**** BEGIN LOGGING AT Fri Jul 06 02:59:57 2007 Jul 06 03:11:00 * * OE Bug 2546 has been RESOLVED (FIXED) by lenehan(AT)twibble.org Jul 06 03:11:02 * *  perl-native does not build on newer distros Jul 06 03:11:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2546 Jul 06 03:31:15 03lenehan 07org.oe.dev * rf73312a7... 10/ (4 files in 3 dirs): Jul 06 03:31:15 perl/perl-native 5.8.8: Fix up asm/page.h include issue. Perl was including Jul 06 03:31:15 it in order to get the definition for getpagesize which has been definied in Jul 06 03:31:15 unistd.h, which it already includes, since glibc 2.1. Some versions of linux Jul 06 03:31:15 libc headers removed the asm/page.h resulting in failures here for some Jul 06 03:31:15 people. Closes #2546. Jul 06 04:15:00 * * OE Bug 2608 has been created by sip.nslu(AT)gmail.com Jul 06 04:15:02 * * asterisk 1.4 bb files Jul 06 04:15:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2608 Jul 06 04:33:48 03lenehan 07org.oe.dev * r0daf2d0b... 10/ (1 packages/perl/perl_5.8.8.bb): Jul 06 04:33:48 perl 5.8.8: A few host paths were being included in the perl for the target. Jul 06 04:33:48 Update the sed magic to catch the files that were being missed. Closes #2575. Jul 06 04:34:00 * * OE Bug 2575 has been RESOLVED (FIXED) by lenehan(AT)twibble.org Jul 06 04:34:02 * *  final perl package has a lot of files poiting erroneously to workdir insted of real bindir , libdir and incdir Jul 06 04:34:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2575 Jul 06 07:31:52 genext2fs: couldn't allocate a block (no free space) reported when the last step for building angstrom-x11-image Jul 06 07:48:00 * * OE Bug 2535 has been REOPENED by Jul 06 07:48:02 * *  PalmLD defconfig for 2.6.21 Jul 06 07:48:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2535 Jul 06 08:06:35 hi all Jul 06 08:07:16 koen|edinburgh : hi Jul 06 08:07:43 hey steliosk Jul 06 08:08:11 bleh, my server at home seems to have dropped of the internet again Jul 06 08:13:00 * * OE Bug 2608 has been marked as DUPLICATE of bug 2540 by Jul 06 08:13:02 * *  asterisk 1.4 bb files Jul 06 08:13:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2608 Jul 06 08:22:54 steliosk: could you mail http://www.openembedded.org/~koen/OE-Changelog-20070630-20070706.txt to the oe-devel mailinglist? Jul 06 08:23:06 * koen|edinburgh can't mail because his mailserver is offline Jul 06 08:25:11 genext2fs: couldn't allocate a block (no free space) reported when the last step for building angstrom-x11-image Jul 06 08:31:44 koen|edinburgh : ok. will do it now Jul 06 08:31:51 thanks Jul 06 08:35:37 morning Jul 06 08:38:46 hey hrw Jul 06 08:39:00 damn, they are taking a long time to get the food done Jul 06 08:43:53 koen|edinburgh: feed-browser in OE has some updates Jul 06 08:51:15 hrw: I noticed :) Jul 06 08:55:03 hard to not notice ;D Jul 06 08:55:58 can it blacklist archs? Jul 06 08:56:14 no Jul 06 08:56:16 I made the list because I only wanted 'real' archs and not the machine archs Jul 06 08:56:47 koen|edinburgh: I am thinking about tables for arch and sections to make queries easier Jul 06 08:57:00 so it will be able to limit amount or archs Jul 06 08:57:43 * wick7xx is away: afk Jul 06 08:58:25 someone remember status of mono in OE? Jul 06 09:00:41 hrw: some stuff works, other stuff SEGV Jul 06 09:00:46 cbrake is looking into it Jul 06 09:02:17 could i adjust my ext2 image manually in host machine? Jul 06 09:03:06 ok Jul 06 09:05:24 koen|edinburgh : mail send but nothing received yet Jul 06 09:05:41 gmane showed it Jul 06 09:05:52 * koen|edinburgh heads off to holyrood park Jul 06 09:05:55 later all Jul 06 09:10:38 hrw could i adjust my ext2 image manually in host machine? Jul 06 09:10:38 the image angstrom build for me is too large, i want to cut some packages Jul 06 09:10:38 or use minimal image and install some packages in. Jul 06 09:13:37 archer: which image? console-image or x11-image? Jul 06 09:15:41 x11-image Jul 06 09:15:41 it's too large Jul 06 09:16:13 archer: create own image recipe based on angstrom-x11-image and remove some packages from it. then rebuild Jul 06 09:18:00 could i use ipkg in host