**** BEGIN LOGGING AT Sat Nov 17 02:59:57 2007 Nov 17 07:18:30 03mwester 07org.oe.dev * r66a10a30... 10/ (1 classes/package.bbclass): Nov 17 07:18:30 package.bbclass: fix some corner cases with the new package copy mechanism. Nov 17 07:18:30 - do not transform a symlink to a directory into a directory on copy Nov 17 07:18:30 - preserve the mode when copying a directory Nov 17 07:18:37 03mwester 07org.oe.dev * refd9b248... 10/ (1 packages/base-files/base-files_3.0.14.bb): Nov 17 07:18:37 base-files: bump PR to force a rebuild of this package with updated package.bbclass Nov 17 07:18:37 (makes /var/volatiles work again) Nov 17 07:48:04 * * OE Bug 3320 has been created by joshua(AT)willowisp.net Nov 17 07:48:06 * * h2200 autobuild images (20071117) don' t ever make it into gpe desktop Nov 17 07:48:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3320 Nov 17 08:29:05 03koen 07org.oe.dev * rd34d09df... 10/ (1 packages/matchbox-common/matchbox-common_0.9.1.bb): matchbox-common: make it bug compatible again Nov 17 09:08:27 Crofton: http://linuxdevices.com/news/NS4338556106.html Nov 17 09:15:40 * Bernardo is back. Nov 17 09:15:52 * Bernardo wonders if the meizu also uses TI parts Nov 17 09:21:05 * * OE Bug 3320 has been RESOLVED (FIXED) by koen(AT)openembedded.org Nov 17 09:21:07 * *  h2200 autobuild images (20071117) don' t ever make it into gpe desktop Nov 17 09:21:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3320 Nov 17 09:23:14 koen: I built yesterday a x11-gpe-image for the akita, and it got stuck after login, could it be the same bug? Nov 17 09:23:24 yes Nov 17 09:23:52 ok, as soon as I finish building this meta-opie I'll rebuild it again Nov 17 09:24:12 * Bernardo would switch right now if there was a iqnotes for gpe Nov 17 09:25:08 Bernardo: you mean notecase? Nov 17 09:25:33 I'll have to try it... I have lots of stuff in iqnotes Nov 17 09:30:28 what format does notecase use to store data? Nov 17 09:30:39 dunno Nov 17 09:30:44 I don't use outliners Nov 17 09:34:20 I use it mostly to store unstructured data Nov 17 09:34:29 meetings anotations Nov 17 09:34:46 memos to self Nov 17 09:34:54 that kind of stuff Nov 17 09:35:08 I'll ask the author Nov 17 09:40:20 what is the best web browser for gpe? minimo? Nov 17 09:52:24 try midori as browser Nov 17 10:02:29 03koen 07org.oe.dev * r449a16e6... 10/ (4 files in 3 dirs): notecase: update to 1.7.2 Nov 17 10:19:40 ok, thanks Nov 17 10:19:43 * Bernardo is away: Ausente por agora. Nov 17 10:50:04 * * OE Bug 3322 has been created by jbs(AT)bandos.homelinux.com Nov 17 10:50:06 * * fortune-mod builds for the host, not for the target Nov 17 10:50:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3322 Nov 17 10:51:04 * * OE Bug 3323 has been created by autobuild(AT)openembedded.org Nov 17 10:51:06 * * notecase-1.7.2-r1-do_compile Nov 17 10:51:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3323 Nov 17 10:51:16 * * OE Bug 3324 has been created by autobuild(AT)openembedded.org Nov 17 10:51:19 * * sysvinit-2.86-r36-do_rm_work Nov 17 10:51:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3324 Nov 17 11:25:18 03pH5 07org.oe.dev * r1c6f6c20... 10/ (1 packages/fuse/fuse-module_2.7.1.bb): fuse-module: add 2.7.1 Nov 17 12:04:24 hi all Nov 17 12:08:10 hi florian Nov 17 12:09:25 howdy Nov 17 12:09:48 anyone seen this before on a fresh install? http://pastebin.ca/778201 Nov 17 12:10:58 not me, hmm Nov 17 12:12:06 I have seen this a _long_ time ago... but I don't remember where. Nov 17 12:12:10 mickeyl|angry: ! Nov 17 12:13:08 morning Nov 17 12:22:59 Stephmw: you have a symlink in your path to the builddir Nov 17 12:23:05 * * OE Bug 3324 has been RESOLVED (FIXED) by koen(AT)openembedded.org Nov 17 12:23:07 * *  sysvinit-2.86-r36-do_rm_work