**** BEGIN LOGGING AT Mon Mar 12 02:59:57 2007 Mar 12 04:20:27 03Laibsch 07org.oe.dev * rba291b1d... 10/ (1 packages/aircrack/aircrack-ng_0.7.bb): Mar 12 04:20:27 aircrack: clean up in accordance with OE policy. Mar 12 04:20:27 removed superfluous (R)DEPENDS lines. Mar 12 04:20:32 03Laibsch 07org.oe.dev * rcf150961... 10/ (6 files in 2 dirs): procps: unify Mar 12 04:20:37 03Laibsch 07org.oe.dev * r4d697e87... 10/ (1 packages/rglueap/rglueap_0.1.bb): glueap: some reordering of fields. Minor additions and clean-up. Mar 12 04:20:43 03rwhitby 07org.oe.dev * r3ba01558... 10/ (1 packages/tasks/task-mokogateway.bb): task-mokogateway: Fixed the PR Mar 12 04:21:02 Laibsc1: you are Laibsch? Mar 12 04:21:34 Laibsc1: while you are "cleaning up" my crappy recipes, mind checking the maintainer file for my name and looking at the rest :-) Mar 12 04:21:41 Laibsc1: I appreciate it Mar 12 04:57:09 can I borrow someone with commit to fix the maintainer section for me? Mar 12 04:57:21 just need a reciepe added to my section Mar 12 06:54:18 what path do I need to use when referencing CONFFILES ? Mar 12 06:54:35 aparently using the target installation location doesn't work Mar 12 06:55:27 *** Error: CONTROL/conffiles mentions conffile /etc/openpbx.org/agents.conf which does not exist Mar 12 06:55:37 what path should I be using here? Mar 12 07:02:33 morn :) Mar 12 07:04:37 morning Mar 12 07:53:17 good morning Mar 12 08:21:02 morning all Mar 12 08:21:11 ~ping NAiL Mar 12 08:21:21 pong NAiL Mar 12 08:23:31 03hrw 07org.oe.dev * rd475ef60... 10/ (4 files in 3 dirs): picocom: add UUCP_LOCK_DIR define and one patch to get it built with gcc4 Mar 12 08:23:35 03hrw 07org.oe.dev * r1f780d8e... 10/ (1 packages/picocom/picocom/nolock.patch): picocom: revert part of my changes, add nolock.patch written by Marek VaÅ¡ut - close #1972 Mar 12 08:32:18 morning Mar 12 08:33:17 hrw: thx for the reply re hostap and madwifi Mar 12 08:33:35 I got 0.5.7 to compile today, so I will look in wpa_supplicant and do the rest like that. Mar 12 08:33:47 I'll give you a look before I commit anything. Mar 12 08:33:51 thx for taking care Mar 12 08:34:00 trouble was that hostap and madwifi don't play well. Mar 12 08:34:09 rwhitby: if it works then commit. I had no time for hostap stuff during last months Mar 12 08:34:23 I get the problem (like others) when I try and do WPA: http://madwifi.org/ticket/316 Mar 12 08:35:09 do you have previous experience with hostapd that I can draw upon to see if it's my setup or the code which is at fault? Mar 12 08:35:15 hi hrw Mar 12 08:35:47 brb, more raid problems Mar 12 08:35:52 I'll try again in the next couple of days with just WEP (although if it only works for WEP, then there's no point in using hostapd for my application( Mar 12 08:38:34 rwhitby: I never used it Mar 12 08:41:39 has anyone seen something like that with perl -> http://rafb.net/p/Uku2Iw61.html Mar 12 08:42:23 hrw: is there someone in OE who does use it, and will care? Mar 12 08:44:00 hrw: at least I now know to change hostapd to suit the madwifi-ng include staging structure, not the other way around ... Mar 12 08:44:03 rwhitby : from experience when hostapd/madwifi don't play well 9 out 10 is madwifi problem Mar 12 08:44:30 rwhitby: you? Mar 12 08:44:36 Ifaistos: you've used them both together before? in OE? Mar 12 08:45:13 rwhitby : not in OE but i have a bit experience with wifi installation/routers Mar 12 08:45:58 hrw: guess it will be. I'm setting up the IOMega as a stand-alone AP with ability to do wireless client mode to an upstream AP, or to do dhcp client to upstream router, and to support downstream dhcp and bridging to clients via wifi and BT and ethernet. Oh, with a hard disk built-in too for storage. Mar 12 08:46:18 nice box Mar 12 08:46:23 An ideal box for training or demos - plug it in, everyone connects to it. Mar 12 08:46:47 both eth0, ath0 and usb0 (neo) bridged on it today. Mar 12 08:46:54 s/both/got/ Mar 12 08:47:23 rwhitby : interesting. does it use xscale ? there was one nas from iomega with powerpc Mar 12 08:47:35 xscale Mar 12 08:47:53 It's basically a slug, with double flash and ram, and a minipci slot with atheros card in it. Mar 12 08:48:09 oh, and an IDE hard disk Mar 12 08:48:14 (and I have two of them) Mar 12 08:48:33 I maintain the upstream kernel board support for it. Mar 12 08:48:52 Runs SlugOS, Angstrom and Debian nicely. Mar 12 08:49:01 rwhitby : does it have any usb ports ? Mar 12 08:49:10 yep, two Mar 12 08:50:22 rwhitby : nice. we are working with a usb fxo/fxs device at the moment Mar 12 08:50:28 although they are only usb 1.1 Mar 12 08:50:44 rwhitby : could make a very nice asterisk pbx ;) Mar 12 08:50:54 Ifaistos: I would just use a yealink handset. Mar 12 08:51:19 yeah, it would make a good asterisk box. I'll probably put that in the mokogateway distro too. Mar 12 08:51:39 D-Link DSMG-600 RevA box is basically the same hardware too. Mar 12 08:51:50 rwhitby : well think about been able to plug a normal telephone and plug it also to the pstn at the same time Mar 12 08:52:11 Ifaistos: got a URL for this usb fxo/fxs? Mar 12 08:52:28 rwhitby : nop. we are working on it Mar 12 08:52:42 ah, ok. cool. Mar 12 08:53:05 send me one when you're done, and we'll get it supported in NSLU2 land. That should be a good market for you. Mar 12 08:53:18 rwhitby : we have one that does g729 onboard also Mar 12 08:54:22 rwhitby : but we are stuck to LE/BE hell at the moment Mar 12 08:55:35 well slugos can be built in either ;-) Mar 12 08:56:23 rwhitby : we try to get a channel for that device in asterisk, and * sends the data in whatever format the host is Mar 12 08:56:50 rwhitby : which can be kind of confusing to hardware Mar 12 08:58:40 has anyone seen anything like that before -> /powerpc-linux/bin/ld: final link failed: Nonrepresentable section on output Mar 12 08:58:46 i am trying to build perl Mar 12 08:59:26 i am using multimachine, it builds for efika, but fails for dht-walnut which uses another pcu Mar 12 08:59:43 Hi Mar 12 09:00:55 Can somebody update the http://www.openembedded.org/wiki/IRC page? Mar 12 09:05:08 are there people working with monotone 0.29? Mar 12 09:09:18 hi! what's the bitbake target to build the toolchain only? Mar 12 09:13:51 gena_: gcc-cross Mar 12 09:19:57 thx Mar 12 09:21:44 hrw_: will it also build glibc? if not, how to build glibc? Mar 12 09:21:51 will Mar 12 09:27:57 good morning Mar 12 09:47:38 morning Mar 12 09:49:07 morning XorA et a.l. Mar 12 09:49:10 -. Mar 12 09:58:17 morning all Mar 12 10:01:30 hey RP Mar 12 10:03:06 hi RP Mar 12 10:04:24 hi RP Mar 12 10:04:25 ;D Mar 12 10:20:52 03rwhitby 07org.oe.dev * r9e34f5ae... 10/ (1 packages/tasks/task-mokogateway.bb): task-mokogateway-wifi: added madwifi for PCI devices, and bridge-utils. Mar 12 10:26:30 NAiL: Hi ! Did that patch for gcc 4.1.1 worked for you ? Mar 12 10:29:44 hrw ! Did that patch for gcc 4.1.1 worked for you ? Mar 12 10:30:15 no time to check sorry Mar 12 10:34:14 Ifaistos: same error, new file. Mar 12 10:34:28 does gcc 4.1.1 build ? Mar 12 10:35:38 same error as with 4.1.2. Mar 12 10:35:47 what is your distro ? Mar 12 10:35:55 ubuntu Mar 12 10:36:02 edgy ? Mar 12 10:36:24 yes Mar 12 10:36:47 (x86_64) Mar 12 10:37:03 i am using the same on an amd64 and compiles fine, never tested 4.1.2 though Mar 12 10:37:59 it will sound stupid but did you do bitbake gcc -c clean or applied the patch and continued from where it stopped ? Mar 12 10:38:24 I did "bb -cclean gcc", yes :) Mar 12 10:39:26 strange. i got it to build both for efika (ppc603) and walnut (ppc405) Mar 12 10:40:03 03florian 07org.oe.dev * r7aa7b871... 