:22:17 no longer working for the guys you were working with? Jul 12 09:22:42 * kergoth nods Jul 12 09:23:06 had enough of the red tape and lack of hardware and everything else that interfered with actually doing work... plus i'm working from home now, which is nice Jul 12 09:23:07 kergoth hm Jul 12 09:23:15 yeah Jul 12 09:23:22 * Crofton understands Jul 12 09:23:58 just need to ensure i keep myself discliplined about it, or i'll get lax and end up with a fucked up sleep schedule or something :) Jul 12 09:24:02 Will you stay in Phoenix? Jul 12 09:24:08 yeah Jul 12 09:24:25 staying in phoenix for probably a year or so, then we'll see Jul 12 09:24:34 not used to having the ability to actually decide where to live Jul 12 09:24:38 hehe Jul 12 09:24:42 cool Jul 12 09:25:14 wtf is open hands url? Jul 12 09:25:29 openedhand.com Jul 12 09:25:33 o-hand.com Jul 12 09:25:56 kergoth: talking about fucked up sleeping schedule. What time is it at your place now ;-) Jul 12 09:26:15 3:45 Jul 12 09:26:19 * Crofton thinks Jul 12 09:26:22 wait Jul 12 09:26:41 2:45 Jul 12 09:27:06 They don't do daylight savings time Jul 12 09:27:25 becasue they want to up b4 the sun so they can go outside in summer :) Jul 12 09:27:27 yeah, we're the same as pacific right now. which is weird Jul 12 09:27:33 hehe Jul 12 09:28:12 I love the oh descriptions of people in their position ads Jul 12 09:28:21 Professional, polite and presentable person. Jul 12 09:28:33 Generally nice and well rounded person. Jul 12 09:32:19 so o-hand has 2x chris? Jul 12 09:33:14 he zecke Jul 12 09:33:16 yeah, but i dont think anyone thinks of me as chris. i'm kergoth@o-hand.com Jul 12 09:33:19 :) Jul 12 09:33:25 heh Jul 12 09:33:45 do they call you kergoth in tele conferences? Jul 12 09:33:53 havent had one yet Jul 12 09:34:03 tim used to call me kergoth in our meetings at ti, confused people Jul 12 09:34:31 kergoth: how is tim doing? is he still at TI? Jul 12 09:34:47 nope, he's elsewhere Jul 12 09:35:00 oddly enough, he's now working on huge clusters of machines Jul 12 09:35:04 talk about a switch from embedded :) Jul 12 09:35:23 huge clusters of machines of small embedded nodes Jul 12 09:35:32 composed of... Jul 12 09:36:08 naw, not embedded Jul 12 09:36:19 zecke, I really hate xerces-c :) Jul 12 09:36:39 Crofton: hehe, I remembered so I didn't ask you if the fix was okay Jul 12 09:36:49 I was on vacatoin anyway Jul 12 09:36:52 Crofton: I hope you are fine with the change Jul 12 09:37:08 It looks like a bertter way of fixing the runConfigure stupity Jul 12 09:37:41 Crofton: only it will fail with 64bit cpus Jul 12 09:37:49 hmm Jul 12 09:38:06 your fix? Jul 12 09:38:08 it would have failed before as well Jul 12 09:38:22 Crofton: it defaults to BITSTOBUILD=32 (what a stupid name) Jul 12 09:38:37 Crofton: and they use autoconf as a portable version of sed Jul 12 09:38:40 ah, the executables fail for 64 bit machines? Jul 12 09:38:58 Crofton: I don't know, I have nothing with 64 bits Jul 12 09:39:12 I use a 64 bit to build stuff on Jul 12 09:39:23 but I haven't actually run xerces on it Jul 12 09:44:26 hu Jul 12 09:44:32 someone new from the fu Jul 12 09:44:35 ~seen CoreDump|afk Jul 12 09:44:38 coredump|afk is currently on #gpe (1d 4h 8m 23s) #oe (1d 4h 8m 23s) #openzaurus (1d 4h 8m 23s), last said: 'hrw|gone: I belive you have forgotten to push pld-onlyarm5.patch for mplayer ;)'. Jul 12 09:45:15 hm??? Jul 12 09:47:26 hi mickeyl Jul 12 09:48:38 morning folks Jul 12 09:49:04 koen: thanks for explaining. you're a pretty good python assistant by now :D Jul 12 09:50:26 :) Jul 12 09:51:57 hi mithro Jul 12 09:52:09 RP: ping Jul 12 09:52:12 hey woglinde Jul 12 09:53:17 RP: you're going to enable the collie/poodle buzzer as a standard audio device like we did in OZ 3.3 ? Jul 12 10:00:12 koen: In case you are interested, my compilation problems seem to have originated from some left-over cruft in $PATH pointing to the old mount point of OE. Sometimes there is a benefit in restarting (or closing a terminal after more than a week) as that gets rid of the cruft ;-) Jul 12 10:00:46 lol Jul 12 10:01:25 Laibsch: d'oh :) Jul 12 10:02:50 ~seen pb_ Jul 12 10:02:56 pb_ was last seen on IRC in channel #oe, 1d 13h 46m 12s ago, saying: 'hi kergoth, zecke'. Jul 12 10:03:18 koen: aren't there ipaq platforms using the HTC bootloader to boot directly into Linux? Jul 12 10:03:36 mickeyl: depends Jul 12 10:04:05 mickeyl: hx4700 and h2200 replace the second stage loader with bootldr and LAB Jul 12 10:04:41 the older ones use bootldr as first stage Jul 12 10:07:25 i see Jul 12 10:07:44 can they dual boot into wince? Jul 12 10:08:25 bye all Jul 12 10:09:33 mickeyl: probably not Jul 12 10:09:46 mickeyl: iirc the bootloader handles suspend/resume Jul 12 10:12:36 /leave Jul 12 10:18:14 Hello! Anybody knows if its possible to tell bitbake that if a library updates it should recompile the packages that depend on that library? Jul 12 10:19:07 unfortunately not Jul 12 10:19:14 such a feature would be accepted :) Jul 12 10:19:36 could you add it to the bugtracker against bitbake? Jul 12 10:19:45 shouldn't be that hard Jul 12 10:19:53 * mickeyl eyes kergoth, RP, and zecke Jul 12 10:21:49 ok. that sucks. Jul 12 10:21:53 thank you :) Jul 12 10:22:40 np Jul 12 10:22:59 koen: the #htc-linux folks are talking about replacing the HTC bootloader Jul 12 10:23:07 koen: and i'm not sure whether this is wise. Jul 12 10:23:22 mickeyl: the first stage or second stage? Jul 12 10:23:24 koen: if they can get the HTC to cooperate w/ linux, it'd be better Jul 12 10:23:28 koen: no idea actually Jul 12 10:23:43 first stage -> brick me Jul 12 10:23:44 the RamdiskRescue looks good. now we only need a "haretless" bootloader. Jul 12 10:23:48 lpotter: why did you guys though putting libqte.so into QTDIR is lame? Jul 12 10:24:21 mickeyl: keep in mind that htc has a zillion different bootloaders Jul 12 10:25:00 anyone fancy updating the libxslt package to 1.1.9? :) Jul 12 10:25:12 tkp: try :) Jul 12 10:27:15 * mickeyl waits for the parcel service again Jul 12 10:27:23 mickeyl: macbook pro? Jul 12 10:27:38 koen: no, i promised to order this not until i've submitted the thesis Jul 12 10:27:49 I ebayed a used HTC universal Jul 12 10:27:54 mickeyl: admit it. You have a rich wife? Jul 12 10:27:55 lol Jul 12 10:28:14 mickeyl: ah, another donation ;) Jul 12 10:28:26 koen: *cough* exactly Jul 12 10:28:38 zecke: not really. and I'm living a bit over my budget in the last year. Jul 12 10:30:17 ohh... that was quite painless Jul 12 10:30:21 aaargh Jul 12 10:30:33 I have no idea if it works Jul 12 10:30:35 somehow I managed to compile busybox in a way that ipkg doesn't work Jul 12 10:30:56 just literally copied the file changed the version number and updated the url Jul 12 10:32:17 tkp: Grand Unified Unit Testing is on my master plan Jul 12 10:32:31 tkp: then we would know this upgrade didn't show any regressions Jul 12 10:35:02 03koen 07org.oe.dev * r8f9357c4... 10/conf/distro/ (angstrom-2006.9.conf angstrom.conf): angstrom: remove some stray slashes from the feeds URIs Jul 12 10:50:23 mickeyl: ever looked at getting a package together for libxml2-python? Jul 12 10:53:29 no, what's that? Jul 12 10:53:49 python bindings for libxml2 Jul 12 10:54:08 is that better than a) the built-in xml package and b) pyxml? Jul 12 10:54:20 tbh, I don't know... Jul 12 10:54:28 *shrug* Jul 12 10:54:31 but a package I'm trying to install requires it Jul 12 10:54:31 * mickeyl notes to todo Jul 12 10:54:40 (kaa canvas) Jul 12 10:55:10 power outages suxx Jul 12 10:55:15 gtg Jul 12 10:55:15 indeed Jul 12 10:55:20 l8er tkp Jul 12 11:05:56 * koen is happy with his native toolchain Jul 12 11:06:26 but why did I choose my c700 to install it on? Jul 12 11:06:37 koen: coz your a muppet Jul 12 11:07:14 that must be it Jul 12 11:07:19 * koen <- muppet Jul 12 11:11:28 tkp: there is no libxml2-python 2.6.22 out. 2.6.21 doesn't compile vs. 2.6.21, looks like we need to wait for it to appear Jul 12 11:13:37 Hi all Jul 12 11:15:50 Hi all Jul 12 11:17:05 I have a problem with the alsa-utils* packages. Can anybody help me please? Jul 12 11:17:24 TEMic: details? Jul 12 11:18:38 i build my distro for a geodelx board and i'm using the geodelx alsa sound driver. when i include the alsa-utils* e.g. speaker-test, it does a segmentation fault Jul 12 11:19:26 TEMic: gcc 4.1.x? Jul 12 11:19:26 sorry, it does the seg. fault when i try to run it. Jul 12 11:19:33 gcc 3.4.4 Jul 12 11:19:40 ah Jul 12 11:19:53 hmm isn't dlopen and friends available in uClinux ? Jul 12 11:19:57 if you had gcc 4.1.x you would have needed to compile it with -O1 Jul 12 11:20:22 compiling qtopia-core fails when trying to link executables against -ldl, it does not exist Jul 12 11:20:59 it seems to bee a problem with a library Jul 12 11:23:06 EricCartman: don't try Qtopia Core Jul 12 11:23:10 EricCartman: it is unlikely to work Jul 12 11:23:25 because ? Jul 12 11:23:48 EricCartman: because even the upstream releases do not compile Jul 12 11:24:01 it compiled the lib Jul 12 11:24:11 EricCartman: and I doubt that it has been tested on any ARM, PPC, MIPS board Jul 12 11:24:21 and that doesn't make any sense, why would trolltech release a broken product Jul 12 11:24:48 EricCartman: why would Trolltech not release Qtopia Platform GPL Jul 12 11:24:56 EricCartman: It doesn't make sense, still they do Jul 12 11:25:18 koen? Jul 12 11:25:31 EricCartman: even if you get it to compile, TouchScreen, Keyboard etc. will be likely to be broken Jul 12 11:26:11 back to my question, is it anyones knowledge if uClinux does support dlopen or not ? apparently there's no libdl.[a|so] for it in uClibc/lib folder but it's an essential part, how come ? Jul 12 11:26:29 EricCartman: dlopen should work Jul 12 11:26:36 EricCartman: check your uclibc .conf file Jul 12 11:26:49 EricCartman: and eventually it doesn't need libdl for dlopen? Jul 12 11:27:12 qmake generated -ldl paramter for linker Jul 12 11:27:20 EricCartman: http://www.uclibc.org/cgi-bin/viewcvs.cgi/trunk/uClibc/test/dlopen/ Jul 12 11:27:22 that's the way on regular linux pc Jul 12 11:27:33 EricCartman: qmake ain't autoconf, nor cmake Jul 12 11:27:42 EricCartman: so fix the QMAKESPEC for uclibc? Jul 12 11:27:52 03koen 07org.oe.dev * r7bd820c4... 