**** BEGIN LOGGING AT Sun Oct 30 02:59:56 2005 Oct 30 03:23:56 zecke: I want to build gpe-image Oct 30 03:24:00 zecke: but it fails Oct 30 03:46:42 lucasvo: well, from the post from yesterday Oct 30 03:46:52 lucasvo: you already had the patch applied, for whatever reason Oct 30 03:49:46 lucasvo: OE/BitBake do not handle the out of disk space case in anyway Oct 30 03:50:05 lucasvo: if applying a patch suceeded but not 'remembering/stamping' that it was done Oct 30 03:50:13 lucasvo: the thing you saw can easily happen Oct 30 03:50:34 (or out of inodes) Oct 30 04:23:35 he koen Oct 30 04:24:09 koen I got ip from the pptp server at the university Oct 30 04:24:24 cool Oct 30 04:24:27 koen but unfornatly I couldnt try if the ip traffic really works Oct 30 04:24:51 it was late and i didnt got a www-browser working Oct 30 04:25:25 hi Oct 30 04:30:05 * france is back (gone 33:20:05) Oct 30 04:30:12 hi france Oct 30 04:30:26 woglinde: hello Oct 30 05:42:30 zecke: I redid the whole thing on a new machine... Oct 30 05:42:43 zecke: it wasn't because of diskspace Oct 30 06:23:40 Is there any release plan for futur familiar release ? I can't found any on familiar.handhelds.org Oct 30 06:24:06 lucasvo: well whatever, let it be permissions, disk quota, reiserfs, out of inodes Oct 30 06:24:24 lucasvo: the stamp that patching was done, wasn't created Oct 30 06:32:03 <[cc]smart> i'd like to make an mpack package, but it can't identify it's license type. it is free to use but it name GPL. what to do ? Oct 30 06:32:09 <[cc]smart> package is here: http://ftp.andrew.cmu.edu/pub/mpack/ Oct 30 06:38:17 zecke: it worked, I had another error now, and I sent a mail to mailing list Oct 30 06:48:30 lucasvo: that mail, might or might not arrive Oct 30 06:55:56 zecke: ping Oct 30 06:59:12 mithro: pong Oct 30 06:59:18 zecke: those are indeed the two possible outcomes Oct 30 06:59:37 mithro: I'm doing my assignments ATM Oct 30 07:07:00 pb_: Potentially, it could only half arrive I guess... Oct 30 07:31:41 zecke: havn't given up and installed the binaries yet? Oct 30 08:25:15 afternoon all Oct 30 08:25:28 RP: did I read right, you got sound working? Oct 30 08:26:22 afternoon. Oct 30 08:27:06 Bernardo: Hi and correct :) Oct 30 08:27:58 RP: pulling now... :) Oct 30 08:29:07 Bernardo: Be warned that the dynamic power management doesn't quite work right so you have to "echo 0x1a 0x1fe > /sys/devices/platform/wm8750/cmd" after you start playback or after you change any of the mixer settings... Oct 30 08:29:17 I need to talk to Liam about that which I will tomorrow Oct 30 08:29:37 that is a strange one... Oct 30 08:29:58 maybe it's time to switch the 2.6 kernel images to acpi instead of apm? Oct 30 08:30:28 Bernardo: You don't quite understand the problem. It'll just be a bug in the dpm code Oct 30 08:30:41 The Zaurii don't have an apm or acpi bios Oct 30 08:30:53 They have an emulated apm layer :-/ Oct 30 08:31:01 Which has nothing to do with this problem... Oct 30 08:31:43 RP: ok Oct 30 08:32:15 too bad for the sd problem I'm seing Oct 30 08:32:30 or I'd be using the 2.6 kernel straight away Oct 30 08:32:52 Bernardo: I have a theory about that - mreimer recently pointed me at a document about different types of delays for SD cards Oct 30 08:32:59 It could be the same issue Oct 30 08:32:59 can i download the 2.6 kernel for the zaurus via http? Oct 30 08:33:13 tanq: for what version of the zaurus? Oct 30 08:33:21 5600 Oct 30 08:33:29 RP: possibly. Thank god now the sd spec is more open Oct 30 08:33:46 tanq: I'd look in the OZ feeds Oct 30 08:33:52 i dont have oe setup but i'm trying to get an image to test Oct 30 08:34:00 its not in the feeds for the 5600 Oct 30 08:34:18 i thought it might be available somewhere on the oe site Oct 30 08:34:26 RP: the sound fixes are in rc4-r4? Oct 30 08:34:38 Bernardo: correct Oct 30 08:34:59 then I already had a correct image built and just trashed it... :( Oct 30 08:35:04 tanq: the source should be on the OZ site Oct 30 08:35:10 Bernardo: I'll try and look into the SD issue soon. As the 2.6.15 merge window has just opened, I need to do other things at the moment... Oct 30 08:35:20 RP: sure Oct 30 08:35:31 RP: get as much as you can into mainline Oct 30 08:35:41 RP: and we'll all be gratefull for ever... :) Oct 30 08:35:45 ~hail RP Oct 30 08:35:46 * ibot bows down to RP and chants, "I'M NOT WORTHY!!" Oct 30 08:35:53 We really need more Zaurus kernel developers as I can only do so much at once :) Oct 30 08:37:03 The pxa irda code has gone into mainline at last, with a bugfix and patches for poodle, corgi and spitz from me following fast on its tail... Oct 30 08:37:20 send me a zaurus :) Oct 30 08:37:56 koen: they only have it for the clamshells i think Oct 30 08:38:14 tanq: all gpl'd source should be on the OZ site Oct 30 08:38:28 RP: if only the day had 24 more hours... :) Oct 30 08:38:36 if it isn't, ask the OZ people Oct 30 08:38:37 koen: I think tanq wants a compiled image... Oct 30 08:38:43 tanq: this isn't the place Oct 30 08:38:47 RP: #openzaurus Oct 30 08:40:36 should and does are two different things. and someone told me I could find an image on oe which i haven't found yet.. Oct 30 08:42:30 that someone was wrong Oct 30 08:42:42 OE has nothing to do with distributing images Oct 30 08:42:55 like I said, #openzaurus Oct 30 08:42:58 OE != OZ Oct 30 08:43:02 RP: 'we' have a SL6000 to give away, if you find some one with enough spare time Oct 30 08:43:25 * tanq yawns Oct 30 08:43:43 s/to give away/on loan for some months/ Oct 30 08:43:55 hi mickeyl Oct 30 08:43:56 hi mickeyl Oct 30 08:44:09 mickeyl: right, to loan to some one trust worthy Oct 30 08:44:28 zecke: I've just been looking at the tosa patches once again and feeling depressed at the bitrot :-( Oct 30 08:44:48 my term started way too early Oct 30 08:44:58 I have a 2.6.14-git kernel running here but none of the tosa patches will apply without work due to the platform device pm changes... Oct 30 08:45:03 * pb_ heads to the office Oct 30 08:45:04 bbiab Oct 30 08:45:09 bummer Oct 30 08:45:21 later pb_ Oct 30 08:45:24 tosa tosa tosa Oct 30 08:45:28 * mickeyl www.tm.uka.de/forschung/SPP1140/ Oct 30 08:45:30 