10/ (4 files in 3 dirs): gpe-conf: Add missing dependency. Add patch to name sound settings script correctly. Mar 12 10:40:45 I've cleaned 4.1.1 manually now, and I'm trying to rebuild with the patch. Let's see how it goes Mar 12 10:40:50 NAiL: The difference is that i am using also the patches for the soft-float mlong-128 support Mar 12 10:41:34 NAiL: -> http://bugs.openembedded.org/show_bug.cgi?id=1918 Mar 12 10:42:58 should they in *any* way affect my build, since I'm building hardfloat? Mar 12 10:43:50 NAiL: not sure. normally not, but i am trying to think what could be different in our builds Mar 12 10:44:44 4.1.1: /home/repvik/slug/foonas/tmp/work/gcc-4.1.1-r11/gcc-4.1.1/gcc/libgcc2.c:1607: error: size of array 'compile_type_assert' is negative Mar 12 10:45:21 NAiL: the thing is that this is general OE problem, simply it does not show up when target and build have the same size long Mar 12 10:45:49 NAiL: haven't seen that one Mar 12 10:47:04 NAiL: do a simple check. go the workdir and in the /gcc subfolder and make a diff from auto-build.h auto-host.h Mar 12 10:47:57 NAiL: auto-build.h should be different especially in the places in defines sizes for variables Mar 12 10:48:54 sizeof long, sizeof void_p, Mar 12 10:49:04 8 vs. 4 Mar 12 10:50:45 thats correct, but i don't know what compile_type_assert checks Mar 12 10:53:20 Ifaistos: this is the same bug for sparc: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20633 Mar 12 10:57:33 NAiL: hmmm will look into it. what i'll do is push all changes i have for gcc 4.1.1/glibc 2.5 for powerpc and after that you can do a clean build and see if its there Mar 12 11:03:12 ok Mar 12 11:19:14 anybody idle ??? Mar 12 11:19:42 Is there any way we can add a embedded system node to a existing cluster? Mar 12 11:24:48 knock knock.. Mar 12 11:27:36 init07: hi Mar 12 11:27:51 init07: what exactly do you want to do? Mar 12 11:28:29 is there anyway way we can connect a embedded system to a already exisisting computer cluster ? Mar 12 11:28:55 sry am a complete newbie to this field.. Mar 12 11:29:20 say a pda to a high performance beowulf cluster Mar 12 11:29:52 init07: As a compute node or control node? Mar 12 11:29:59 init07: sure, you can always use a wireless network, cf ethernet cards or even bluetooth Mar 12 11:30:16 as a control node ... Mar 12 11:30:30 how will the process migration takes place? Mar 12 11:30:37 init07: Just install whatever tools you'd use on a normal control node onto the pda.. hook it up via USB/CF/BT networking Mar 12 11:30:57 process migration? from a control node? Mar 12 11:31:06 Wouldn't that occur between compute nodes? Mar 12 11:31:23 Either way.. process migration won't happen between architectures AFAIK Mar 12 11:31:41 linux to embedded linux Mar 12 11:32:04 vice versa .. any idea? Mar 12 11:32:55 init07: Embedded linux isn't really "special" as opposed to normal linux, apart from the packages installed typically being smaller variants.. i.e. busybox as opposed to gnu coreutils Mar 12 11:33:43 (Ignoring uclinux/linux, and uclibc/glibc differences, if the former is used) Mar 12 11:35:09 so isnt there any way for distributed computing with embedded system Mar 12 11:36:37 init07: It really depends on what you're using for distributed computing. You'd be better off talking with the developers of the cluster software about compatibility between linux on $ARCH_A and linux on $ARCH_B Mar 12 11:38:39 init07: OE doesn't really do anything "special" to linux to make it "embedded". In essence, it's a cross-compiled OS with a few patches, and a different set of base packages that allow the system to be slimmer than a typical desktop/server linux box. Mar 12 11:39:15 * NAbyss dons an asbestos suit Mar 12 11:39:29 k thx for that .. Mar 12 11:39:58 np Mar 12 12:37:58 is it OK to set PREFERRED_PROVIDER_virtual/kernel in local.conf? (for some reason it's not being taken) Mar 12 12:41:45 I think either DISTRO or MACHINE overrides it in conf/*/* Mar 12 12:48:02 gena_, in local.conf, you can do anything ;-) (assuming you know what you do) Mar 12 12:49:39 oxo: DISTRO/MACHINE overrides this? even if priority of local.conf is higher than anything else? Mar 12 12:50:38 I'm not an expert at this, it was just my guess :-) Mar 12 12:50:49 oxo: ok, thx anyway Mar 12 12:56:12 so, is there any obvious reason why PREFERRED_PROVIDER_virtual/kernel in the local.conf (collection with hithest priority) is NOT taken? Mar 12 12:59:07 gena_: machine has own set Mar 12 12:59:23 local.conf, machine conf, distro conf - this is how bitbake read configs Mar 12 13:01:21 hrw: sorry to reask 20 times, need to be sure. Does this mean that stuff from machine/distro.conf will ALWAYS override same-named settings of local conf REGARDLESS of the local collection priority? Mar 12 13:04:10 hrw: > "machine has own set" Does it mean, some settings are only valid in machine.conf? and they get ignored in local.conf? Mar 12 13:06:20 settings can be in any config. but each file can override already set vars Mar 12 13:06:27 bitbake read local.conf (and auto.conf, site.conf), parse them. then from it get MACHINE and parse machine config. then it get DISTRO and parse distro config Mar 12 13:06:44 machine config can override local.conf, distro config can override machine and local Mar 12 13:06:53 where local == local.conf/site.conf/auto.conf Mar 12 13:09:17 got it. THX A LOT!!! Mar 12 13:50:21 hrw, any news on that ROAD's version of konq-emb? Mar 12 13:50:44 no Mar 12 13:52:13 psokolovsky: I do not work on opie related stuff Mar 12 13:52:34 hrw, you told you'd be interested in that Mar 12 13:53:08 psokolovsky: no, one company asked me about it Mar 12 13:55:32 hrw: Would you be willing to try and troubleshoot the nunome issues together? Would be nice if you can provide further hints as to what needs to be fixed. Mar 12 13:56:13 Laibsch: not before