10/packages/bitbake/bitbake_1.4.2.bb: bitbake: update to 1.4.2 Jul 12 11:27:56 EricCartman: oh well Jul 12 11:28:08 EricCartman: it looks like uclibc has libdl Jul 12 11:28:24 EricCartman: I would suggest to look at your .config file to see if it is disabled Jul 12 11:28:35 I'm just curious: What do you try to do with QtopiaCore? Jul 12 11:28:51 engin@bzzzt-dell:~/embedded/uClinux-dist/uClibc/extra/Configs$ grep -i dl Config.bfin* Jul 12 11:28:55 yields nothing Jul 12 11:29:18 I'm looking for possibilities that I can use for a project Jul 12 11:30:57 EricCartman: I don't want to black mail QtopiaCore. But to get it working, more than compiling is needed Jul 12 11:32:11 there is a reason why it isn't in OE Jul 12 11:32:20 it was in once Jul 12 11:32:31 but i had to patch the guts out of it Jul 12 11:32:50 then i found it lacking drivers everywhere Jul 12 11:33:01 drivers which we had submitted to TT ages ago Jul 12 11:33:03 no doubt QtopiaCore is promising, but there is plenty to do Jul 12 11:33:21 looks like getting QtopiaCore4 working need purchasing license from TT and it does not give any warranty that it will build/work Jul 12 11:33:30 and as mickeyl pointed out, it is unlikely that any sort of patch will flow back to Trolltech Jul 12 11:33:36 I have evaluation license Jul 12 11:33:37 so Qtopia* is very unlike Qt Jul 12 11:34:13 well, that is why there's evaluation Jul 12 11:34:15 EricCartman: oh well, they might release working versions to possible paying customers Jul 12 11:35:30 mickeyl, RP, do13: I will add 2.6.17 into .oz354x (def_pref=-1) Jul 12 11:36:02 hrw|work: good Jul 12 11:36:17 EricCartman: okay, I just checked. uclibc has dlopen code Jul 12 11:36:29 EricCartman: check if dlopen symbols are in your library Jul 12 11:38:10 mickeyl, RP, do13: what do you think about making 2.6.17 default/2.6 for tosa/poodle/collie? those machines does not get any release with 2.6.16 so it does not change anything for them Jul 12 11:38:50 hrw|work: you could bump it up to 3.5.4.2 and make it default for all machines :) Jul 12 11:39:13 hrw|work: good plan Jul 12 11:39:35 zecke: there's no libdl, you mean it's supposed to be embedded in libc or something ? Jul 12 11:39:50 EricCartman: in the source tree there ldso/libdl Jul 12 11:39:59 koen: 2.6.17 is something which I want to leave for those who know how to use bitbake Jul 12 11:40:02 EricCartman: it could be embedded, use nm to see if you see a symbol Jul 12 11:40:17 already did Jul 12 11:40:18 it's not there Jul 12 11:40:31 EricCartman: then change your uclibc config Jul 12 11:41:21 what's the related parameter Jul 12 11:41:22 e.g. have you shared library support turned on? Jul 12 11:41:27 dunno Jul 12 11:41:33 and how did I build tons of stuff without dlopen in the first place Jul 12 11:41:52 build includes 'linking' Jul 12 11:41:59 ? Jul 12 11:42:08 need to find some food Jul 12 11:42:08 of course Jul 12 11:42:22 it didn't use dlopen then Jul 12 11:42:29 or at least didn't specify -ldl Jul 12 11:42:46 without -ldl it gives unresolved symbol errors for dlopen and friends Jul 12 11:43:37 EricCartman: well, enable Shared Library support in uclibc? Jul 12 11:44:10 CIA-9: ping Jul 12 11:44:22 03hrw 07org.oe.oz354x * r4d517470... 10/packages/python/python-pyephem_3.7b.bb: python-pyephem: added 3.7b from .dev Jul 12 11:44:27 03hrw 07org.oe.oz354x * r4ad555d2... 10/packages/linux/ (19 files in 3 dirs): linux-openzaurus: added 2.6.17 (not default) Jul 12 11:44:49 CIA-9: pöng Jul 12 11:45:04 zecke: uhmm I'm not sure which config I have used... but there's Config.bfin.default with HAVE_NO_SHARED=y Jul 12 11:45:33 btw, isn't there dlopen in Qt3 ? I've compiled Qt 3.2 embedded with this toolchain and run examples on the board Jul 12 11:46:03 EricCartman: that depends on your config, I guess Jul 12 11:46:05 no Jul 12 11:46:11 EricCartman: QCOM should use dlopen Jul 12 11:46:23 EricCartman: but if QCOM/QLibrary was compiled, I can't tell Jul 12 11:46:36 qcom in Qt 3.2 ? Jul 12 11:46:56 03mickeyl 07org.oe.dev * r188d7346... 10/packages/libxslt/ (libxslt_1.1.8.bb libxslt_1.1.7.bb libxslt_1.1.12.bb): libxslt add 1.1.2, remove 1.1.7, remove 1.1.8 Jul 12 11:47:00 03mickeyl 07org.oe.dev * rc691913c... 10/packages/libxml/ (4 files): libxml2: bring 2.6.22 up to modern standards, remove 2.6.7, 2.6.9, 2.6.10 Jul 12 11:47:04 03mickeyl 07org.oe.dev * r87830c48... 10/packages/ (3 files in 2 dirs): add python-lxml, a set of Pythonic bindings for the libxml2 and libxslt libraries. Jul 12 11:47:20 3.3.5 btw, not 3.2 Jul 12 11:47:47 koen: do you have an idea about the alsa problem? Jul 12 11:47:56 mickey|thesis: why remove version 1.1,7 and 1.1.8 abd add 1.1.2? Jul 12 11:47:59 TEMic: no idea, sorry Jul 12 11:48:06 zecke: mmh... typo, sorry :/ Jul 12 11:48:12 1.1.12 Jul 12 11:48:13 TEMic: try asking lpotter, he maintains the also stuff Jul 12 11:48:15 EricCartman: trial the TT first level support Jul 12 11:48:30 ok, thanx. :-( Jul 12 11:48:33 s/also/alsa/ Jul 12 11:48:38 zecke: actually, good idea Jul 12 11:49:07 EricCartman: I expect they will say "Works here" or "uclibc is not supported" Jul 12 11:50:19 TEMic: i recommend using gdb and strace to get more information about the SIGSEGV Jul 12 11:51:19 mickey|thesis: gdb shows an error in libpthread Jul 12 11:51:36 sounds like you should take this up with upstream Jul 12 11:51:36 zecke: well, I should first figure this dlopen thing out Jul 12 11:53:14 EricCartman: findout when libdl gets build Jul 12 11:53:30 EricCartman: I think it will be build when libso gets built Jul 12 11:54:59 whats upstream? Jul 12 11:56:39 hi all Jul 12 11:56:46 when i run it on my development machine with its libs, alsamixer and speaker-test are running. chrooting to rootfs and using the target libs crashes both apps Jul 12 11:56:58 * EricCartman is frustrated Jul 12 11:57:06 gdb 6.4 works with eabi. whom can I ask to have a look at the uclibc patch to update it if it's still needed? Jul 12 11:58:14 TEMic: upstream means "the guys who actually wrote it" Jul 12 11:58:39 how to do a feature request? Jul 12 11:59:52 ah ok Jul 12 12:01:05 or where to do a feature request - right here? Jul 12 12:09:26 03mickeyl 07org.oe.dev * r9d666f4a... 10/packages/alsa/ (9 files): alsa: remove wrong MAINTAINER Jul 12 12:10:20 03mickeyl 07org.oe.dev * r56a0dbda... 10/packages/python/python-lxml_1.0.2.bb: python-lxml 1.0.2 add DEPENDS Jul 12 12:10:22 03mickeyl 07org.oe.dev * r5029b708... 10/packages/alsa/ (13 files): alsa: add HOMEPAGE, remove 1.0.8 and 1.0.10, mark alsa-driver as BROKEN Jul 12 12:12:06 mr_nice_slacker: bugtracker Jul 12 12:14:55 03xora 07org.oe.dev * r8d2a59cf... 10/packages/tinymail/tinymail_svn.bb: tinymail_svn.bb : tinymail moved URL Jul 12 12:54:16 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Jul 12 12:54:25 morning Jul 12 12:54:41 hey chouimat Jul 12 12:55:04 hi chouimat Jul 12 13:05:27 'Contact form' - how it will be in deutsch? Jul 12 13:07:02 <_law_> hhhcz, Kontakt Formular Jul 12 13:07:12 <_law_> hrw|work, Kontakt Formular Jul 12 13:07:28 thx Jul 12 13:12:31 _law_: is it one word or two? Jul 12 13:13:05 aynway Jul 12 13:17:22 oops, now i have an other problem Jul 12 13:17:46 my glibc-2.3.5 broke with Assembler messages Jul 12 13:18:00 03mickeyl 07org.oe.dev * r84570769... 10/packages/qte/qte-common_2.3.10.inc: qte all reenable RTTI, must have been lost during a merge Jul 12 13:18:38 03mickeyl 07org.oe.dev * r8b41db42... 10/packages/konqueror/ (4 files in 3 dirs): konqueror 20060404 add gcc4 patch, remove 20050322 Jul 12 13:19:04 goxboxlive: that should be it Jul 12 13:19:13 symbol definition loop encountered at '__divdi3_internal' Jul 12 13:19:33 thx mickey Jul 12 13:20:28 it happend after updateing my oe.db Jul 12 13:21:38 mickey|thesis: how goes konqueror integration? Jul 12 13:21:58 how can i solve this error? Jul 12 13:22:47 hrw|work: no time to improve, still two issues left. a) need to remove the hard-coded Helvetica b) need to use OPIEDIR as KDEDIR on startup Jul 12 13:23:16 ok Jul 12 13:23:34 probably just 2 hours of focused work... Jul 12 13:23:39 mickey|thesis: how goes thesis? much left? Jul 12 13:24:43 hrw|work: only small steps. i now hope to be able to get it accepted by supervisor in August and submit in September. Not sure on these dates though :/ Jul 12 13:25:03 the last chapter took me much longer than expected Jul 12 13:25:09 mickey|thesis: happens Jul 12 13:25:10 and it's still not finished Jul 12 13:25:12 *nod* Jul 12 13:25:22 mickey|thesis: happens, I know you will succeed Jul 12 13:25:30 my fault anyway. it was the hardest. i should've started with the hardest, not with the easiest Jul 12 13:25:32 zecke: :) Jul 12 13:25:32 mickey|thesis: I do a website now - planned to end it in last friday Jul 12 13:27:09 i'm glad i don't run a company. my time estimates are horribly. need to improve those. Jul 12 13:27:57 mickey|thesis: oh well... Jul 12 13:37:45 * mrz80 is back (gone 00:43:29) Jul 12 13:38:32 * mickey|thesis submits a bitbake enhancement to the bugtracker Jul 12 13:38:48 TEMic: you need binutils 2.16.X not 2.17 Jul 12 13:39:57 #1163 Jul 12 13:42:57 TEMic: you should include something like the following line in your local.conf: PREFERRED_VERSION_binutils-cross = "2.16" Jul 12 13:46:28 mickey|thesis : still some bugs: Qte-mt compiled just fine, but not konq/E http://pastebin.ca/86056 Jul 12 13:47:14 strange. it works here Jul 12 13:47:22 ok Jul 12 13:49:43 I'm having some dependency grief Jul 12 13:49:51 mickey|thesis: hey, that's not the real Konqueror, is it? Jul 12 13:49:58 RROR: Cannot satisfy the following dependencies for task-bootstrap-unionroot: Jul 12 13:49:59 define "real" Jul 12 13:50:08 tinylogin modutils-initscripts linux-hotplug angstrom-version coreutils dropbear ipkg-collateral ipkg-link modutils-depmod unionfs-modules unionfs-utils unionroot unionroot-utils Jul 12 13:50:12 mickey|thesis: not some ancient crap dubbed "Embedded" Jul 12 13:50:23 mikearthur: add PREFERRED_PROVIDERS_task-bootstrap = "task-bootstrap" Jul 12 13:50:27 mickey|thesis: the browser that beats FireFox Jul 12 13:50:44 Luke-Jr: it is a new embedded port of konqueror 3.x Jul 12 13:50:46 Luke-Jr: Firefox you mean? Jul 12 13:50:51 dunno whether that qualifies as "real" for you Jul 12 13:50:58 mickey|thesis: to where? Jul 12 13:51:02 mickey|thesis: hm, does it suck like the old one? =p Jul 12 13:51:05 mikearthur: some .conf Jul 12 13:51:08 Luke-Jr: no Jul 12 13:51:11 nice Jul 12 13:51:21 lol Jul 12 13:51:46 local., $DISTRO, site. ... whereever you like Jul 12 13:51:47 mikearthur: conf/local.conf for easiness Jul 12 13:52:06 is there a Qt/X11 image-bb in OE? :) Jul 12 13:52:10 mickey|thesis: so.. you fancy trying to get these python xml bindings into oe? :p Jul 12 13:52:12 n1, cheers Jul 12 13:52:38 tkp: i tried. they don't compile. libxml2 is one version ahead of the python bindings Jul 12 13:52:48 hmm Jul 12 13:52:49 tkp: especially for OE, I imagine libffi would do better than bindings Jul 12 13:52:49 which means we'll wait for the matching release Jul 12 13:53:03 darn :( Jul 12 13:53:06 mickey|thesis: #1162 - can you look? Jul 12 13:53:30 presuming libffi supports the CPU you use Jul 12 13:53:57 hrw|work: that needs fixing in opie i'm afraid Jul 12 13:54:25 mickey|thesis: or hardcode in opie-i18n probably Jul 12 13:54:50 mickey|thesis: does pyxml work? Jul 12 13:55:01 jkp: sure. Jul 12 13:55:12 i've also added python-lxml today Jul 12 13:55:17 which is a different binding to libxml2 Jul 12 13:55:20 which do you prefer? Jul 12 13:55:23 ah - ok Jul 12 13:55:26 mickey|thesis: nice... I just did that too :) Jul 12 13:55:28 pyxml because I know it Jul 12 13:55:30 we are using kaa - do you know that? Jul 12 13:55:37 no, that's new to me Jul 12 13:55:40 kaa uses libxml2 by default Jul 12 13:56:13 i dont know pythonwell, how do you do the equivilent of #ifdef'ing in python? Jul 12 13:56:23 mickey|thesis: http://urandom.ca/mebox/canvas/ Jul 12 13:56:41 you could do Jul 12 13:56:43 try: Jul 12 13:56:45 import foo Jul 12 13:56:50 jkp: if false: ? Jul 12 13:56:51 except ImporError: Jul 12 13:56:56 ahhh Jul 12 13:56:57 import bar Jul 12 13:56:58 ok :) Jul 12 13:57:12 what if you wanted to conditionally disable something? Jul 12 13:57:17 mickey|thesis: I'd not thought about the buzzer on collie/poodle. Is it that useful? Jul 12 13:57:17 i guess that would work Jul 12 13:57:20 wrap it in """ Jul 12 13:57:31 oh, conditionally Jul 12 13:57:33 then 