machine? Jul 06 09:19:35 archer: proper OE way is to create recipe not hack resulting image Jul 06 09:19:36 i want just use ipkg to install packages to images, that would be a easy way Jul 06 09:20:08 archer: OE do that Jul 06 09:20:21 ok, i will try Jul 06 09:20:36 archer: you really should do own recipe for image - it will save your time with new images Jul 06 09:22:26 i have a failed experience on create image recipe... Jul 06 09:24:39 ~seen mickeyl Jul 06 09:24:42 mickeyl was last seen on IRC in channel #oe, 1d 14h 18m 20s ago, saying: 'wb'. Jul 06 09:24:52 is there a tutorials on that? Jul 06 09:26:27 it is really easy Jul 06 09:27:51 archer: all you need is: RDEPENDS = "what you want to install", export PACKAGE_INSTALL = "${RDEPENDS}" and "inherit image" Jul 06 10:24:01 moin polyonymous Jul 06 10:24:09 heya Jul 06 10:24:21 no sleep for you? Jul 06 10:25:18 just woke up Jul 06 10:25:27 hehe Jul 06 10:25:28 okay Jul 06 10:25:42 you don't call that "no sleep", do you? :) Jul 06 10:26:06 But well, I think I didn't sleep until 4am at least. Jul 06 10:36:22 Hi! Jul 06 10:36:33 polyonymous: any news on opie building? Jul 06 10:36:35 hi psokolovsky Jul 06 10:37:05 psokolovsky, I'm going to build it today :) Jul 06 10:37:13 cool ;-) Jul 06 10:37:19 * woglinde too Jul 06 10:37:30 after so much changing yesterday Jul 06 10:37:40 I've been fixing x stuff yesterday. I'm leaving in a couple of hours and will start build before I leave. Jul 06 10:38:23 psokolovsky, does it have /usr by default or do I adjust it in local.conf ? Jul 06 10:39:37 polyonymous: by default of course. and more refactoring goes on, towards glorious aim of having 0 machine-dependent packages in opie ;-). I'd appreciate if you could keep eye on that ;-) Jul 06 10:40:41 psokolovsky, "of course" :) I'm not sure if it's the right thing to do, because I also have opie "in production". But well, I didn't feel the need to upgrade, so I think I'll refrain from upgrading it. Jul 06 10:41:14 psokolovsky, well, I'll familiarize myself with recent changes today. Jul 06 10:41:37 polyonymous: yep, you really want to start with reviewing my changes first ;-) Jul 06 10:42:17 psokolovsky, no, first I'll just see if everything works and then I'll be fixing stuff if it doesn't :)) Jul 06 10:42:46 opie is not a kind of code one may want to look into unless he has to :) Jul 06 10:43:27 polyonymous: I mean, make sure you saw my commits, if you haven't yet ;-) Jul 06 10:43:37 and the look is mostly in .bb anyway ;-) Jul 06 10:43:52 psokolovsky, I sure have, of course. Jul 06 10:45:08 ok then Jul 06 10:57:56 03pfalcon 07org.oe.dev * rbb59a70d... 10/ (5 files in 4 dirs): linux-handhelds-2.6: Add defconfigs for htcapache & htcsable, defconfigman r322. Jul 06 10:58:01 03pfalcon 07org.oe.dev * rf3bfa89a... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom-2007.1.conf" Bump PREFERRED_VERSION_linux-handhelds-2.6 to "2.6.21-hh10". Jul 06 11:11:22 morning Jul 06 11:13:25 gm Jul 06 11:13:33 mickeyl, you get the email I sent you about u-boot? Jul 06 11:23:20 Crofton: yeah, thanks. that should fix half of the issues. I'm trying that out asap (weekend hopefully) Jul 06 11:23:32 the other half is the linker symbols which are unaffected Jul 06 11:23:36 (by said patch) Jul 06 11:25:49 ok Jul 06 11:25:55 I was wondering about that Jul 06 11:26:32 I'll keep watching the u-bbot list for helpful info Jul 06 11:26:34 probably 80% of the traffix is ppc related Jul 06 11:26:35 this part is where I'm totally in the dark. I have zero knowledge about linker foo Jul 06 11:26:49 Crofton: thanks, I'm sure eventually they'll get to that problem as well Jul 06 11:26:55 s/m sure/hope/ Jul 06 11:26:57 ;) Jul 06 11:27:27 once we fixed these issues in a clean way we can revamp the u-boot directory in OE Jul 06 11:27:39 this mess keeps crying "clean me up, please" Jul 06 11:27:42 I saw your note about srting through the u-boot files Jul 06 11:27:45 yeah Jul 06 11:28:03 All the u-boot_*.bb files are used somewhere Jul 06 11:28:15 we need to see about getting people to try later versions Jul 06 11:28:21 yeah Jul 06 11:28:28 i'm afraid a lot of patches aren't submitted upstream