next week Mar 12 13:56:24 OK, let me know. Mar 12 13:56:55 I will be in .jp again the week after that. Mar 12 13:58:23 is there a way to disable ccache in a recipe? Mar 12 14:00:42 Laibsch, did you see my RFC regarding glibc's locale gen's plucking? Mar 12 14:08:57 yes Mar 12 14:09:09 I was tempted to comment that it works fine for me. Mar 12 14:09:43 But then again I was not sure what I have gotten from you is still the 100% the same as you are proposing now. Mar 12 14:09:47 Laibsch, please do, or it will ignored ;-) Mar 12 14:09:56 All I can say is that it works fine for me. Mar 12 14:10:03 psokolovsky: OK, I will send a message. Mar 12 14:10:17 Laibsch, nope, if you like it, please find time to try the latest patch, and commentthen ;-) Mar 12 14:10:35 Laibsch, tell also taht you need it ;-) Mar 12 14:14:03 psokolovsky: Made a comment. Mar 12 14:14:48 Laibsch, btw, now is good tiem to apply that patch - glibc's PR was bumped anyway ;-) Mar 12 14:15:51 thanks Mar 12 14:16:14 morning Mar 12 14:39:47 psokolovsky: I will just wait until your RFC is approved. Mar 12 14:40:21 hrw: Do you have access to the setup of the BTS? I suggest to add an entry for openmoko under "Version" Mar 12 14:40:32 Laibsch, well, RFC's need support to be approved ;-) Mar 12 14:40:36 no Mar 12 14:41:06 asked this at openzaurus, but maybe fits better here: anyone else hit by this ${OPIE_CVS_PV} buglet since this change on Saturday: http://cia.navi.cx/stats/project/openembedded/.message/3aed3a? Mar 12 14:41:15 psokolovsky: I supported it. Jan supported it. So far no opposition. IMO it looks good. Mar 12 14:41:29 hrw: Who does? Mar 12 14:42:22 treke? Mar 12 14:42:57 Laibsch: I was wondering what does the glibc-package.bbclass patch you were talking about do? Mar 12 14:47:12 hvontres|poodle, http://bugs.openembedded.org/show_bug.cgi?id=1966 . please give it a try ;-) Mar 12 14:50:05 good morning all Mar 12 14:50:34 hi koen Mar 12 14:50:46 hrw: mickeyl and I should be bugzilla admins as well Mar 12 14:50:56 but my connection is too sucky to do bugzilla admin Mar 12 14:52:52 koen|galinhas: ok Mar 12 15:02:34 koen|away: Thanks, I will contact the three of you. Mar 12 15:20:08 cu Mar 12 15:30:04 Hello Mar 12 15:30:39 Can anyone help me with sharprom compiling. Rather, getting it to work properly? Mar 12 15:40:57 Anyone here? Mar 12 15:43:32 Glombool: perhaps you can post your real question/problem , if somebody knows they will answer(the no meta-question thingy). I will not help because I don't know anything :p Mar 12 15:44:43 Oh, ok... Mar 12 15:45:28 Glombool: I think it's just the way most irc channel work Mar 12 15:45:39 Right Mar 12 15:46:01 * mwester also knows nothing.... Mar 12 15:46:33 Crap, I have to restart so I can boot into linux, so I can compile, so I can get the error again Mar 12 15:46:56 * mwester recommends vmware Mar 12 15:47:39 Lol, yeah. Mar 12 15:47:43 That would be nice. Mar 12 15:48:14 some simpad ppl here? Mar 12 15:48:53 VMware costs money. I don't have that luxury. Mar 12 15:49:03 ;P Mar 12 15:49:19 isnt there anyone that are good at simpad's `? Mar 12 15:49:43 Really , the wmplayer is free , I use qemu Mar 12 15:50:00 qemu is too slow for me Mar 12 15:50:29 even with kqemu Mar 12 15:51:08 some one knows where i can get drivers for my simpad SL4 (windows drivers) ? Mar 12 16:04:50 Glombool: VMware server is now a free download: http://www.vmware.com/download/server/ I use it on several systems. Mar 12 16:05:10 cool Mar 12 16:06:44 some one knows where i can get drivers for my simpad SL4 (windows drivers) ? Mar 12 16:47:07 hi Mar 12 16:48:49 03Sleep.Walker 07org.oe.dev * rdd55aab1... 10/ (1 conf/machine/palmtt3.conf): palmtt3.conf: New machine, Palm Tungsten T3. Mar 12 16:48:53 03Sleep.Walker 07org.oe.dev * rf5008760... 10/ (4 files in 4 dirs): linux-hackndev-2.6 svn: Add support for palmtt3. Mar 12 16:48:58 03pfalcon 07org.oe.dev * rbf5aab80... 10/ (1 packages/linux/linux-hackndev-2.6/palmtt3/defconfig): Mar 12 16:48:58 linux-hackndev-2.6 svn: palmtt3/defconfig: Turn off all input dev interfaces Mar 12 16:48:58 except for generic EVDEV, which all contemporary embedded software uses. Mar 12 16:49:01 03pfalcon 07org.oe.dev * r4ab41d30... 10/ (4 files in 3 dirs): Mar 12 16:49:01 opie-taskbar: Be sure to export full set of OPIE envvars to user environment Mar 12 16:49:01 * from init.d/opie Mar 12 16:49:06 03ifaistos 07org.oe.dev * r96ed498a... 10/ (1 packages/perl/perl-5.8.7/config.sh-powerpc-linux): packages/perl/perl-5.8.7/config.sh-powerpc-linux: Don't use share library for powerpc targets Mar 12 16:49:10 03ifaistos 07org.oe.dev * rc4cc94bf... 10/ (3 files in 3 dirs): Mar 12 16:49:10 packages/gcc/gcc_4.1.1.bb: Make sure that the build machine uses the correct site files and not cached values which Mar 12 16:49:12 contain target's data Mar 12 16:49:14 03ifaistos 07org.oe.dev * r577d4786... 10/ (4 files in 3 dirs): Mar 12 16:49:16 packages/gcc/gcc_4.1.1.bb :The following set of 2 patches adds support for soft-float and mlong-128 to powerpc Mar 12 16:49:18 targets without fpu for gcc 4.1.1/glibc 2.5 Mar 12 16:49:22 03ifaistos 07org.oe.dev * r53b0d427... 10/ (6 files in 3 dirs): Mar 12 16:49:24 packages/glibc/glibc_2.5.bb: The following set of patches adds support for soft-float to powerpc targets Mar 12 16:49:26 without fpu for gcc 4.1.1/glibc 2.5 Mar 12 16:49:28 03ifaistos 07org.oe.dev * rc08ed680... 10/ (1 packages/speex/speex_1.1.12+1.2beta1.bb): packages/speex/speex_1.1.12+1.2beta1.bb: Make sure speex compiles with fixpoint support on walnut(ppc405) Mar 12 16:49:33 03ifaistos 07org.oe.dev * r31b24108... 10/ (1 packages/linux/linux-efika_2.6.20.bb): packages/linux/linux-efika_2.6.20.bb: Add full duplex ethernet to efika Mar 12 16:49:38 03ifaistos 07org.oe.dev * r37c48b9a... 10/ (1 site/powerpc-linux): site/powerpc-linux: Add powerpc definitions Mar 12 16:49:41 03ifaistos 07org.oe.dev * r450f9d22... 10/ (4 files in 3 dirs): conf/distro/oplinux.conf: Initial commit of oplinux and oplinux-uclibc Mar 12 16:49:46 03ifaistos 07org.oe.dev * rd53da6d6... 