ups, sorry Oct 30 08:45:33 Its easy enough to fix but I don't know whether to do it or wait for dirk to do it... Oct 30 08:45:44 ~seen do13 Oct 30 08:45:52 do13 was last seen on IRC in channel #oe, 34d 2h 29m 31s ago, saying: 'zecke_: To build an opie-image you need this patch http://www.do13.de/oz/temp/defconfig-tosa.patch'. Oct 30 08:46:07 Personally, I think dirk should push the core tosa code that works into mainline rather than waiting for it all to be complete... Oct 30 08:46:54 mickeyl: You saw the news on cxx00 sound? Oct 30 08:51:16 Oh, and I have quite a nice patch from the the fbdev maintainer which will please cxx00 owners. It's going to encourage me to update to 2.6.14-git1 so we can apply it easily Oct 30 08:52:23 RP: yeah, excellent. i have some testing to do with it, i have a lot of sound applications and APIs to fire on it Oct 30 08:53:14 mickeyl: Its not totally user ready yet as you'll read in scrollback. I see these as minor issues though - the core looks good :) Oct 30 08:56:06 cool. i guess i should make myself familiar with alsa sooner than later Oct 30 08:56:30 hrw|gone: wifi seems to have been broken lately, ever since I tried the hentges rom I've had the same kind of problems you're having being unable to connect Oct 30 08:56:46 On 2.6.14-git1, pxa27x ohci suspend/resume doesn't crash anymore :). It unmounts all the usb disks but I can live with that :) Oct 30 08:57:10 Based on that, I'll push my patches for ohci... Oct 30 08:57:18 hrw|gone: seems the culprit is udhcp Oct 30 08:57:27 good to hear. that unmounting is not an issue Oct 30 08:57:51 i guess you didn't decide on a LED abstraction yet right? ;) Oct 30 08:58:12 mickeyl: I know what I want to do, just need to code it Oct 30 08:58:44 mickeyl: We're getting there though. I'll run out of other things to do soon ;-) Oct 30 08:59:03 sure Oct 30 08:59:15 LED is really not important, but i like reminding you about that ;) Oct 30 08:59:27 I haven't forgotten ;-) Oct 30 08:59:43 RP: the 2.6 keymap seems to have switched the calendar and the address book buttons around (in opie) Oct 30 09:01:00 Bernardo: Interesting. I wonder if I've messed that up in the kernel driver :-/ Oct 30 09:01:16 no idea... :) Oct 30 09:01:24 I just notice that now Oct 30 09:01:58 I'll might look at it next time I fiddle with the keyboard driver... Oct 30 09:02:04 and I can't associate with my ap, but that was happening with the oz-3.5.3 based hentges rom I was using too Oct 30 09:02:42 do I need to do anything special to enable sound? Oct 30 09:02:58 Bernardo: Lots of modprobing as sound is all modular Oct 30 09:04:15 Bernardo: http://www.rpsys.net/openzaurus/temp/inssound Oct 30 09:04:28 Something like that but s/wm8731/wm8750/ Oct 30 09:05:00 That list is a bit overkill as well... Oct 30 09:05:31 well, it seems no snd-* modules got built for me Oct 30 09:05:50 time to try and find out what did I do wrong pulling the dev branch... Oct 30 09:06:38 Bernardo: You'll probably have to manually install them as the image doesn't know about them yet Oct 30 09:06:59 Keep in mind this is bleeding egde - the stable kernel knows nothing about these yet Oct 30 09:07:33 yep, they're there in the ipk dir Oct 30 09:07:59 ok, I'll try that later, have to go now. Thanks a lot! Oct 30 09:45:05 hi Oct 30 09:45:25 hi hrw Oct 30 09:46:12 hrw: Check the dev branch commits for some c7x0 sound news... Oct 30 09:46:25 I saw them Oct 30 09:46:36 will build and test Oct 30 09:47:51 I'm curious do they pass my opie test :) Oct 30 09:48:01 hrw: You can control speaker/earphone enable/disable from sysfs now Oct 30 09:48:18 (temporarily until we get them into the alsamixer Oct 30 09:48:34 hrw: I'm equally curious :) Oct 30 09:48:44 RP: no headphones here Oct 30 09:48:53 but I could connect speakers instead ;) Oct 30 09:49:15 I'm happy the controls work - I think it was XorA that wanted them... Oct 30 09:49:25 mickeyl: could you "chmod 775 *" in 3.5.4-RC dir? Oct 30 09:49:28 yes - it was him Oct 30 09:52:18 hrw, you got some time Oct 30 09:52:42 bam_: ask Oct 30 09:53:18 regarding overclocking values Oct 30 09:53:24 batteryplus Oct 30 09:53:43 bam_: iirc we dont have it in OE Oct 30 09:53:58 oh Oct 30 09:54:09 bam_: and probably wont have Oct 30 09:54:28 bam_: which Z you have? Oct 30 09:54:34 3100 Oct 30 09:55:07 bam_: OZ use 2.6 kernel on it (you can help RP with testing btw) Oct 30 09:55:37 rp? Oct 30 09:56:30 bam_: and we rather will have cpufreq support to underclock then for overclock Oct 30 09:57:22 bam_: 3100 should now work but 2.6 hasn't been tested much by anyone... Oct 30 09:58:04 im surprised Oct 30 09:59:14 OE/OZ only has one active kernel developer (me) looking after 6 models of which I have two :-/ Oct 30 09:59:44 We don't have cpufreq support under 2.6 yet for the pxa27x... Oct 30 09:59:57 Someone needs to write it... Oct 30 10:00:02 RP: first we need stable 2.6 on pxa27x Oct 30 10:00:06 ;) Oct 30 10:00:22 hrw: It is perfectly stable! Oct 30 10:00:44 RP: ok Oct 30 10:01:02 building 2.6.14-rc4-r4 now Oct 30 10:01:03 Its not fully featured but it the core kernel is bullet proof Oct 30 10:01:18 sensitive, but then again I too am a programmer, I understand Oct 30 10:01:25 o.. 2.6.14 released. Oct 30 10:02:14 hrw: dev branch is looking likely to move to 2.6.14+git1 soon ;-) Oct 30 10:02:39 * chouimat is away: weekly dishes time :( Oct 30 10:03:23 so there is a testing release of OZ for 3100 Oct 30 10:04:40 re Oct 30 10:04:54 my desktop machine is too unstable last times Oct 30 10:05:22 time for a reinstall perhaps? Oct 30 10:05:36 my ubuntu work very nice Oct 30 10:05:52 bam_: rather for hardware reinstall Oct 30 10:06:07 ouch Oct 30 10:06:21 bam_: PSU and cpu cooling probably Oct 30 10:06:49 I need to build a nice desktop system soon Oct 30 10:07:07 my laptop is fine but.....desktop would be nicer Oct 30 10:07:32 hrw: I "fixed" my firewall psu with a compressed air spray Oct 30 10:07:43 the fan started again and is working perfectly since... Oct 30 10:07:52 Bernardo: that one is clean and have new fan (zalman 80mm one) Oct 30 10:08:18 Bernardo: its just cheap PSU (Codegen 400X) which just worked ok before Oct 30 10:11:57 