either Jul 06 11:28:28 and work out the u-boot tool situation Jul 06 11:28:33 yeah Jul 06 11:29:09 well, if I am succesfull getting u-boot going on the board I have, hopefully I can figure out the process to get stuff moved upstream Jul 06 11:29:43 mickeyl: can you enable slapin's mtn key? he sent it to you and koen ~month ago Jul 06 11:29:47 crofton stands all on the uboot wiki side Jul 06 11:30:07 stylegide too Jul 06 11:30:38 hrw: yes, i can do that once i can dig it out again... Jul 06 11:30:46 Crofton: great! Jul 06 11:30:47 ok Jul 06 11:36:49 re Jul 06 11:36:56 hi jineld Jul 06 11:37:04 hey Jul 06 11:37:41 hm who is busybox maintaine`r? Jul 06 11:41:40 i'm afraid no one really is commited to that. koen and hrw update it once and then Jul 06 11:41:50 hm Jul 06 11:42:37 got some love care and tenderness for it? Jul 06 11:42:40 go ahead ;) Jul 06 11:42:51 beware of global breakage though Jul 06 11:42:57 bb is kind-of critical Jul 06 11:43:03 busybox suxx Jul 06 11:43:16 hrw hm but there are no alternative Jul 06 11:43:17 1.6.0 is broken in other way then before etc Jul 06 11:43:50 woglinde, there's heck of a lot of alternatives - bash, util-linux, findutils, gzip, tar, etc. :) Jul 06 11:43:52 or you have to compile a lot of stuff more Jul 06 11:44:02 exactly :) Jul 06 11:44:19 hm space isnt a prolem anymore Jul 06 11:44:21 than Jul 06 11:44:28 woglinde, depends on the machine. Jul 06 11:44:35 thats the point Jul 06 11:44:42 But I'd definitely want most of the "real" stuff on my borzoi Jul 06 11:45:01 and ssh instead of dropbear too :) Jul 06 11:45:12 polyonymous: I've done that with my borzoi :) Jul 06 11:45:24 NAbyss, so have I, but not in the initial image. Jul 06 11:45:43 polyonymous: If you're using angstrom.. in local.conf, use ANGSTROM_EXTRA_PACKAGES Jul 06 11:46:14 NAbyss, would that work for openssh? I mean, dropbear would still be installed? Jul 06 11:46:47 mickeyl: key was June 9th Jul 06 11:46:48 why you need a full ssh? Jul 06 11:46:53 Yeah, you have to manually remove dropbear iirc Jul 06 11:47:02 I haven't yet found the option that'll override the distro sshd setting Jul 06 11:47:15 there is a distro-specific conf setting that'll change that iirc Jul 06 11:47:24 woglinde, because of missing features, but I don't remember wich ones. Jul 06 11:47:34 hm? Jul 06 11:47:52 which feature should that be for embedded device Jul 06 11:48:10 hm is there now virtual/ssh? Jul 06 11:48:11 woglinde: Some of us use it not as an embedded device, but as a mini laptop.. Jul 06 11:48:20 now=no Jul 06 11:48:26 woglinde, I don't remember. But it's not an "embedded device" in this sense. Jul 06 11:48:33 as NAbyss already said :) Jul 06 11:48:41 yeah I know Jul 06 11:51:07 mickeyl, hi! Jul 06 11:52:04 copying cd over wifi takes time.. Jul 06 11:52:21 I've recently copied 8G over wifi. Jul 06 11:52:27 but overnight :) Jul 06 11:52:31 polyonymous: ;D Jul 06 12:01:33 hrw: slapin has access now Jul 06 12:02:14 * polyonymous hails slapin Jul 06 12:02:27 (first I typed "nails", what did I mean now?) Jul 06 12:04:37 mickeyl: thx Jul 06 12:05:15 hahaha Jul 06 12:05:19 polyonymous: freudian slip? Jul 06 12:05:21 i hope not Jul 06 12:05:42 of course if you want to... go ahead. Jul 06 12:05:43 mickeyl, I sure didn't mean to nail him. But you never know with psychoanalysts :) Jul 06 12:05:43 :D Jul 06 12:05:48 hahaha Jul 06 12:05:49 right Jul 06 12:06:06 hi slapin. welcome on board Jul 06 12:06:13 slapin: what are you going to work on? Jul 06 12:06:19 They always think of your subconscious desire to do all the nasty things even if you don't actually make any slips :) Jul 06 12:06:40 "sometimes a cigarette is nothing but a cigarette" Jul 06 12:06:58 * polyonymous guesses slapin will fix locale probs in x Jul 06 12:07:22 by the amount of cigs I smoke I can confirm that cigarette is almost always just a cigarette :) Jul 06 12:07:33 good! missing locales in X are (one of) the reason(s) that X core fonts are not working right now Jul 06 12:07:36 heh Jul 06 12:08:36 I also look forward for fixes in this realm and I still haven't got into it. I'm pretty sure slapin is more familiar with the field. Jul 