10/ (10 files in 3 dirs): packages/binutils/binutils_2.17.50.0.12.bb: Add latest binutils Mar 12 16:56:02 * hvontres|poodle is not looking forward to tonights pulll :( Mar 12 17:09:43 damn, /. had something useful today Mar 12 17:09:51 http://video.google.nl/videoplay?docid=-4216011961522818645 Mar 12 18:00:57 * psokolovsky finally hacked "turn off dependency handling" patch for bitbake 2 days ago and has fun since then Mar 12 18:05:13 hi all Mar 12 18:05:54 hi Mar 12 18:06:04 Hi! Mar 12 18:06:42 ~lart ipkg-make-index. lart psokolovsky's broken patches to it. Mar 12 18:06:42 * ibot changes ipkg-make-index. lart psokolovsky's broken patches to it.'s permissions to 0777 and tells the world Mar 12 18:08:05 yep, something like that Mar 12 18:08:38 psokolovsky: it works that good? Mar 12 18:08:55 hvontres|poodle, what exactly? Mar 12 18:09:24 psokolovsky: your ipkg-make-index mods :) Mar 12 18:10:21 hrw|gone: What distro are you using for your Linksys wrt? I just learned that the openwrt people are braindead enough to be proud that "ipkg update && ipkg upgrade" completely fucks up your system. Mar 12 18:10:32 actually, appears it works not that bad ;-I. it's all core i-m-i screwedness Mar 12 18:10:35 * Laibsch wonders about an alternative. Mar 12 18:11:12 Laibsch, we need to add tehir kernels to OE, and rock on from that Mar 12 18:11:51 re Mar 12 18:11:59 Laibsch: openwrt whiterussian rc5 Mar 12 18:12:59 Laibsch: I think that openwrt is probably the best you can get for linksys wrt54. works reliable, OE can build compatible packages.. Mar 12 18:13:17 Laibsch: and models with 8M flash can have lot of functionality Mar 12 18:14:43 I just feel really bad having my internet connectivity depend on such pricks. Mar 12 18:15:01 That gives me a BIG headache Mar 12 18:15:24 The attitude of mbm in #openwrt is unbelievable. Mar 12 18:16:23 Laibsch: keep in mind that if you want to maintain own distro for wrt you will sooner or later work with him.. Mar 12 18:17:11 Laibsch: thats why I decided to just use rc5 (two releases since rc5) as it works for me Mar 12 18:17:42 if a package misses obvious shlib dependency, what to blame? Mar 12 18:18:14 psokolovsky: any binary in package needs it? or just do dl() for it? Mar 12 18:18:27 * hrw goes to coffee Mar 12 18:20:09 hrw, well, there's shlib in package itself, and it depends on more shlibs. Does our shlib walker works on exe's only? Mar 12 18:22:33 * Laibsch shudders at the thought of having to deal with somebody as unstable as mbm Mar 12 18:23:05 not that I ever really considered being sole maintainer for any kind of distro Mar 12 18:24:10 ?? Mar 12 18:25:07 chouimat: read hrw about seven minutes ago Mar 12 18:26:53 Laibsch: the openwrt asshole ... Mar 12 18:28:27 Laibsch: I did the mistake to ask them a question once ... no wonder why some people had forked openwrt :) Mar 12 18:28:53 yes, amazing. Mar 12 18:29:06 Unfortunately no alternative yet. Mar 12 18:29:44 hrw, please remind me, why can be "Couldn't find shared library provider" Mar 12 18:29:45 Laibsch: look at www.freewrt.org Mar 12 18:29:49 Makes me feel real nervous now. I need Internet connection and I was lucky to have another router at hand to replace the still broken Linksys right away. Mar 12 18:30:34 chouimat: I know about it and after the nice greeting in #openwrt it was my first stop. But it looks like they still have not made any release. Mar 12 18:30:37 Or have they? Mar 12 18:30:45 chouimat: Would that make him a forkitg asshole ???? :) Mar 12 18:31:06 Laibsch: openwrt briked my wrtsl54gs ... so I got my old p200mmx ... and I'm also thinking about dd-wrt but it's not as flexible as openwrt or freewrt Mar 12 18:31:09 /s/frkitg/forking/ Mar 12 18:31:35 hvontres|poodle: don't know ... Mar 12 18:31:40 hvontres|poodle: I think you are mixing mbm from openwrt with the freewrt dude? Mar 12 18:31:57 chouimat: how? Mar 12 18:32:05 chouimat: Repairable without jtag? Mar 12 18:32:07 Laibsch: if I believe the stuff I saw on the openwrt forum ... they don't care about the linksys anymore ... Mar 12 18:32:07 I had dd-wrt running at university some time ago, imho they use it for the wireless network there now. Mar 12 18:32:29 chouimat: So what are they about then? Mar 12 18:32:32 It wasnÄt that flexible, but stable and easy to use. Mar 12 18:32:41 Laibsch: I need to put the jtag and the serial on my wrtsl54gs and my slug Mar 12 18:32:45 florian: yup Mar 12 18:33:09 * Laibsch wonders if the dd-wrt will offer enough flexibility for the basic needs he has. Mar 12 18:33:21 Laibsch: it's basicly linsys firmware with some modification and packages support ... Mar 12 18:33:55 Laibsch: this one is also interesting www.thibor.co.uk but no package support Mar 12 18:34:00 chouimat: Do you have a JTAG? How did you get it into that state? Is my Linksys in the same state after "ipkg update && ipkg upgrade"? Mar 12 18:34:15 Laibsch: well its a suoed up wrt basicaly Mar 12 18:34:17 Laibsch: bad openwrt flash Mar 12 18:34:33 plus a web gui Mar 12 18:35:25 Laibsch: I still have wrtoe+ on my to do list but I have to get the jtag working first ;) Mar 12 18:35:44 jtag is easy to make, its a few passive components Mar 12 18:35:45 ifaistos_ppc: What? suoed? Mar 12 18:36:12 chouimat: If you have the skills, open up a bounty for it. Mar 12 18:36:32 * Laibsch talked to some ubuntu guy at FOSDEM to try and create a better platform. Mar 12 18:36:33 ifaistos_ppc: I know but I have to wait for my friend (who have way better soldering skill than me) to come back from Barbados Mar 12 18:37:05 souped up, but my fingers are really big on this keyboard Mar 12 18:37:10 ifaistos_ppc: I simply doesn't want to burn the house :) Mar 12 18:37:23 ifaistos_ppc: I still do not understand Mar 12 18:37:37 beefed up? Mar 12 18:37:58 Laibsch: good luck ... mbm is "the WRT GOD" ;) Mar 12 18:38:05 chouimat: send me an email with your address i'll send you one Mar 12 18:38:28 Laibsch: yes :) Mar 12 18:38:42 ifaistos_ppc: :) Mar 12 18:39:11 ifaistos_ppc: I have the cables ... I just need to put the connectors on the boards Mar 12 18:39:48 ok Mar 12 18:40:25 ifaistos_ppc: thanks anyway :) Mar 12 18:40:30 is there a svn argument for SRC_URI? if so, anyone able to point me to any examples of it's use? Mar 12 18:40:59 dullard: "rgrep SRC_URI packages/|grep -I svn" ? Mar 12 18:41:02 ;-) Mar 12 18:41:11 fair nuff... I'll go away Mar 12 18:42:12 np. i have made a few of them, as people tend to brick wrt's when they switch distro's Mar 12 18:42:44 basically the nvram part gets garbled Mar 12 18:43:16 ifaistos_ppc: In my case is I forgot to put boot_wait(or somtething like this) at 1 Mar 12 18:44:12 Maybe I am thick. But can somebody please confirm that my Linksys is bricked and that it can only be revived with JTAG? Mar 12 18:45:20 and then its the dreaded blinking light Mar 12 18:45:20 anyone going to CEBIT? Mar 12 18:45:40 Somebody should backport kexec to mips 2.4 kernels, then one will be able to hack on latest 2.6, etc and not be afraid of such bricking Mar 12 18:45:54 :) Mar 12 18:46:54 no blinking light here. Mar 12 18:47:00 DMZ light is solid Mar 12 18:48:29 RP, hrw: I have package in recipe's DEPENDS. it doesn't appear in ipk's Depends:. what can be wrong? it's opie-appearance if you care Mar 12 18:49:27 psokolovsky: DEPENDS are only build dependencies. if shlibs doesn't take care of adding it to runtime depends, you'll have to add it to RDEPENDS