may I suspend a build with bitbake ? Oct 30 10:12:03 then resume it later ? Oct 30 10:12:12 magnet: you can Oct 30 10:12:18 very nice. Oct 30 10:12:46 bitbake looks like a lot to gentoo's package management tool. Oct 30 10:12:59 so crtl + c is ok ? Oct 30 10:13:02 yes Oct 30 10:13:10 thanks. Oct 30 10:13:13 gentoo doesn't have packages Oct 30 10:13:17 I'll try to use distcc. Oct 30 10:13:22 and hence no package management tools Oct 30 10:13:26 koen: it's ports. Oct 30 10:13:49 but linux people understand better the concept of 'package management'. Oct 30 10:14:18 also note you can make binary packages on gentoo. Oct 30 10:16:23 * chouimat is back. Oct 30 10:18:31 hrw: did you get an url from cobelius? Oct 30 10:19:47 koen: no Oct 30 10:20:10 OZ 3.5.4-rc is nice fetchable via scp Oct 30 10:21:06 now time to be not nice to my c760... Oct 30 10:21:56 in other words: I will make attempt to get wifi working Oct 30 10:49:41 RP: starting 2.6.14-rc4-r4 Oct 30 10:50:31 * pb_ returns! Oct 30 10:51:08 03hrw 07org.oe.dev * re7172473... 10/: opie-appearance 1.2.1: updated CVS backport of icon size settings Oct 30 10:52:24 ~lart my wrt or me or my Z for fscking problems with get wifi connected and working Oct 30 10:52:24 * ibot flings poo at my wrt or me or my Z for fscking problems with get wifi connected and working Oct 30 10:57:19 RP: ping Oct 30 10:58:59 hrw: pong Oct 30 10:59:24 http://pastebin.ca/27077 after some flooding to driver Oct 30 10:59:36 flooding = qss + screen clicks sounds Oct 30 10:59:46 I will reboot and try without qss Oct 30 11:00:50 hrw: I guess we should be thankful it doesn't oops :) Oct 30 11:01:00 ;) Oct 30 11:01:36 I think the response from Liam will be that with oss, we can only have one program using the sound system at once so the behaviour is sort of correct Oct 30 11:01:49 I'm not sure it should get that deep into the sound code before it decides that though Oct 30 11:03:11 I have a lovely bit of code for cxx00 owners thanks to Antonino Daplas :) Oct 30 11:03:23 who is? Oct 30 11:03:32 fbdev maintainer Oct 30 11:03:55 ah. Oct 30 11:04:06 let me guess - fb rotate? Oct 30 11:04:10 Someone else has written code to further the Zaurus effort for a change though which makes me happy :) Oct 30 11:04:18 got it in one :) Oct 30 11:05:04 mickeyl is going to kill me as it breaks the elpp patch... Oct 30 11:05:47 fsck elpp Oct 30 11:05:55 :) Oct 30 11:06:37 if it is not in kernel then ignore it - working kernel is more important Oct 30 11:07:51 I know. I can probably make elpp apply anyway but we'll see... Oct 30 11:08:11 Having the console the right way around on spitz is great :) Oct 30 11:09:29 RP: leave elpp for mickeyl - it is his baby Oct 30 11:09:50 heh.. having wifi working would be great... Oct 30 11:10:14 someone have not needed pci or cf wifi card? :) Oct 30 11:11:37 I have to test my AP with other wifi equipment... my c7x0 is not unable to connect in <1minute (no WEP, no WPA) Oct 30 11:14:38 hrw: MAC filter? Oct 30 11:15:12 zecke: none - all security things turned off. Oct 30 11:15:22 I will look how fast collie will connect Oct 30 11:19:10 collie (2.4.18-r22, hostap 0.3.9) get dhcp lease in <10 seconds from inserting wifi card... Oct 30 11:20:37 hrw: Very strange :-/ Oct 30 11:22:55 anyway after resume it is scked too Oct 30 11:27:20 RP: After this release I'll ditch elpp - instead I'll make a fullscreen "Ângström Booting - Please wait" Logo and then start the boot manager + init progress manager which has the same look as the boot logo in userland. That way it looks even more professional and we have everything but the standard linux logo out of kernel space Oct 30 11:28:09 hi mickeyl Oct 30 11:28:16 mickeyl: Ok, that sounds good to me. I can lose elpp from the dev kernel then? Oct 30 11:28:48 yep, go ahead Oct 30 11:28:49 hi hrw Oct 30 11:28:56 collie flashing.. 3.5.4-rc Oct 30 11:29:22 mickeyl: ssh mml@shell.sf.net;chmod 775 3.5.4-RC - could you? Oct 30 11:29:30 hrw: sure... hold on Oct 30 11:29:45 mickeyl: cardmgr is also totally broken now under recent (> 2.6.14-rc1) kernels. I'll force a move to pcmciautils in the dev branch and we'll just have to pick up the pieces (mainly network device setup) Oct 30 11:30:34 we should also drop c7x0 support from oz-pxa 2.4-crapix Oct 30 11:30:34 -bash-2.05b$ chmod 775 -R 3.5.4-RC Oct 30 11:30:34 chmod: changing permissions of `3.5.4-RC/c7x0,c8x0/MD5SUMS': Operation not permitted Oct 30 11:30:34 chmod: changing permissions of `3.5.4-RC/sl6000/MD5SUMS': Operation not permitted Oct 30 11:30:34 chmod: changing permissions of `3.5.4-RC/all-md5sums.txt': Operation not permitted Oct 30 11:30:43 mickeyl: ok - they are mine Oct 30 11:30:45 RP: yes, that sounds like the way to go Oct 30 11:30:58 mickeyl: will update them Oct 30 11:31:01 thanks Oct 30 11:31:19 mickeyl: It isn't supposed to be broken btw but it is :-( Oct 30 11:32:04 mickeyl: done Oct 30 11:32:37 hold fingers for me - starting 3.5.4-rc on collie Oct 30 11:32:48 * mickeyl crosses fingers Oct 30 11:33:14 the multibuild thing didn't work for me which is why it took so long to build/upload all variants of the RC Oct 30 11:33:43 mickeyl: what machine you use for builds? Oct 30 11:34:05 someone offered cpu power for OE recently Oct 30 11:35:27 mickeyl: users complain about lack of SD in pcmciaapplet.. Oct 30 11:36:07 hm 0.6M free on / - small.. Oct 30 11:37:20 hrw: i use a dual xeon 2.4 Ghz Oct 30 11:37:59 hrw: i have no time (nor motivation) to add Sharp SDcontrol support to pcmciaapplet. We better ship cardmon for the kernel 2.4 devices Oct 30 11:38:16 ok Oct 30 11:39:17 i hope we can finally get this thing out next month. What do you think? Oct 30 11:39:22 ipsec/openswan must go off from collie Oct 30 11:39:26 I hope so Oct 30 11:40:06 the branch will either be frozen then or become yet another point release if there's enough interest. Oct 30 11:40:16 that'll make it easier for people to recreate 3.5.4 Oct 30 11:40:28 bbl, washing dishes Oct 30 11:40:32 frozen for bugfixes (feed/upgrades) Oct 30 11:41:12 I know WHY people get broken font in opie-console Oct 30 11:41:13 ;) Oct 30 11:43:29 mickeyl: can qt/e have font substitutions? Oct 30 