Nov 17 12:23:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3324 Nov 17 12:23:17 * * OE Bug 3323 has been RESOLVED (FIXED) by koen(AT)openembedded.org Nov 17 12:23:19 * *  notecase-1.7.2-r1-do_compile Nov 17 12:23:21 * * http://bugs.openembedded.org/show_bug.cgi?id=3323 Nov 17 12:23:42 it does something like if(foo) != basename(pwd) && -e ./configure ; error Nov 17 12:25:58 koen: that's the cause of the error? ok, moving things around Nov 17 12:26:07 * Stephmw grumbles Nov 17 12:27:34 mount -o bind is a wonderful option Nov 17 12:28:40 mickeyl|angry: did your sofa break down? Nov 17 12:30:35 haha Nov 17 12:30:36 no :)) Nov 17 12:30:52 there, you're smiling again :) Nov 17 12:31:15 thanks :) all i can say is i'm utterly dissatisfied with the way how clueless managers try to ruin long term visions Nov 17 12:31:31 and there's nothing more i can say, so lets get back to other topics Nov 17 12:32:34 * koen guesses the neo will now start using a mv 2.4.2 kernel with binary-only drivers Nov 17 12:33:03 koen: nono, om will drop their UI layer and use Android... Nov 17 12:33:06 * Stephmw hides Nov 17 12:33:34 no, android with a java vm instead of dalvik Nov 17 12:33:45 haha Nov 17 12:33:52 you would be surprised Nov 17 12:34:02 i heard the term MontaVista dropping here Nov 17 12:34:20 kergoth: no pun intended ;) Nov 17 12:34:22 *sigh* Nov 17 12:34:37 give me $40k and I'll rename angstrom to "Carrier Grade Angstrom" Nov 17 12:34:40 hahaha Nov 17 12:34:45 ya Nov 17 12:35:05 well, if all fails, I'm going back to academics Nov 17 12:35:08 I wonder what the xtreemos dudes are going to do with angstrom Nov 17 12:36:32 ya, me too Nov 17 12:37:05 ~lart ubuntu for breaking 'mount' to always use fuse Nov 17 12:37:05 * ibot takes a rusty axe and swings it violently, taking ubuntu's head off for breaking 'mount' to always use fuse Nov 17 12:37:07 koen: looks like the build is flying now, thanks Nov 17 12:37:25 gutsy doesn't like fuse in 2.6.23.... Nov 17 12:37:44 * koen wonders how to fix that when /boot and /OE can't be mounted Nov 17 12:56:20 koen: do you know whether we build the latest and greatest HAL and OHM in OE yet? Nov 17 12:56:40 next week i want to finally start evaluating thse two Nov 17 12:56:47 we don't Nov 17 12:57:11 we use a very recent release of hal, but you need to merge ohm from poky to get the latest bits Nov 17 12:57:19 i see, k Nov 17 12:58:25 hmm, ok. two hours. Nov 17 12:58:32 * mickeyl decides to attempt an E upgrade Nov 17 12:58:35 koen, awesome, I think Nov 17 13:00:51 I wonder if the tools are usabel on anythiong other than the neuros ... Nov 17 13:01:44 03jeremy_laine 07org.oe.dev * r7b1dd968... 10/ (3 files in 3 dirs): iptables: backport fix for "getsockopt failed strangely" (Closes: #3260) Nov 17 13:01:57 woot. ~praise jeremy_laine Nov 17 13:23:27 hiya Nov 17 13:24:04 mickeyl : I seem to remember some time ago you wrote a LEDs driver for EZX, is that correct? Nov 17 13:27:05 jeremy_laine: yep, that was me Nov 17 13:27:34 the LEDs framework is prettycool Nov 17 13:28:10 I have written a driver for the MPC8323E-RDB board, it's on its way to OE Nov 17 13:28:12 ya, that's the stellar work of Richard 'RP' Prudie Nov 17 13:28:19 s/Prudie/Purdie/ Nov 17 13:28:28 cool Nov 17 13:28:58 steliosk is also playing with such a board, wasn't going too well though last I heard Nov 17 13:41:51 good morning Nov 17 13:42:47 hey sakoman_ Nov 17 13:43:02 hey koen! Nov 17 13:43:22 are there docs soemwhere on site.conf usage? Nov 17 13:43:46 sakoman_: it's the same as local.conf Nov 17 13:43:57 as is auto.conf Nov 17 13:44:14 sakoman_: have a look a the end of conf/bitbake.conf in OE Nov 17 13:44:36 hmm . . . tried creating one and it doesn't seem to be recognized. Can