'if' =p Jul 12 13:57:34 coz it wouldnt hit the code at all if you had a bool protecting it Jul 12 13:57:52 hrw|work: I'm fine with a 2.6.17 kernel for 3.5.4.2 for all models. Its certainly fine for tosa/collie/poodle for now Jul 12 13:57:57 RP: the quality sucks, but having a full loudspeaker is always better than the arbitrary restriction to 4 presampled sounds (which suck as well) Jul 12 13:58:06 yeah, im used to thinking in C where you couldnt compile it at all if the library wasnt included Jul 12 13:58:09 RP: so in other words... yes, please :) Jul 12 13:58:17 * RP plans to remove the default preference -1 on 2.6.17 soon Jul 12 13:58:38 mickey|thesis: Is this the speaker connected to the amp off the wm8731 or something else? Jul 12 13:59:00 mickey|thesis: I'm certainly not planning presampled sounds :) Jul 12 13:59:06 RP: i have no idea. but i have something better... a patch, albeit against 2.4.18 Jul 12 13:59:27 RP: sound-2.4.18r2.patch Jul 12 13:59:39 RP: tc35143af is audio iirc Jul 12 13:59:51 in packages/linux/openzaurus-sa-2.4.18-rmk7-pxa3-embedix20030509 Jul 12 14:02:14 hrw|work: tc35143af is audio used in collie Jul 12 14:02:38 mickey|thesis: I'm not sure poodle has this hardware? Jul 12 14:03:04 RP: oh, ok. i have never seen one, sorry :) Jul 12 14:03:14 RP: poodle was improved when it comes to sound rather Jul 12 14:03:22 mickey|thesis: you will see Jul 12 14:03:26 ya ;) Jul 12 14:03:40 mickey|thesis: poodle gather dust near collie Jul 12 14:03:46 poodle is very like corgi but wil an inbuilt mic Jul 12 14:04:14 I do remember some buzzer driver for the pxas but I can't remember the specifics... Jul 12 14:05:04 Right, poodle emulated the buzzer but passed the data to the wm8731 Jul 12 14:05:17 So we don't care about that as its a userspace issue under 2.6 :) Jul 12 14:05:24 righto Jul 12 14:11:17 um Jul 12 14:11:34 just FYI, that article on the GPL's source distribution requirement isn't 100% accurate Jul 12 14:11:57 Part 3c applies Jul 12 14:13:12 "c) Accompany it with the information you received as to the offer to distribute corresponding source code. (This alternative is allowed only for noncommercial distribution and only if you received the program in object code or executable form with such an offer, in accord with Subsection b above.)" Jul 12 14:14:05 (obviously there's limited application, but it *is* relevant) Jul 12 14:14:22 Luke-Jr: please propose a better phrasing Jul 12 14:14:52 :-)) works! Thankx Jul 12 14:14:56 zecke: well, in the case of the example, provided MEPIS are not building their own packages, but using Debian's, they need only include Debian's offer for source Jul 12 14:15:22 Luke-Jr: I have never seen such offer from Debian :} Jul 12 14:15:40 (just out of interest) where is this debian offer? Jul 12 14:16:00 so is Debian in violation? Their binaries don't include source code Jul 12 14:16:21 Luke-Jr: the source is beside the binary on the ftp site Jul 12 14:16:28 don't they have "source debs"? Jul 12 14:16:36 thus an implicit offer for the source code Jul 12 14:16:36 Luke-Jr: so supplied on the same medium Jul 12 14:16:51 ah okay Jul 12 14:16:56 this GPL makes me nuts :) Jul 12 14:17:32 TEMic: Thank XorA and zwelch -- they figured it out Jul 12 14:18:44 hey guys Jul 12 14:18:48 3a) is what debian does, but its a suckilly written clause Jul 12 14:19:07 ive jujst been having second thoughs about our hardware strategy Jul 12 14:19:22 oh well Jul 12 14:19:39 i was wondering if you guys have any recommendations of boards or suppliers for embedded systems? Jul 12 14:19:43 3a would be met by supplyuing the software printed in a book Jul 12 14:19:54 we're doing a STB remember Jul 12 14:20:04 either way, how is it not compliance to link to sources off-site? Jul 12 14:20:10 provided that link remains valid Jul 12 14:20:29 Luke-Jr: thats not accompanying Jul 12 14:20:39 3a doesn't say it needs to be on the same medium as the binaries-- it says it needs to be on a customary medium Jul 12 14:21:17 03florian 07org.oe.dev * r94291eae... 10/packages/gpe-timesheet/gpe-timesheet_0.31.bb: gpe-timesheet: Add 0.31 Jul 12 14:21:24 03florian 07org.oe.dev * rcbb2b499... 10/packages/glib-2.0/glib-2.0_2.8.6.bb: glib: add 2.8.6 Jul 12 14:21:28 Luke-Jr: but it does say it must accompany, Id certainly think the English for accompany discounts someone elses site Jul 12 14:21:31 03florian 07org.oe.dev * r596b10cd... 10/packages/glib-2.0/ (glib-2.0_2.10.1.bb glib-2.0_2.10.2.bb): glib: drop broken 2.10.1 and 2.10.2 Jul 12 14:21:43 anyway, GPL compliance is easy if you just keep the build directory linked from the page ;) Jul 12 14:22:20 if someone doesn't want to go upstream, they can dig through the uncompressed source directories Jul 12 14:22:42 (or better yet, just throw a big blob of the build dir up as tar.bz2 ;) Jul 12 14:23:05 and throttle the bandwith back to 0.1kB/s Jul 12 14:23:11 as some companies do Jul 12 14:23:15 heh Jul 12 14:23:26 the current debate is wether linking or refering back is valid Jul 12 14:23:41 I don't see why not Jul 12 14:23:51 YOUR not providing source access Jul 12 14:23:53 as long as the links are ensured to be valid Jul 12 14:24:10 therfore YOUR violating the license Jul 12 14:24:17 if it a violation to use a free webhost, then? Jul 12 14:24:19 Luke-Jr: or print it in a book and ship it to customer, PGP proved books are machine readable Jul 12 14:24:19 its a semantic debate Jul 12 14:24:20 rule of thumb: "be in control of the sources yourself" Jul 12 14:24:52 XorA: "a medium customarily used for software interchange" Jul 12 14:25:09 Luke-Jr: again, pgp used books for software interchange Jul 12 14:25:26 Luke-Jr: magazine used to print code in them every month Jul 12 14:25:29 ... doesn't make it customary Jul 12 14:25:33 its a ruling that will expose more than 75% of linux distros Jul 12 14:25:40 !!! Jul 12 14:25:41 Luke-Jr: I have no frelling idea! Jul 12 14:25:43 my library is probably full of old book scontaining source Jul 12 14:26:06 my daughter got ahold of a Kubuntu thing Jul 12 14:26:17 managed to get the CD out and in her mouth before i noticed... Jul 12 14:26:24 oh well Jul 12 14:26:30 you need to buy her a toy Jul 12 14:26:34 lol Jul 12 14:26:42 she doesn't want the toy, it seems Jul 12 14:27:31 so what happened to the OE page with status on 2.6 cxx0 support? Jul 12 14:27:37 i wish i would remember where the article was ... Jul 12 14:28:26 Luke-Jr: http://rpsys.net/openzaurus/ Jul 12 14:28:37 you know, a bug in the supposed application of the ruling-- someone says the users need the exact source to fix bugs... if the bugs aren't in the newer version, then WTF do they need fixing? Jul 12 14:28:45 koen: thanks Jul 12 14:29:29 Luke-Jr: although it is a bit out of data Jul 12 14:29:42 probably stems from the fact in certain cases the newest version is not the valkid on Jul 12 14:29:47 valid* Jul 12 14:30:27 koen: hm, so the sound stuff was completed? including microphone/headset? Jul 12 14:30:35 hrw|work: doesn't oz contain gpsdrive? Jul 12 14:30:40 Luke-Jr: ask RP Jul 12 14:31:01 RP: Jul 12 14:31:23 koen: ./x11/gpsdrive_2.10pre2-r1_arm.ipk Jul 12 14:31:51 hrw|work: ok, so pavel is really as dumb as a brick.... Jul 12 14:32:34 sounds like the chip-mic-headset is done Jul 12 14:32:40 speakers still need work Jul 12 14:32:58 emte: re cxx0? Jul 12 14:33:26 jkp: you might want to ask your question in #elinux Jul 12 14:33:29 i'd have to look in logs sec Jul 12 14:33:44 Kerwood: ok, thanks for the recommendation Jul 12 14:35:59 Jul 11 15:50:31 CIA-9 rpurdie org.oe.dev * rc5b3cb5a... /packages/linux/ (4 files in 3 dirs): linux-oz-2.6.17 + 2.6.17+git: Add poodle audio patch. Gives working playback via headphones and working mic support. The speaker will probably also work when we fix the corgi speaker bug. Jul 12 14:36:26 so which ever Zaurii poodle corresponds to Jul 12 14:36:32 !poodle Jul 12 14:36:32 poodle is sharp sl-b500/5600, or a dog Jul 12 14:36:46 another bot? Jul 12 14:36:56 shrug, isn't that one pretty old? Jul 12 14:37:07 I recognize the response format, anyhow Jul 12 14:37:18 hmm its dupy's Jul 12 14:37:28 nice to know someone is using it Jul 12 14:37:29 243/843 revs written wee Jul 12 14:37:39 monotone needs to go faster Jul 12 14:38:06 i should just add it to cron or something Jul 12 14:38:30 Luke-Jr: see http://openembedded.org Jul 12 14:39:27 koen: eh? Jul 12 14:40:01 Luke-Jr: monotone migration Jul 12 14:40:01 should just build a montone-chip Jul 12 14:40:09 Luke-Jr: mtn 0.27 switch Jul 12 14:40:09 dedicated processor Jul 12 14:40:23 FPGA for a SCM? Scary. Jul 12 14:40:26 doesn't change the fact that I need to update, does it? O.o Jul 12 14:40:27 lol Jul 12 14:40:31 or would upgrading make it faster? Jul 12 14:40:44 would use less system CPU time Jul 12 14:40:55 faster ... doubtful Jul 12 14:41:27 actually ... Jul 12 14:41:32 yeah it would be a bit faster Jul 12 14:41:33 Luke-Jr: 0.26+ are much faster then 0.25- Jul 12 14:41:58 about 10 times faster Jul 12 14:42:10 oh, that's pretty significant ;) Jul 12 14:42:23 but the db are not backwards compat Jul 12 14:42:35 so I'll need a fresh checkout? Jul 12 14:42:43 or convert it Jul 12 14:42:57 Luke-Jr: read the article Jul 12 14:42:58 I will wget new database Jul 12 14:43:03 hm, guess I'll need to unmask monotone Jul 12 14:43:31 zecke: "The read only users will need to get a new database once the migration is done." Jul 12 14:43:49 I will miss most of migration on beach :) Jul 12 14:44:01 hrw|work: poor boy ;) Jul 12 14:44:09 hrw|work: just make sure to push all your stuff before you leave Jul 12 14:44:12 erm Jul 12 14:44:18 monotone 0.25 is latest in Portage Jul 12 14:44:24 Luke-Jr: Sound is basically sorted for all models except collie Jul 12 14:44:31 !collie Jul 12 14:44:32 collie is the codename for the Sharp Zaurus 5000 and 5500 PDA's, or a dog Jul 12 14:44:37 RP: incl mic? Jul 12 14:44:44 Luke-Jr: Including mic Jul 12 14:44:49 The speaker issue is some minor bug in ASoC Jul 12 14:44:50 yay voip Jul 12 14:46:28 zecke: I will Jul 12 14:46:57 zecke: hopefully most (if not all) of them are pushed Jul 12 14:47:38 zecke: and rest are out-of-tree kept Jul 12 14:48:32 can you specify that svn should pull down the latest revision? Jul 12 14:48:39 ie HEAD Jul 12 14:48:45 hi Jul 12 14:48:51 tkp: see bitbake manual Jul 12 14:49:12 zecke: I have looked... can I use the phrase HEAD for revision Jul 12 14:49:33 I felt out the revision flag, and by default it is pulling -r {20060712} Jul 12 14:49:33 tkp: go to doc type make manual Jul 12 14:49:38 tkp: look at the fetcher chapter Jul 12 14:50:49 RP: I shall try and get that debug info to Liam tomorrow, been rushed off my feet here today Jul 12 14:56:03 tkp: I might have forgotten to update it on the server, but you can generate it Jul 12 14:56:17 tkp: basicly JustinP patched BitBake to allow SRCDATE = "now" Jul 12 14:56:28 tkp: so you could write SRCDATE_yourpn = "now" Jul 12 14:56:30 i used rev=HEAD... seemed to work Jul 12 15:01:14 XorA: ok, thanks. If not, I'll probably get around to looking at it at some point :) Jul 12 15:02:04 RP: although one strange thing is you 2.6.17 with ASOC update patch is nice and quite, Liams makes lots of static and popping for me, so I am thinking you are doing something slightly different than him Jul 12 15:03:49 XorA: sounds like a toolchain issue Jul 