11:44:40 hrw: because fixed is missing Oct 30 11:44:53 zecke: exactly Oct 30 11:45:19 and probably there is no way to tell qt/e which font should be used instead.. Oct 30 11:45:44 not dynamically Oct 30 11:45:47 anyway it works on collie out-of-box - no /home/ here Oct 30 11:46:01 zecke: only in fontdir by duplicating line? Oct 30 11:49:00 IIRC yes+ Oct 30 11:50:14 ok. so will leave it as it is Oct 30 11:50:15 hey all! Oct 30 11:50:31 anybody know where I can find a qvfb package? Oct 30 11:50:35 hrw!!!!!!!!!!!!!!!!!!!!! Oct 30 11:50:44 vanille.de/tools Oct 30 11:50:45 hrw: just the man I wished to talk to.. Oct 30 11:51:07 hrw: I'm trying to get kdepim working on gpe. Is this the work you had done but not packaged (bug 371)? Oct 30 11:51:41 orangey: it was just started as I dont use kdepim/pi anymore Oct 30 11:51:49 orangey: feel free to finish it Oct 30 11:51:50 hrw: what do you use? Oct 30 11:52:13 hrw: I would love to finish, but I'm relatively new to OE. Just got my hx4700 a few weeks ago. Oct 30 11:52:31 orangey: hey, what do you want to achieve? Oct 30 11:52:35 orangey: standard opie pim Oct 30 11:52:43 zecke: kdepim/pi x11 Oct 30 11:52:46 hrw: A couple of questions: 1. What is the compilation platform? Usually I compile on the platform that I use, but that seems silly here.. Oct 30 11:53:04 zecke: I want to be able to sync remote calendars using some secure method (e.g., scp) Oct 30 11:53:31 hrw: what platform do you use? Oct 30 11:53:43 x86 for builds Oct 30 11:54:01 hrw: OK, so if you use OPIE, why were you trying to get kdepim working with x11? Oct 30 11:58:09 because he's nice and this has been requested Oct 30 11:58:16 by GPE folks Oct 30 11:58:18 ooooooh. Oct 30 11:58:22 OK. Oct 30 11:58:24 cool. Oct 30 11:58:35 so, I guess I'm one of the people who would love to see this happen : ) Oct 30 11:58:44 but I'm not sure exactly where to start contributing.. Oct 30 11:59:09 orangey: I would first get it compiled natively (e.g. your desktop machine) Oct 30 11:59:22 orangey: and then starting using OE to cross compiler for your ARM device Oct 30 11:59:59 zecke: OK, fair enough if I knew what cross compiling were, but I'll look that up.. Oct 30 12:00:13 zecke: So, hrw has actually "finished" this work. Oct 30 12:00:34 zecke: but I'm not sure how to apply things. Oct 30 12:00:42 he has patches based on 2.2.3 Oct 30 12:01:03 http://bugs.openembedded.org/show_bug.cgi?id=371 Oct 30 12:01:31 orangey: I guess he did not Oct 30 12:04:07 zecke: well, it's not packaged.. Oct 30 12:04:36 OK.. is there any documentation on how to "cross-compile using OE"? Oct 30 12:05:08 orangey: oe.handhelds.org Oct 30 12:05:22 orangey: OE is a build system, it helps to create Distributions easily Oct 30 12:05:53 orangey: e.g bitbake gpe-image would (cross)compile and package everything necessary Oct 30 12:06:20 orangey: I will update to 2.2.7 later Oct 30 12:06:28 hrw: update what? Oct 30 12:06:31 and downgrade oz354fam083 version to 2.2.0 Oct 30 12:06:33 kdepim/pi Oct 30 12:06:36 in OE Oct 30 12:07:12 hrw: OK. Well, Could you tell me what's left to be done? I don't mind finishing the work now, even if it involves a lot of studying to figure out how OE works. Oct 30 12:07:28 orangey: creating pacakges probably Oct 30 12:07:32 hrw: it sounds like all that is left is packing it. Oct 30 12:08:37 sure Oct 30 12:09:12 hrw: This must sound stupid, but could you point me to any documentation on how to create an ipkg pack? Oct 30 12:09:18 but packages have to be named like kopi-x11 kapi-x11 etc.. libs can be named like they are Oct 30 12:09:25 orangey: http://www.openembedded.org/ Oct 30 12:09:57 orangey: kdepimpi/qte libs will fsck kdepimpi/x11 and viceversa probably but we can live with it Oct 30 12:10:20 hrw: Could you pack up and send me the whole directory so I don't have to figure out how to cross compile? Oct 30 12:10:26 and then I'll do the packaging Oct 30 12:11:42 You know what.. it just occurred to me that I don't even remember the last time I used DCC. Oct 30 12:11:57 orangey: I cannot as I dont have one now Oct 30 12:12:22 orangey: and I have 170kbps upload so I prefer to not send anything Oct 30 12:12:38 mickeyl: http://openzaurus.org/wordpress/releases/354/ started Oct 30 12:12:49 hrw: OK.. So, I figure if I'm going to go "from scratch", I may as well try to make 2.2.5 work.. advice? Oct 30 12:13:19 orangey: current 'stable' is 2.2.0, current 'testing' is 2.2.7 Oct 30 12:13:40 it says stable is 2.2.5 Oct 30 12:13:47 orangey: kdepimpi need some patches for x11 (I wrote them already) which need to be sent upstream (which I will do soon) Oct 30 12:13:58 really? nice Oct 30 12:14:22 http://sourceforge.net/project/showfiles.php?group_id=104103&package_id=112604 Oct 30 12:14:33 so I will do 2.2.7 for .dev and 2.2.5 for .oz354fam083 Oct 30 12:14:43 orangey: and pacakging does not need patching sources Oct 30 12:15:03 hrw: but compiling the binaries which will be packaged does, right? Oct 30 12:16:06 orangey: patches are written now Oct 30 12:16:37 I understand.. but you said that "packaging does not need patching sources". Does that mean that it doesn' need new patch work? or that it doesn't need patch work period? Oct 30 12:16:53 orangey: building != packaging Oct 30 12:17:21 hrw: I understand. but packaging can't happen unless building has happened. Oct 30 12:17:33 hrw: and since nobody has a binary, I'm going to have to build, right? Oct 30 12:17:50 orangey: use OE, build it and then add packaging part Oct 30 12:18:14 hrw: It'll take me all day (if I'm lucky) to get that simple command done : ) Oct 30 12:18:18 I'm gonna go try! Oct 30 12:22:14 goodluck Oct 30 12:29:49 NOTE: package kdepimpi-2.2.7-r0: task do_build: completed Oct 30 12:30:16 now pwmpi has to build Oct 30 12:31:30 hi piku Oct 30 12:32:24 woohoo! Oct 30 12:32:28 hrw: How are you doing all that? Oct 30 12:32:42 hrw: Are you making the X 11 packs, or is this the regular stuff? Oct 30 12:33:32 NOTE: package pwmpi-2.2.7-r0: task do_build: completed Oct 30 12:33:36 orangey: qt/e Oct 30 12:34:05 hrw: : ) Oct 30 12:34:19 orangey: how I do it? "bitbake kdepimpi pwmpi" Oct 30 12:35:26 hrw: COol. I'm reading the wiki now Oct 30 12:36:13 03hrw 07org.oe.dev * r189674c0... 