BBFILES collection stuff go there or is that not OK? Nov 17 13:45:08 OK, I'll look at bitbake.conf Nov 17 13:45:18 bitbake.conf just does 'include local.conf ; include site.conf ; include auto.conf' Nov 17 13:45:28 so it doesn't really matter where you put stuff Nov 17 13:45:46 Got it! Probably operator error then :-) Nov 17 13:49:04 * * OE Bug has been RESOLVED by jeremy.laine(AT)bolloretelecom.eu Nov 17 13:49:07 * *  iptables fails with error "getsockopt failed strangely" Nov 17 13:49:09 * * http://bugs.openembedded.org/show_bug.cgi?id= Nov 17 14:36:26 mickeyl: I think adding --disable-xauth to EXTRA_OECONF in ohm is all we need in OE Nov 17 14:44:28 wow TI finally opened their DSP tools? open == general available Nov 17 14:44:36 not really .. Nov 17 14:44:38 sort of Nov 17 14:44:50 only hte linux c-compiler for specific CPU Nov 17 14:44:58 still, a start Nov 17 14:45:10 Hopefully they work on the OSK Nov 17 14:52:43 koen: sounds good Nov 17 14:53:34 koen: NOTE: build 200711171229: completed :D Nov 17 14:53:44 Stephmw: great! Nov 17 14:54:07 so that was -minimal Nov 17 14:54:24 * Stephmw jumps back into #angstrom Nov 17 14:54:46 Crofton|home: http://www.angstrom-distribution.org/unstable/autobuild/ Nov 17 14:55:10 * mwester sits down with his cup of cofee, and scans the logs to see if the commit he made *waaaay* to late last night ruined anyone's life... Nov 17 14:55:26 Good Mornin' Nov 17 14:55:42 'ello Nov 17 14:57:55 koen, thanks Nov 17 15:26:35 koen, i think fortran is included in gnueabi-cc. We only need to patch the make file to let octave pass do_configure. Am I on the right track? Nov 17 15:29:53 koen, maybe not. fortran is enabled for gcc but not gcc-cross Nov 17 15:30:31 it's enabled for gcc 4.2.x Nov 17 15:30:43 IOW, angstrom 2008 should have fortran for arm Nov 17 15:30:53 koen, I'm trying to generate another pub key Nov 17 15:31:05 Howcan I specifiy my ientity Nov 17 15:31:11 the fm isn't helpful Nov 17 15:31:17 for monotone? Nov 17 15:31:24 balister@localhost.localdomain is not hepfli Nov 17 15:31:25 ssh Nov 17 15:31:31 for angstrom work Nov 17 15:32:00 no idea Nov 17 15:32:04 <- ssh n00b Nov 17 15:32:07 heh Nov 17 15:32:18 I thought you know everything :) Nov 17 15:32:29 -i ? Nov 17 15:32:48 -C Nov 17 15:32:58 that's compression Nov 17 15:33:09 even though it say that only works for RSA1 keys Nov 17 15:33:11 -i path/to/key Nov 17 15:33:17 comment Nov 17 15:36:04 koen: I see. thanks Nov 17 15:37:55 ssh-keygen Nov 17 15:41:00 boot Nov 17 15:41:11 oops Nov 17 15:41:15 bbl Nov 17 15:42:59 * XorA|gone does as he is told and sends a size 10 boot florians way :-) Nov 17 15:51:18 ssh-keygen -C philip@balister.org is what I wanted :) Nov 17 15:51:25 ssh-keygen -t dsa -C philip@balister.org is what I wanted :) Nov 17 15:54:14 koen: while you are on it (fuse), why not bump the version of gmailfs to 0.8.0 as well? I'm pretty sure there is no big issue. Let me know if you want me to test or submit a patch Nov 17 16:02:17 what is an x11-pimlico image? Nov 17 16:11:22 xjqian: ph5 is the fuse master Nov 17 16:11:58 Crofton|home: an x11 image with pimlico apps installed Nov 17 16:12:07 Crofton|home: http://pimlico-project.org/ Nov 17 16:12:55 I need to find the OSK with the screen Nov 17 16:19:42 I'm by no means fuse master, but that doesn't stop me from bumping gmailfs to 0.8.0 :P Nov 17 16:38:11 03jeremy_laine 07org.oe.dev * ra1d56e48... 10/ (4 files in 3 dirs): linux-2.6.23: add LEDs driver for MPC8323E-RDB boards Nov 17 16:38:17 03mickeyl 07org.oe.dev * r1e012394... 10/ (1 packages/gsm/gsmd.inc): gsmd: add new plugin gta01 Nov 17 16:38:22 03pH5 07org.oe.dev * readfc810... 