12 15:04:15 RP: I am using same toolchain for both Jul 12 15:07:29 XorA: I'm not sure there should be any difference then. Certainly I've not made any changes that should influence that... Jul 12 15:07:51 XorA: Have you tried diffing the patches? Jul 12 15:07:53 tkp: well whatever Jul 12 15:08:04 tkp: either you ask and listen, or you don't ask at all Jul 12 15:08:15 tkp: HEAD will download your stuff exactly once Jul 12 15:08:27 tkp: if you rebuild this package next year, it will still use the same tarball Jul 12 15:08:38 zecke: right ok Jul 12 15:08:50 sorry... got distracted for a minute Jul 12 15:09:28 CoreDump|home: Morning Jul 12 15:10:00 hvontres|work: I heard the good news about the sound driver ;) Jul 12 15:10:19 zecke: so I add SRCDATE_python-kaa = "now" Jul 12 15:10:30 and leave off rev from the svn url? Jul 12 15:11:24 CoreDump|home: Yes !!! I have built a bootstrap image and now I am waiting to see if my opie-image build finishes... I started it at ~1am :) Jul 12 15:11:54 good luck Jul 12 15:12:41 RP: not yet, on tuesday bitbake was doing wierd shit, and today Ive bew rushed off feet at work Jul 12 15:13:02 CoreDump|home: Any plans for building new images? I would love to be a tester for a new hentges(tm) image...:) Jul 12 15:13:21 * RP knows all about wierdness in builds. I need a clear out of mine... Jul 12 15:13:37 /quit Jul 12 15:19:31 RP: It looks like the g_ether driver also works with Windblows XP ...:) Jul 12 15:20:09 hvontres|work: which Z? Jul 12 15:20:15 hvontres|work: Yay :) Jul 12 15:20:27 JustinP: poodle, not cxx00 Jul 12 15:21:52 RP: :-( oh well Jul 12 15:22:14 * JustinP has a wireless card finally so RNDIS isn't as important....would still be nice, though Jul 12 15:22:36 ah yes, and one fo those SD cards with builtin USB.....that's really nice for copying images around ;-) Jul 12 15:22:39 hvontres|work: yeah why not. Is the patch in .oz already? Jul 12 15:23:01 CoreDump|home: care to test an e-image? Jul 12 15:23:08 JustinP: Its on the todo list Jul 12 15:23:13 JustinP: sure Jul 12 15:23:23 RP: I know. You have notes in your inbox ;-) Jul 12 15:23:52 CoreDump|home: The kernel patch is. If you wait 5 mins, I'll add the module autoloading bits Jul 12 15:24:27 RP: excellent Jul 12 15:25:00 CoreDump|home: 2.6.17 is in .oz354x Jul 12 15:27:08 Here are the instructions on how to get Windblows to work : http://www.handhelds.org/moin/moin.cgi/WindowsXpUsbNetworkHowTo Jul 12 15:27:50 speaking about autoloading... do we autoload oss emu? Jul 12 15:28:10 hrw|work: yes Jul 12 15:28:47 mickey|thesis: As you probably know, I'm trying to write a package to install python-kaa Jul 12 15:28:57 this stuff includes some c bindings Jul 12 15:29:16 and I don't think distutils alone is installing them Jul 12 15:29:17 hvontres|work: I've started a build, will take a few hours probably Jul 12 15:29:46 do you have a good example of one of your python packages that also installs c bindings? Jul 12 15:30:05 more dependency grief :s Jul 12 15:30:12 03rpurdie 07org.oe.dev * r1a01264e... 10/ (3 files in 2 dirs): poodle: Autoload sound modules (and cleanup modules for other models that should be included via moudule dependencies) Jul 12 15:30:13 Cannot satisfy the following dependencies for task-bootstrap-unionroot: Jul 12 15:30:17 tinylogin modutils-initscripts linux-hotplug angstrom-version coreutils dropbear ipkg-collateral ipkg-link modutils-depmod unionfs-modules unionfs-utils unionroot unionroot-utils Jul 12 15:30:24 any ideas? Jul 12 15:30:32 mikearthur, build them? Jul 12 15:31:02 "Hey I cannot find these programs that task-bootstrap-unionroot requires" Jul 12 15:31:39 Kristoffer: whatever happened to automatic dependency resolution? Jul 12 15:31:40 mikearthur: PREFERRED_PROVIDER_task-bootstrap=task-bootstrap Jul 12 15:31:48 koen: thank you Jul 12 15:32:49 ~lart mc implementation of ftpfs Jul 12 15:32:49 * ibot cuts mc implementation of ftpfs into thin stripes Jul 12 15:33:31 03rpurdie 07org.oe.oz354x * r0c38f65e... 10/ (2 files in 2 dirs): poodle 2.6: autoload sound modules (from .dev) Jul 12 15:35:20 grrr Jul 12 15:35:25 bitbake is broken in OE Jul 12 15:35:52 on my z, python is not in #!/data/build/koen/OE/build/tmp/angstrom/staging/i686-linux/bin/python Jul 12 15:36:53 really? Jul 12 15:37:08 and the RDEPENS are wrong Jul 12 15:37:14 ImportError: No module named glob Jul 12 15:38:15 03koen 07org.oe.dev * rbc5f5478... 10/conf/distro/angstrom-2006.9.conf: angstrom-2006.9: setp preferred provider for task-bootstrap Jul 12 15:43:56 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Jul 12 15:43:56 koen: python stuff need RDEPENDS handwriting Jul 12 15:44:12 hrw|work: filling it in right now Jul 12 15:44:27 install ipkg, update .bb, run bitbake, repeat Jul 12 15:45:18 I'm current;y at RDEPENDS += "python-shell python-lang python-stringold python-math" Jul 12 15:47:41 o_O my pda runs this Jul 12 15:48:34 cant be Jul 12 15:48:49 ? Jul 12 15:52:30 * raduga yow Jul 12 15:52:54 * raduga just suddenly came into possession of some... ancient hw Jul 12 15:53:08 not "ancient" maybe, but pretty old :/ Jul 12 15:53:23 WallStreet era powerbook, jornada 540 Jul 12 15:53:40 i think both have dead batteries Jul 12 15:54:03 root@c7x0:/usr/src/OE$ bitbake Jul 12 15:54:04 Nothing to do. Use 'bitbake world' to build everything, or run 'bitbake --help' Jul 12 15:54:06 finally Jul 12 15:54:41 Cyorxamp, your PDA runs OE? How long does a build take? Jul 12 15:54:50 *g* Jul 12 15:54:55 3 months Jul 12 15:55:04 nah Jul 12 15:55:05 it's a Sharp Zarus SL5500... running OpenZaurus Jul 12 15:55:11 oh Jul 12 15:55:11 my slug can do it in a few days :) Jul 12 15:55:27 Cyorxamp: pssst, this is #openembedded, not #openzaurus Jul 12 15:55:27 Does OE build images that run on any digital cameras? Jul 12 15:55:39 my camera runs DOS :) Jul 12 15:56:07 koen, how nice... why don't you run around and spread your contempt for others to the rest of the community Jul 12 15:56:15 Then we can all be one angry group! Jul 12 15:56:17 koen a lot runs dos Jul 12 15:56:19 Cyorxamp, cool it! Jul 12 15:56:28 Cyorxamp: this is #oe not #openzaurus Jul 12 15:56:37 koen: does it really? Jul 12 15:56:38 koen: DOS? please hand in your Geek Badge on the way out, thank you! Jul 12 15:56:44 And openzaurus takes parts of oe Jul 12 15:56:48 Luke-Jr: yes Jul 12 15:56:58 CoreDump|home: better than a proprietary OS that doesn't have a replacement! Jul 12 15:57:02 koen: what camera? Jul 12 15:57:04 Cyorxamp: read the topic: "his is still not a distribution support channel" Jul 12 15:57:08 Luke-Jr: canon eos Jul 12 15:57:11 Cyorxamp: OpenZaurus is *built* from OE Jul 12 15:57:20 dos is cheap and fits the rt stuff in this case Jul 12 15:57:21 koen: eos is a full model #? Jul 12 15:57:23 Luke-Jr: most canon cameras use dos (not ms-dos) Jul 12 15:57:26 Cyorxamp is just trying to get banned from as many channels as he can. Jul 12 15:57:34 koen: is it replacable by FreeDOS? Jul 12 15:57:39 bye Jul 12 15:57:53 Luke-Jr: no idea, it works well enough for me Jul 12 15:57:54 woglinde: wait Jul 12 15:57:55 Seriously... Jul 12 15:58:02 Cyorxamp is quite a nut xase Jul 12 15:58:05 *case Jul 12 15:58:06 IMO Jul 12 15:58:17 Luke-Jr: it it were a 350d I'd reflash it :) Jul 12 15:58:24 koen: yes, but can you hack the code to do new crap Jul 12 15:58:32 tkp: well behaved python extensions should do that automatically, e.g. see python-pyxml. distutils just builds the C extension automatically. Jul 12 15:59:09 mickey|thesis: it just seems to be trying to link to stuff on my host system Jul 12 15:59:24 tkp: that's not good. then you need to patch setup.py Jul 12 15:59:31 koen: any idea if the CPU can handle 32-bit mode? Jul 12 15:59:47 Luke-Jr: no idea, it's a custom cpu Jul 12 16:00:15 koen: aww, you seriously have no interest in getting OE to run on it? =p Jul 12 16:00:22 I don't even know what architecture it uses :) Jul 12 16:00:22 mickey|thesis: would you mind taking a quick look at this .bb file just to make sure I'm not doing something stupid? Jul 12 16:00:27 Lu :p Jul 12 16:00:30 tkp: sure, pastebin it Jul 12 16:00:36 http://rafb.net/paste/results/rCeE1I33.html Jul 12 16:00:56 koen: x86-compatible Jul 12 16:01:01 Google says Jul 12 16:01:14 03koen 07org.oe.dev * r55d39127... 10/packages/bitbake/bitbake_1.4.2.bb: bitbake 1.4.2: add correct RDEPENDS, shebang is still pointing to staging Jul 12 16:01:41 mickey|thesis: any idea on how to fix the shebang of the bitbake ipkg? Jul 12 16:01:50 ? Jul 12 16:02:13 mickey|thesis: and setup.py looks like this: Jul 12 16:02:13 http://rafb.net/paste/results/1GV0WC66.html Jul 12 16:02:30 koen: ah. no idea, i'll take a look Jul 12 16:02:32 mickey|thesis: the first line of /usr/bin/bitbake is "#!/data/build/koen/OE/build/tmp/angstrom/staging/i686-linux/bin/python" Jul 12 16:02:55 koen: i see. will take a lok Jul 12 16:03:01 thanks Jul 12 16:03:02 tkp: builds perfectly here Jul 12 16:03:09 Packaged contents of python-kaa into /local/pkg/oe/c7x0/tmp/deploy/ipk/python-kaa_cvs-02_armv5te.ipk Jul 12 16:03:10 try running : Jul 12 16:03:13 python Jul 12 16:03:19 import kaa.canvas Jul 12 16:03:36 i have no working arm device accessable Jul 12 16:03:51 hmm Jul 12 16:04:03 see the contents of the package Jul 12 16:04:41 basically, when I run that, it complains 'ImportError: cannot import name display' Jul 12 16:04:53 hmm? i thought it's a build-time problem Jul 12 16:04:58 and in the traceback it lists files from the host system Jul 12 16:05:23 http://rafb.net/paste/results/FhcOXP87.html Jul 12 16:06:17 http://rafb.net/paste/results/omXwYx75.html Jul 12 16:06:55 i'm not quite sure what you are doing there Jul 12 16:06:59 you build for x86? Jul 12 16:07:04 mickey|thesis: yes Jul 12 16:07:16 aha. you installed the package into a chroot or something? Jul 12 16:07:23 http://wiki.openzaurus.org/RoadMap Jul 12 16:07:30 I'm simply trying to use the kaa module on thee target system Jul 12 16:07:44 mickey|thesis: I have tried from a chroot Jul 12 16:08:23 and? Jul 12 16:08:23 I'll try actually booting into the system rather than running it from chroot Jul 12 16:08:33 that error is from chroot Jul 12 16:08:48 hmm... it doesn't look like that Jul 12 16:08:54 it looks like you're running stuff from your build directory Jul 12 16:09:02 doesn't look like you installed a package into / Jul 12 16:09:25 into / ? Jul 12 16:09:30 ipkg install Jul 12 16:09:32 or something Jul 12 16:09:58 what package... all the kaa stuff is in the rootfs Jul 12 16:10:26 so you built an image and are chrooting into $BUILDDIR/tmp/rootfs ? Jul 12 16:10:49 well actually, I mounte an ext2 file and chrooted into that Jul 12 16:11:00 but that's the same thiing prety much Jul 12 16:12:10 that path in the traceback doesn't look like you're running python from a chroot Jul 12 16:12:19 exactly! Jul 12 16:12:23 yet I am Jul 12 16:12:35 aha Jul 12 16:12:57 well, i have not enough context information to be of any help here. i will try to build that package and run it when i have a chance. Jul 12 16:13:21 ok, that would be appreciated Jul 12 16:13:43 is there anything else I can tell you that may give you more of an idea? Jul 12 16:13:52 I'm at a los here :( Jul 12 16:17:57 I'm really confused as to how that traceback can refer to files outside of the chrooted environment Jul 12 16:23:52 hrw|work: the oz-logo in the wiki has a different