10/: kdepimpi: upgraded to 2.2.7 (current testing) Oct 30 12:36:49 mickey|TV: what do you think about one change in pcmciaapplet: set 'activate' as 1st choice. also maybe rename 'activate' to 'do nothing' which is better (or other more descriptive) Oct 30 12:37:19 hrw: awesome : ) Oct 30 12:37:31 orangey: awesome? OE is awesome Oct 30 12:37:51 hrw, activate means do nothing? Oct 30 12:37:59 see. it should be do nothing ;) Oct 30 12:38:34 piku: basically 'activate' means 'do nothing in applet - let cardmgr do its job nevermind what it is' Oct 30 12:39:17 piku: iirc - I dont use that applet Oct 30 12:50:53 hi koen Oct 30 12:51:13 koen: can I give patch for #350 with something which is not in .dev? Oct 30 12:51:34 koen: I want to add kdepimpi 2.2.5 (current stable) into .oz354fam083 - we have 2.2.7 in .dev Oct 30 12:52:01 I think so Oct 30 12:52:46 next one will be harder stuff.. reorganized collie.conf which will not use handhelds-common.conf to gain some / space (no-ipsec) Oct 30 12:53:15 makes sense Oct 30 12:53:20 collie only has 14MB Oct 30 12:53:30 and most stuff was optimzed of 16MB Oct 30 12:53:40 (16MB - kernel to be exact) Oct 30 12:54:21 0.6M free is a bit too low Oct 30 12:54:36 especially with jffs2 Oct 30 12:54:41 anyway users will complain even when we will give them 3M free Oct 30 12:54:57 I always keep 1MB free to be safe Oct 30 12:55:17 I usually move /opt/ to sd card so have ~8M free Oct 30 12:55:39 does jffs2 need a % free space or a fixed number of kB? Oct 30 12:55:49 /opt? Oct 30 12:55:56 koen: opie Oct 30 12:55:57 * koen doesn't have that ;) Oct 30 12:56:13 koen: not only you;) Oct 30 12:57:02 ever since SuSE fscked my system with /opt voodoo I avoid it like the plague Oct 30 12:57:59 I dont have /opt on debian Oct 30 12:58:44 /opt makes it too easy to have bad packagemanagement Oct 30 12:58:59 *cough* RPM *cough* Oct 30 12:59:11 /opt is part of LSB and i belive KDE uses it heavily by default ... Oct 30 12:59:31 emte: if compiled by user then maybe it use /opt/ Oct 30 13:00:08 any non-core packages are "supposed" to go into /opt instead of /usr/local or /usr/share accroding to LSB Oct 30 13:00:13 distro stuff should go in /usr, local stuff in /usr/local and /opt may be used for evil closed source binaries Oct 30 13:01:01 i dislike most of the LSB policies myself Oct 30 13:03:14 i forget how it goes exactly but its soemthing like: /usr - install time /usr/local - post install core /usr/share - post install core with user access /opt - all non-core packages Oct 30 13:17:58 hrw: Is there a "proper" way to define my sourceforge mirror? or should I just hand-edit the file? Oct 30 13:18:17 orangey: oe wiki: UsingMirrors Oct 30 13:19:56 Ciao all Oct 30 13:21:12 hrw: thanks Oct 30 13:23:23 NOTE: The included, default BB base.bbclass does not define a useful default task. \ NOTE: Try running the 'listtasks' task against a .bb to see what tasks are defined. \ Build statistics:\ Attempted builds: 1 Oct 30 13:23:30 hmmm. OK, seems like there's one last obstacle here (that one) Oct 30 13:23:50 that's with "bitbake kdepmpi" or "bitbake kdepimpi-x11" Oct 30 13:25:47 kdepimpi-x11 ? Oct 30 13:26:13 http://bugs.openembedded.org/show_bug.cgi?id=371 Oct 30 13:26:19 kdepimpi for x11 Oct 30 13:26:47 bitbake is a really amazing build system.. I'm actually quite surprised. Oct 30 13:28:49 orangey, it is not yet in oe, isn't it ? Oct 30 13:29:01 Pigi: That's what we're trying to change.. Oct 30 13:29:05 I'm trying to build and package it now. Oct 30 13:29:15 however, I'm obviously an idiot : ) Oct 30 13:29:28 so it may take a while. Oct 30 13:29:54 let me try to vompile it. I do love kdepimpi Oct 30 13:30:02 compile even Oct 30 13:30:11 pigi: please go ahead : ) Oct 30 13:30:47 how do you build this stuff? I tried bitbake kdepimpi -c listtasks Oct 30 13:30:54 and it showed me "do_build", but it wont' take it. Oct 30 13:31:14 it should be as easy as bitbake kdepimpi-x11 I suppose Oct 30 13:31:30 riiiiight : ) Oct 30 13:31:37 that gives the message as above Oct 30 13:32:23 [14:23] NOTE: The included, default BB base.bbclass does not define a useful default task. \ NOTE: Try running the 'listtasks' task against a .bb to see what tasks are defined. \ Build statistics:\ Attempted builds: 1 Oct 30 13:32:28 apologize, I just joined. Oct 30 13:32:51 Pigi: any ideas? Oct 30 13:33:15 just the time that my pc gives the error also to me. Oct 30 13:33:35 that's a wrong BBPATH - it picks up the bitbake base.bbclass, not the OE base.bbclass. Oct 30 13:33:35 pigi : ) Oct 30 13:33:49 mickey|TV: huh? Oct 30 13:33:59 mickey|TV: so where do we put the right "OE base.bbclass"? Oct 30 13:34:07 I put BBPATH=/stuff/bittake/ Oct 30 13:34:10 bitbake, even Oct 30 13:34:13 yeah, that's wrong Oct 30 13:34:19 reread gettingstarted Oct 30 13:34:20 absolutelly wrong :) Oct 30 13:34:23 orangey: BBPATH=builddir:oedir Oct 30 13:34:31 woah. Oct 30 13:34:33 OK. moment. Oct 30 13:34:39 mickey|TV: what do you think about my proposition to pcmciaapplet change? Oct 30 13:34:59 hey mickey|TV Oct 30 13:35:11 "do nothing", while being technically accurate does sound worse than "activate" to me Oct 30 13:35:18 heya mickey Oct 30 13:35:18 but I rethink it Oct 30 13:35:21 hi ggilbert Oct 30 13:35:23 hi koen Oct 30 13:35:23 how goes the phd? Oct 30 13:35:41 ggilbert: ever got my mail about admin rights for the bugtracker? would like to add some categories and restructure stuff Oct 30 13:35:47 phd goes... slowly, but progressing. Oct 30 13:35:48 let me add you Oct 30 13:35:52 thanks Oct 30 13:35:53 There was a small problem Oct 30 13:35:59 The mail server was offline :p Oct 30 13:36:01 mickey|TV: nevermind how it will be named - make it first choice instead of suspend Oct 30 13:36:05 ggilbert: bugzilla is working great! Oct 30 13:36:07 hrw: *nod* Oct 30 13:36:08 and the monitoring script erm... couldnt mail me :p Oct 30 13:36:12 heh Oct 30 13:37:11 I'll add you as an admin Oct 30 13:37:17 Anyone else who should be added? Oct 30 13:37:20 koen, please Oct 30 13:37:24 ok Oct 30 13:37:28 perhaps hrw as well Oct 30 13:37:35 that should be enough for now Oct 30 