10/ (4 files in 2 dirs): gmailfs: add 0.8.0 Nov 17 17:37:52 * Bernardo is back. Nov 17 18:22:04 * * OE Bug 3325 has been created by autobuild(AT)openembedded.org Nov 17 18:22:06 * * glib-2.0-2.14.1-r2-do_compile Nov 17 18:22:07 * * http://bugs.openembedded.org/show_bug.cgi?id=3325 Nov 17 18:51:47 not sure if anyone is around... Nov 17 18:51:55 but I am trying to get fusepod compiled... Nov 17 18:52:06 and it depends on libgpod, taglib, and fuse. Nov 17 18:52:18 when it compiles libgpod, I get this error: Nov 17 18:52:19 | /oe/build/tmp/staging/arm-angstrom-linux-gnueabi/lib/libtag.so: undefined reference to `uncompress' Nov 17 18:52:30 but it never says it depends on zlib although I put it in the depends argument Nov 17 18:52:43 anyone out there have any suggestions on this? Nov 17 18:56:42 christopher: Since you get the error during build time, the dependency should be specified to be not on zlib, but rather on zlib-native Nov 17 19:09:03 * * OE Bug 3326 has been created by watashi.metan(AT)gmail.com Nov 17 19:09:05 * * vim: arrows doesn't work in insert mode Nov 17 19:09:07 * * http://bugs.openembedded.org/show_bug.cgi?id=3326 Nov 17 19:09:16 * * OE Bug 3327 has been created by watashi.metan(AT)gmail.com Nov 17 19:09:18 * * gvim: syntax highliting doesn't work Nov 17 19:09:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3327 Nov 17 19:15:45 is a monotone specialist online? I've made a type in --branch=blah while committing a boatload of changes. The changes are now in the wrong branch. How can I move the content of branch A into branch B? Nov 17 19:16:33 CoreDump: be hardcode and kill the rev, propagate Nov 17 19:16:43 CoreDump: propagate should be pretty easy Nov 17 19:17:04 hmm Nov 17 19:20:11 mwester: so, would I have to recompile libtag or just libgpod? Nov 17 19:20:26 mwester: I tried with libgpod to have that dependency with no difference. I'll try both Nov 17 19:21:30 You may have to remove the stamps to force a full rebuild of libgpod -- it probably already thinks the dependencies are satisfied? Nov 17 19:21:31 mwester: yeah, same thing Nov 17 19:21:47 I did a bitbake -crebuild libgpod Nov 17 19:21:50 Is it building zlib-native? Nov 17 19:22:09 I guess it already did Nov 17 19:22:11 in the past Nov 17 19:22:14 I didn't see that fly by Nov 17 19:22:20 Check in i686 to see if its there Nov 17 19:22:32 If so, then it's not a dependency issue. Nov 17 19:22:38 (any longer, anyway) Nov 17 19:22:49 yep, it is there Nov 17 19:23:03 yeah, I don't know what is going on. I had libgpod building before... Nov 17 19:23:20 but then I added in taglib for the building of fusepod Nov 17 19:23:23 and now this. Nov 17 19:23:44 I'd like to have something that is written for ipods that can run in userspace and run fairly quickly Nov 17 19:23:50 I'm not sure if this is it or not Nov 17 19:23:51 Do you know that uncompress is an artifact of zlib? Nov 17 19:24:09 I do not Nov 17 19:24:24 I was just thinking that due to the error, and from what I saw from google, that it could be zlib Nov 17 19:26:06 what's weird is that I rebuild libgpod and it is referencing libtag.so Nov 17 19:26:18 but it wasn't a dependency before that I know of Nov 17 19:26:24 I didn't think libgpod had a dependency on it Nov 17 19:26:29 but maybe it does Nov 17 19:26:32 Ok, SlugOS also builds zlib-native, but there is no uncompress binary created. Nov 17 19:26:58 So you'll have to find whatever provides uncompress, either in OE or in your native distro, and install that. Nov 17 19:27:20 hmmm. interesting. Nov 17 19:28:34 do you have an ipod? Nov 17 19:29:15 Depends how you mean. I have one on my desk right now that my kid tells me I need to fix. Nov 17 19:29:40 hehe :) Nov 17 19:29:53 hitting menu and center button to reboot/refresh Nov 17 19:29:54 :) Nov 17 19:30:07 and you use it with linux? do you