**** BEGIN LOGGING AT Thu Sep 28 02:59:56 2006 Sep 28 06:02:52 around ? Sep 28 06:37:54 hi Sep 28 06:39:06 hi hrw Sep 28 06:45:34 hi Marex Sep 28 07:22:09 good morning all Sep 28 07:23:00 morning all Sep 28 07:26:09 morning Sep 28 07:56:26 hi Sep 28 07:58:43 morning Sep 28 08:03:25 hi XorA Sep 28 08:08:37 hi Sep 28 08:10:25 * XorA has a new dog Sep 28 08:10:59 terrier? Sep 28 08:11:06 does it bites? Sep 28 08:11:06 hrw|work: yup Sep 28 08:11:13 hrw|work: only in japanese Sep 28 08:12:21 hey. morning. Sep 28 08:13:52 XorA: haven't you flashed it yet? Sep 28 08:14:23 koen: needed to find japanese to uk power adapter Sep 28 08:14:46 XorA: use other-zaurus power Sep 28 08:14:53 XorA: or ipaq charger Sep 28 08:14:56 or ipaq-power Sep 28 08:15:58 hrw|work: I got japanese to uk adapter Sep 28 08:16:05 hrw|work: all other Z's here are USA ones Sep 28 08:16:23 bad luck Sep 28 08:16:25 koen: official Ipaq ones are too narrow for Z Sep 28 08:16:42 works with my c700 :) Sep 28 08:17:03 koen: I tried 5550 and 3630 and they didnt fit my c7x0 Sep 28 08:17:42 I have sharp original 220V one, noname ipaq one, one universal Delta for Z, one US Delta for Z Sep 28 08:19:22 XorA: I bet lrg is happy to have a cxxxx 'available' for testing as well Sep 28 08:19:46 koen: he not here yet Sep 28 08:20:00 koen: but that gives us 5600 6000 c7x0 cxx00 Sep 28 08:20:34 sound support should be solid when we get cracking Sep 28 08:22:01 XorA: only 5500 is lacking Sep 28 08:22:11 hrw|work: does that have a wm chip? Sep 28 08:22:35 Every come across a RUNSTRIP error? Sep 28 08:22:54 Or indeed, any easy way of turning the STRIP process off (part of packaging) Sep 28 08:23:28 XorA: nope. Sep 28 08:23:40 blaster8: read local.conf.sample Sep 28 08:24:08 I've read it, but I use NSLU2-Linux Sep 28 08:24:27 XorA: right noy cky is blasting out of another WM chip over here :) Sep 28 08:24:30 got it Sep 28 08:24:31 (ipod) Sep 28 08:31:39 In foobar-image.bb, for what packages would I use DEPEND (build dep) and when RDEPEND (run dep)? Sep 28 08:33:10 likewise: since an image is runtime you only need RDEPENDS Sep 28 08:33:55 koen: and iff any DEPENDs are there, they are *native, required to build the image? Sep 28 08:34:32 koen: good morning and thanks for the zen-less approach this morning :-) Sep 28 08:34:35 yes Sep 28 08:34:38 :) Sep 28 08:35:35 DEPENDS in foo-image.bb are usually only for do-funky-stuff-for-dumbass-bootloader operations Sep 28 08:36:09 Q2: also, sometimes I see += " bla" and sometimes I see += "bla" in the .bb and docs. (note the space before bla). Will += introduce a space? Sep 28 08:40:08 re Sep 28 08:41:21 IMAGE_FSTYPES = "jffs2 tar" > does the jffs2 and tar are valid bootable system ? Sep 28 08:46:06 likewise: += introduces a space, _append does not Sep 28 08:46:58 I sometimes to += " foo " to make catch people _append'ing without a leading space Sep 28 08:47:03 s/to/do/ Sep 28 08:47:49 hrw|work: ewi is up again Sep 28 08:48:04 k Sep 28 08:49:13 re Sep 28 08:53:47 good morning Sep 28 08:54:04 hi florian_kc Sep 28 08:54:19 hey florian_kc Sep 28 08:54:27 well oz3.5.4.1 installed Sep 28 08:58:03 what variable should I use to cross compile a simple package (not autotools, just source code)? Sep 28 08:58:09 ${CC}? Sep 28 09:08:09 morninig Sep 28 09:08:12 morning, as well Sep 28 09:10:20 blaster8: that should work, yes Sep 28 09:10:28 morning mickeyl Sep 28 09:11:50 hi mickeyl RP mithro Tom Sep 28 09:12:31 IMAGE_FSTYPES = "jffs2 tar" > this create a bootable system ? Sep 28 09:13:23 yes Sep 28 09:13:44 if you board can boot jffs2 or tar files Sep 28 09:13:52 my workstation can't :) Sep 28 09:14:07 there are a command to create jffs2 and tar files ? Sep 28 09:14:09 ~irclog Sep 28 09:14:14 i guess irclog is apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" Sep 28 09:15:03 http://ibot.rikers.org/stats/oe.html.gz Sep 28 09:15:10 cosmicpenguin has a nice quote there Sep 28 09:15:44 I rule with 126+81 Sep 28 09:17:58 humm .. so its no :) Sep 28 09:18:52 yo hrw|work Sep 28 09:18:57 cheers mithro_, how are things? Sep 28 09:19:43 hey mickeyl Sep 28 09:20:29 can i hack bitbake in order to just display me the gcc compile messages so i see my warnings ?? Sep 28 09:20:52 rob_w: bitbake blah -D -D -D ? Sep 28 09:23:03 http://lickwid.net/~tlp/images/youMeanGNU.jpg Sep 28 09:23:06 morning all Sep 28 09:23:15 hi Liam Sep 28 09:23:23 hey lrg Sep 28 09:23:54 koen, yeah that gives me DEBUG for all task`s but i would love to only see the gcc compile and link stuff Sep 28 09:24:34 03xora 07org.oe.dev * rb4422f8e... 10/ (1 packages/dbus/dbus-native_0.92.bb): dbus-native_0.92.bb : add libxml2-native and expat-native to DEPENDS Sep 28 09:29:29 plz how i can generate jffs2 and tar ? there are a special command or there are somthing missing in my conf ? (and there are nothing about that in the wiki gettingstarted) Sep 28 09:36:52 gumz: set IMAGE_FSTYPES and then bitbake bootstrap-image Sep 28 09:37:01 gumz: ALL needed stuff will get build Sep 28 09:37:10 XorA: dbus 0.93 released Sep 28 09:38:24 thanks! Sep 28 09:39:50 rob_w: all logs are in WORKDIR/temp/ Sep 28 09:40:17 hrw|work: Im sticking well clear of dbus Sep 28 09:40:21 ok,, so i could grep through them Sep 28 09:44:33 omg... openzaurus.org ... 