06 12:10:05 hi mr:nice Jul 06 12:14:19 hrw: hello! Jul 06 12:16:06 hi Christopher Jul 06 12:16:24 hrw: So, I got the sarge stuff and built a completely new project directory. Building with gcc 4.1.2 (although sarge was 4.1.1) and it fails along the way asking for libiconv.so . so, I export the LD_LIBRARY_PATH to it and it goes along till it fails on gconvert in glib saying that iconv.h is from libiconv but gnu libiconv is not in use. This is a fresh build. Any idea what I could be doing incorrectly? Just rebuild without the L Jul 06 12:16:24 D_LIBRARY_PATH and take away the -liconv I saw in the build params? Jul 06 12:16:28 bye all Jul 06 12:17:04 Christopher: which exactly you was building and which distro used? Jul 06 12:17:05 hi woglinde Jul 06 12:17:16 mr_nice till later Jul 06 12:17:25 I got initrd working Jul 06 12:17:26 I used sarge distro (same happened with angstrom), and I was building glib-2.0 Jul 06 12:17:38 cool :) Jul 06 12:17:38 but have to twek the initrd anf busybox a little bit Jul 06 12:17:51 Christopher: do not use sarge distro written by McGregor but angstrom Jul 06 12:18:01 wb koen Jul 06 12:18:22 mr_nice could you mail me account for the wike.opensimpad side so I can document it there Jul 06 12:18:28 or should I say "hey koen" here :) Jul 06 12:18:36 hi all Jul 06 12:18:38 :) Jul 06 12:18:40 he koen Jul 06 12:18:55 koen its foggy and raininig? Jul 06 12:18:55 I'm sure slapin will fix X, E, and and lotsa other stuff too ;-) Jul 06 12:19:03 woglinde: just create one yourself. it's free for anyone :) Jul 06 12:19:08 woglinde: no fog, only rain :) Jul 06 12:19:14 mr_nice okay Jul 06 12:19:15 but at least hope he'll start with cleaning up palm machines support Jul 06 12:19:19 so now hurry Jul 06 12:19:21 hrw: so, when I get an error where it can't find libiconv.so.2, do I remove the -liconv from the build params? it should be using glibc's iconv, right? Jul 06 12:21:28 Christopher: no, you broke your build somehow Jul 06 12:22:12 hrw: ok. I don't know what or how I could have done that. I changed one thing and that was gcc 4.1.1 to 4.1.2 Jul 06 12:22:22 hrw: after I asked what everyone was using. Jul 06 12:22:27 hrw: should i use 3.4.5 ? Jul 06 12:23:23 mickeyl, psokolovsky: I will do a slow start, mostly from small fixes in palm land, then I'm interested in locales. Jul 06 12:23:35 slapin: cool Jul 06 12:24:06 Christopher: remove tmp; backup your local.conf. set DISTRO="angstrom-2007.1" MACHINE="sarge-at91" in local.conf and do build Jul 06 12:24:29 pardon my ignorance, slapin, can I make any use of tungsten with oe? Jul 06 12:25:07 polyonymous: tt1? Jul 06 12:25:24 polyonymous, depends on model Jul 06 12:25:31 hrw: will do and I'll let you know where it breaks. thx Jul 06 12:25:34 slapin, tungsten t Jul 06 12:26:18 polyonymous, ask Marex, he should know Jul 06 12:26:34 ok. I really doubt I can, but wondering anyway. Jul 06 12:26:36 * Marex knows nothing O:-) Jul 06 12:26:50 Marex, but you should :) Jul 06 12:26:56 * Marex hides :b Jul 06 12:27:03 just a sec ... gotta read wassup Jul 06 12:27:26 hmm ... I managed to get familiar running on T1, yeah Jul 06 12:27:54 but it was _Slow_ as hell ... not talking about the mmc driver ... that in one big piece of crap Jul 06 12:28:05 Marex, you also pardon my ignorance, does it take some hardware intervention or how do you put it into the device at all? Jul 06 12:28:09 even though I added some things to speed it up and fix some basics, it is still crappy a lot Jul 06 12:28:17 yeah, okay, I think I shouldn't bother :) Jul 06 12:28:34 polyonymous, what ignorance are you talking about :O Jul 06 12:28:41 doesnt matter ... no, it is simple Jul 06 12:28:49 Marex, MY ignorance! :) Jul 06 12:28:55 I probably missed it ;] Jul 06 12:29:50 as for "slow", that's why we drag OPIE around, no? Jul 06 12:29:54 polyonymous, just get some _MMC_ (this is important ... single look at the driver explains everything - it doesnt support SD), download bootpack, copy it all to the card, backup your data you have in palmos (ram will be wiped), tap Garux, tap BOOT Jul 06 12:30:23 psokolovsky, opie is slow too :T the mmc driver is really bloody ... and since you have TT2, you can fix it :b Jul 06 12:30:27 aha, thanks. maybe one day Jul 06 12:31:20 polyonymous, feel free to contact me about it ;] Jul 06 12:31:55 sure. my wife uses it as an ebook reader, once we have a backup solution... :) Jul 06 12:33:04 :) it is great for that purpose, yeah Jul 06 12:33:50 no, having opie-reader would be better :) Jul 06 12:34:16 if not REALLY slow :) Jul 06 12:35:35 CIA-3: speak up! Jul 06 12:47:03 03mickeyl 07org.oe.dev * rfc1bdba1... 