explicitly. Mar 12 18:49:43 psokolovsky: I think DPENDS only apply at build time and RDEPENDS affect packaging..and ph5 types faster than me :) Mar 12 18:49:51 hvontres|poodle: :) Mar 12 18:50:02 psokolovsky: just like guys told Mar 12 18:50:08 pH5, what you mean by "shlibs doesn't take care of adding it to runtime depends" - we have automagic scanning, no? Mar 12 18:50:15 can you ping it ? Mar 12 18:50:15 /me wishes he had brought its notebook while waiting Mar 12 18:50:31 psokolovsky: if you need xyz to build abc - you add it into DEPENDS. Mar 12 18:50:45 psokolovsky: if you want to use xcv with abc - you add it into RDEPENDS Mar 12 18:50:47 hrw, don't we have means to transfrom DEPENDS to RDEPENDS? Mar 12 18:51:12 psokolovsky: no. what is needed to BUILD goes to DEPENDS. Mar 12 18:51:22 psokolovsky: DEPENDS === Build-depends Mar 12 18:51:22 hrw, heck, why? if I need something to build, apparently, I need it to run, unless it's native Mar 12 18:51:38 hrw, that's perfectly understood Mar 12 18:51:44 psokolovsky: I need jpeg to build qte - is it depends or rdepends? Mar 12 18:52:03 hrw, but doesn't RP did lotsa work to fish out any runtime dependencies out of anything? Mar 12 18:52:21 psokolovsky: and it was done, we cleaned DEPENDS a lot Mar 12 18:52:28 hrw, it should go to ipk depends eventually Mar 12 18:52:35 psokolovsky: I think it was the other way around, RDEPENDS can turn into DEPENDS automagically Mar 12 18:52:39 psokolovsky: gpe-conf need pcmcia-cs. but only on device so RDEPENDS Mar 12 18:53:01 hvontres|poodle: right. DEPENDS are calculated from DEPENDS and RDEPENDS Mar 12 18:53:27 hrw, do you wnat to say taht for every flaming X apps I'm going to put dozen of X libs to RDEPENDs? Mar 12 18:53:33 but DEPENDS should contain *everything* needed to get recipe built. Mar 12 18:53:38 psokolovsky: NO! Mar 12 18:53:49 psokolovsky: you need to put them into DEPENDS Mar 12 18:53:53 hrw, why? and I want that, too! Mar 12 18:54:14 psokolovsky: example: gpe-conf. Mar 12 18:54:16 hrw, that's what I have with opie-appearance! Mar 12 18:54:25 hrw, example - opie-appearance Mar 12 18:55:23 psokolovsky: to build gpe-conf you need: gtk (in DEPENDS), x11 (but this is in gtk recipe so no need to repeat), glib-2.0 (needed by gtk also), libgpewidget (DEPENDS) Mar 12 18:55:54 hrw, could you please look at opie-appearance recipe? soemthing wrong with it? Mar 12 18:56:01 I can - moment Mar 12 18:56:10 need to pull Mar 12 18:57:35 hrw, there was no changes to it Mar 12 18:58:27 DEPENDS = "libqtaux2" Mar 12 18:58:28 this? Mar 12 18:59:37 hrw, yes Mar 12 19:00:03 psokolovsky: try to build opie-appearance from scratch and you will hit into this Mar 12 19:00:41 hrw, my q: why don't I get libqtaux in resulting ipk's Depends: ? Mar 12 19:00:44 . Mar 12 19:01:19 psokolovsky: good question. Mar 12 19:02:17 hrw, you ironic or serious? I would expect it to go there by one or another mean. and yet it is skipped (other opie libs are there) Mar 12 19:02:32 serious Mar 12 19:02:57 I added that dependency because it needed this. I did not checked dependencies of package after that Mar 12 19:04:00 #asterisk Mar 12 19:04:44 hrw, psokolovsky : so maybe make opie-appearance RDEPEND on libqtaux2 ? Mar 12 19:04:48 hvontres|poodle: no Mar 12 19:04:50 psokolovsky: are the shlibs files for libqtaux2 in place? (in staging/arm-.../shlibs) Mar 12 19:05:16 hrw, ok. if you don't have quick tip, what that can be, I'm pacthing bbclasses with logging... Mar 12 19:05:42 pH5, yes, checked that already ;-) they are Mar 12 19:05:57 hvontres|poodle, that shouldn't be necessary Mar 12 19:06:02 psokolovsky: so where are those turned into dependencies? Mar 12 19:06:17 pH5, package.bbclass apparently Mar 12 19:08:14 psokolovsky: of course. read_shlibdeps? Mar 12 19:08:36 bbiab Mar 12 19:08:42 pH5, yep, tracing thru that Mar 12 19:09:06 there's so much magic in bitbake/OE that I never had to care about... :) Mar 12 19:09:37 you are confusing magic with witchcraft Mar 12 19:09:44 common mistake Mar 12 19:11:34 No no no -- there's no witchcraft in software; it's all done with Voodoo. It's chicken entrails on the keyboard, not eyes of newts, after all. :-D Mar 12 19:13:40 heh Mar 12 19:13:45 hello all Mar 12 19:13:55 hey likewise Mar 12 19:13:55 ifaistos_ppc: that smells like an EFIKA there :-) Mar 12 19:14:03 Jin^eLD: Hi, long time no see. Mar 12 19:14:13 Crofton: hi Mar 12 19:14:14 yeah.. well, I've been around, just sort of silent - working :) Mar 12 19:14:50 Jin^eLD: good :-)( Mar 12 19:14:58 likewise: actually is my pocket pc telephone :( Mar 12 19:15:09 ifaistos_ppc: on powerpc? Mar 12 19:15:10 hi Mar 12 19:15:30 back in .nl? Mar 12 19:15:57 ifaistos_ppc: Got my im? Mar 12 19:17:40 Crofton: yes, was very nice though, being in .ca Mar 12 19:17:43 Laibsch: yes replied, but i am on gprs/3g connection and gets cut so maybe you did not get it Mar 12 19:17:57 Crofton: I had a question about s/w radio. Mar 12 19:18:01 heh, you mean ca.us? Mar 12 19:18:05 shoot Mar 12 19:18:11 Crofton: ehm, .ca.us yes :-/ Mar 12 19:18:27 Crofton: hold on, have to find an URL Mar 12 19:18:30 Hopefully none of the Canadians around noticed that little mistake :) Mar 12 19:18:55 btw likewise I am already starting to check airline prices for next years FOSDEM Mar 12 19:19:11 although tickets for that time frame won't go on sale for a few more weeks Mar 12 19:19:36 likewise: it's a qtek 9000, not sure what is running on. probably xscale Mar 12 19:19:55 Crofton: I was a bit disappointed by the lack of space at FOSDEM. They *really* need a new venue... Mar 12 19:20:06 yeah Mar 12 19:20:16 but, they provide really good value for money Mar 12 19:20:26 it was a little less insane on Sunday Mar 12 19:20:40 I'm confused somewhat about adding packages using DISTRO_EXTRA_RDEPENDS. e.g. "libjpeg62" is a ipk package name, but I get "Nothing provides runtime dependency libjpeg62". I'm sure it's just something simple I'm missing. Mar 12 19:21:01 (I did do the task-base -crebuild) Mar 12 19:22:08 Crofton: I (just for fun) bought a very cheap remote controlled plane ($25 or so with tx), and its digital control protocol has been hacked. Mar 12 19:22:14 likewise, I've been to high conferences in better facilities with lower quality talks Mar 12 19:22:19 Laibsch: hmmm this nick is not registered so that's why you donlt see my replies Mar 12 19:22:31 tnb: I think it should be jut "jpeg" Mar 12 19:22:36 cool Mar 12 19:22:50 so you want to make a SW tc for it? Mar 12 19:22:58 er tx Mar 12 19:23:16 likewise: got a URL so we can get one too? Mar 12 19:23:36 Crofton: exactly: http://home.versanet.de/~b-konze/xtwin/xtwin_en.htm Mar 12 19:23:55 Google for X-Twin (European brand) and let me look-up the US brand for it... Mar 12 