274b/s ... :) Sep 28 10:10:10 zecke: the tinderbox is up again Sep 28 10:11:02 on which machine and why? Sep 28 10:11:20 ewi546 and because the move has been completed Sep 28 10:11:44 which move? I thought you lost the machine completely? Sep 28 10:11:50 hi zecke Sep 28 10:11:55 long time no see Sep 28 10:11:59 zecke: how was aKademy? Sep 28 10:12:06 mickeyl: I'm still there ;) Sep 28 10:12:09 zecke: that's still in the queue, but they had to move it to a new location first Sep 28 10:12:10 ah Sep 28 10:12:21 koen: ah okay Sep 28 10:12:25 ~lart linuxdevices.com for comparing qemu with nslu2 Sep 28 10:12:25 * ibot turns linuxdevices.com into a lifesized tux doll for comparing qemu with nslu2 Sep 28 10:13:34 when used on one of the fastest PCs you can buy, it outperforms a $50 gadget Sep 28 10:13:37 wow Sep 28 10:14:28 a 4 ghz pc can outperfrom a 133MHz arm chip Sep 28 10:16:10 koen: wow 8ghz might outperform a 166mhz chip Sep 28 10:16:21 :) Sep 28 10:16:34 and it wastes so much more power as well Sep 28 10:16:44 that makes it fit perfectly for debian :) Sep 28 10:16:55 in fact, why do we bother with out silly embedded devices at all Sep 28 10:17:13 we should just drag around full size PCs and emulate their functionality Sep 28 10:18:44 hehe, actually I have hooked up people with qemu arm as well Sep 28 10:18:59 but just because it is the 'cheapest' ARM you can by and just works Sep 28 10:19:05 by/buy/ Sep 28 10:19:37 mickeyl: I think my talk was okay, it is even covered in one german linux online publication but I'm too shy to read it Sep 28 10:19:52 mickeyl: I will give a small lightning talk about poky+X+display migration this evening Sep 28 10:20:00 mickeyl: and ATM I'm playing with GCov Sep 28 10:20:24 cool! Sep 28 10:20:27 ouch BenQ is going out of business Sep 28 10:20:32 *nod* Sep 28 10:20:33 bummer Sep 28 10:20:42 If I run "mount -a" several times on device, I end up with several /var mounts :-( Sep 28 10:21:22 same here on my desktop Sep 28 10:21:31 Hi! Sep 28 10:21:42 mickeyl, Did you have a look http://bugs.openembedded.org/show_bug.cgi?id=1443 ? Sep 28 10:21:45 mickeyl: My desktop doesn't do that... Sep 28 10:22:26 mickeyl, glibc breakage after your last update of it... Sep 28 10:23:28 psokolovsky: that's disgusting. without the patches, it doesn't build on x86. let's make it _x86 for now Sep 28 10:23:32 can you do it? Sep 28 10:23:43 SRC_URI_x86 += , that is Sep 28 10:24:01 mickeyl: ok, will do in couple of hours. thanks Sep 28 10:24:10 np, sorry 'bout that. Sep 28 10:24:39 np Sep 28 10:25:38 zecke: do you have url ? Sep 28 10:26:09 lrg: about BenQ? Sep 28 10:26:17 zecke: yes Sep 28 10:26:31 lrg: online www.heise.de which is sadly german. But these are two lines Sep 28 10:27:05 zecke: thanks. Sep 28 10:28:11 lrg: The german BenQ Mobile GmbH is likely to declare bankruptcy. The german headquarter and german manufacturing are likely to go out of business Sep 28 10:28:26 mickeyl: I fear we can't have launch at the place I wanted to go :} Sep 28 10:28:38 zecke: ah Sep 28 10:30:14 zecke: heh. too bad. does that have any negative consequences for your friends? Sep 28 10:30:51 mickeyl: I don't know, I really don't know Sep 28 10:31:07 zecke: do you have a link to that article about your presentation? Sep 28 10:31:45 koen: http://www.linux-community.de/Neues/story?storyid=20941 I didn't read that, so If if blew up just rephrase it nicely :) Sep 28 10:36:55 http://static.kdenews.org/jr/akademy-2006-group-photo.html Sep 28 10:37:04 search for 'Hoger' :) Sep 28 10:37:13 haha Sep 28 10:42:15 ah, thus proving that KDE people shoukd be kept in dark rooms ;) Sep 28 10:42:21 should Sep 28 10:42:24 man that with BenQ sucks badly, I will get myself a Guinnes or two now Sep 28 10:42:31 ade|desk: thanks Sep 28 10:42:35 :) Sep 28 10:42:41 guinness Sep 28 10:42:46 hmmmm Sep 28 10:42:48 nice Sep 28 10:45:26 zecke: where are you in the photo Sep 28 10:45:38 "Hoger" even If I'm Holger Sep 28 10:51:00 do that. Sep 28 10:51:06 (guiness) Sep 28 10:51:14 can't wait until thursday Sep 28 10:51:16 :D Sep 28 10:51:23 double n, double s Sep 28 10:51:31 right Sep 28 10:51:34 ~spell guinness Sep 28 10:51:36 possible spellings for guinness: Guinness Guinness's guineas gunners Guinna's Guinea's guinea's Gaines's Ginnie's Gunner's gunner's Inness Gaines quines Ginni's Ginny's Quinn's gainers gayness gingers gunny's gurneys Guenna's Guiana's Gwynne's gunnel's Ginger's Kinney's gainer's ginger's grinner's Sep 28 10:51:40 03pfalcon 07org.oe.dev * rb3b16abc... 10/ (1 packages/glibc/glibc_2.4.bb): Sep 28 10:51:40 glibc 2.4: Apply recently added patches only fo x86. Sep 28 10:51:40 * Fixes breakage on ARM. Sep 28 10:51:40 * Don't bump PR (previous commit neither did), to save long rebuilds. Sep 28 10:51:40 Few people affected should use "bitbake glibc -c rebuild". Sep 28 10:51:40 * Fixes #1443. Sep 28 10:51:42 * Approved by Michael 'Mickey' Lauer . Sep 28 10:52:10 thanks psokolovsky Sep 28 11:13:36 hmm i can't remember if I took my drugs today Sep 28 11:16:19 ~kill phpbb Sep 28 11:16:25 * ibot shoots a magneto-ionized meson gun at phpbb Sep 28 11:18:41 poor phorums Sep 28 11:20:29 XorA: we have one nonprofit website. they have phpbb forum with some weird hackset applied Sep 28 11:20:39 now they want fixes, changes etc.. Sep 28 11:20:42 ofcourse free Sep 28 11:20:52 hrw|work: thats the free software way Sep 28 11:21:05 XorA: I do not work in foss company Sep 28 11:28:35 mickeyl: Were the sysvinit /proc/progress changes added by you for elpp? Sep 28 11:29:19 RP: yes. Sep 28 11:29:58 mickeyl: What are the current thoughts on elpp? Are we likely to go back to it? Sep 28 11:30:22 pfalcon: ping Sep 28 11:30:33 http://www.openembedded.org/viewmtn/revision.psp?id=b3b16abcf910c9b1e8e145776848c3272f9d2f6b Sep 28 11:30:45 blaster8, pong Sep 28 11:30:46 did you forget to add the patches to the workspace? Sep 28 11:31:13 see URL Sep 28 11:31:27 blaster8, I only moved patch references around Sep 28 11:31:36 hrm Sep 28 11:31:44 RP: I would dump it - there are userspace splash solutions Sep 28 11:31:46 psokolovsky: x86 isn't the only x86 machine around :( Sep 28 11:32:12 blaster8, patches where there since previous commit Sep 28 11:32:21 ok, sorry to fuss Sep 28 11:32:27 blaster8, so what's exact issue? Sep 28 11:33:27 none, it's ok :) Sep 28 11:33:33 koen, was suggested by Mickey. My aim was to fix build for arm. Anyone knowing better should improve that fix ;-) Sep 28 11:33:37 ok Sep 28 11:34:43 someone mentioned a fixed version of the patches on irc or the mailinglist Sep 28 11:35:13 RP: definitly not. our plan should be going to initramfs or another user space solution. i wouldn't support any kernel based solutions any more. for now we could yank all /proc/process traces Sep 28 11:36:17 mickeyl: ok, thanks. It might not hit .dev immediately but things are moving on that front in poky and .dev might want to copy at some point :). I'll just remove that code, Sep 28 11:40:34 03blaster8 07org.oe.dev * r78ec2315... 10/ (5 files in 4 dirs): slugos-init: improve module loading at boot Sep 28 11:40:40 03blaster8 07org.oe.dev * r442cdb37... 10/ (1 packages/tcpdump/tcpdump_3.9.3.bb): tcpdump: tcpdump is compiled without crypto, so remove openssl RDEPENDS Sep 28 11:40:51 03blaster8 07org.oe.dev * rba21e3a1... 10/ (6 files in 4 dirs): Sep 28 11:40:51 mii-diag: add new package Sep 28 11:40:51 * Used for debugging MII, needed for fixing NSLU2 LE Sep 28 11:41:46 this is not possible to have a TAR of whole system ? other than all .tar.gz ? (yes i abuse :p) Sep 28 11:45:06 RP: excellent. thanks Sep 28 11:45:06 morning ~! Sep 28 11:49:19 gumz: tar as 'non-compressed tar'? Sep 28 11:50:13 gumz: it is not possible. 'tar' type was tar.gz in past and we left it as is Sep 28 11:52:01 bother Sep 28 11:52:09 busybox build dies on me last night Sep 28 11:52:24 arm-linux-gcc: Internal error: Killed (program cc1) Sep 28 11:52:24 Please submit a full bug report. Sep 28 11:53:18 mickeyl: >>> os.path.join("data/kdelibs", "/CMakeTmp/cmTryCompileExec") Sep 28 11:53:19 '/CMakeTmp/cmTryCompileExec' Sep 28 11:53:22 is that a bug? Sep 28 11:54:09 no Sep 28 11:54:22 os.path.join expects a sequence, doesn't it? Sep 28 11:54:26 Crofton: looks like OOM Sep 28 11:55:08 hrw|work: i mean all files of rootfs in 1 tar.gz, actually all package are separed... Sep 28 11:55:16 zecke: hmm, wait Sep 28 11:55:37 koen, very likely Sep 28 11:55:51 Crofton: no swap present on that box? Sep 28 11:56:05 total used free shared buffers cached Sep 28 11:56:06 Mem: 444756 421288 23468 0 8100 94344 Sep 28 11:56:06 -/+ buffers/cache: 318844 125912 Sep 28 11:56:06 Swap: 884728 572556 312172 Sep 28 11:56:32 zecke: either it is a bug or it's illegal using path separators within arguments to os.path.join Sep 28 11:56:56 gumz: build image and look into tmp/deploy/images/ Sep 28 11:56:59 2nd argument is not allowed to start with /, then it works Sep 28 11:58:56 hrw|work: i juste have the jffs2 file in images Sep 28 12:02:53 gumz: IMAGE_FSTYPES = "jffs2 tar.gz" set and rebuild bootstrap-image Sep 28 12:05:38 hrw|work: thanks i try now ... BTW, mysql are working now on my gumstix .. OE OWNZ! :) YOU ALL rules... i'm happy to have OE, and finally, its work ! :) Sep 28 12:06:03 ~hail oe Sep 28 12:06:05 * ibot bows down to oe and chants, "I'M NOT WORTHY!!" Sep 28 12:09:44 anyone have any bright ideas how to build busybox on a machine with "limited" ram? Sep 28 12:09:59 more RAM? Sep 28 12:10:02 :p Sep 28 12:10:07 heh Sep 28 12:10:09 more swap? Sep 28 12:10:15 well Sep 28 12:10:21 I though abut more swap Sep 28 12:10:37 but the svn server slows to a crawl during the build Sep 28 12:10:53 i use swap on CF card when i need more ram ... Sep 28 12:11:03 I just find it amusing that a key package for an embedded system has such compile requirements Sep 28 12:16:38 hrw|work: always just JFFS2 file.. but its ok, i just have to make a little script who ungzip all files, and make all in one Sep 28 12:18:19 Crofton: XIP? Sep 28 12:19:43 ~xip Sep 28 12:19:47 ? Sep 28 12:19:52 it has been said that xip is pretty sweet though - since the stuff is already in "rom". or eXecute In Placem which is no good for a compressed r/w FS in ROM Sep 