just use rhythmbox? Nov 17 19:30:38 i just don't know why this isn't building when it was before Nov 17 19:30:55 I figured my issue would have been with fusepod, not libgpod. it's like it got rebuilt and now has a dependency issue Nov 17 19:30:59 where it didn't before Nov 17 19:31:42 She uses iTunes and Windows - almost the way Steve Jobs intended (I don't do macs here, Windows is enough sys admin for me). Nov 17 19:32:02 Try adding a dependency on gzip-native, and seeing if that adds uncompress to your staging area. Nov 17 19:32:35 mwester: how does building a x86 binary solve missing symbols in an arm library? Nov 17 19:33:12 Did I read the error incorrectly? Nov 17 19:33:25 | /oe/build/tmp/staging/arm-angstrom-linux-gnueabi/lib/libtag.so: undefined reference to `uncompress' Nov 17 19:33:33 christopher: Listen to koen: he knows what he's doing. :) Nov 17 19:33:42 /oe/build/tmp/staging/arm-angstrom-linux-gnueabi/lib/libtag.so: undefined reference to `uncompress' Nov 17 19:34:13 mwester: yeah, he's solved a lot of issues before. I just want this thing to build and it built fine on my pc Nov 17 19:34:16 I tested that first Nov 17 19:34:17 could be CFLAGS being blown away Nov 17 19:34:19 * mwester is indeed guilty of not reading the error correctly. Nov 17 19:34:19 just to be sure Nov 17 19:34:39 try adding CFLAGS_append = " -lz " or something similar to the recipe Nov 17 19:34:39 * mwester goes to "fix" an iPod. Nov 17 19:35:27 would I add that to just the libgpod or whatever creates libtag.so? Nov 17 19:36:12 try adding it to libgpod first Nov 17 19:36:38 assuming 'uncompress' is a symbol in libz Nov 17 19:36:43 ok. with that, I get a configure error saying that the c compiler cannot create executables Nov 17 19:37:26 ah Nov 17 19:37:27 koen: weird thing is, I had gtkpod up and running... Nov 17 19:37:33 but, it was slow as molasses. Nov 17 19:37:43 so, I was trying to find something else and wanted to see how fusepod worked. Nov 17 19:37:54 same libgpod dependency of course, but also one for taglib and fuse Nov 17 19:38:06 taglib built fine, fuse built fine. Nov 17 19:38:10 libgpod built fine before. Nov 17 19:38:12 but now, it won't Nov 17 19:38:39 I did use gtkpod + fuse sshfs to restore a backup last week Nov 17 19:38:48 but lets not get into how much OSX sucks Nov 17 19:39:24 ok. :) what do you think would work well with an arm? Nov 17 19:39:29 that would be fast enough. Nov 17 19:39:34 no idea :) Nov 17 19:39:38 I just want to get a listing and have a user tell it to play Nov 17 19:39:42 ok. Nov 17 19:39:55 I suspect the data transfer over usb is killing the speed Nov 17 19:40:26 well, from what I have heard, gtkpod builds a database of the whole ipod instead of just getting a list of let's say, 10 to start or whatever... Nov 17 19:40:27 the usb controllers on arm chips aren't known for being well designed or efficient Nov 17 19:40:31 so, it takes a while there. Nov 17 19:40:44 I can play mp3s and such from usb sticks with no issues Nov 17 19:41:11 I'm just hoping to send a command over to the ipod, have it play the song I want and have it pipe through an audio amp Nov 17 19:41:24 | creating test-init-ipod Nov 17 19:41:24 | creating get-timezone Nov 17 19:41:24 | /oe/build/tmp/staging/arm-angstrom-linux-gnueabi/lib/libtag.so: undefined reference to `uncompress' Nov 17 19:41:24 | collect2: ld returned 1 exit status Nov 17 19:41:29 it's like this is at the end of the compile. Nov 17 19:41:40 gah Nov 17 19:41:44 I made a thinko Nov 17 19:41:54 s/CFLAGS/LDFLAGS/g Nov 17 19:42:15 can you convert your thinko to more words? Nov 17 19:43:39 LDFLAGS_append = " -lz " Nov 17 19:44:25 i can try that. still confused of how it would have compiled before. Nov 17 19:44:30 but I'll give