background colour as wiki (246/246/246), I suspect this stems from the palette of the gif Jul 12 16:24:30 CoreDump|home: Do you have any .bb files for genrating your opie and cardfs images for poodle? Jul 12 16:27:09 hvontres|work: Its in CVS, yes Jul 12 16:27:17 hold on Jul 12 16:27:31 so... anybody heard of getting embedded linux onto a digital camera? Jul 12 16:27:40 CoreDump|home: push them into oe Jul 12 16:27:52 I go home Jul 12 16:27:53 ARgh, will every distro need to keep their own sources to be GPL compliant? Jul 12 16:28:15 Kristoffer: it looks that way. yes. Jul 12 16:28:17 Kristoffer: that's probably the easiest method Jul 12 16:28:31 CoreDump|home, feels like my webspace is running lower and lower :D Jul 12 16:28:36 Kristoffer: you can store them on cd/dvd and keep at home Jul 12 16:28:43 right Jul 12 16:28:46 Kristoffer: then charge anyone who want a copy from you Jul 12 16:28:47 or someone can just host source and offer agreements to keep hosting it for the 3 year period Jul 12 16:28:48 Kristoffer: just don't forget the offer Jul 12 16:28:52 Aha :D Jul 12 16:28:59 hrw|work: then you need to support it 3 years Jul 12 16:29:35 Luke-Jr: I hope that in 2009 I will still be in OE business Jul 12 16:29:36 Kristoffer: or only give out sources for specific stuff on request, that's basically what I do. Jul 12 16:29:45 CoreDump|home, oki, good idea Jul 12 16:29:52 I think it'd work out best for the big mirrors (the ones who mirror lots of distros) to just offer source providing agreements ;) Jul 12 16:30:00 CoreDump|home: you have mirror of 3.5.4.1 sources Jul 12 16:30:09 CoreDump|home: contacted ibiblio maybe? Jul 12 16:30:15 hrw|work: you hope to be using the exact version of software from today in 2009? Jul 12 16:30:17 hrw|work: I was thinking about hentges ;) Jul 12 16:30:27 hrw|work: nope. We have enough mirrors by now IMO Jul 12 16:30:52 I'll add one tomorrow and we have another offer Jul 12 16:30:53 grrr Jul 12 16:31:04 ~lart mediawiki for erasing half of my fixes Jul 12 16:31:04 * ibot hooks into a hydrant and hoses mediawiki down for erasing half of my fixes Jul 12 16:31:15 ibot made fixes? Jul 12 16:31:28 ~resolve feeds.openzaurus.org Jul 12 16:31:37 ~nslookup feeds.openzaurus.org Jul 12 16:31:38 I can't find feeds.openzaurus.org in DNS Jul 12 16:31:45 kergoth`zzz: ARGH Jul 12 16:31:51 hehe Jul 12 16:32:09 CoreDump|home: saw RoadMap in OZ wiki? Jul 12 16:32:23 hrw|work: not yet Jul 12 16:32:47 * koen had just finished fixing spelling in grammar in the roadmap Jul 12 16:32:59 s/in/and/ Jul 12 16:33:55 koen: thx Jul 12 16:34:21 hrw|work: but somehow media wiki thrashed half of them Jul 12 16:34:26 koen: ;( Jul 12 16:36:08 argh Jul 12 16:36:12 * JustinP lost his 2.4 kernel Jul 12 16:36:23 03kristoffer 07org.oe.dev * r1fa6c1c3... 10/ (18 files in 4 dirs): Jul 12 16:36:23 (KERNEL CHANGES) Jul 12 16:36:23 conf/distro/jlime-donkey.conf : minor updates Jul 12 16:36:23 linux/linux-jlime-arm-2.6.17/* : Fixes/updates of patches and config Jul 12 16:36:23 linux/linux-jlime-sh3-2.6.17/LinuxSH-2.6.17 : Update to CVS 20060712 Jul 12 16:36:24 * update of cvs from linuxsh, fixes some backlight/suspend issues. Jul 12 16:36:26 linux/linux-jlime-sh3-2.6.17/defconfig_jlime : Minor update to reflect patch. Jul 12 16:37:33 CoreDump|home: Now that we have USB and Sound working, are there any major problems left with 2.6 on poodle? Jul 12 16:37:58 nothing major IIRC Jul 12 16:38:16 charging works? Jul 12 16:38:36 hrw|work: the rest of the spelling stuff in in /Talk Jul 12 16:38:39 hrw|work: As far as I can Tell yes :) Jul 12 16:39:05 hvontres|work: http://oz.hentges.net/openembedded/hentgescvs/daily-snapshot/2006/July/hentgescvs_12-Jul-2006.tar.gz Jul 12 16:39:27 there's a anonymous CVS server, too. Jul 12 16:39:40 CoreDump|home: Thanks. I will try that out later :) Jul 12 16:39:45 lpotter: can you explain why putting libqte.so into the Qtopia lib directory is what your customers expect? Jul 12 16:40:13 hi zecke Jul 12 16:40:32 CoreDump|home: pssst: http://openzaurus.org/wordpress/kernel-26-on-poodle-status-page/ Jul 12 16:40:51 koen: psst, I wrote that ;) Jul 12 16:40:55 anyone know what package libgif and libjpeg are part of? Jul 12 16:42:31 koen: whats wrong with 'it is' that you change it to "it's"? Jul 12 16:42:36 CoreDump|home: and it's outdated Jul 12 16:42:46 hrw|work: personal preference Jul 12 16:43:10 RP: What did you use to test the sound on poodle? I am trying to get speaker-test to work... but I think I might be missing something. Jul 12 16:43:32 hvontres|work: madplay foo.mp3? Jul 12 16:44:00 koen: will try later.. installing vorbis-tols right now. Jul 12 16:44:11 hvontres|work: aplay test.wav? Jul 12 16:45:01 koen: added Jul 12 16:45:23 cu Jul 12 16:59:08 * mrz80 is back (gone 01:15:12) Jul 12 17:13:23 I'm working on getting an LCD working with the Linux Frame buffer, once I have something that I think works, what is a good way to test it out, is there a simple app in OE that I can use to just send an image? Jul 12 17:14:13 Gerrath: cat /dev/urandom > /dev/fb0 Jul 12 17:14:25 that should give you a very colourfull picture Jul 12 17:14:59 koen, thanks thats exactly what I was looking for, something very simple like that. Jul 12 17:15:50 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Jul 12 17:20:01 RP: should the pcm-oss module still be loaded ? Jul 12 17:35:27 hvontres|work: yes, its the oss emulation - lots of apps need it Jul 12 17:37:10 hvontres|work: I use madplay and aplay for testing Jul 12 17:37:25 hvontres|work: Use headphones, not the speaker as that doesn't work quite right yet Jul 12 17:38:49 RP: Actually, I managed to get sound out of the speaker, but not the headphones while the headphones were plugged in ;0 Jul 12 17:39:31 hvontres|work: Thinking about it, the jack detection isn't wired up atm ;-) Jul 12 17:39:37 (in software) Jul 12 17:43:54 I'm trying to bitbake libx11-native, and it's defaulting to -6.2.1+cvs20060712-r6, which fails... can I tell it to not use cvs? Jul 12 17:44:30 OK, something is still not quite right. Now glibc will not build. http://rafb.net/paste/results/PIdVxt73.html I am at my unimpressive wisdom's end. What patch is quilt trying to apply? If I knew that maybe I could find out the reason it is failing. Jul 12 17:45:38 I trieed setting PREFERRED_VERSION_libx11-native = "6.2.1", but it still seems to try it with +cvs.. Jul 12 17:46:02 Laibsc1: try the run.xxx file Jul 12 17:55:37 woglinde: hey Jul 12 17:55:46 hey :( Jul 12 17:59:20 hm vmware server is final Jul 12 18:04:24 JustinP: Thank you for the information about the runXXX script. I ran it. It of course fails. There is no more information than during the bitbake phase. Looking at the lines with quilt in the runXXX script does not reveal the problem. I can call that line fine from the shell. Jul 12 18:04:37 It just tests whether any patches are applied it seems. Jul 12 18:05:09 But when the line is called from the runXX script it seems that the syntax is somehow wrong. Jul 12 18:05:28 I still fail to see what is going wrong here :-( Jul 12 18:07:29 is there a package for libx11-dev? Jul 12 18:08:17 Laibsch: did you try running it with the quilt in your staging dir? Jul 12 18:08:25 Laibsch: check the PATH in there and see if you can find another quilt Jul 12 18:17:23 or... does anyone know which package provides glx.h? Jul 12 18:18:49 re koen Jul 12 18:20:40 hey woglinde Jul 12 18:21:00 *sigh* Jul 12 18:21:20 I got zecke sick Jul 12 18:24:14 woglinde: you should bathe more often :) Jul 12 18:24:21 joking, of course Jul 12 18:25:38 kerwood sorry Jul 12 18:25:43 that istn really funny Jul 12 18:26:31 woglinde: Forgive me, it seemed liked zecke was OK Jul 12 18:27:19 kerwood wasnt here Jul 12 18:31:10 koen: can you do me a favour? Jul 12 18:31:48 goxboxlive: I can try Jul 12 18:32:11 do you have konq/E the 2006 version based on qte-mt? Jul 12 18:32:47 nope Jul 12 18:32:54 * koen has no qt/e software Jul 12 18:32:57 ko, then you cant do me a favour :-) Jul 12 18:33:01 i c Jul 12 18:33:18 * koen hides his EZX phone Jul 12 18:45:14 RP: does alsamixer save it's settings? Jul 12 18:45:33 hvontres|work: They stay active. use alsactl store to save Jul 12 18:45:35 hvontres|work: 'alsactl store; Jul 12 18:45:51 RP: thanks Jul 12 18:46:11 and 'alsactl restore' to get them back Jul 12 18:46:20 but iirc zaurusd does that for you Jul 12 18:46:31 * koen looks at RP Jul 12 18:48:46 Why not use unversioned distros? Jul 12 18:49:00 use them for what? Jul 12 18:49:05 building images Jul 12 18:49:13 you can if you want to Jul 12 18:49:20 but don't complain about breakage Jul 12 18:49:23 Luke-Jr: because it's not likely to work Jul 12 18:49:24 well, someone put a warning on it Jul 12 18:49:31 JustinP: it used to... Jul 12 18:49:38 once upon a time Jul 12 18:49:42 soemtimes it will, sometimes it won't Jul 12 18:49:55 they tend to be broken in one way or another Jul 12 18:49:59 oh well Jul 12 18:50:05 are the -unstable kept semi-recent? Jul 12 18:50:07 hmmm looks like 2.6.16 is still the default for .oz. I guess I should hard-wire it to 2.6.17 for poodle huh? Jul 12 18:50:09 'unpredictable' is a better word Jul 12 18:50:24 CoreDump|home: didn't hrw set 2.6.17 for collie,poodle and tosa? Jul 12 18:50:32 iirc he wanted to Jul 12 18:50:40 just rebuilt poodle, .16 got pulled in Jul 12 18:50:46 do I need bitbake trunk for the massive speedups I hear about? Jul 12 18:50:59 no Jul 12 18:51:03 Luke-Jr: 1.4.2 release Jul 12 18:51:13 the release-1.xx should have it Jul 12 18:51:16 I'm using branches/1.4 and it's still slow Jul 12 18:51:21 * koen feels like a ricer Jul 12 18:51:28 I have bitbake running on my z Jul 12 18:51:33 lol Jul 12 18:51:57 svn://svn.berlios.de/bitbake/branches/bitbake-1.4 to be exact Jul 12 18:52:06 rev 548 Jul 12 18:52:38 * CoreDump|home wonder where 2.6.16 is set Jul 12 18:52:54 http://prdownload.berlios.de/bitbake/bitbake-1.4.2.tar.gz to be exact Jul 12 18:52:55 oh well, -i works Jul 12 18:53:05 CoreDump|home: def_pref = -1 in 2.6.17? Jul 12 18:53:25 ah lol Jul 12 18:53:31 Luke-Jr: bitbake-1.4 branch is *far* faster than it used to be Jul 12 18:53:35 CoreDump|home: you could do DEFAULT_PREFENCE_poodle = "1" Jul 12 18:53:37 Luke-Jr: even first run of parsing is fast Jul 12 18:53:49 JustinP: well, it took a few minutes here Jul 12 18:53:54 Luke-Jr: if it's not really fast for subsequent parses you're editing local.fonc probably Jul 12 18:53:57 koen: I'll do that, thanks Jul 12 18:54:03 not that I'm complaining... -i works, and it used to be a ton slower Jul 12 18:54:06 Luke-Jr: a few minutes is better than 10 minutes Jul 12 18:54:17 Luke-Jr: subsequent parse should be a few seconds Jul 12 18:54:21 ~8mb cache is better than 400MB Jul 12 18:54:23 koen: okay Jul 12 18:54:31 CoreDump|home: add PREFERRED_VERSION_linux-openzaurus = "2.6.17" Jul 12 18:54:33 ibot: forget 8mb Jul 12 18:54:33 koen: i didn't have anything called '8mb' to forget Jul 12 18:54:41 ibot: forget 8mb cache Jul 12 18:54:42 i forgot 8mb cache, koen Jul 12 18:54:43 yes, a reparse is fast :) Jul 12 18:55:06 hvontres|work: that's even better =) Jul 12 18:55:43 CoreDump|home: works for me... opie image almost done.... after ~11 hrs :( Jul 12 18:55:58 just finished rebuilding gpe and opie Jul 12 18:56:22 CoreDump|home: heh, need to feed hamsters better :) Jul 12 18:57:10 