28 12:20:34 So busybox is compiles to xip? Sep 28 12:20:54 Crofton: I don't know how it works nowadays. On Agenda Vr3 you had a special kernel Sep 28 12:21:04 Crofton: and abused the sticky bit on executables to XIP Sep 28 12:27:15 spitz machines dont charge from usb? Sep 28 12:45:19 XorA: correct, AFAIK spitz machines only charge from their DC-IN port Sep 28 12:46:06 XorA: it makes sense, since the USB intf is on-the-go Sep 28 12:47:15 XorA: at any rate, I'm delighted at the news of your new acquisition :) Sep 28 12:51:56 kerwood: that sucks, +5v readilly available and I have to lug around a power brick as well Sep 28 12:54:57 XorA: well, a few moments with a soldering iron, an old USB cable, and a matching power connector should simplify that, eh? Sep 28 13:07:38 kerwood: :-D Sep 28 13:13:56 hi Crofton :) Sep 28 13:15:18 hi bigd0g Sep 28 13:15:38 we have fixed the kernel build for the OSK to track with improvements in git Sep 28 13:15:46 of course I haven't been able to test ... Sep 28 13:16:07 <- test dummy ;) Sep 28 13:16:25 i've been busy developing software we needed for our project Sep 28 13:16:29 so the board's been on a back burner Sep 28 13:16:33 that sounds important :) Sep 28 13:16:37 but... i'm ready to go back to playing with it Sep 28 13:16:40 cool Sep 28 13:17:10 I am finsihing some papers, I was hoping to check the images, but my busybox build died for lack of memeory Sep 28 13:17:23 hrw has also been helping with machine and distro issues Sep 28 13:17:24 ouch Sep 28 13:17:41 apparently I can project management with emacs Sep 28 13:17:43 excellent, excellent. i really appreciate the updates you guys have put into the omap Sep 28 13:18:05 the kernel is builfing of cutrent git atm Sep 28 13:18:18 since the dspgw updates are only there Sep 28 13:18:39 i see Sep 28 13:18:57 you need the dspgw? Sep 28 13:19:07 cut rent git, sounds the opposite to me ;) Sep 28 13:19:38 nah, i'm not using the dsp Sep 28 13:19:45 ...so won't be needing the dspgw Sep 28 13:19:51 good Sep 28 13:19:58 I need to make it work at soe point Sep 28 13:20:17 the board is actually more powerful than what we need. but it was the cheapest dev board available Sep 28 13:20:41 it's just a proof-of-concept design, anyway. minimization can be for another dev team ;) Sep 28 13:20:43 cool Sep 28 13:20:57 random question: Sep 28 13:21:12 do you know where i can get the driver for the lcd controller on the osk ? Sep 28 13:21:52 it is in the kernel Sep 28 13:22:07 one upon a time I made opie work on the osk Sep 28 13:22:48 nice Sep 28 13:23:04 Do you need it soon? Sep 28 13:23:19 I will need to go check that we build the modules Sep 28 13:23:27 the lcd driver ? Sep 28 13:23:30 yeah Sep 28 13:23:55 yeah... it's my next task to do. now that the lcd pcb has been built, they need me to draw stuff on it Sep 28 13:24:02 ah Sep 28 13:24:13 i didn't realize it had to be built with something Sep 28 13:24:15 when I am over at school I'll look at the kconfig Sep 28 13:24:17 i thought i could do it afteerwards Sep 28 13:24:34 ok, cool. much appreciated. Sep 28 13:25:10 i'm using a custom lcd ... so i'm not sure what modifications i'll need to make Sep 28 13:25:12 i have an idea, tho Sep 28 13:25:17 heh Sep 28 13:25:30 I do not know much about this sort of thing at all Sep 28 13:26:03 are the files already in org.oe.dev somewhere so that i can look them over ? Sep 28 13:26:05 bigd0g: look at 2.6.12-rc3 config Sep 28 13:26:12 ding Sep 28 13:26:14 hrw|work: thanks Sep 28 13:27:44 looks like there is some fb support in the current bb file Sep 28 13:27:46 i see rc2, not rc3. was it added today? i'll update the repo again Sep 28 13:27:59 bigd0g: look in 2.6.12-rcX anyway Sep 28 13:28:32 that kernel has the advantage of having worked at some point :) Sep 28 13:28:35 Crofton: ok. ideally i'd like to use just the framebuffer. but if i need to, i can run gtk via x11 Sep 28 13:28:43 Crofton: haha ;) Sep 28 13:29:08 I think I had to use the micrdrive to actually get opie running on the OSK :) Sep 28 13:29:32 at least it worked, right ? Sep 28 13:29:41 right :) Sep 28 13:30:23 celf use 2.6.12-rcX Sep 28 13:30:46 i'll take a look-see as soon as the mtn update finishes Sep 28 13:30:53 I am not so familiar with celf Sep 28 13:32:42 hrw|work: i only see 'linux-omap1-2.6.12-rc2' ... i don't see an rcX Sep 28 13:32:56 X = 2 Sep 28 13:32:56 hrw|work: is it not in packages/linux? (i just updated org.oe.dev) Sep 28 13:32:57 X might mean he can't remember two Sep 28 13:33:01 ohhh Sep 28 13:33:01 heh Sep 28 13:33:19 confusing because some of the bb files actually don't have a number ;) Sep 28 13:33:36 2.6.x+git you mean :) Sep 28 13:33:40 i.e. linux-omap1_2.6.x+git.bb Sep 28 13:33:43 precisely ;) Sep 28 13:34:40 Crofton: I hope that 2.6.18-omap1 will be soon released Sep 28 13:34:48 yeah Sep 28 13:34:53 the patch? Sep 28 13:35:28 hrm. "linux-omap1_2.6.12-rc2.bb" and "linux-omap1_2.6.12-rc2.bb~" show up empty in emacs Sep 28 13:36:36 stuff here ..... Sep 28 13:37:42 hrm? Sep 28 13:38:00 weird Sep 28 13:38:13 do you have 2.6.18-rc ... Sep 28 13:38:32 koen, looks like dspfs is intended to replace dspgw Sep 28 13:38:35 