13:37:39 oe bugzilla? Oct 30 13:37:41 yep Oct 30 13:37:47 mickey|TV: In think the bootstrap part in the branch is good to go Oct 30 13:38:06 I'm familiar with it at reporter level and let it stay like that for now Oct 30 13:38:20 in other words: current status is ok for me Oct 30 13:38:23 koen: ok, leaves some subtle issues in opie and gpe to fix before the release Oct 30 13:38:27 hrw: k Oct 30 13:38:43 koen|770, I've tried g_ether on h39 but it seems that it doesn't work. Oct 30 13:39:26 Pigi: ok, maybe patrick has some hints Oct 30 13:40:11 who's patrick ( irc nick ) ? Oct 30 13:40:40 pohly on irc Oct 30 13:41:02 he responded to your bug Oct 30 13:41:30 and he's the one who backported g_ether Oct 30 13:42:25 I see. I will read on the bug, but it seems that g_ether is not sending events on usb port. Oct 30 13:46:57 pigi: any luck with that compile? Oct 30 13:47:38 not completely. It is complaining now about "moc" not found. Oct 30 13:48:03 maybe there is some missing dependencies on the .bb file. Oct 30 13:48:17 pigi: hrw suspected as much Oct 30 13:48:20 http://sztuka-reklamy.info/tv/others/IWantWalkYouHome2.avi roxx Oct 30 13:48:32 orangey: really? Oct 30 13:48:49 hrw, what's that link ? Oct 30 13:49:10 Pigi: lets name it 'commercial' Oct 30 13:49:10 hrw: When you spoke of not spending time on getting the dependencies right Oct 30 13:49:21 ok. Oct 30 13:49:38 orangey: no - I was talking that it is not packaged Oct 30 13:49:51 hrw: OK. I'm out of ideas then : ) Oct 30 13:49:56 hrw: any thoughts on the 'moc' problem? Oct 30 13:50:38 orangey: build qt3 Oct 30 13:51:32 with bitbake, or on the system? Oct 30 13:51:35 hrw, I suspect that the kdepimpi-x11 build would take a lot of space :) Oct 30 13:51:45 orangey: its #oe so with bitbake Oct 30 13:52:05 Pigi: less then few G ;) Oct 30 13:52:11 time for movie.. Oct 30 13:52:25 hrw: OK... Oct 30 13:52:26 today it will be: The Chronicles of Riddick Oct 30 13:52:44 hrw|afk, heh Oct 30 13:53:23 Pigi: I dont know how much - I had gpe-image built before Oct 30 13:53:56 ok. I will thest then Oct 30 13:56:04 hrw|afk, when you say "build qt3", you probably mean qte-x11 ? Oct 30 13:57:14 I meant qt-free-x11 3.x (or how it is called in OE) Oct 30 13:57:28 qt-x11-free_3.3.5.bb Oct 30 13:57:30 ok. I suspected. Sorry Oct 30 13:57:39 np Oct 30 13:58:01 Pigi: so what does that mean? I do bitbake qt-x11-free , THEN the other one? would it keep them both in the same "environment" so one can use the other? Oct 30 13:58:16 orangey, yes Oct 30 13:58:36 pigi: and where do I grab the qt .bb? Oct 30 13:59:37 orangey: you need BitBake + the OpenEmbedded tree Oct 30 13:59:38 is in org.openembedded.dev, so you should have it on your machine Oct 30 13:59:46 hi zecke Oct 30 13:59:50 orangey: and then packages/qt/* Oct 30 14:00:03 zecke: OK. Oct 30 14:00:11 monotone is getting it now. Oct 30 14:00:16 how many megs is the tree? Oct 30 14:00:26 ~ewi Oct 30 14:00:30 i heard ewi is ewi546.ewi.utwente.nl - main backup server for OE monotone. Hosts database snapshot: http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Oct 30 14:00:36 orangey: you might get the complete tree from ewi Oct 30 14:00:52 orangey: this goes way faster than pulling/cloning the db Oct 30 14:01:00 orangey, 126 megs for bb files, and 190+ for oe.db Oct 30 14:01:33 wow. Oct 30 14:01:35 OK. Oct 30 14:01:42 and then I update by checking out or something? Oct 30 14:02:12 orangey: by reading wiki again and again Oct 30 14:02:18 then you will know ;) Oct 30 14:02:55 hrw: Come on : ) I don't want to become an OE programmer.. I just want to do this one thing. Oct 30 14:03:45 reading GettingStarted scares you? it has all that info Oct 30 14:03:56 and more.... Oct 30 14:04:09 hrw: I am going through it.. Oct 30 14:04:19 hrw: But different people have different modalities by which they learn best. Oct 30 14:04:21 orangey: you cannot start working with OE without reading and understanding GettingStarted Oct 30 14:04:43 anyway I will go back to Riddick. cu in ~2h Oct 30 14:05:07 we shou Oct 30 14:05:26 ld have .debs for bitbake Oct 30 14:07:42 koen|770: ./setup.py dist_bin; alien... Oct 30 14:18:46 * emte should work on my installer more ... Oct 30 14:27:55 Is anyone else missing oe-commits emails btw? Oct 30 14:31:30 * zecke drops the coing to decide if he start a port upgrade Oct 30 14:33:07 RP: a lot of your mails are stuck in the hh.org mailq Oct 30 14:33:25 your->to you Oct 30 14:34:26 run 'mailq' after logging to hh.org Oct 30 14:35:55 koen|770: strange. I can't get to hh.org :-/ Oct 30 14:36:16 -- 10016 Kbytes in 1389 Requests. Oct 30 14:36:21 nice amount of mails Oct 30 14:36:59 I suspect dns problem Oct 30 14:37:01 the new MIT ip is blacklisted by spamcop Oct 30 14:37:14 I wonder why qte-x11 depends on mysql ( which could not be found on oesources.org ) Oct 30 14:37:15 host handhelds.org Oct 30 14:37:15 ;; connection timed out; no servers could be reached Oct 30 14:37:25 and a lot isps use that list... Oct 30 14:37:38 I am my own mailserver so its not that... Oct 30 14:38:13 and dns is still not quite right Oct 30 14:38:24 Pigi: iirc, because qt includes some kind of funky sql bindings. Oct 30 14:38:31 Pigi: MySQL SQL plugin Oct 30 14:38:41 What are the ip numbers for hh.org dns servers? Oct 30 14:38:41 ah ok. Oct 30 14:38:42 Pigi: That is inherited from schurig Oct 30 14:38:47 Hi koen. Oct 30 14:39:14 hey lamikr Oct 30 14:40:01 I got finally up the new kernel patches that uses udev and brings some new drivers. Do you have time to apply those? Oct 30 14:40:04 ~nslookup handhelds.org Oct 30 14:40:08 handhelds.org is 128.31.0.25 Oct 30 14:40:32 ~nslookup ns.handhelds.org Oct 30 14:40:34 ns.handhelds.org is 128.31.0.21 Oct 30 14:41:28 lamikr: maybe next week Oct 30 14:42:12 * koen|770 has two weeks full of exams in front of him Oct 30 14:43:15 Ok... Is 0.8.3 happening soon? I could try to add more locales to gpe's timezone selection list. Oct 30 14:43:54 is there someone that can upload a mysql 4.1.12 on oesources, or provide a link for d/loading ? Oct 30 14:47:07 pigi: any luck with kdepimpi-x11? Oct 30 14:47:29 not yet. Searching a mysql ATM Oct 30 14:47:52 I think the