CoreDump|home: have you tried altboot from an image file on SD yet? Jul 12 18:57:17 hvontres|work: I had a flying start so to speak. I only had to rebuild changes from an acient pre .1 version Jul 12 18:57:26 hvontres|work: not lately no Jul 12 18:57:51 CoreDump|home: And you are probably not using a "blazing" dual PII 333 :) Jul 12 18:58:17 hehe Jul 12 18:58:35 well, a single P4 2.8 actually Jul 12 18:58:48 * CoreDump|home sets PREFERRED_VERSION_linux-openzaurus_poodle = "2.6.17" on wonders if that'll work Jul 12 18:59:29 CoreDump|home: zecke could probably tell you. I can't remember whether that was a one of the problem cases :) Jul 12 18:59:40 =) Jul 12 19:00:02 RP: Works for me :) Jul 12 19:03:03 NOTE: package linux-openzaurus-2.6.17: started Jul 12 19:04:56 * JustinP builds his third glibc in as many days Jul 12 19:06:09 JustinP: uhgg, what made you 'WANT' to do that? Jul 12 19:06:10 * koen closes a bitbake bug Jul 12 19:06:28 hi RP Jul 12 19:06:39 thanks for the .17 updates :) Jul 12 19:06:49 hi raduga, np :) Jul 12 19:09:54 rp: is the pxafb_changeres-r1 coming close to working yet? Jul 12 19:12:18 weeee Jul 12 19:12:21 bash works Jul 12 19:17:43 re tmbinc Jul 12 19:22:01 * CoreDump|home charges poodle Jul 12 19:24:42 JustinP: There is another quilt under staging/i686-linux/bin/. It is an older version than the installed one. That is also on the path when calling runXX script. Hardcoding the call the staging bin does not help. Jul 12 19:29:20 what's the quickest way to generate a site file when you have autofoo installed on your target? Jul 12 19:29:40 site file? Jul 12 19:30:17 woglinde: http://www.openembedded.org/filebrowser/org.openembedded.dev/site Jul 12 19:30:58 sorry dont get it Jul 12 19:31:41 woglinde: those file have cached information for autotools on arch specific stuff Jul 12 19:31:45 or maybee I never done something or here about site Jul 12 19:31:48 like sizof(int) Jul 12 19:31:54 hm ah Jul 12 19:32:23 sorry never worked with that Jul 12 19:32:36 only Makefile.am configure.ac Jul 12 19:33:09 ok, I'm gone Jul 12 19:33:12 bye bye ! Jul 12 19:33:20 iirc you can get them from config.log, but you'd need to compile a lot of programs to get a complete list Jul 12 19:33:25 petn-randall: cu Jul 12 19:33:34 cya petn-randall Jul 12 19:33:38 I was hoping someone here would have a hint on how to be lazy Jul 12 19:33:55 * koen looks at kergoth`zzz Jul 12 19:33:56 hm I am gooling for it Jul 12 19:41:20 ah Jul 12 19:41:26 its autoconf cache file Jul 12 19:42:17 koen http://www.gnu.org/software/autoconf/manual/autoconf-2.57/html_node/autoconf_86.html Jul 12 19:42:48 hvontres|work: flashing poodle now Jul 12 19:42:51 that's what I feared Jul 12 19:43:05 CoreDump|home: =) Jul 12 19:45:14 yay, new OZ logo is enabled for poodle, too Jul 12 19:45:43 he coredump Jul 12 19:47:17 * mrz80 is back (gone 02:31:28) Jul 12 19:47:46 CoreDump|home: Are you going to upload images tonight? Jul 12 19:47:47 hmm why doesn't work Fn+LeftArrow anymore Jul 12 19:48:05 CoreDump|home: ?????????????? Jul 12 19:48:06 hvontres|work: depends on how they look Jul 12 19:48:18 CoreDump|home: it works with bootstrap Jul 12 19:48:21 hvontres|work: Fn-[left|right]Arrow is supposed to change VTs Jul 12 19:48:34 heh works now... Jul 12 19:49:15 CoreDump|home: are you running opie or gpe right now? Jul 12 19:49:20 opie Jul 12 19:49:30 gpe needs some more love IIRC Jul 12 19:49:58 and yet *again*, Jul 12 19:50:09 ~curse this god damn medium mount dialog crap Jul 12 19:50:11 May you be reincarnated as a Windows XP administrator, this god damn medium mount dialog crap ! Jul 12 19:50:42 can't that be disabled or something? Jul 12 19:50:50 heh I wish Jul 12 19:50:55 IMO that beats the keylaunch crap from gpe Jul 12 19:51:09 koen: keylaunch works fine now Jul 12 19:51:27 this medium mount crap annoys users for years now Jul 12 19:52:37 why don't we remove it? Jul 12 19:52:40 2.6 has udev Jul 12 19:53:01 CoreDump|home: can't the medium-mount box be suppressed by disabling it in the config file which is put in the image? Jul 12 19:53:23 koen: it doesn't really "mount" the cards but configures what opie does w/ them :\ Jul 12 19:54:02 Kerwood: there is no such setting to my knowledge Jul 12 19:54:25 grrr Jul 12 19:54:34 gcc has no fortran support Jul 12 19:55:08 lol Jul 12 19:55:14 my ts doesn't work Jul 12 19:55:42 g77? Jul 12 19:55:59 g77, f90, f95, etc Jul 12 19:56:51 what do you mean by 'gcc has no fortran support' Jul 12 19:57:09 'bitbake gcc' and look for fortran packages Jul 12 19:57:09 do you mean, that g77 is not supported by arm? Jul 12 19:57:18 oh, you mean bitbake doesn't know about it Jul 12 19:57:19 ok Jul 12 19:57:24 OE disables it Jul 12 19:57:31 because its broken? Jul 12 19:57:38 or for some other reason? Jul 12 19:57:43 * koen dunnos Jul 12 19:57:46 hvontres|work: sorry no images today Jul 12 19:57:50 * nchip has allways wondered why so many autofooed applications test for a dozen or different fortran compilers Jul 12 19:58:06 CoreDump|home: np :) Jul 12 19:58:16 nchip: :) Jul 12 19:58:46 CoreDump|home: tslib problem? Jul 12 19:58:57 hvontres|work: yeah Jul 12 19:59:08 no where is that cfg again Jul 12 19:59:45 /etc/profile.d? Jul 12 20:00:05 CoreDump|home: well, I guess at least all the other stff will have already built when I get home tonight :) Jul 12 20:00:32 koen: nah, the once which sets the "driver" for the ts Jul 12 20:00:37 s/once/one Jul 12 20:00:54 /usr/share/tslib? Jul 12 20:01:50 that's it ;) Jul 12 20:01:56 hey Jul 12 20:02:11 iirc it can be anywhere TSLIB_TSCONFFILE points to Jul 12 20:02:15 I'm having a little trouble with the libx11-native-6.2.1 package Jul 12 20:02:19 (or whatever that var is called) Jul 12 20:02:22 hey tkp Jul 12 20:02:45 hmmm looks right Jul 12 20:02:48 the source directory contains include/X11/Xlib.h Jul 12 20:02:53 which I need Jul 12 20:03:11 yet it doesn't seem to be getting installed into my staging dir Jul 12 20:04:19 koen: was there something wrong with the faad2 patch i posted via pastebin? Jul 12 20:04:42 (or were you waiting for me to paste one without the bumped PR?) Jul 12 20:04:45 tmp/staging/i586-oe-linux/include/X11/ contains loads of the .h files... but not that one :( Jul 12 20:05:48 I am right in thinking that tmp/staging/i586-oe-linux is where all the native stuff goes right? Jul 12 20:06:58 tkp: if your host is i586, yes Jul 12 20:13:26 hmmm tslib.sh for poodle is bugged Jul 12 20:13:55 CoreDump|home: ts.conf probably doesn't have module 'input' select Jul 12 20:15:22 koen: it has but is missing TSLIB_CONFFILE=/usr/share/tslib/ts.conf for some reason Jul 12 20:15:49 put ts.conf in /etc and it should be alright Jul 12 20:16:09 the /usr/share/tslib is there to select between 2.4 and 2.6 configs for ipaqs Jul 12 20:16:16 koen: where do I need to specify PREFERRED_VERSION_ for it to take effect? Jul 12 20:16:42 tkp: local.conf or $DISTRO.conf Jul 12 20:17:05 ahh.. cool. I was trying to do it in a packege's bb file Jul 12 20:17:11 and it wasn't taking effect Jul 12 20:17:41 tkp: AFAIK .bbs only affect their own 'scope' Jul 12 20:17:48 tkp: I could be wildly wrong Jul 12 20:20:29 koen: seems to have done the trick thanks :) Jul 12 20:26:12 Laibsch: maybe it's calling the wrong "patch".....? Jul 12 20:27:14 hvontres|work: I don't want to....I just had to. First for armv5te e-images, second for arm (collie) e-images, third to rebuild some old stuff Jul 12 20:30:29 03coredump 07org.oe.oz354x * r6085cf4d... 10/packages/tslib/ (tslib/poodle/tslib.sh tslib_cvs.bb): tslib: Fix Poodles tslib.sh for kernel 2.6 Jul 12 20:30:33 ~lart glibc-package Jul 12 20:30:34 * ibot tries to shut glibc-package up Jul 12 20:33:06 JustinP: I am not sure. I am going to clean out tmp *again* and rebuild overnight in the hope that I can actually use OE tomorrow. Jul 12 20:34:04 Is it true that using NFS mounted OE directory as /path/1 on computer A and /path/somewherelse on computer B will break things? Jul 12 20:34:21 Seems like some stuff is hardcoded in ./tmp. Jul 12 20:36:09 Laibsch: not sure about nfs, but i use some symlinked dirs on my pc to help spread the load Jul 12 20:36:35 so far, oe has not complained Jul 12 20:38:04 hi thejapa Jul 12 20:38:18 woglinde: hi! :) Jul 12 20:41:44 thejapa: NFS itself is possible. Is used it successfully so far. But it does not seem possible to have a different path on machine A and B when accessing the same OE tree. Moving around the OE dir in the filesystem also seems to wreak havoc (which is unfortunate). Jul 12 20:43:46 * JustinP nods Jul 12 20:50:52 Laibsch: my .dev tree is symlinked from my homedir and each dir in my /stuff/build/tmp is symlinked to other paths... Jul 12 20:50:59 s/to/from Jul 12 20:51:21 thejapa: Laibsch means the tmpdir Jul 12 20:51:26 bild dir Jul 12 20:51:29 build Jul 12 20:52:13 mikey|thesis: python-curses is complaining about libncurses5 eventhough libncuses is installed :( Jul 12 20:56:08 thejapa: why /stuff/? Jul 12 20:56:41 I don't want to know how many people called it /stuff Jul 12 20:56:54 I'm trying to install libx11, and having some trouble (again!) Jul 12 20:57:17 it's complaining that it can't find Xtrans.h... Jul 12 20:57:38 yet I have the header file in the staging include dir Jul 12 20:57:51 located at: Jul 12 20:57:52 tmp/staging/i586-oe-linux/include/X11/Xtrans/Xtrans.h Jul 12 20:58:13 what's the best way to go about adding include dirs? Jul 12 20:58:17 zecke: hehe, i don't like it too, but i find it makes easier to talk about OE tree this way. Jul 12 20:58:35 I don't quite understand why this dir isn't being included already Jul 12 20:59:26 not really sure what the bitbake way is with regards to setting include dirs Jul 12 21:00:29 I can see from the debug log that the top level include dir is being included Jul 12 21:00:29 -I/home/tom/stv/oe-stv/build/tmp/staging/i586-oe-linux/include Jul 12 21:00:58 but it doesn't seem to be finding the header which is in a sub directory of this (X11/Xtrans/Xtrans.h) Jul 12 21:02:08 the full error is: http://rafb.net/paste/results/MMUtn974.html Jul 12 21:02:50 hvontres|work: file a bug in OE and I'll take a look Jul 12 21:03:27 g'night Jul 12 21:03:36 nite mickeyl Jul 12 21:04:58 thejapa: Do you speak Japanese, BTW? Jul 12 21:05:55 thejapa: I try to avoid symlinks like the plague when dealing with OE. The wiki says OE does not deal easily with symlinks. So I do not use them. If I need to I sometimes use "mount -o bind" Jul 12 21:06:34 Laibsch: no, the nick is "the japa" because I'm brazilian, and asiatic ppl here are commonly called japs Jul 12 21:07:07 * thejapa must RTW again Jul 12 21:07:17 Well, I also checked your /whois, that is why I asked. Jul 12 21:11:16 03mickeyl 07org.oe.dev * re67ed465... 10/packages/libdvdread/libdvdread_0.9.6.bb: add libdvdread, a library to read dvds (sic!) Jul 12 21:11:24 03mickeyl 07org.oe.dev * r11d8497b... 10/packages/libzvbi/libzvbi_0.2.22.bb: Jul 12 21:11:24 add libzvbi, a library that provides functions to capture and decode VBI data. It is written in plain ANSI C with few Jul 12 21:11:24 dependencies on other tools and libraries. Jul 12 21:18:00 hmmm. so it seems that ffmpeg doesn't build for me. also, their cvs server has been retired (after being migrated to svn), so the cvs version is dead Jul 12 21:18:43 zwelch than migrate the .bb file to svn Jul 12 21:18:58 there are enough examples in the packages files Jul 12 21:19:46 oh, the "how" is not necessarily a problem; the observation was made in the hope that someone would say, "oh, we know, it's bug #n, and so-and-so is fixing it." ;) Jul 12 21:20:08 03mickeyl 07org.oe.dev * r74842567... 