http://rafb.net/paste/results/AU2sXN84.html Sep 28 13:39:29 hmm, I must have never cmmited the .2.6.18 based off git tag Sep 28 13:39:38 likely becuase it didn't work .... Sep 28 13:39:38 morning Sep 28 13:39:48 hey chouimat Sep 28 13:45:41 oh, another random question. how can i submit documentation to OE ? Sep 28 13:46:02 wiki? Sep 28 13:46:11 i've generated a pretty hefty how-to through the course of this project and want to give it to you guys when i'm done Sep 28 13:46:39 http://www.openembedded.org/documentation Sep 28 13:46:40 i don't know if it's wiki-able ... it's more or less and idiot's guide to getting OE up-and-running, building everything, and getting the osk working Sep 28 13:46:43 bigd0g: look into org.openembedded.documentation branch Sep 28 13:47:05 bigd0g: in other words: you have 'Use Case' Sep 28 13:47:14 also called 'Case Study' Sep 28 13:47:20 yeah Sep 28 13:47:35 is there a place you prefer those get dropped? Sep 28 13:47:39 the thing which we lack Sep 28 13:48:00 mhm... sort of why i made it as i stumbled around and bothered everyone :) Sep 28 13:48:04 bigd0g: look into org.openembedded.documentation branch - I do not remember does usermanual has section for it Sep 28 13:48:44 ok. Sep 28 13:48:50 * bigd0g reads the org.oe.doc README Sep 28 13:48:58 oh Sep 28 13:49:02 it's 1-line long. rofl Sep 28 13:50:25 ;) Sep 28 13:51:45 well, it's not terribly important right now. i just figured it'd be cool to "give back" somehow Sep 28 13:52:18 ~seen minipanda Sep 28 13:52:32 minipanda was last seen on IRC in channel #oe, 14d 2h 13m 34s ago, saying: 'crap, in $OESTUFF/build/tmp/rootfs/ have /dev, but in bootstrap-image.tar.gz, no!'. Sep 28 14:59:20 hi.. Sep 28 14:59:51 there seems to be a problem with the python-lxml package Sep 28 15:00:48 what kind of problem? Sep 28 15:01:05 sory... phone.. Sep 28 15:01:14 hi hrw|work Sep 28 15:01:16 its not python-libxml2.. :) Sep 28 15:01:19 sorry... Sep 28 15:01:26 ImportError: /usr/lib/python2.4/site-packages/lxml/etree.so: undefined symbol: xmlC14NDocSaveTo Sep 28 15:01:58 this happens when trying to import the etree module Sep 28 15:02:23 I'm not sure if this problem is lxml of libxml2 Sep 28 15:02:26 I suspect the later Sep 28 15:03:19 I notice libxml2 is being compiled with '--without-python' Sep 28 15:03:27 but I don't think that should matter Sep 28 15:03:56 mickeyl: you about? Sep 28 15:07:39 I have the libxml2 dev packages installed on the target too Sep 28 15:07:49 tkp: in debian it need libxml2, libxslt, python-pyrex Sep 28 15:07:59 python-pyrex... Sep 28 15:08:06 hmm... thats not in the deps Sep 28 15:09:25 from reading the lxml site, I think that is need for compilation only and not runtime Sep 28 15:09:52 tkp: debian libxml2 contain xmlC14NDocSaveTo Sep 28 15:10:12 libxml2 2.6.26 Sep 28 15:10:31 oe is on 2.6.22 Sep 28 15:10:44 I know Sep 28 15:11:04 I wonder if thats the problem Sep 28 15:11:54 I'm checking Sep 28 15:12:36 it was sugested to me in #python that the problem was one of versioning Sep 28 15:12:56 03pfalcon 07org.oe.dev * r8cccd2b6... 10/ (1 packages/linux/handhelds-pxa-2.6/h4000/defconfig): h4000: Update defonfig, enable audio (statically). Sep 28 15:13:13 tkp: debian 2.6.22 also contain that symbol Sep 28 15:13:21 hm, delays with cia Sep 28 15:13:35 03blaster8 07org.oe.dev * r8dcac6d1... 10/ (1 packages/linux/ixp4xx-kernel_2.6.18.bb): Sep 28 15:13:35 ixp4xx-kernel: SVN Revision bump Sep 28 15:13:35 * now includes patches allowing LE ethernet functionality Sep 28 15:13:44 psokolovsky: yeah, oe.org was hanging on contacting the cia server Sep 28 15:13:54 nslu2-linux too Sep 28 15:14:23 tkp: OE one does not have it Sep 28 15:14:31 hmm Sep 28 15:14:48 hrw|work: any ideas why not? Sep 28 15:14:53 tkp: remove --disable-c14n Sep 28 15:15:38 tkp: then rebuild and try Sep 28 15:15:47 rebuilding now... Sep 28 15:15:59 the c14n modules provides a "Canonical XML" implementation Sep 28 15:16:09 xmlC14NDocSave is part of it Sep 28 15:16:25 I see Sep 28 15:16:36 fingers crossed then Sep 28 15:16:45 ;) Sep 28 15:17:47 hm siemens mobile is dead Sep 28 15:19:41 hrw|work: that seems to have fixed it :) Sep 28 15:21:26 tkp: how bigger libxml is? Sep 28 15:22:28 it's 984K now... Sep 28 15:22:34 I'm not sure what it was before Sep 28 15:22:50 arm? Sep 28 15:22:53 libxml2-dev is noe 3.3M Sep 28 15:22:57 now* Sep 28 15:23:00 x86 Sep 28 15:23:10 aha Sep 28 15:23:22 can you rebuild it later without c14n and check? Sep 28 15:23:38 sure... give me 5 Sep 28 15:23:52 tx Sep 28 15:27:07 c Sep 28 15:27:32 libxml2: 984K -> 976K Sep 28 15:28:10 libxml2-dev is now 3296k... need to recompile again to get a more accurate size reading for what it was before! Sep 28 15:28:27 libxml2 shrank! Sep 28 15:29:47 03blaster8 07org.oe.dev * rfdaad165... 10/ (5 files in 5 dirs): slugos-init: Further improvements to module loading Sep 28 15:35:12 tkp: then push change to not disable c14n (and describe why you enable it), put python-lxml Sep 28 15:35:54 libxml2-dev: 3296K -> 3328K Sep 28 15:41:39 a+ Sep 28 16:03:41 03blaster8 07org.oe.dev * r5342a583... 10/ (1 packages/udev/udev_100.bb): udev: Enable udev-100 on SlugOS Sep 28 16:03:46 03blaster8 07org.oe.dev * rd3f331b7... 