10/ (14 files in 4 dirs): openmoko-base: merge w/ openmoko svn overlay Jul 06 12:47:16 thank you, CIA-3 :) Jul 06 12:47:52 here we are Jul 06 12:48:02 but there should be more IIRC Jul 06 12:48:10 03mickeyl 07org.oe.dev * r84dbea41... 10/ (8 files in 3 dirs): add openmoko-terminal Jul 06 12:48:17 03mickeyl 07org.oe.dev * r348fa7b8... 10/ (2 files in 2 dirs): openmoko-panel-mainmenu: add more dependencies Jul 06 12:48:24 03mickeyl 07org.oe.dev * rbd0a54b8... 10/ (23 files in 5 dirs): openmoko-pim: sync with openmoko svn overlay Jul 06 12:48:30 03mickeyl 07org.oe.dev * r25d05449... 10/ (1 packages/openmoko-apps/openmoko-calculator_svn.bb): add openmoko-calculator Jul 06 12:48:37 03mickeyl 07org.oe.dev * r87a802e7... 10/ (1 packages/tasks/task-openmoko.bb): task-openmoko: merge with OpenMoko svn overlay Jul 06 12:48:43 03mickeyl 07org.oe.dev * r23109292... 10/ (3 files in 3 dirs): remove openmoko-mainmenu, it has been integrated into openmoko-panel-mainmenu Jul 06 12:51:52 mickeyl: did you see that evas(-native) is broken when building from scratch? Jul 06 12:52:13 seen your remark Jul 06 12:52:15 surprised me Jul 06 12:52:21 let me try to reproduce that Jul 06 12:52:23 what did you build? Jul 06 12:52:27 evas Jul 06 12:52:30 ehm Jul 06 12:52:32 ok Jul 06 12:52:35 ecore-(native) Jul 06 12:52:45 evas-native isn't staging some headers Jul 06 12:52:51 ok, anyways, let me build edje-viewer Jul 06 12:52:55 that needs a lot Jul 06 12:53:54 evas had a custom do_stage that installed the x11(-16) engine headers manually before your rework Jul 06 12:54:02 I'm affraid that manual bit is still needed Jul 06 12:54:16 ok, i'll look into that Jul 06 12:54:42 the build breaks on not finding Evas_Engine_Software_16_X11.h Jul 06 12:56:03 well, the x11-16 shoudln't be included anyway Jul 06 12:56:18 iirc the snap we're building don't has it yet Jul 06 12:56:22 doesn't even Jul 06 12:56:34 038 or 039 has it Jul 06 12:56:43 we were able to stage that header before the rework :) Jul 06 12:56:44 ok, anyways. I'm builindg it now Jul 06 12:57:01 thanks Jul 06 12:58:52 mickeyl: shall we discuss at fosdem which features from poky to put in OE before branching openmoko? Jul 06 12:59:47 fosdem next year? ;) Jul 06 12:59:50 or rather guadec Jul 06 12:59:57 aargh Jul 06 12:59:59 hehe Jul 06 12:59:59 guadec :) Jul 06 13:00:07 anyways, yes, please put this on the agenda Jul 06 13:00:10 (you have one?) Jul 06 13:00:10 :) Jul 06 13:00:27 I have none atm Jul 06 13:00:39 my calendaring server at home went down yesterday Jul 06 13:00:45 ouch Jul 06 13:00:48 murphy tends to visit when I'm away from home Jul 06 13:01:16 * koen|edinburgh heads out again Jul 06 13:01:19 later all Jul 06 13:09:30 heh, indeed opie-image build fails. Jul 06 13:34:58 hi all Jul 06 13:39:03 how do i fix those messages from make config at a kernel ? : trying to assign nonexistent symbol Jul 06 13:42:10 mrdata hi Jul 06 13:53:58 03polyonymous 07org.oe.dev * rd4a4e1ae... 10/ (3 files in 2 dirs): opie-icon-reload: fix packaging. I suppose it will need more once I try the image. Jul 06 13:54:22 rob_w|mis: ignore them? Jul 06 14:20:35 mickeyl: pull Jul 06 14:26:14 hi Jul 06 14:39:34 ~seen mallum Jul 06 14:39:37 mallum was last seen on IRC in channel #oe, 64d 23h 21m 58s ago, saying: 'hey koen'. Jul 06 14:49:51 zecke: catch him via jabber or mail Jul 06 14:57:47 * wick7xx is back (gone 1193041:15:33) Jul 06 15:07:32 <_law_> psokolovsky: any ideas how we can manage the 2 different models oft the aximx50 ? Jul 06 15:08:02 <_law_> 2 defconfig man versions ? and to machine.conf in OE Jul 06 15:10:52 _law_: what are reasons for 2 models? Jul 06 15:11:16 <_law_> psokolovsky: vga and nonvga version Jul 06 15:12:13 _law_: ok. let's start with kernel side first. what are reasons for having 2 separate kernels? Jul 06 15:12:23 hi Jul 06 15:14:29 <_law_> psokolovsky: X50_VGA shouldnt be set Jul 06 15:15:11 _law_: so, rewrite it to do runtime detection instead, and voila - the same kernel serves both Jul 06 15:16:12 <_law_> psokolovsky: ok i am not kernel hacker but i ?ll take a look :-) Jul 06 15:17:00 * * OE Bug 2609 has been created by hma(AT)syd.odn.ne.jp Jul 06 15:17:02 * * xserver-common should have /etc/X11/shepherd.xmodmap Jul 06 15:17:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2609 Jul 06 15:17:40 _law_: please do. so, you'd need to find a way to detect vga vs qvga. If you won't find a way, then well, just register separate machine ID for another resolution version, and dispatch on that. Jul 06 15:17:42 <_law_> psokolovsky: but in oe there should be a own machine.conf or? Jul 06 15:18:29 _law_: as for OE, then yep, we discussed that, you'd need to have 2 machines, but config of one should just include other's, and override screen resolution. Jul 06 15:18:45 <_law_> psokolovsky: right Jul 06 15:19:12 <_law_> psokolovsky: but i should the kernel stuff before.... Jul 06 15:19:36 _law_: well, those are independent things, feel free to start in any order Jul 06 15:23:02 re Jul 06 15:29:03 woglinde: hi Jul 06 15:29:47 woglinde: obexftp give me also today an error while building opie-image Jul 06 15:30:59 mrdata can you pastebin he error? Jul 06 15:31:02 woglinde: undefined reference to libiconv for libobexftp.so Jul 06 15:31:09 hm Jul 06 15:31:19 that shouldnt happen Jul 06 15:31:25 woglinde: i will make it Jul 06 15:31:35 because libiconv is present Jul 06 15:31:49 zecke fixed uclibc compilig a while ago Jul 06 15:32:21 woglinde: yes, libiconv is available Jul 06 15:37:33 woglinde: obexftp build error -> http://www.pastebin.ca/605951 Jul 06 15:46:37 mrdata hm Jul 06 15:48:25 woglinde: config.log in obexftp say also this error Jul 06 15:48:38 mrdata which image you are building Jul 06 15:48:47 woglinde: opie-image Jul 06 15:48:55 in 16mb isnt openobex Jul 06 15:49:12 but I rememeber darkly it has this error Jul 06 15:49:47 woglinde: opie-image-16mb will also build obexftp for task-opie-irda Jul 06 15:51:31 hm Jul 06 15:51:51 I will try here Jul 06 15:51:58 hm right Jul 06 15:52:04 saw here it Jul 06 15:52:06 but it builds Jul 06 16:02:20 koen: Ping Jul 06 16:06:41 woglinde: did you buy the new tocotronic cd yet? Jul 06 16:09:25 oh mrdata my fault Jul 06 16:09:37 I have a patch for openobex Jul 06 16:09:43 I should push it Jul 06 16:11:13 args Jul 06 16:11:19 no only obexftp Jul 06 16:11:36 damnit Jul 06 16:11:42 woglinde: we need to change odevice_simpad.cpp a lot. Do you know a clever way for testing it? Building opie with oe and manipulate the influenced .bb file? crosscompile? Jul 06 16:11:43 ups Jul 06 16:12:05 mr_nice: a) check the README in the opie treee on how Jul 06 16:12:16 a) opie tree on how to use OE staging dir Jul 06 16:12:21 hm yes and obexftp is what fails Jul 06 16:12:23 okay Jul 06 16:12:24 b) just copy libopiecore over to the device Jul 06 16:12:39 args I put them always together openobex obexfto Jul 06 16:12:43 obexftp Jul 06 16:12:51 zecke|math: thx Jul 06 16:20:37 * RP returns Jul 06 16:20:53 hi RP Jul 06 16:20:59 hi rp Jul 06 16:21:10 woglinde: i waiting for push ... ;-) Jul 06 16:21:13 RP: wb Jul 06 16:21:14 yeah Jul 06 16:21:16 its done Jul 06 16:21:18 hi florian, woglinde, sirfred Jul 06 16:21:19 * florian is happy - gpe-scap works again Jul 06 16:21:21 in a moment Jul 06 16:21:36 i added libiconv in the dependencies Jul 06 16:21:53 woglinde: no stress Jul 06 16:23:03 pushed Jul 06 16:23:28 than i will make a pull ;-) Jul 06 16:23:51 and thx for that quick help Jul 06 16:25:02 yes soory Jul 06 16:25:11 I should have pushed it yesterday Jul 06 16:25:34 mrdata ntp will fail too Jul 06 16:25:46 but I have no real good fix for it now Jul 06 16:27:10 woglinde: yes i know that Jul 06 16:27:33 woglinde: yesterday i change one line in a file Jul 06 16:28:10 woglinde: define of struct txc Jul 06 16:28:21 hm yes Jul 06 16:28:30 