it a go here Nov 17 19:45:07 same Nov 17 19:46:16 SECTION = "libs" Nov 17 19:46:16 LICENSE = "GPL" Nov 17 19:46:16 PR = "r0" Nov 17 19:46:16 SRC_URI="http://easynews.dl.sourceforge.net/sourceforge/gtkpod/libgpod-0.6.0.tar.gz" Nov 17 19:46:33 that's basically all that's in the recipe. I have this as well: Nov 17 19:46:35 do_stage() { Nov 17 19:46:35 autotools_stage_all Nov 17 19:46:35 } Nov 17 19:47:14 and 'inherit autotools', right? Nov 17 19:47:32 yep Nov 17 19:47:33 inherit autotools pkgconfig Nov 17 19:47:55 this @#*$)@*#$ thing worked before I tried to build fusepod. somehow that messed it up Nov 17 19:48:15 brb Nov 17 19:48:17 a bitbake -crebuild does a full clean and rebuild right? Nov 17 19:48:18 ok. Nov 17 19:53:04 * * OE Bug 3234 has been RESOLVED (FIXED) by joshua(AT)willowisp.net Nov 17 19:53:06 * *  bluetooth not working under h2200 with CSR radio (bcsp) Nov 17 19:53:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3234 Nov 17 19:53:16 * * OE Bug 3328 has been created by joshua(AT)willowisp.net Nov 17 19:53:18 * * sound broken for h2200 in 20071117 autobuild Nov 17 19:53:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3328 Nov 17 20:46:04 * * OE Bug 3329 has been created by joshua(AT)willowisp.net Nov 17 20:46:06 * * h2200 images (20071117 and earlier) have a media 'folder' in the gpe menu with no contents Nov 17 20:46:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3329 Nov 17 20:47:04 * * OE Bug 3330 has been created by watashi.metan(AT)gmail.com Nov 17 20:47:06 * * epdfview: error while loading libpoppler-glib.so.1 Nov 17 20:47:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3330 Nov 17 21:04:44 join #openmoko Nov 17 21:16:01 join #openmoko Nov 17 21:16:04 :-) Nov 17 21:16:10 hey likewise Nov 17 21:16:17 Jin^eLD: hey. Nov 17 21:18:38 Jin^eLD: that way, we can work rather independently (because we are too busy to work in tandem). Nov 17 21:19:07 sounds good Nov 17 21:38:19 anyone paying attention? Nov 17 21:38:42 dahitokiri: try your luck ;> Nov 17 21:39:21 i have experience programming in c/c++/asm on the pc, but i've been looking to get into embedded systems programming too. can you guys recommend any good intro resources for me? books/articles/sites? Nov 17 21:40:10 dahitokiri: what exactly do you want to do? stuff like kernel hacking and drivers or application programming? I switched from PC to embedded and did not notice much difference on application level Nov 17 21:40:19 the former Nov 17 21:40:26 i've done embedded apps Nov 17 21:40:30 well, sure, less resources and stuff, but the app I was developing for the PC suddenly crosscompiled and ran on ARM without any modifications :) Nov 17 21:40:59 Jin^eLD, i experienced the same thing Nov 17 21:41:08 that's why i want to do kernel/driver stuff Nov 17 21:41:10 close to hardware Nov 17 21:41:42 I started there recently, and discovered that actually understanding how hardware works is helpful :) so far I am pretty much lacking that knowledge Nov 17 21:41:57 ah Nov 17 21:42:05 * * OE Bug 3331 has been created by watashi.metan(AT)gmail.com Nov 17 21:42:07 * * ping: permission denied. (are you root?) Nov 17 21:42:08 if you want to write kernel drivers, there is a good and free book Nov 17 21:42:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3331 Nov 17 21:42:10 i have a basic knowledge of EE stuff Nov 17 21:42:12 Linux Device Drivers Nov 17 21:42:20 Jin^eLD, aye, reading that atm. :) Nov 17 21:42:22 you can download a PDF Nov 17 21:42:29 :) Nov 17 21:42:36 anything beyond that, though? Nov 17 21:43:01 well, maybe get started on some small and more or less easy driver Nov 17 21:43:41 I'm in the learning process too now, and I figured there are various "levels" there, some really deep