10/ (1 packages/images/slugos-image.bb): slugos: Revert to using udev (mdev functions, but turnup breaks) Sep 28 16:05:40 does any of the OE machines have a kernel module written to map all the PXA270 I/Os to /proc/gpio? Sep 28 16:28:04 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Sep 28 17:37:00 03pfalcon 07org.oe.dev * r600d1cf3... 10/ (1 packages/glibc/glibc_2.4.bb): glibc_2.4.bb: Typo fix in comment. Sep 28 17:42:12 is there any plan to branch OE into a stable branch. I love OE but it seems like everytime I update OE I have to go back through a fix a couple broken packages. Sep 28 17:46:23 Gerrath: .....org.oe.oz354x Sep 28 17:46:42 Gerrath: of course, it's only for Zaurii/OZ Sep 28 17:47:01 Gerrath: you're of course welcome to do a stable branch of your own Sep 28 17:49:16 JustinP, it just seems to me that it would make sense to do 1 branch of OE for all boards and only apply patches to that branch that have been highly reviewed and tested. Maybe its just a pipe dream :-) Sep 28 17:49:55 Agreed, I find that I cannot build anything more complicated than bootstrap-image Sep 28 17:50:14 i suspect if we had the manpower, something like that would be viable. Sep 28 17:51:27 kergoth, I'm not trying to be critical.. I think what you guys have done with the little man power is amazing, Sep 28 17:52:05 and i'm explaining that the required manpower increases substantially with every additional line of development.. which is true for any project Sep 28 17:52:26 hi Sep 28 17:52:41 hey Sep 28 17:52:49 kergoth, I see your point. Sep 28 17:53:00 JustinP: I do not agree with you. .oz354x was made for openzaurus but there are no problems with using it for other stuff. Sep 28 17:53:32 Gerrath: use distro with 'sane-srcdates' include to keep more stuff usable Sep 28 17:55:02 Gerrath: creating stable branch need few devels and limiting to few targets only. but branch will stagnate if those devs will do it in free time. Sep 28 17:56:22 Gerrath: I have that situation with .oz354x branch. I maintain it but it lack many things from .dev (and viceversa too ;( but it is kind of stable. for first experiencies with OE it can be used but change from .oz354x to .dev can be hard then Sep 28 17:57:13 hrw, thats a real good point. You would end up stuck on an isolated island (stable-release) :-) Sep 28 17:58:29 hrw, honestly I was unaware of the sanesrcdates, that seems like a happy medium approach. Sep 28 17:58:35 hrw, if I took the oz354 branch, use the local.conf.openzaurus-3.5.4, what do I tell bitbake to build? Sep 28 17:58:39 how i can contact the one who have make gumstix.conf in conf/machine ? Sep 28 17:58:51 ofcourse there are other projects then OE in embedded world. but how many of them are made by passionats? Sep 28 17:58:59 hrw: bitbake Sep 28 17:59:10 T0mW: bitbake your-selected-target Sep 28 17:59:17 hrw: gpe? Sep 28 17:59:21 T0mW: I need to update local.conf.. Sep 28 17:59:24 T0mW: gpe-image Sep 28 17:59:25 hrw, I have used some of the other projects and they can't come close to OE. Sep 28 17:59:41 hrw: not found Sep 28 17:59:45 buildroot, t2, timesys Sep 28 18:00:00 T0mW: I build gpe-image from .oz354x about 1-2 weeks ago Sep 28 18:00:19 hrw: I guess I have to copy the task + image dirs from org.openembedded.dev ? Sep 28 18:00:28 dont forget scratchbox, which kicks our ass as a development tool, but is mediocre at best at the build and distro side of things Sep 28 18:01:36 T0mW: no Sep 28 18:01:43 T0mW: .oz354x contain gpe-image Sep 28 18:02:05 hrw, bitbake complains about TARGET_ARCH not being set, sorry. Sep 28 18:02:06 kergoth: yep - I have to use it one day Sep 28 18:02:18 T0mW: which distro/machine combo? Sep 28 18:02:26 zaurus Sep 28 18:02:34 T0mW: there is no such machine Sep 28 18:03:32 hrw, I used the org.openembedded.oz354x/conf/local.conf.openzaurus-3.5.4 as the local.conf Sep 28 18:03:44 03blaster8 07org.oe.dev * r11eafc7d... 10/ (3 files in 3 dirs): slugos-init: Add delay to disk booting to cope with module loading Sep 28 18:03:50 T0mW: and it does not have machine set Sep 28 18:04:10 hrw: true, one could use it, but one would have to copy in the machine/distro configs and kernel recipes from dev if not building for OZ.... Sep 28 18:04:21 hrw, oh, ok, I thought that it was all ready go Sep 28 18:04:59 hrw: can you suggest what would work or what works for you? Sep 28 18:05:34 T0mW: MACHINE="c7x0" DISTRO="openzaurus-3.5.4.1" - rest is your local stuff Sep 28 18:06:03 T0mW: my simplest local.conf contain only DISTRO. MACHINE is set in auto.conf, paths are in site.conf Sep 28 18:06:18 hrw: nod, thanks Sep 28 18:06:51 hrw: I desperately need to move beyond bootstrap-image, but all my builds with org.openembedded.dev for gpe-image are badly broken Sep 28 18:07:21 hrw: so, if some form of zaurus is buildable, I'll try from there. Sep 28 18:09:05 T0mW: use .dev with openzaurus-unstable distro? Sep 28 18:09:21 yeah Sep 28 18:09:58 03hrw 07org.oe.oz354x * rffbb21ec... 10/ (3 files in 2 dirs): added example local.conf for 3.5.4.1, dropped one for 3.5.4 as this distro version is not supposed to be used Sep 28 