hh.org domain is listing its nameservers as handhelds.org addresses rather than IP numbers? Can anyone confirm that? Oct 30 14:49:48 Depends what you mean. Oct 30 14:49:59 whois only lists the names, but the .org nameservers are handing out the glue ok. Oct 30 14:50:27 "dig @TLD1.ULTRADNS.NET. handhelds.org" gives 140.211.167.116 (snowmane) / 128.31.0.21 (firefoot). Oct 30 14:53:57 I can talk to firefoot but not snowmane. I don't understand why I'm seeing the errors I am. Something is wrong :-/ Oct 30 14:54:09 What are you seeing? Oct 30 14:54:19 I don't think snowmane has been installed yet. The DNS should work fine (albeit a bit slowly) with only firefoot, though. Oct 30 14:54:38 If you're actually getting wrong answers, you must have some stale cached data somewhere. Oct 30 14:55:08 50% of queries timing out isn't good though. Oct 30 14:55:16 pb_: I'm getting no answers at all Oct 30 14:55:21 Noodles: heh, indeed. Oct 30 14:55:40 I can get them if I manually query firefoot... Oct 30 14:56:10 http://www.squish.net/dnscheck/ is lovely for such things btw. Oct 30 14:56:47 03jbowler 07org.oe.dev * re9974d38... 10/: (log message trimmed) Oct 30 14:56:47 nslu2-kernel: updated patch set and defconfig in 2.6.14 Oct 30 14:56:47 - 16-ixp4xx-ioremap is removed (l-a-k wants the warning in there) Oct 30 14:56:47 - 25-nslu2-arch-reset is removed (doesn't seem necessary in 2.6.13+) Oct 30 14:56:47 - 10-ixp4xx-copy-from update to remove compiler warning Oct 30 14:56:47 - 18-ixp4xx-io-h-addr better patch - removes casts Oct 30 14:56:49 - defconfig - update from CVS Oct 30 14:56:58 mm, cute Oct 30 15:03:41 hmmm. so if I want to make a kernel using bitbake, it'll know what modules to pick for my hx4700? Oct 30 15:10:27 ;))) futuristic movie, made in US, and status display in prison is in Russian ;) Oct 30 15:27:27 Why does my local dns always try to lookup snowmane.hh.org and fail? :-/ Oct 30 15:29:59 Has it cached a failure? Oct 30 15:30:25 Does "dig +trace www.handhelds.org" help at all? Oct 30 15:30:41 Oh. Interesting. Oct 30 15:30:46 That fails on snowmane for me too. Oct 30 15:32:25 Noodles: That's basically my problem. http://pastebin.com/411481 Oct 30 15:32:46 Yup. That's what I get on the same command. Oct 30 15:33:37 It would seen that firefoot has no A record for snowmane. Oct 30 15:33:42 s/seen/seem/ Oct 30 15:33:51 So there's a glue record for snowman, but that's all. Oct 30 15:34:06 Who's responsible for the DNS? Oct 30 15:34:55 france: ping Oct 30 15:35:18 * Noodles sleeps. Oct 30 15:35:19 hrw: I have the following problem now: NOTE: package kdepimpi-x11-2.2.3-r0: task do_patch: started \ NOTE: Applying patch 'libkcal.patch' \ ERROR: function do_patchcmd failed Oct 30 15:35:41 Noodles: You're right, that's the problem. The A record is missing... Oct 30 15:39:16 hmm. Oct 30 15:39:22 could somebody help me with the meaning of "file:// Oct 30 15:39:31 "? Does that go from cwd? Oct 30 15:40:04 orangey, look at tmp/work/kdepimpi-x11/temp/log_do something. Oct 30 15:40:13 Pigi: empty Oct 30 15:41:10 oh. I may be missing quilt.. Oct 30 15:41:20 it could be. Oct 30 15:41:34 that was in /temp/run.do... Oct 30 15:41:52 yep. worked. Oct 30 16:21:13 nite all Oct 30 16:32:15 RP: hi Oct 30 16:33:21 hi Oct 30 16:33:28 Noodles: hi Oct 30 16:34:39 RP: snowmane is on a truck heading to OSU, firefoot is the 2ndary dns server. We will have to live with the 2ndary until the primary arrives and is setup. Oct 30 16:35:52 france: Can you add an A entry to point snowmane at firefoot for now or something? You can no doubt see the problems that are occuring :-/ Oct 30 16:37:10 RP: yes, I can do that, but in theory things should just automagicly fail over, also, the bigger problem is that the ip address for snowmane is set at a level higher than me with the domain name provider. Oct 30 16:37:34 RP: I would have to change it with network solutions which takes about 5 days, by which time the server should be up. Oct 30 16:38:21 france: The problem is in that pastebin I showed earlier - the fallover doesn't work when both entries don't have IPs. The IPs don't have to work but they do need to have them by the looks of it :-/ Oct 30 16:38:47 france: I know the network solutions one is wrong but that's not the one causing the problem... Oct 30 16:39:16 RP: after dinner maybe I can do some ip alias magic and setup a fake snowmane at OSU. Oct 30 16:40:16 currently to resolve hh.org NS you need hh.org NS? Oct 30 16:40:21 re Oct 30 16:40:21 france: I suspect even an A entry for snowmane with firegoot's ip address would do the trick for now... Oct 30 16:41:20 I would doubt it, as most folks get the ip from a much higher level. Oct 30 16:42:32 you have to get the ip of snowmane before you can reach snowmane. Oct 30 16:43:05 and that's the problem, it asks firefoot for snowmane's ip address which fails the query Oct 30 16:43:32 If it handed out an address for snowmane's ip address, it would fail to contact it and retry on firefoot Oct 30 16:44:22 or if it hands out firefoot's address, it will connect and get the data without a retry... Oct 30 16:45:14 RP: added an A record for snowmane pointing to OSU, this will get ride of the snowmane not found error Oct 30 16:45:34 RP: I will setup something up at OSU after dinner, (a temp fix, until the servers arrive.) Oct 30 16:45:49 RP: pointing it to firefoot will just cuase issues one level up. Oct 30 16:46:42 RP: TTL is 20 minutes, lets see what happens then. Oct 30 16:47:59 cu - have a good night, day, evening Oct 30 16:49:21 france: Ok, thanks. I'll probably be in bed by then :) Oct 30 16:51:32 RP: I will try to find somebody else with the same issue, as dns works perfectly from my house Oct 30 16:51:56 RP: it appears that only about 1 in 10 people are having issues. Oct 30 16:53:02 france: "mailq" on hh.org should give an idea if its resolved... Oct 30 16:53:37 RP: that is only part of the problem many of those are bounces from spamcop Oct 30 16:53:57 france: From what I can see here, the tld*.ultradns.net servers are looking more healthly now so it might be working... Oct 30 16:54:14 RP: they put that server on their block list as they are having issues