but thats not the only one Jul 06 16:28:51 could you provide me the txc patch? Jul 06 16:29:14 when i make it again today, no problem i think Jul 06 16:29:21 thanks Jul 06 16:29:26 than i can push it Jul 06 16:29:40 hm will now push the simpad kernel stuff Jul 06 16:29:53 new patches and a defconfig Jul 06 16:29:57 hurray Jul 06 16:30:50 woglinde: wait a minute. I will put mrdatas ucb1x00 patch Jul 06 16:31:07 hm the latest he send per email? Jul 06 16:31:12 I already included it Jul 06 16:31:18 mr_nice: woglinde have my latest file Jul 06 16:31:38 woglinde: yes, this one from my e-mail Jul 06 16:31:51 yepp I included it yesterday Jul 06 16:32:06 and thats the kernel I use now with initramfs Jul 06 16:32:19 woglinde: ah,ok please also add the battery patch from testing dir Jul 06 16:32:30 I already have Jul 06 16:32:36 usbgadet too Jul 06 16:32:37 woglinde: big thx Jul 06 16:32:39 but not activated Jul 06 16:37:41 03florian 07org.oe.dev * r77ccc19a... 10/ (1 packages/gpe-scap/gpe-scap_1.3.bb): gpe-scap: Add fixed version 1.3 Jul 06 16:37:46 03woglinde2 07org.oe.dev * r66573a06... 10/ (3 files in 3 dirs): (log message trimmed) Jul 06 16:37:46 obexftp needed same autotools-love because it depends on libiconv Jul 06 16:37:46 when compiling with uclibc. Jul 06 16:37:48 iconv.patch Jul 06 16:37:50 fix the Makefile.am so it uses libiconv in linker stage Jul 06 16:37:52 obexftp_0.20.bb Jul 06 16:37:56 bump PR Jul 06 16:48:04 mr_nice, mr_data its out Jul 06 16:48:10 now you can test Jul 06 16:48:21 if I did something wrong Jul 06 16:48:42 Hmm, I suppose I'm missing something Jul 06 16:48:50 sirfred? Jul 06 16:48:58 If I ask for 'bitbake -s | grep gstreamer' I got: Jul 06 16:49:03 gstreamer 0:0.10.13-r0 0:0.10.12-r1 Jul 06 16:49:24 But if I run 'bitbake gstreamer' 0.10.13 doesn't get built. Jul 06 16:50:02 What could be the cause of that? Jul 06 16:50:02 why the hack libiconv now bails out Jul 06 16:50:19 seems I should recompile all Jul 06 16:51:35 woglinde: thx, I am myself have no working oe machine atm :(. but mrdata will test it I hope :). Jul 06 16:51:37 I'm going to force with -b path/to/gstreamer_0.10.13.bb but... Jul 06 16:51:46 * mr_nice repartitioning my swap Jul 06 16:51:52 sirfred try bitbake yourimage -g nad bitbake yourimage -e Jul 06 16:51:59 and look at the outputs Jul 06 16:52:41 woglinde: Thanks. let's see Jul 06 16:53:59 woglinde: What is that? Jul 06 16:55:09 Hmm, perhaps there's some preferred gstreamer in angstrom that is avoiding 0.10.13 generation. Jul 06 16:55:10 woglinde: obexftp will build right Jul 06 16:55:21 PREFERRED_VERSION_gstreamer="0.10.12" Jul 06 16:55:22 woglinde: after your commits ;-) Jul 06 16:55:27 catched Jul 06 16:58:04 03woglinde2 07org.oe.dev * rebb88027... 10/ (7 files in 3 dirs): Jul 06 16:58:04 Simpad updates taken from Jul 06 16:58:04 http://slackpad.svn.sourceforge.net/viewvc/slackpad/trunk/SIMpad-Kernel-2.6/2.6.21/patches/ Jul 06 16:58:04 amd tweaking the defconfig a little bit Jul 06 17:00:28 hm now back to my initrd Jul 06 17:00:52 :) Jul 06 17:05:13 * wick7xx is away: afk Jul 06 17:34:42 Hmmm, Nokia just released a flash 9 plugin for the N800. I if it might work with any of the Zaurii. Jul 06 17:37:12 hm Jul 06 17:39:21 jo florian Jul 06 17:42:34 re Jul 06 17:56:34 woglinde: ntp build without error now Jul 06 18:21:01 woglinde: the kernel does not compile some file is missing http://pastebin.ca/606177 Jul 06 18:26:20 woglinde: no new patch is in linux/simpad -> http://pastebin.ca/606189 Jul 06 19:45:20 how do I run qemuarm, anyway? Jul 06 19:45:45 it 'could not initialize sdl, exiting'... Jul 06 19:51:44 polyonymous: you need X ;-) Jul 06 19:51:59 psokolovsky, you will be surprised, but I do have it :) Jul 06 19:52:36 psokolovsky, I'm trying to run hosts's qemu (gentoo) and... well, it does initialize SDL... Jul 06 19:53:00 well, sth broken then ;-) Jul 06 19:53:06 good guess :) Jul 06 19:53:53 I'd blame it on amd64, but I really doubt it has anything to do with it... Wish it provided more diagnostics than just "couldn't initialize sdl" Jul 06 19:59:12 hmm... my system qemu now seems to boot... Jul 06 20:12:38 mickey|away, ping Jul 