18:11:10 I want 'mtn cp' command... Sep 28 18:12:04 hrw: maybe for xmas if you're a really good boy ;) Sep 28 18:12:12 chouimat: ;D Sep 28 18:12:29 chouimat: few versions ago I got pluck so who knows... Sep 28 18:12:38 njs: did you read it? :D Sep 28 18:13:01 * mrz80 is back (gone 01:44:57) Sep 28 18:13:04 someone used via integrated gfx? Sep 28 18:13:36 ARRRGGGG ... I hate this they call that the develop is getting late but the main development box is 500km away and it's down today and nobody are there to answer the phone ... you have to love distributed development Sep 28 18:14:19 chouimat: I will soon have situation where I will work on machine -9h from me Sep 28 18:14:49 hrw: it's not bad ... except when it got down and you can't reach anybody Sep 28 18:15:17 chouimat: too bad that my work hours == their sleep hours Sep 28 18:15:33 hrw: yeah this sucks Sep 28 18:19:27 03kergoth * r613 10bitbake/lib/bb/taskdata.py: Minor bugfix in taskdata, the dump function was passing a list to bb.msg.debug where it expected a string. Use ', '.join() to create a string. Sep 28 18:19:36 chouimat: or vice versa Sep 28 18:21:29 hrw, where is this machine? Sep 28 18:21:39 Crofton: US Sep 28 18:21:47 hrw, check out the sane-srcdates in more detail.. thats really a good idea. I think that will address most of my issues :-) Thanks. Sep 28 18:21:48 big place :) Sep 28 18:22:20 If it is close, I could kick it for you if you have problems :) Sep 28 18:22:30 hrw, /s/check/checked Sep 28 18:22:38 of course the chances of it being close to me are slim Sep 28 18:22:59 Crofton: the changes to be able to get it are slim too Sep 28 18:23:35 hrw: -9, is that Pacific time then? Sep 28 18:24:39 San Jose, CA Sep 28 18:25:22 be back in ~hour probably Sep 28 18:25:38 too far to drive .. Sep 28 18:25:46 heh Sep 28 18:29:04 Crofton: you are about 100 miles closer than I am, go for it! Sep 28 18:34:46 kergoth: That would be a leftover from when the dump function used print :) Sep 28 18:35:18 :) i figured Sep 28 18:36:05 he zecke Sep 28 18:36:09 hey Sep 28 18:36:16 woglinde: BenQ is... :( Sep 28 18:36:21 zecke beny-siemens is dead Sep 28 18:36:28 yes Sep 28 18:36:34 I feared that Sep 28 18:36:59 woglinde: well, one always knew that but "Insolvenz" is pretty bad Sep 28 18:37:21 what was benw? Sep 28 18:37:24 er benq Sep 28 18:37:33 yes benq Sep 28 18:37:56 benq bought the cellphone production from siemens Sep 28 18:38:01 ah Sep 28 18:38:13 and now siemens stopped the paying Sep 28 18:38:30 and benq called chapter 7 Sep 28 18:38:33 or so Sep 28 18:38:35 dont know Sep 28 18:38:48 in german its "insolvenz" Sep 28 18:39:17 hm insolvency Sep 28 18:54:53 03blaster8 07org.oe.dev * r31316236... 10/ (6 files in 5 dirs): Sep 28 18:54:53 ixp4xx-kernel: Add auto-byteswapping patch Sep 28 18:54:53 ixp4xx-npe: Remove byteswapping behaviour Sep 28 18:54:53 slugos-init: lengthen delay slightly Sep 28 19:10:03 hmm.. I just realised that oedem is in NEXT week, not week after.. Sep 28 19:11:33 someone used VIA integrated gfx? Sep 28 19:11:50 03hrw 07org.oe.oz354x * rf63a3b7b... 10/ (1 conf/local.conf.openzaurus-3.5.4.2): added local.conf for OZ 3.5.4.2 Sep 28 19:11:52 in x86 desktop machines, not epias Sep 28 19:30:37 03pfalcon 07org.oe.dev * r2885bcf7... 10/ (1 packages/rocksndiamonds/rocksndiamonds_3.2.0.bb): rocksndiamonds 3.2.0: Pass target AR & RANLIB to make. Sep 28 20:00:30 cu Sep 28 20:09:58 is http://www.oesf.org/forums official forum ? Sep 28 20:10:37 SD840_: no, that has nothing todo with OE Sep 28 20:10:45 except that they stole our name Sep 28 20:11:02 ok Sep 28 20:11:11 And my ham sandwich Sep 28 20:11:45 ERROR: Nothing provides dependency kernel-module-jfs Sep 28 20:11:46 grrrrr Sep 28 20:13:53 i don't know what but OE clearly hate me when i want to do a distro! :) Sep 28 20:18:00 ah, right Sep 28 20:18:18 someone added a RDEPEND on a kernel module Sep 28 20:18:22 that's a bug Sep 28 20:18:40 ./conf/machine/gumstix.conf:22:BOOTSTRAP_EXTRA_RDEPENDS = "kernel-modules-mmc kernel-modules-usb kernel-modules-fat kernel-modules-bluetooth kernel-module-jfs udev lrzsz" Sep 28 20:18:50 ok i don't think that went well... Sep 28 20:19:43 i think my initrd is borked :( Sep 28 20:20:13 well time to figure out whats wrong Sep 28 20:21:55 03koen 07org.oe.dev * r3529a286... 10/ (1 conf/machine/gumstix.conf): Sep 28 20:21:55 gumstix.conf: fix typos for kernel modules Sep 28 20:21:55 * this file still (R)DEPENDS on kernel modules, which is bug, one should only RRECOMMEND them, leaving that to the maintainer or to the convert-to-task-base-squad Sep 28 20:22:53 SD840_: the quick fix is to remove the kernel-module- from conf/machine/gumstix.conf Sep 28 20:23:21 yep its what i do Sep 28 20:23:40 fat and jfs don't look very 'needed' to me Sep 28 20:23:45 gmustix (and not gumstix??) in distro don't work.. gumstix don't boot Sep 28 20:23:51 it's not like the board won't boot without them :) Sep 28 20:23:55 so now i try with generic distro Sep 28 20:24:14 gmustix is a distro from the George Mason University iirc Sep 28 20:24:20 hence the GMU Sep 28 20:24:24 ok Sep 28 20:24:36 so nothing to do with gumstix? :) Sep 28 20:24:39 ~seen TMM1 Sep 28 20:24:40 lol Sep 28 20:24:51 koen: In the packaged-staging branch, are there many changes to the packages themselves? Sep 28 20:24:53 SD840_: yes, it's for the gumstix Sep 28 20:24:54 tmm1 was last seen on IRC in channel #oe, 5d 33m 17s ago, saying: 'ah yea'. Sep 28 20:25:03 03blaster8 07org.oe.dev * rbb69f40c... 10/ (1 packages/linux/ixp4xx-kernel_2.6.18.bb): ixp4xx-kernel: One last revbump, cleaning up the patch directory Sep 28 20:25:05 RP: there should be none Sep 28 20:25:18 RP: only a EXCLUDE_FROM_SHLIBS for glibc-i iirc Sep 28 20:25:45 koen: ok, I know of one more :) Sep 28 20:26:00 * koen tries to remember Sep 28 20:26:12 gnu-config - missing install -d Sep 28 20:26:39 ah Sep 28 20:26:52 I'll push to .dev Sep 28 20:27:00 thanks Sep 28 20:27:23 I was just suddenly worried I was going to wipe out a load of stuff Sep 28 20:27:55 I'm without netconnection tomorrow at work so it might take sometime for me to responds to questions :) Sep 28 20:28:24 np. I'm doing this out of hours anyway and just when I get a few minutes Sep 28 20:28:41 I doubt that an army base will have insecure wifi :) Sep 28 20:28:52 Nothing would surprise me... Sep 28 20:28:56 yeah Sep 28 20:29:09 koen : ah ? btw, dont work, gumstix don't boot, init error... Sep 28 20:29:16 the army dudes keep misplacing usbsticks in rental carms Sep 28 20:29:22 cars* Sep 28 20:30:40 RP: I intended to look at syncing classes/ with .dev this morning but ubuntu corrupted /boot again Sep 28 20:30:55 koen: :-/ Sep 28 20:31:05 koen: I have enough here to make it work with bitbake trunk... Sep 28 20:31:07 a 20MB /boot is too small Sep 28 20:31:35 any diff in classes/ is considered spurious Sep 28 20:31:58 03rpurdie 07org.oe.dev * r91c0213e... 10/ (1 packages/gnu-config/gnu-config-native_20050701.bb): gnu-config: Add missing install -d (from packaged-staging) Sep 28 20:32:08 the changes needed are in .dev as well, as people have noticed with the tasks being changed around Sep 28 20:32:59 koen: I noticed - I also switched the tasks in poky Sep 28 20:33:03 I suspect you can drop packaged-staging.bbclass into poky as well Sep 28 20:33:21 btw, you need ipkg + ipkg-utils on your host Sep 28 20:33:31 I never got around to adding an ipkg-bootstrap .bb Sep 28 20:33:50 ipkg-native isn't enough? Sep 28 20:34:13 yes, but a clean rebuild needs ipkg before quilt is built Sep 28 20:34:53 ah, I see Sep 28 20:35:03 This could be interesting :) Sep 28 20:35:14 you can bitbake -b ipkg ipkg-utils if you want Sep 28 20:35:22 but having them on your host is easier Sep 28 20:35:39 zecke recommended a ipkg-bootstrap package for that Sep 28 20:36:12 Yes, its a good idea Sep 28 20:36:29 or absorb ipkg-build.sh into OE Sep 28 20:36:48 which I think poky did for the .deb changes Sep 28 20:36:49 Is that all we need it for? We should probably just pinch it ;-) Sep 28 20:36:56 right :) Sep 28 20:37:03 I've not looked too closely at that yet Sep 28 20:37:06 ipkg-build, ipkg-make-index and ipkg-cl Sep 28 20:37:26 I was already hacking bits of make-index... Sep 28 20:37:27 I was temped to use the sh version of ipkg Sep 28 20:38:39 anyway, send ideas/suggestions etc by mail or to the mailinglist so I won't forget them :) Sep 28 20:38:42 'night all Sep 28 20:58:57 * v8jlene is back Sep 28 21:04:53 03lenehan 07org.oe.dev * rf8394551... 10/ (6 files in 3 dirs): Sep 28 21:04:53 perl: Update cpan.bbclass to set CCFLAGS to CFLAGS so that the compiler Sep 28 21:04:53 settings for the current OE build are used instead of the made up values in Sep 28 21:04:53 the perl config (which were manually generated since we are cross Sep 28 21:04:53 compiling). Remove this from each of the perl modules that was manually Sep 28 21:04:53 setting this. Sep 28 21:06:53 03lenehan 07org.oe.dev * rb66084df... 10/ (1 packages/perl/liblocale-gettext-perl_1.05.bb): Sep 28 21:06:53 perl: Add the perl Locale::gettext module, which is used for gettext Sep 28 21:06:53 style message handling. Sep 28 22:09:29 * XorA|gone really should stop ejecting the internal microdrive on spitz Sep 28 22:12:03 RP: my hostap crash is prism3 only, prism2 card works fine Sep 28 22:12:35 XorA|gone: Not something I can debug then I'm afraid :-/ Sep 28 22:12:54 RP: its cool, when I get the serial lead made, I shall have the oops Sep 28 22:13:05 :) Sep 28 22:13:12 RP: and now its part of my job Liam might allow me time to fix it Sep 28 22:13:48 I wish it worked like that :) Sep 28 22:13:54 RP: 3 times today I have done pccardctl eject and wondered why device locked Sep 28 22:14:17 XorA|gone: You'll get used to it eventually. Maybe in a few months ;-) Sep 28 22:14:46 * XorA|gone can see a script calling pcccardctl.real Sep 28 22:24:55 * XorA|gone goes to sleep **** ENDING LOGGING AT Fri Sep 29 00:05:08 2006 **** BEGIN LOGGING AT Fri Sep 29 00:12:12 2006 **** ENDING LOGGING AT Fri Sep 29 00:19:30 2006 **** BEGIN LOGGING AT Fri Sep 29 00:24:12 2006 Sep 29 00:57:43 * v8jlene is away Sep 29 02:16:45 hi, mtn failed to update workspace , error message " misuse: path _MTN/tmp/1 already exists" Sep 29 02:17:34 any idea of this issue? **** ENDING LOGGING AT Fri Sep 29 02:59:57 2006