10/packages/zroadmap/zroadmap_cvs.bb: zroadmap cvs add dependency on expat Jul 12 21:21:38 one question that is on my mind is: what is the best means of developing something like that and then getting it merged? the proper path to take involves a file rename Jul 12 21:22:46 at the worst, i'll just create a mt diff style patch; i seem to recall reading that being listed as the proper method Jul 12 21:23:17 * zwelch shrugs and wanders off to fix ffmpeg Jul 12 21:44:40 * JustinP gets his 2.4 kernel back Jul 12 21:44:42 heh Jul 12 22:24:30 Applying patch 'kernel-keymap.patch' fails Jul 12 22:25:56 bye all Jul 12 22:26:55 quit Jul 12 22:35:44 * kergoth sighs Jul 12 22:35:50 build keeps failing at rootfs population Jul 12 22:42:30 kergoth hm Jul 12 22:42:47 what error? Jul 12 22:44:40 nite all Jul 12 22:45:37 at first, i kept getting ipks shoveled into the morgue Jul 12 22:45:49 and we all know how fun that is Jul 12 22:47:47 :\ Jul 12 22:56:11 03florian 07org.oe.dev * r7c2fa582... 10/conf/machine/htcuniversal.conf: htcuniversal.conf: Update to latest version by goxboxlive. Jul 12 22:56:15 03florian 07org.oe.dev * r263e786a... 10/packages/linux/handhelds-un-2.6_2.6.16.bb: handhelds-un-2.6: New description for up to date Universal kernel. Courtesy by goxboxlive@gmail.com. Jul 12 22:56:20 03florian 07org.oe.dev * r895e42f5... 10/packages/tslib/ (tslib/htcuniversal/tslib.sh tslib_20060703.bb): tslib: Update universal bits and specific dependencies. Jul 12 22:56:24 03florian 07org.oe.dev * rd09f0621... 10/packages/base-files/ (base-files/htcuniversal/fstab base-files_3.0.14.bb): base-files: Add htcuniversal fstab by goxboxlive. Jul 12 23:12:43 * Kerwood_ wonders at all the sighing melancholia Jul 12 23:23:13 Hey I tried bitbaking opie-image, and I got pretty far, but received the following error Jul 12 23:23:20 http://pastebin.ca/86523 Jul 12 23:23:32 Any help would be greatly appreciated Jul 12 23:27:49 | CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/libffi Jul 12 23:27:49 | cc1: internal compiler error: in add_path, at c-incpath.c:362 Jul 12 23:27:56 NOTE: package gcc-cross-initial-4.1.1-r5: task do_compile: failed Jul 12 23:33:12 i have a fun compiler error too: http://rafb.net/paste/results/eSzEOL40.html Jul 12 23:39:42 Any good way to specify different gcc version for machine specific? Jul 12 23:40:02 like setting orinoco modules to be compiled by CCSUFFIX=xxx but only for machine=x Jul 12 23:40:45 SonicvanaJr: do you have zlib staged? it looks like zlib.h is missing Jul 12 23:42:04 erm I'm not sure Jul 12 23:42:06 Luke-Jr: do you have CFLAGS or something in your environment? Jul 12 23:43:31 How can I set if statements in .bb files? Jul 12 23:45:43 zwelch how would I check ? Jul 12 23:46:06 SonicvanaJr: bitbake zlib-native Jul 12 23:46:11 if it builds it, it wasn't installed Jul 12 23:46:55 Kristoffer: at the top level? you don't, but there are conditional mechanisms documented in the bitbake manual Jul 12 23:48:21 and regarding the gcc for a specific machine... i wonder if you can do something like PREFERRED_VERSION_gcc-cross_machine = "version" with different "machine" tags Jul 12 23:48:23 ah thanks zwelch Jul 12 23:52:51 http://pastebin.ca/86545 < New errors :( Jul 12 23:53:22 zwelch, thx for idea Jul 12 23:54:10 Kristoffer: np, but since no one has verified/refuted it, let me know how it goes Jul 12 23:54:54 zwelch, will do. I used CCSUFFIX when compiling my kernel, but that was "my" .bb file which no one else uses, in this case others will use it also so I cannot set a 100% CCSUFFIX, but instead check against machine type Jul 13 00:00:45 hi, is this the place to ask general questions about OE vs. other linux dist plz? Jul 13 00:01:20 i am curious about the difference bewteen OE and embeddedUbuntu or gentoo-emb for example? Jul 13 00:01:38 genbie: what do you want to do with these? Jul 13 00:02:09 i want to get involved in OS development zwelch Jul 13 00:02:17 okay, in what capacity? Jul 13 00:02:37 what's your prior experience? where do you see yourself in 5 years? Jul 13 00:02:38 as developer Jul 13 00:03:05 5 years? i dont know! Jul 13 00:03:08 hehehe Jul 13 00:03:22 well, where do you see your growth as a developer taking you? Jul 13 00:03:24 i am just curious Jul 13 00:03:34 what areas of OS development would interest you? Jul 13 00:03:39 what sparked your curiousity? Jul 13 00:04:07 i'm asking to try to understand your perspective, so i can communicate the differences Jul 13 00:04:16 i think there is a great advantage to use free software on billion plus devices out there! Jul 13 00:04:31 nobody asked me such things i'm jealous Jul 13 00:04:56 thejapa: jump in now, then Jul 13 00:05:12 is OE like a generic framwork wheras emb-gentoo and emb-ubuntu are specific ones? Jul 13 00:05:27 genbie: that's one difference, sure Jul 13 00:05:48 so you are like trying to not reinvent the wheel as it were? Jul 13 00:05:49 * zwelch calls bitbake a meta-distro, in that it is designed to produce/manage entire distributions Jul 13 00:06:12 it's more like a wheel maker Jul 13 00:06:19 ppl use it to reinvent the wheel Jul 13 00:06:20 well, someone great once said "the best thing about standards is that there are so many to choose from" Jul 13 00:06:23 or something like that Jul 13 00:06:32 which means you end up with lots of incompatible wheels Jul 13 00:06:36 zwelch any advice on my new errors ? Jul 13 00:07:38 i am very much impressed by the success of ubuntu on the desktop market and now on server and i was wondering if they coluld achieve the same in the embedded realm but they dont have aroom here!! Jul 13 00:08:03 SonicvanaJr: don't fear the reaper? Jul 13 00:08:07 genbie: here = embedded? Jul 13 00:08:18 SonicvanaJr: sorry, i'm clueless on that one :) Jul 13 00:08:21 no i mean this chat server Jul 13 00:08:28 here = FreeNode Jul 13 00:08:36 thejapa Jul 13 00:09:20 can any programmer join your effort or do you require experience with embedded dev? Jul 13 00:09:44 genbie: anyone can join, but that only becomes meaningful when you start to contribute to the project :) Jul 13 00:10:02 cool Jul 13 00:10:13 i mean, i consider myself part of the effort by nature of the fact that i am here, participating, and sending in patches from time to time Jul 13 00:10:37 yes Jul 13 00:11:06 now, official status (e.g. w/ write privs) is another thing entirely... but open source is pretty resiliant when it comes to fluxuating membership Jul 13 00:11:59 the question about which project is best for you, given your aspirations for contributing, is probably the one that offers you the most for the platforms that you want to target Jul 13 00:12:15 you can start from zero too, if you are very persistent. :) i'm about to submit my first patches for official devs for revision Jul 13 00:12:19 very good point zwelch Jul 13 00:12:27 i.e. what platforms are you building on and for? which tools/distro give you the best support for those platforms? Jul 13 00:12:51 how can i decide that? Jul 13 00:13:13 genbie: thejapa just alluded to a good point; you should definitely try to do "linux from scratch" at some point Jul 13 00:13:33 genbie: there are people that make a living selling their time to other people to help them make exactly that kind of decision Jul 13 00:13:50 i'd like to play with developing for mobile phones for example... but i am only a developer not an embedded developer.. so i guess i will be starting from zero like thejapa said Jul 13 00:14:21 me for example, own a ipaq h2200. i can install linux into it, but making new apps run into it, is oe business. Jul 13 00:14:53 if i want something existing, i go for the existing handheld distros Jul 13 00:15:24 the problem is deciding which project to join.. i understand why people dont like comparisons, but one needs to start somewhere! Jul 13 00:17:31 so, you want to develop for mobile phones running linux? Jul 13 00:18:03 yes thejapa Jul 13 00:18:15 or install linux on them Jul 13 00:20:16 zwelch: don't think so Jul 13 00:20:26 zwelch: nope Jul 13 00:20:48 CPATH is... Jul 13 00:21:26 CPATH==/usr/include/libffi Jul 13 00:21:38 Luke-Jr: i had a problem earlier where my configure script was detecting the speex library in /usr/lib.. i solved it by adding OECONF_EXTRA += "--with-speex-includes=${STAGING_INCDIR} --with-speex-libraries=${STAGING_LIBDIR}" Jul 13 00:21:51 you might need a similar addition to the .bb file that you are using Jul 13 00:21:56 zwelch: GCC doesn't use speex Jul 13 00:21:58 i'm justing tossing out ideas Jul 13 00:22:12 I'm trying w/ CPATH unset Jul 13 00:22:15 oh yeah, your problem was there.... i'm clueless. :) Jul 13 00:22:28 ah, maybe that was it? Jul 13 00:22:35 seeing that the error referred to the pathh in CPATH Jul 13 00:22:47 yep Jul 13 00:22:48 yah, it's awfully conspicuous isn't it? :) Jul 13 00:22:51 it just finished, lol Jul 13 00:22:55 cool Jul 13 00:23:02 wonder why it only complained at the end Jul 13 00:23:19 genbie: well, to install linux in a mobile phone oe might help. but it's main task is to automate cross-compiling and it's dependencies and building the toolchain for it. Jul 13 00:23:26 * zwelch finds it is sometimes better not to ask such difficult questions in face of a gift horse ;) Jul 13 00:23:49 thejapa: any idea how simple porting to a digicam would be? Jul 13 00:24:09 to me it would be very difficult Jul 13 00:24:24 Luke-Jr: if you have the hardware specs, easy. otherwise, very very very hard Jul 13 00:24:27 heh Jul 13 00:24:53 * Luke-Jr stabs Ricoh for porting Linux and never releasing it Jul 13 00:25:05 zwelch: what if the SoC has a serial port? Jul 13 00:25:43 Luke-Jr: if you can get the source code, i'd consider it vaguely possible; if not, i wouldn't want to play that game Jul 13 00:26:00 they never released anything other than "here's what we did and how" Jul 13 00:26:09 with the "how" being vague English Jul 13 00:26:29 so who's suing them for GPL violations? ;) Jul 13 00:26:31 they do make mention to the devices used tho Jul 13 00:26:35 they didn't violate it Jul 13 00:26:48 they didn't release *anything*, not even a binary Jul 13 00:26:55 is the binary on the camera? Jul 13 00:27:03 the camera binary is Vxworks Jul 13 00:27:14 their R&D just ported Linux to it Jul 13 00:27:22 ah, for giggles Jul 13 00:27:23 vxworks does some linksys firmwares, IIRC Jul 13 00:27:33 thejapa: is that relevant? O.o Jul 13 00:27:34 * zwelch understands Jul 13 00:27:37 zwelch: not quite Jul 13 00:27:40 I think they wanted to release Jul 13 00:27:56 * zwelch points to ARM/Linux machine 28 and commiserates Jul 13 00:27:58 last thing in their paper mentioned talking to their legal dept Jul 13 00:28:45 camera is RDC-i700-- about $250 today Jul 13 00:28:55 it has PCMCIA slots for stuff like GPS Jul 13 00:29:05 it sucks when internal corporate pressures prevent the release of such R&D results Jul 13 00:30:15 http://www.ukuug.org/events/linux2004/programme/paper-SOuchi-1/linux-digitalcamera.pdf Jul 13 00:30:16 does anyone know of a mobile phone device onto which developers could actually flash a new (working) Linux