not be able to match the forward and reverse dns entries. Oct 30 16:54:30 france: The rpsys.net ones (my server) are DNS failures. If they clear up, you know its working :) Oct 30 16:54:31 RP: I hired a lawyer in CA to go after them on Monday. Oct 30 16:54:52 RP: very nice. Oct 30 16:55:59 Hopefully spamcop might recheck the DNS and drop you from the list again :-/ Oct 30 16:57:20 RP: they did not respond to e-mail :-( Oct 30 17:11:57 has anyone noticed zlib-1.2.3 not packaging the shared libs ? Oct 30 17:12:28 I get -dev and -doc but no zlib package Oct 30 17:14:01 'night all Oct 30 18:15:42 hrw: still around? Oct 30 18:15:44 hrw: still around? Oct 30 18:16:33 in any case, I'm trying to install qt-x11-free with bitbake qt-x11-free Oct 30 18:16:45 however I get: NOTE: package qt-x11-free-3.3.5-r10: task do_configure: failed Oct 30 18:19:23 don't you have others error message ? Oct 30 18:19:41 magnet: Indeed I do.. just a moment. I'm trying something out and I"ll post a more complete error in a moment. Oct 30 18:19:49 magnet: I just thought maybe there was a common error that caused such a thing Oct 30 18:20:20 I'm new to use bitbake too. it's a strange tool :) Oct 30 18:20:35 magn: yeah.. but quite nice. Oct 30 18:20:42 magnet: much like emerge's tools Oct 30 18:21:00 yep Oct 30 18:21:08 I was telling this earlier. Oct 30 18:22:07 heh, probably because it was originally based on portage Oct 30 18:22:23 jacques: that would explain it indeed : ) Oct 30 18:22:47 I didn t knew that, trough bb files had ebuild favor . Oct 30 18:23:10 magnet: I wonder why not every source-based thing is like this Oct 30 18:23:18 this just makes so much sense! Oct 30 18:24:40 sometimes bitbake behavior is weird. Oct 30 18:26:14 here is my error with qt: http://oe.pastebin.com/411628 Oct 30 18:28:03 umm. Oct 30 18:29:28 oh. I guess I don't have qmake-native Oct 30 18:29:34 which should I install? 1.xx or 2.xx? Oct 30 18:29:58 1.xx : ) Oct 30 18:45:44 I think I'm missing something here.. Oct 30 18:46:23 if my qt-x11-free depends on qmake-native , then I can do: "bitbake qmake-native" then "bitbake qt-x11-free", and that'll work? Oct 30 20:15:19 WOW! Oct 30 20:15:27 reading the "getting started" really *IS* a good idea! Oct 30 20:17:03 nah Oct 30 20:17:21 emte: heheh ; ) Oct 30 20:24:25 * hardwire is installing debian on his qube2 Oct 30 20:24:26 heh Oct 30 21:01:27 hmmm. where can I change the "target system" so I don't get: Oct 30 21:01:28 checking target system type... Invalid configuration `arm-INVALID': system `INVALID' not recognized Oct 30 21:02:13 you or your target machine didnt define the TARGET_OS Oct 30 21:03:51 hmm. Oct 30 21:04:17 generally "linux" or "linux-uclibc", depending on your choice of C library Oct 30 21:04:23 what are you trying to build for? what device? Oct 30 21:04:45 TARGET_ARCH = "arm" \ TARGET_OS = "linux" Oct 30 21:04:51 kergoth: it's a pxa270 Oct 30 21:04:54 the hx4700 Oct 30 21:06:56 then clearly the file you're defining those in isnt being parsed, or you have a syntax error in it. Oct 30 21:07:02 confirm the sanity of the .conf, and check your BBPATH Oct 30 21:07:14 kergoth: OK. Oct 30 21:07:19 I'm cleaning out the /tmp and trying again too Oct 30 21:15:32 I have BOOTSTRAP_EXTRA_DEPENDS and BOOTSTRAP_EXTRA_RDEPENDS it builds the DEPENDS but doesn't install the RDEPENDS Oct 30 21:15:55 hmmm. Oct 30 21:16:06 so even tho oe builds all the extra packages in BOOTSTRAP_EXTRA_DEPENDS it only installs the default ones - nothing from BOOTSTRAP_EXTRA_RDEPENDS Oct 30 21:16:18 kergoth: what happens if I delete /STUFF/bitbake/tmp ? does that undo any previously compiled packages or something? Oct 30 21:16:38 it starts you with a clean slate. removing everything from the build. Oct 30 21:16:45 kergoth: doh! Oct 30 21:16:58 is there a way to "clean" a single build so I can restart it? Oct 30 21:17:20 bitbake -c clean package_i_want_to_clean Oct 30 21:17:28 kergoth: thank you Oct 30 21:17:29 or use the interpreter and use the rebuild command, iirc Oct 30 21:17:30 np Oct 30 21:17:41 nicole kidman? Oct 30 21:17:50 no problem Oct 30 21:18:05 no no.. the interpreter = nicole kidman? Oct 30 21:18:11 it was a bad joke. I'm sorry Oct 30 21:18:15 ah, yep, it was Oct 30 21:18:16 ;) Oct 30 21:18:53 hahah : ) Oct 30 23:26:45 error=>make[2]:glibc-2.3.5+cvs20050627-r0/build-i386-linux/iconv/iconvconfig Oct 30 23:26:51 (MACHINE = "x86" | TARGET_ARCH = "i386" | TARGET_OS = "linux" | DISTRO = "openzaurus-3.5.4" | KERNEL_VERSION = "2.6") Oct 30 23:28:01 could experts give me some info.. Oct 30 23:34:11 yeah Oct 30 23:34:33 why would OZ build for x86? Oct 30 23:35:04 I wanna test it in my x86 Oct 30 23:35:24 your going to have to remove quite a few patches i think Oct 30 23:36:41 should work fine. Oct 30 23:36:51 theres a reason we keep distro and machien stuff seperate Oct 30 23:37:01 about something patches?please Oct 30 23:37:22 kergoth, yeah? Oct 30 23:38:30 would have thought OZ might ignore taht directive since it is targeted for specific hardware Oct 31 02:22:35 morning Oct 31 02:54:46 morning all Oct 31 02:55:16 hey RP Oct 31 02:56:10 If I want to jump from one kernel version to another that's supposedly a lesser version (2.6.14-rc4 to 2.6.14-git2), how do I convince ipkg to "upgrade" the modules? Oct 31 02:57:14 If I manually remove and install each one it works. If I try to use the force options, I see "ERROR: Package x is not available from any configured source" Oct 31 02:57:16 RP: remove newer modules from feed, then for each one you have to ipkg install -force-downgrade Oct 31 02:57:33 RP: there is no swishy way Oct 31 02:57:55 I have them on a CF card and am pointing at then with a *.ipk Oct 31 02:58:15 I see the above error with -force-downgrade :-/ Oct 31 02:58:27 RP: oh :-( Oct 31 02:58:46 RP: I managed the that trick using grep, ipkg list_installed last time I tried Oct 31 02:59:31 RP: for your kernel you might need to do what fedora and debian do and just stick a build number at front of kernel before the kernel version Oct 31 02:59:49 RP: as long as your build number always goes forward then upgrade will work **** ENDING LOGGING AT Mon Oct 31 02:59:56 2005