stuff and some higher levels in the kernel Nov 17 21:44:19 higher in the sense that you have some API, like video 4 linux or something like that and you write a driver for it Nov 17 21:44:48 that seemed easier to me Nov 17 21:45:04 iirc, v4l runs in userspace Nov 17 21:45:14 most graphics APIs do, it seems Nov 17 21:45:31 well, if you write support for some card, you write it acc to the v4l spec Nov 17 21:45:38 yeah Nov 17 21:46:08 so internally you have stuff that you can and have to use, v4l headers, certain way of defining callbacks for handlign ioctl calls and stuff like that Nov 17 21:46:15 heh, to tell you the truth, i'm just looking for more and more things i can do w/ the languages Nov 17 21:46:34 anything besides GUI, i want to do, basically :P Nov 17 21:46:51 so you build your driver on top of something, that's what I meant... it's different if you, say, have a new board and have to make a kernel run there :) that's really lowest stuff which is still a mystery to me :) Nov 17 21:47:04 Jin^eLD, yup Nov 17 21:47:38 heh, manipulating registers individually...and intentionally, that sounds like fun. beyond me atm though Nov 17 21:47:57 I was not so amused first time I tried it hehe Nov 17 21:48:05 why's that? Nov 17 21:48:20 because I have no hw background Nov 17 21:48:30 so I had a hard time dealing with all that stuff Nov 17 21:48:53 ahh, i see Nov 17 21:49:19 there are a few good books that deal with EE stuff for programmers Nov 17 21:49:26 in fact, a lot of asm books teach that stuff Nov 17 21:49:50 I was doing some asm in the old DOS times, but that was like, 10 years ago Nov 17 21:50:01 ah Nov 17 21:50:07 * dahitokiri <3 asm Nov 17 21:53:26 what's annoying is the lack of resources on the subjects though Nov 17 21:53:34 the closer you get to hardware, the less info there is Nov 17 21:53:43 :) Nov 17 21:53:48 there are a million books on how to do "game" programming Nov 17 21:54:01 well, luckily I can always ask the guys from our hw department :) Nov 17 21:54:02 but like...10 on systems level stuff Nov 17 21:54:08 nice Nov 17 21:55:34 I never actually learned javascript :) Nov 17 21:55:36 oops wrong window Nov 17 21:55:38 hehe Nov 17 21:55:54 nice switch from close hw talk hehe Nov 17 21:57:09 web development FTL Nov 17 21:57:16 luckily not :) Nov 17 21:57:45 I use spidermonkey in my UPnP server project Nov 17 21:57:53 i started on that stuff when i was 9, got out by 11. by then it was too much about what looks better than what has more content Nov 17 22:37:43 upgrading kernel modules on "ipkg upgrade" is _really_ bright on systems with flashed kernel grrrr Nov 17 22:39:09 coredump *g* Nov 17 22:39:16 /boot Nov 17 22:43:37 "morning" all Nov 17 22:43:40 * dahitokiri wishes bash scripts were capable of multi-threading Nov 17 22:43:44 hi rp ;) Nov 17 22:43:53 yo RP Nov 17 22:47:52 03rpurdie * r983 10bitbake/lib/bb/fetch/perforce.py: perforce.py: Typo fix from Embedded Alley Nov 17 22:55:00 03rpurdie * r984 10/ (15 files in 6 dirs): Fetcher SRCREV handling updates, improvements and fixes from Poky, typo fix for bin/bitbake Nov 17 23:32:42 what does "gnu-configize" mean? Nov 17 23:52:34 the default fortran compiler is defined as arm-angstrom-linux-gnueabi-g77 after gnu-configize, how can I let it know it should actually be arm-angstrom-linux-gnueabi-gfortran Nov 18 00:19:20 Who on earth added meta-toolchain-openmoko?! :/ Nov 18 00:19:28 what OE machine confs are for cpu: e300, chip: MPC83xx ? Nov 18 00:20:00 RP: I think florian was working on the meta-toolchain things, not sure about openmoko though Nov 18 00:20:31 Jin^eLD: meta-toolchain is something I created ;-) Nov 18 00:20:55 aha, then I got it all wrong again :) Nov 18 00:21:38 but since you mention it... ;) I had