kernel/rootfs? Jul 13 00:30:35 ... and which could still be used on the mobile networks for which they were designed? :) Jul 13 00:31:42 certainly, i'm not aware of such where the vendor has provided full disclosure, as is the spirit of the GPL (if not the letter) Jul 13 00:32:37 genbie: the point here is that you might need to consider an intermediate target for your development training :) Jul 13 00:33:26 * zwelch wonders what the current "favorite" targets are for OE developers Jul 13 00:33:46 zwelch: who are you talking to? genbie isn't saying anything...O .o Jul 13 00:33:47 i.e. devices that they might recommend for others to use for developing Jul 13 00:33:59 http://openezx.org/ Jul 13 00:34:08 Luke-Jr: the last couple of comments were @ the room Jul 13 00:34:49 ljp: hey you Troll! Jul 13 00:34:55 ljp: question :) Jul 13 00:34:56 ljp: thanks for that reference :) genbie should value that link too Jul 13 00:35:05 ljp: I hear TrollTech doesn't like patches for Qt? Jul 13 00:35:16 thanx ;-) Jul 13 00:35:21 http://wiki.xda-developers.com/index.php?pagename=Xanadux Jul 13 00:35:23 however, it only confirms my suspicion that the vendors don't want to touch such an open project with a 10' pole Jul 13 00:35:47 you would have heard wrong Jul 13 00:35:55 that was for luke-jr Jul 13 00:36:08 ljp: oh, so there's a good chance a patch for QDir would be accepted? Jul 13 00:36:18 in particular, a GetSpecialDir static func? Jul 13 00:36:38 is it expensive to develop for embedded systems btw? i mean do i have to buy all expensive toolkits? Jul 13 00:36:39 zwelch: you are correct. Jul 13 00:37:00 ljp: so sad :( Jul 13 00:37:01 genbie: not if you publish GPL sources Jul 13 00:37:09 luke-jr: depends on what version, and what the patch does, and what the patches license is Jul 13 00:37:38 ljp: Qt4, returns directories for eg DataDir or Pictures, BSD-license Jul 13 00:37:45 Luke-Jr: but if you write it, you can give TT the liberty they need to incorporate it Jul 13 00:37:56 yup Jul 13 00:38:09 zwelch: I don't approve of proprietary exceptions-- my usual compromise is a general BSD license Jul 13 00:38:22 which Digium seems to be ok with Jul 13 00:38:30 genbie: if you use OE, you can develop for many many embedded systems for free Jul 13 00:38:37 bsd is good, as long as there is no advertizing clause Jul 13 00:38:39 same for other tools... open source rocks :) Jul 13 00:38:45 ljp: right, that's the one I meant Jul 13 00:39:03 cool Jul 13 00:39:34 ljp: What if it depends on an otherwise GPL library? Jul 13 00:40:15 ljp: eg, I as author of the library license it for non-GPL use by my Qt patch Jul 13 00:43:23 trolltech has to be able to sell the code in non GPL commercial license Jul 13 00:43:27 maybe most of you have seen this already but OE features very well in its very first year... http://www.linuxdevices.com/articles/AT4036830962.html Jul 13 00:46:11 ljp: Do they need to be able to sell the code of my library under a non-GPL license outside of Qt? Jul 13 00:46:26 ljp: eg, would an exception to the GPL applying only to Qt be acceptable? Jul 13 00:46:50 not real sure, would have to run that by legal Jul 13 00:47:08 ok, would you mind? Jul 13 00:49:09 could you send me an email? Jul 13 00:50:08 ljp@trolltech.com ? Jul 13 00:50:18 lorn.potter@trolltech.com :) Jul 13 00:50:23 ok Jul 13 00:50:32 | ../sysdeps/unix/sysv/linux/bind.S: Assembler messages: Jul 13 00:50:33 | ../sysdeps/unix/sysv/linux/bind.S:5: Error: symbol `__bind' is already defined Jul 13 00:50:33 (glibc-2.3.5+cvs20050627) Jul 13 00:50:34 :/ Jul 13 00:53:16 Luke-Jr: you are using binutils-2.17 Jul 13 00:53:24 add the following line to your local.conf: Jul 13 00:53:25 Luke-Jr: This came up several times on IRC today. By adding the line 'PREFERRED_VERSION_binutils-cross = "2.16"' to your local.conf, you can make it go away. Thank zwelch for this bit of info Jul 13 00:53:25 no idea lol Jul 13 00:53:33 right, what Kerwood said :) Jul 13 00:53:35 whoops Jul 13 00:53:40 lol thanks Jul 13 00:53:43 hehe Jul 13 00:54:11 zwelch, I have to say I was underwhelmed by the regulars' lack of interest in conveying this tidbit Jul 13 00:54:15 * zwelch supposes he should report this in bugzilla Jul 13 00:54:31 heh, the regulars always have better things to do that give free support Jul 13 00:54:47 * zwelch suddenly realized that he does too and disappears again Jul 13 00:55:10 toodles! Jul 13 01:06:56 Kerwood_, you misunderstand some things Jul 13 01:07:17 I know Jul 13 01:07:31 I haven't been here long at all Jul 13 01:07:37 its generally guess that new people intially use a stable release or base things off a stable release Jul 13 01:07:45 guessed* Jul 13 01:08:09 and certain versions of a package that works for one project may or may not work for any other Jul 13 01:08:43 regular users in most cases can only offer general guidance Jul 13 01:09:03 unless you happen to be using the same targets and distros as them Jul 13 01:09:08 I understand that it's trouble when people complain while using the unstable configuration Jul 13 01:09:19 I understand that Jul 13 01:10:11 i can tell you from personal experiance that i spent 4 days chasing a problem that went back to binutils changes for eabi support Jul 13 01:10:29 it showed up in glibc as a gcc library error Jul 13 01:10:40 But after saying the same thing about the binutils version 3 times today, I thought I might hear a suggestion about which wiki it would be most appropriate to put this info, understanding that they main developers are too busy to do so Jul 13 01:11:03 instead I heard silence Jul 13 01:11:11 how else shall I interpret it? Jul 13 01:11:26 that is very common if your not working on an OZ distro Jul 13 01:11:56 but that has to do with the amount of active people on the project Jul 13 01:12:57 An already agreed-upon major issue with OE right now is communication; another one is the top developers' lack of time Jul 13 01:13:05 i would guess that over half the OE devs are OZ developers Jul 13 01:13:07 so why wouldn't they delegate somrthing like this Jul 13 01:13:30 I know hrw does it Jul 13 01:13:43 there is no board, commitee, or project leader Jul 13 01:13:46 Kerwood_: fyi, a lot of the major devs are in europe. its, what, 4 am there right now? Jul 13 01:14:00 if you have something to add to the wiki, do it, or wait till they're conscious Jul 13 01:14:32 kergoth: nice try, but the events I'm talking about happened during the european daytime Jul 13 01:14:38 and? Jul 13 01:14:47 Kerwood_, i agree with one of your issues9its been long standing) but not the lack of dev's time Jul 13 01:14:53 people are busy. they're doing this with their own personal time Jul 13 01:14:55 deal with it. Jul 13 01:15:07 if you have an improvement to make to the wiki, get off your ass and do so. its a wiki. thats the point. Jul 13 01:15:13 wow, kergoth, I don't think you read what I just said. Jul 13 01:16:04 kergoth: there are strong opinions about what should go in the wiki and where. it was reasonable for me to wait for a pointer. Jul 13 01:16:07 i dont knwo anyone that actually gets paid specifically to work on OE Jul 13 01:16:18 if you put it in the wrong place, someone will fix it. Jul 13 01:16:29 just for gods sake stop whining Jul 13 01:16:39 have you sent an email to the list? irc isn't the best place to ensure issues are tracked Jul 13 01:16:44 is there a bug in the bts? Jul 13 01:17:39 so how goes the new job kergoth ? Jul 13 01:18:24 pretty good, still in the initial phase of getting up to speed on things, getting my development environment happy, that sort of thing. looking forward to getting past that Jul 13 01:18:27 brb Jul 13 01:20:48 sorry, I was saying gn to the kid Jul 13 01:21:19 well, I regret my words were taken as whining; that wasn't the intent Jul 13 01:21:37 and I won't feed these flames any further Jul 13 01:29:08 I hate to be a begging noob, but I'm really stumped here. When trying to bitbake opie-image I receive the following errors http://pastebin.ca/86606 Jul 13 01:29:25 If anyone could offer any insight, I would be forever greatful Jul 13 01:32:06 SonicvanaJr: which branch of oz are you building from, .dev or .oz354x? Jul 13 01:32:14 .oz354x Jul 13 01:32:52 well, there go all my ideas.... Jul 13 01:32:57 lol Jul 13 01:33:29 My only idea was that maybe I was missing some libraries on the machine, since it is ubuntu, and it dosen't come with anything for dev installed Jul 13 01:33:44 you on amd64? Jul 13 01:33:51 no Jul 13 01:34:09 P4 2.8 Jul 13 01:34:57 sorry I didn't see it before -- yeah, ubuntu might be missing a useful tool Jul 13 01:36:11 such as ? Jul 13 01:36:45 actually, I don't know, I'm just guessing at this point Jul 13 01:37:02 * thejapa using dapper drake. but i'm using .dev Jul 13 01:37:35 hmm maybe I should install .dev Jul 13 01:37:39 * thejapa goes fetch notebook notes Jul 13 01:37:39 and stick with it Jul 13 01:39:01 Actually, working from .dev brings with it random variables associated with constant changes. you sure you want all that too? Jul 13 01:39:11 no :( Jul 13 01:39:19 i cannot find any particular grumbles with ubuntu Jul 13 01:39:36 this means i must have only followed the wiki to the letter Jul 13 01:39:52 i remember having some success after i started with a fresh local.conf Jul 13 01:40:06 SonicvanaJr: it appears it can't compiled QEMU Jul 13 01:41:35 yeah Jul 13 01:44:12 mithro so how do "fix" it, for a lack of a better word Jul 13 01:44:41 SonicvanaJr: you could disable locale building Jul 13 01:46:48 * thejapa nods Jul 13 01:48:23 ok how would I go about doing that ? Jul 13 01:49:49 check comments in your local.conf, about QEMU breaking, near line 90 Jul 13 01:50:16 ok Jul 13 01:59:31 SonicvanaJr: That's odd -- the path to my qemu work dir is .../tmp/work/armv5te-linux/qemu-native-0.7.0-r1 Jul 13 02:00:07 Kerwood_ where did you see mine was different Jul 13 02:00:22 yours is missing /armv5te-linux/ Jul 13 02:01:43 you both using akita/oz 0.3.5.4.1 on .oz354x ? Jul 13 02:01:55 no, mine's a spitz Jul 13 02:02:18 but it still should have put it in a dir named with the "arm" string, right? Jul 13 02:05:39 Kerwood_: well, it makes sense. Jul 13 02:06:20 Oh mithro I didn't see anything the QEMU breaking in the conf Jul 13 02:06:26 I don't see how it could happen or even why it would cause problems, but it seemed out of place Jul 13 02:06:38 but i have no clue as to why it is building outside arm or if it causes problems Jul 13 02:06:47 I'm going to bed...gn & gl Jul 13 02:07:22 SonicvanaJr: look in the wiki Jul 13 02:07:49 there is something you can set which will cause it to not building qemu Jul 13 02:09:39 mithro the only thing that pops up while searching for QEMU is something about setting the machine to "qemuarm" so that it can run under the QEMU arm emulation Jul 13 02:10:16 yeah thats it Jul 13 02:36:12 there is also the binary_local_generation flag Jul 13 02:36:30 which disables using qemu Jul 13 02:36:43 ? Jul 13 02:37:20 ENABLE_BINARY_LOCALE_GENERATION = "0" Jul 13 02:37:34 disables using qemu Jul 13 02:37:43 or atleast it used to Jul 13 02:45:53 yeah emte not in there anymore :( Jul 13 02:49:31 SonicvanaJr: just paste it into your local.conf Jul 13 02:49:40 ah Jul 13 02:49:41 ok Jul 13 02:49:45 sorry :( Jul 13 02:50:00 np. :) i'm newbie too **** ENDING LOGGING AT Thu Jul 13 02:59:56 2006