19:24:15 ifaistos_ppc: I see. Mar 12 19:25:23 Laibsch: do this for a quick fix, provided that nvram is not corrupted Mar 12 19:25:29 hvontres|poodle: where are you from? Mar 12 19:25:37 For the US: http://www.airhogs.com/ (needs flash...) Mar 12 19:26:43 Crofton: it's 27 MHz, quite "standard" for R/C stuff, but I wondered if it makes any sense to use a s/w tx for it? Probably not Mar 12 19:26:47 ah Mar 12 19:26:47 Laibsch: power off, press the reset switch and keep it pressed while powering up again Mar 12 19:26:54 so you know the carrier? Mar 12 19:27:05 well, not really Mar 12 19:27:19 It appears all you need to do is control a CW source Mar 12 19:27:39 which you should be able to do easily enough from a "computer" Mar 12 19:28:41 Laibsch: this will put it in re-flash mode. then use the linksys tools to reflash it Mar 12 19:28:45 You could do it using SDR practices, but it would be pricey for what you need to do :) Mar 12 19:29:33 psokolovsky: opie-appearance link with libqtaux in opie cvs: appearance2.pro Mar 12 19:29:41 Crofton: OK, well that's the answer I expected :-) Mar 12 19:30:09 In other words, I have a board I could use, but I wouldn't suggest you by one just for this :) Mar 12 19:30:31 Crofton: I just (simply) figured if I was to use an embedded computer already, what would SDR add to the h/w requirements? Mar 12 19:30:41 aaaah, finally some better internerd Mar 12 19:30:50 ~usrp Mar 12 19:31:01 methinks usrp is a Universal Software Radio Peripheral or http://comsec.com/wiki?UniversalSoftwareRadioPeripheral Mar 12 19:31:05 koen|galinhas__: hey Mar 12 19:31:11 I think I could outout pulsed 27MHz with that Mar 12 19:31:16 I'd need a PA Mar 12 19:31:27 Crofton: I have been talking rishi from TI about OSK and dsp stuff Mar 12 19:31:33 Crofton: the buildboards that 'listen' to the radio station a car is tuned in use SR i assume Mar 12 19:31:53 I suspect they are hybrids Mar 12 19:32:07 there are software radios and reconfigurable radios Mar 12 19:32:21 and yes, I could clog the channel talking about it :) Mar 12 19:32:31 hey likewise Mar 12 19:32:47 koen|galinhas_, I need free tools for DSP SW dev :) Mar 12 19:32:49 Crofton: he knew your name :) Mar 12 19:32:55 interesting Mar 12 19:33:05 Crofton: they are going to free the dsp tools Mar 12 19:33:11 I don't know many at TI Mar 12 19:33:14 wow Mar 12 19:33:18 likewise: any davinci questions? Mar 12 19:33:35 Crofton: but dspgateway and dspfs are going away "soon" Mar 12 19:33:37 koen|galinhas: need HDTV output on the framebuffer device 1920x1080 Mar 12 19:33:47 Does he know someone called "Carlos Aguayo" Mar 12 19:34:01 koen|galinhas, will they be replaced? Mar 12 19:34:18 they are gpl'ing dsp-bridge right now Mar 12 19:34:22 ah Mar 12 19:34:26 the TI solution Mar 12 19:34:37 right Mar 12 19:34:40 Good to hear they are starting to see the light! Mar 12 19:34:42 koen|galinhas: nice. And the tooling is VisualDSP, or am I confusing the Analog Devices stuff now? Mar 12 19:34:43 the ti dsp bridge is being gpl'd? wow Mar 12 19:34:44 nokia is going to use it as well later on Mar 12 19:34:55 likewise: vdsp is analog iirc Mar 12 19:35:08 koen|galinhas: Ah yes, that came with the Blackfins. Mar 12 19:35:15 kergoth: the arm part should be available already Mar 12 19:35:25 I would really like access to a linux based command line dsp compiler so we could build code in OE Mar 12 19:35:29 for that half Mar 12 19:35:42 Crofton: the cmdline compiler is free as in beer Mar 12 19:35:48 yeah Mar 12 19:35:59 free beer is not too bad :) Mar 12 19:36:27 koen|galinhas: DaVinci boards for OE would be nice :-) Mar 12 19:36:44 Crofton: As long as it's not Natrural Light or Buttwiper :) Mar 12 19:36:45 kergoth: hi Mar 12 19:36:50 ifaistos_ppc: Thanks. I am glad I can quickly revive this beast. I just wish I did not have to do all reconfiguration and customization. Mar 12 19:36:51 heh Mar 12 19:37:01 hey Mar 12 19:37:02 hi all Mar 12 19:37:13 I wonder if I should try out a different distro for my WRT. Mar 12 19:37:15 koen|galinhas: when will you be in the US? First back to .nl? Mar 12 19:37:18 Crofton: ohh, wait you said beer, not paturized american beer food product :) Mar 12 19:37:29 I need to beg some money from TI/Lyrtech to get Linux going on the SFF SDR board Mar 12 19:37:31 Laibsch: there is not too much of them... Mar 12 19:38:25 * likewise now really develops an interest in that beer that is in the fridge. Mar 12 19:39:09 Laibsch: openwrt WR (stable), openwrt kamikaze (devel), freewrt (fork or openwrt WR), X-wrt (also fork or even repack), dd-wrt (license problems recently, small development) Mar 12 19:39:19 koen|galinhas: how's the locale over there? Good atmosphere? Mar 12 19:40:01 likewise: nl, 17 april -> us Mar 12 19:40:13 likewise: we're in a resort, so we won't get kidnapped and robbed Mar 12 19:40:58 hrw: Yes, that was my impression. And it sucks. Mar 12 19:41:11 * Laibsch dreams about an OE-built distro for the wrt. Mar 12 19:41:18 koen|galinhas: hmm, yes, you are worth a million, at least here :-) Mar 12 19:41:37 likewise: I feel like the floortje dessing of the OE project :) Mar 12 19:42:03 koen|galinhas: Haha! Mar 12 19:42:15 Laibsch: dd-wrt has a web interface if you don't want to get your hands dirty Mar 12 19:42:57 ifaistos_ppc: openwrt/freewrt/xwrt have x-wrt interface which roxx Mar 12 19:48:32 http://stz-softwaretechnik.com/~ke/touchscreen/evtouch.html - Evtouch is a Touchscreen-Driver for X Mar 12 19:52:52 Hello Mar 12 19:52:59 hrw: too bad it doesn't use tslib... Mar 12 19:53:02 hi sirfred Mar 12 19:53:37 koen|galinhas: yep Mar 12 19:54:14 hrw: It doesn't seem to have an easy way to calibrate it Mar 12 19:56:40 koen|galinhas: so do mt510 used in progear Mar 12 19:58:16 woglinde: hi Mar 12 19:59:41 woglinde: can you please send me the mq200 related files from mrdata? I can't get them compiled and I think I am missing a changed file. Mar 12 20:04:02 Crofton, likewise: mail rishi@ti.com about your request for omap/davinci stuff Mar 12 20:04:28 good evening all :) Mar 12 20:05:06 http://videsignline.com/showArticle.jhtml?printableArticle=true&articleId=193501506 Mar 12 20:05:28 hey gremlin[it] Mar 12 20:05:47 Crofton: yep, that's the same Rishi Mar 12 20:06:05 happy sun,beach & sea koen|away :) Mar 12 20:06:27 Crofton: he was interested in your SDR work Mar 12 20:06:42 Crofton: maybe you can get some stuff donated to your uni if you ask nicely :) Mar 12 20:07:50 I've just added a bug for the new Ximageon release. Mar 12 20:08:01 Bug #1983 Mar 12 20:10:31 cool Mar 12 20:10:36 I'll drop him an email Mar 12 20:10:55 Let him know we support Angstrom on the OSK via OE :) Mar 12 20:11:25 someone with commit access that could fix my maintainer section real quick please? Mar 12 20:12:13 koen|galinhas, where did you run into him? Mar 12 20:12:30 Zero_Chaos, what needs fixing? Mar 12 20:12:48 hvontres|poodle: that gets rid of the errors at least. Now I'll see if it grabbed the right packages, as well. Thanks. Mar 12 20:12:51 Crofton: Rick Farina, add "rglueap" to my reciepe list if you don't mind Mar 12 20:13:16 Crofton: was a small oversight on last commit. Mar 12 20:13:33 not a huge deal, but I want people to know who to harrass if it doesn't work Mar 12 20:14:38 Done, working on pushing now Mar 12 20:15:47 Crofton: thanks for the help Mar 12 20:15:59 I wish I had looked mine over before pushing :) Mar 12 20:16:35 heh Mar 12 20:23:09 03crofton 07org.oe.dev * r9f410686... 