one oddness there, one sec Nov 18 00:22:40 if you change the SDK_DIR from /usr/local to /opt or something else Nov 18 00:22:55 then it still packages an empty /usr directory Nov 18 00:23:18 I had to rm -rf ${SDK_OUTPUT}/usr before tarring Nov 18 00:23:28 Jin^eLD: I remember you mentioning it, I don't know why... Nov 18 00:23:38 ah, ok, so I complained already =) Nov 18 00:23:40 Jin^eLD: Someone needs to find out which package thats creeping in through Nov 18 00:24:00 Yes and I stuck my head in the sand ;-) Nov 18 00:24:06 me too ;) Nov 18 00:24:17 I wonder if that could be done in a somewhat automated way Nov 18 00:24:27 I mean not the head in the sand thing, but the figuring out about this directory ;) Nov 18 00:26:05 RP, can you point me to a machine conf that is for a Freescale e300/MPC83xx based cpu? Nov 18 00:26:20 Jin^eLD: I have no idea... Nov 18 00:26:30 doh, I guess I'll have to wait till tomorrow Nov 18 00:26:58 Jin^eLD: To find the directory you need to look through the sdk .ipk files Nov 18 00:28:09 it creates ipk files? Nov 18 00:28:19 I thought it just does the toolchain Nov 18 00:28:30 which is not run on the target system Nov 18 00:28:35 The toolchain is packaged into ipks Nov 18 00:28:52 huh? I got a .tar.gz as the result of bitbaking meta-toolchain Nov 18 00:29:06 meta-toolchain then takes the ipks and turns them into a tarball, a bit like normal image generation Nov 18 00:29:13 aha.. Nov 18 00:29:34 but those ipk's are wiped out in the process? Nov 18 00:29:52 or are they stored in work ? Nov 18 00:31:29 Jin^eLD: They're in deploy/ipk Nov 18 00:32:26 indeed, found it Nov 18 00:33:06 I'll have a look and get back to you on this Nov 18 00:33:10 not today though :) Nov 18 00:33:50 I'll be sleeping in a minute too... Nov 18 00:34:05 happy suspending then Nov 18 00:34:09 I'm off too Nov 18 00:34:10 :) Nov 18 00:34:56 'night Jin|away Nov 18 01:01:36 03rpurdie 07org.oe.dev * rf685529b... 10/ (12 files in 2 dirs): classes: Sync staging layout cleanups from Poky Nov 18 01:01:41 03rpurdie 07org.oe.dev * r6bfc2ae8... 10/ (1 conf/bitbake.conf): bitbake.conf: Set verion dependencies of -dbg and -dev packages more correctly (from Poky) Nov 18 01:01:46 03rpurdie 07org.oe.dev * r3fad086a... 10/ (1 conf/bitbake.conf): bitbake.conf: Extend PATH in anticipation of staging changes, add ability to switch OOB layout for yaffs2 images (from Poky) Nov 18 01:01:50 03rpurdie 07org.oe.dev * r64520e9d... 10/ (1 conf/bitbake.conf): bitbake.conf: Fix TERMCMD handling and a typo (from Poky) Nov 18 01:01:56 03rpurdie 07org.oe.dev * rbd9d2f1f... 10/ (1 packages/atk/atk.inc packages/atk/atk_1.20.0.bb): atk: DEPENDS on gtk-doc-native, not gtk-doc (from Poky) Nov 18 01:02:03 03rpurdie 07org.oe.dev * r3850fbf8... 10/ (4 files in 3 dirs): autoconf: Apply patch to stop bogus x includes creeping in through standard tests (from poky) Nov 18 01:02:07 03rpurdie 07org.oe.dev * re34628e7... 10/ (1 packages/eds/eds-dbus_svn.bb): eds-dbus: Sync some changes from Poky (mainly for staging fix) Nov 18 01:02:12 03rpurdie 07org.oe.dev * r7da1c93d... 10/ (3 files in 2 dirs): glibc: Various tweaks from Poky Nov 18 01:02:17 03rpurdie 07org.oe.dev * r60125320... 10/ (4 files in 3 dirs): gnutls: Add patch to fix bogus -L and -I compiler options being used (from poky) Nov 18 01:02:22 03rpurdie 07org.oe.dev * r24f8a3e0... 10/ (1 packages/meta/meta-toolchain.bb): meta-toolchain: Sync with Poky Nov 18 01:02:27 03rpurdie 07org.oe.dev * r548dc7f0... 10/ (9 files in 2 dirs): Remove meta-sdk* and replace with meta-toolchain stubs Nov 18 01:02:33 03rpurdie 07org.oe.dev * r3cffb6f8... 10/ (20 files in 11 dirs): packages: Remove various hardcoded staging layout assumptions **** ENDING LOGGING AT Sun Nov 18 02:59:56 2007