10/ (1 MAINTAINERS): Update entry for Rick Farina in MAINTAINERS. Mar 12 20:23:25 Zero_Chaos, hopefully I did it right, can you check? Mar 12 20:23:29 Crofton: rishi is setting ~1m away from me :) Mar 12 20:23:37 heh Mar 12 20:23:41 he's also speaking at this conference Mar 12 20:23:43 Crofton: pulling now Mar 12 20:23:47 what are you doing in Brazil anyway? Mar 12 20:24:01 speaking about OE at a conference Mar 12 20:24:08 well, let him know I would love to talk to him about SDR on TI stuff Mar 12 20:24:10 cool Mar 12 20:24:24 Crofton: I already said as much ;) Mar 12 20:24:39 Crofton: isn't in alphabetical order, but I'll somehow live :-) Mar 12 20:24:41 Crofton: thanks Mar 12 20:24:52 well, it is there Mar 12 20:25:14 the order can be corrected later, at least people know who to blame :) Mar 12 20:25:17 Crofton: I'm teasing you, it doesn't matter, I just wanted people to know who to blame for the ugly reciepe. Thanks ;-) Mar 12 20:25:34 I suppose I need to updat emy list of stuff to blame for Mar 12 20:25:58 koen|away, tell rishi when I am done with my thesis I will have a lot more time to do cool stuff Mar 12 20:26:04 ok Mar 12 20:26:10 Crofton: only if your work is as sloppy as mine... I was given a link to the style guide after my last submission ;-) Mar 12 20:26:10 like mess with omap family stuff Mar 12 20:26:15 heh Mar 12 20:26:34 on friday i will be at cebit ,,, any oe presence there ? Mar 12 20:26:35 03manuel.teira 07org.oe.dev * rcc36c0fc... 10/ (3 files in 3 dirs): kdrive: update imageon patch, closes #1983 Mar 12 20:26:39 I went and fixed the style guide to drop MAINTAINER from the bb file example :) Mar 12 20:26:44 sirfred: there you go Mar 12 20:26:50 koen|away: Thanks a lot Mar 12 20:27:04 ( I didn't test, so lart me if I broke it) Mar 12 20:27:20 koen|galinhas: I'm going to test Mar 12 20:27:54 (I only pushed it to .dev) Mar 12 20:28:04 koen|away: ok Mar 12 21:11:39 Laibsch: how goes spitz usage? Mar 12 21:13:53 03hrw 07org.oe.dev * r308b1dd4... 10/ (1 MAINTAINERS): MAINTAINERS: edited my entry: removed qtopia apps/fonts Mar 12 21:18:54 http://www.hrw.one.pl/2007/03/12/programming-tools-which-i-use/ Mar 12 21:21:35 hrw: basket is cool ... I use it in kontact ... and my kontact plugin for trac is cool too (not yet released) Mar 12 21:22:05 chouimat: I do not like its integration in kontact Mar 12 21:22:24 version 1.0 is okish ... Mar 12 21:22:30 I like to be able to ctrl-shift-alt-w to get basket on screen Mar 12 21:22:49 without switching to desktop 3 and then to proper kontact component Mar 12 21:23:10 ~change 100brl to usd Mar 12 21:23:12 that doesn't look right Mar 12 21:23:17 ~change 100 brl to usd Mar 12 21:23:25 100.00 Real (BRL) makes 48.0077 United States Dollar (USD) (from http://www.xe.com/) Mar 12 21:23:31 hrw: I like to have my trac info in kontact sumary view in the morning ... need to implement a complete (and standalone apps) and maybe a kdevelop 3.4/4 plugin for it Mar 12 21:26:13 chouimat: that post has been for over month in queue to be finished ;d Mar 12 21:26:41 hehe Mar 12 21:27:18 hrw: spent the weekend integrating Tinyerp with kontact ... Mar 12 21:28:47 hrw: lazyness is a key component of mankind ... Mar 12 21:29:11 yes... Mar 12 21:30:30 this is hysterical Mar 12 21:31:07 one ti guy emailing the omap list about a passworf from MV and another TI guy suggesting he look at the OSK for dummies page to get staarted Mar 12 21:31:08 * mwester is full of key components of mankind, then Mar 12 21:33:49 if I set PV = "6.5.2+git${SRCDATE}" in a bb file which uses a git fetcher, and also SRCDATE = "20070307" and PREFERRED_VERSION_mesa = "6.5.2+git${SRCDATE}" in distro.conf... shouldn't that tell bitbake to fetch the messa source from the git repo from that specific date? Mar 12 21:34:53 I ask because every time | try to bake the new mesa recipie, it seems to try and update the already fetched git archive Mar 12 21:34:53 rather than just ising the one I already have Mar 12 21:34:53 using* Mar 12 21:35:24 tkp: DEFAULT_PREFERENCE probably.. Mar 12 21:37:32 Crofton: sigh Mar 12 21:37:52 tkp: git doesn't support dates Mar 12 21:38:11 Crofton: I wonder if all this TI/MV 'synergy' prevents any good OSS support from happening Mar 12 21:38:48 hrw: picocom look interesting Mar 12 21:38:54 * likewise really should get Compiz/Beryl going before someone installs Vista in my neighbourhood. Mar 12 21:39:26 heh Mar 12 21:39:39 It sounds like koen|away is hering some positive things Mar 12 21:39:59 I suspect the OSK issue is that the SW that MV provides in the box is outdated Mar 12 21:40:23 good afternoon all Mar 12 21:40:25 I hope most people know to find stuff on the net. i've heard of people using the images I made with OE Mar 12 21:43:18 Ifaistos: ping? Mar 12 21:44:26 I have this mostly-working recipe http://www.pastebin.ca/392487 however, it doesn't seem to want to AddTask. No error is reported durring te package build but when I install it on a target platform, it becomes evident that the do_moreinstall task is not being run, or do I have to overload postinst in that it's the only special case task that gets run when the packaged is unpacked? Mar 12 21:45:01 hillct: s/do_moreinstall/do_install_append and remove addtask Mar 12 21:45:19 hillct: and there is no such task as do_postinst Mar 12 21:45:47 ok. thanks Mar 12 21:46:22 what's the name of the postinstall task? I thought that's what update-rc.d used Mar 12 21:46:55 anyway, I'll go with your method. thanks Mar 12 21:47:34 hillct: pkg_postinst iirc Mar 12 21:47:45 hillct: and this is not a task Mar 12 21:48:59 I see. Mar 12 21:49:15 thanks Mar 12 21:51:16 damn, I finally taught ipkg-make-index to check ipk timestamps Mar 12 21:53:57 hrw, I bet the could only get 36 phones to work :) Mar 12 21:54:06 hrw: is there a corresponding pkg_postinst_append ? Mar 12 21:55:15 Crofton: no. 14 developers decided to skip p0 phone Mar 12 21:55:25 Ifaistos: NOTE: package gcc-4.1.1: completed Mar 12 21:55:34 well, it is more fun to make up stuff :) Mar 12 21:55:35 hillct: iirc yes - look at classes/update-rc.d.bbclass Mar 12 21:55:40 Ifaistos: gcc 4.1.2 still fails though. But that doesn't matter much ;-) Mar 12 21:56:18 hrw: checked it over last night. will look again. I was late... Mar 12 21:56:27 er it was late Mar 12 21:57:21 !oebug 1951 Mar 12 21:57:22 * * Bug 1951, Status: REOPENED, Created: 2007-03-07 15:45 Mar 12 21:57:23 * * nail(AT)nslu2-linux.org: gcc fails to build on x86_64 Mar 12 21:57:24 * * http://bugs.openembedded.org/show_bug.cgi?id=1951 Mar 12 21:58:54 I'm trying to write an initscript to update few packages on a machine... Mar 12 21:58:59 The machine currently has xserver-kdrive installed, and I want to switch to xserver-xorg in the initscript. Mar 12 21:59:24 hrw: nope. there doesn't seem to be a pkg_postinst_prepend or append. I guess I can just call updatercd_postinst from my own pkg_postinst Mar 12 21:59:45 would I need to specifically remove xserver-kdrive and install xserver-xorg? Or can I use virtual/xserver somehow? Mar 12 22:00:24 hillct: do it in proper way Mar 12 22:00:40 hillct: inherit update-rc.d etc.. Mar 12 22:00:52 yeah Mar 12 22:00:56 I'm doing that Mar 12 22:01:17 but I have to set directory perms imediately before updatercd_postinst is run Mar 12 22:01:53 and I have to create a user account on the target box before setting perms Mar 12 22:02:38 so I must do it from postinst, so the only way I can do that and take advantage of update-rc.d.class is to call updatercd_postinst from within my pkg_postinst Mar 12 22:02:47 unless I'm missing something Mar 12 22:03:34 hrw, fyi, that crap I had appears to be another ipkg-utils "features". I'm sick of that, and going to patch up them till the end finally ;-) Mar 12 22:04:02 hrw: I'm still new to this, so if I'm wrong about that, let me know... Mar 12 22:04:05 03nail 07org.oe.dev * r328dfe9d... 10/ (1 conf/distro/foonas.conf): foonas: update preferred gcc to 4.1.1 since 4.1.2 fails to build Mar 12 22:04:34 hillct: iirc update-rc.d append to pkg_postinst? Mar 12 22:04:49 that doesn't seem to be possible Mar 12 22:04:54 hillct: and update-rc.d stuff is better to be done during image creating instead of on device Mar 12 22:05:04 according to my reading of update-rc.d.class Mar 12 22:05:35 hillct: look into generated package? Mar 12 22:06:26 how can you start a service durring image creation? that would never work. that's why update-rc.d does it durring pkg_postinst. the problem is there seems to be no pkg_postinst_prepend available Mar 12 22:06:38 are you saying there is? Mar 12 22:06:45 update-rc.d do symlinks only Mar 12 22:06:59 not so Mar 12 22:07:17 well Mar 12 22:07:51 hrw, also, of possible, q regarding opie UI. on old opie release (1.2.0?) I so multikey's layout indicator in "tray" area. what does that? Mar 12 22:07:53 unless starting someting from the init.d structure is being done using some other automated process I haven't identified Mar 12 22:08:18 psokolovsky: allow to change keymapping without using multikey prefs Mar 12 22:09:07 hrw, umm, is that just setting in multikey or some standalone applet? Mar 12 22:09:32 <_law_> is there any way to cleanup the sources tree? remove old versions.... Mar 12 22:09:35 applet iirc Mar 12 22:09:43 <_law_> s/old/unused Mar 12 22:09:48 _law_: DL_DIR? Mar 12 22:09:54 <_law_> hrw: yes Mar 12 22:10:03 _law_: I do that by hand Mar 12 22:10:29 <_law_> hrw: ok, a script would be nice :-) Mar 12 22:10:55 hrw, ok, so I assume that's opie-multikeyapplet then ;-) Mar 12 22:11:26 yes Mar 12 22:11:37 _law_, find | xargs rm ? Mar 12 22:12:04 <_law_> psokolovsky: but what are my criterias ? :-) Mar 12 22:12:46 _law_, yep, that's the most hard question, that's why OE doesn't cleanup that dir behind your back ;-) Mar 12 22:13:04 <_law_> :-) Mar 12 22:16:24 how can I get ipkg to switch a package to a different virtual provider? Mar 12 22:17:54 ie. I have diet-x11 installed on a machine, and want to update it to use libx11 Mar 12 22:18:36 diet-x11 build libx11 package Mar 12 22:18:46 libx11-full also build libx11 Mar 12 22:19:04 yes... so if I want to switch on my machine? Mar 12 22:19:06 and we keep both recipes PR synced Mar 12 22:19:14 ipkg install it via hand Mar 12 22:19:15 surely the PR of all providers needs to be kept ion sync no? Mar 12 22:19:53 need Mar 12 22:21:32 _law_: Note that recent fetchers touch the md5 file timestamp when a file is accessed Mar 12 22:22:07 ok, so I just need to bump the PR of both, rebuild, and ensure the libx11 package in my feed is the one from libx11 and not the one from diet-x11 Mar 12 22:22:07 <_law_> RP: ok thanks Mar 12 22:22:48 is there an easy way to fond all providers of a bitbake package? Mar 12 22:23:00 RP: remember our talk that ipkg-make-index should track mtime of ipk's? I added separate file for that, to not overload Packages. sounds good? Mar 12 22:23:11 (so I can ensure I update all PR's) Mar 12 22:31:20 03pfalcon 07org.oe.dev * r9e8066c3... 10/ (4 files in 2 dirs): opie-multikey: RRECOMMEND opie-multikeyapplet, as it's barely real-world usable without it. Mar 12 22:31:40 psokolovsky: for me it is nice usable without that applet Mar 12 22:32:01 hrw, poor you ;-) Mar 12 22:32:49 I'd rather kill of handwriting from default image Mar 12 22:40:15 Night Mar 12 22:45:52 hrw, regarding bug #1931, is there some flag I should set to prevent /proc/net/wireless beign checked? Mar 12 22:46:51 Crofton: check interfaces does it has eth0 or wlan0 with wifi settings as auto Mar 12 22:46:57 <_law_> n8 all Mar 12 22:47:08 where do I do this check? Mar 12 22:47:24 the image comes from DISTRO = angstrom and machine omap59120sk Mar 12 22:47:47 I am trying to move from openomap to Angstrom as my base distro Mar 12 22:50:48 Crofton: hm. does this happen on your omap too? Mar 12 22:50:57 this is on the omap Mar 12 22:51:09 MACHINE = omap5912osk Mar 12 23:26:16 how do I convince a configure script not to produce test executables? Mar 12 23:27:49 dullard: that depends Mar 12 23:28:01 hmmm Mar 12 23:28:29 if its an AC_TRY_RUN/AC_RUN_IFELSE thats only being done to sanity check the environment, you can just not execute it when crosscompiling. if the script is making decisions based on it, you're best off having it cache the result of the test, and supplying that test result in the oe site file Mar 12 23:30:00 kergoth, thanks... that just tells me I'm out of my depth.. time do some reading up I guess Mar 12 23:39:08 kergoth, te configure script insists on trying to find the size of int itself Mar 12 23:49:09 ..which is already given to it in an oe site file Mar 12 23:53:24 hmmm... AC_REQUIRE?? Mar 13 01:22:15 dullard: did you properly inherit autotools in the .bb you're dealing witht there? **** ENDING LOGGING AT Tue Mar 13 02:59:56 2007