06 20:12:50 * Crofton curses time zones ... Jul 06 20:13:43 re Jul 06 20:14:31 crofton mickey said he is back at midnight Jul 06 20:19:16 polyonymous: hi, thx for your fix to opie-icon-reload, now my do_rootfs for opie-image was successfull ;-) Jul 06 20:19:19 hmm Jul 06 20:19:30 I should be drinking by then ... Jul 06 20:19:37 mrdata|afk, you're welcome :) Jul 06 20:20:07 I did remember someone complain about it, but couldn't remember who was that :) Jul 06 20:20:15 re mrdata Jul 06 20:21:28 woglinde: your commits for simpad miss some patch-files Jul 06 20:22:28 woglinde: only three there now Jul 06 20:23:02 -> http://www.openembedded.org/filebrowser/org.openembedded.dev/packages/linux/linux/simpad Jul 06 20:31:38 03marek.vasut 07org.oe.dev * rb2b4eebd... 10/ (10 files in 3 dirs): qemu-native: amd64-enabled 0.9.0 cvs snapshot from bug #2589. Jul 06 20:38:04 mrdata hu? Jul 06 20:38:13 args Jul 06 20:38:16 hm Jul 06 20:38:27 I forget to add Jul 06 20:38:28 right Jul 06 20:38:41 woglinde: no problem Jul 06 20:39:26 no not good Jul 06 20:39:34 I should be more carefully Jul 06 20:43:34 mrdata hm I am now at ntp with the txc problem Jul 06 20:43:56 woglinde: i make a patchf Jul 06 20:44:01 woglinde: file Jul 06 20:44:09 yes Jul 06 20:44:13 use quilt ;) Jul 06 20:44:17 its the easiest Jul 06 20:44:21 polyonymous: is that you who pushed new qemu? Jul 06 20:44:48 woglinde: no so easy for me Jul 06 20:45:39 mrdata type quilt new blafo.patch Jul 06 20:45:49 quilt edit fileIwanttochange Jul 06 20:46:05 edit it with he editor which opened Jul 06 20:46:17 and then quilt refresh patchname Jul 06 20:46:23 the patch is than in the dir patches Jul 06 20:46:33 from where you type quilt new Jul 06 20:47:25 psokolovsky, yes. bad idea? Jul 06 20:47:56 polyonymous: where's DEFAULT_PREFERENCE=-1? also, what about those PV hacks? Jul 06 20:48:52 psokolovsky, yeah, I thought about default preference, should've added it probablyl. I will then. Jul 06 20:49:23 polyonymous: unless you're sure it works for everyone ;-) Jul 06 20:49:48 psokolovsky, Well, I'm pretty sure it works for more people than the old one, but of course, it's just "sure", not "SURE" :) Jul 06 20:50:46 and yes, I didn't nocided PV weirdness but why do you call it hacks? :) Jul 06 20:52:10 polyonymous: well, if it's cvs recipe, it should use SRCDATE, with it pinned in some config file. Jul 06 20:52:29 psokolovsky, yes, I agree. I just wouldn't call it hacks :) Jul 06 20:52:37 polyonymous: it for sure shouldn't be called "0.9.0" if it's actually not that, but a random cvs snapshot Jul 06 20:52:38 Okay, I'll fix it now along with defprefs Jul 06 20:52:54 polyonymous: it's just botched recipe from marex, yeah Jul 06 20:53:16 psokolovsky, yes, otherwise I wouldn't be able to --author him :) Jul 06 20:53:21 polyonymous: someone would definitely need to fix it, now that you picked it up, well ... ;-) Jul 06 20:53:29 psokolovsky, wtf? ;-E Jul 06 20:53:39 I feel responsible for the stuff I push, yes. Jul 06 20:53:49 Marex: when will you learn to submit stuff in good shape? Jul 06 20:54:12 one day he will Jul 06 20:54:31 psokolovsky, right after the hell freezes ;-E Jul 06 20:54:31 hm Jul 06 20:54:36 what happened? Jul 06 20:54:51 And I had a strong urge to --author marek.wassup him :) Because whenever you highlight hm he says "wassup" :) Jul 06 20:55:39 psokolovsky, but I really do think that it should be 0.9.0 cvs and SRCDATE should go into .bb itself. Jul 06 20:55:40 hahaha Jul 06 20:55:41 dnl The tick/tickadj sections were written by Skippy, who never learned Jul 06 20:55:41 dnl that it's impolite (horridly gross) to show your guts in public. Jul 06 20:55:48 thats from configure.ac Jul 06 20:55:51 of ntpd Jul 06 20:55:58 and that is 100% true Jul 06 20:57:32 03woglinde2 07org.oe.dev * ra8dc61f2... 10/ (8 files in 2 dirs): Jul 06 20:57:32 shame on me, I forget to added the new patches Jul 06 20:57:32 here now they are Jul 06 20:58:04 psokolovsky, because the patchset depends on the snapshot Jul 06 20:58:43 glibc-intermediate build is failing for me Jul 06 20:58:54 configure: error: Need linker with .init_array/.fini_array support. Jul 06 20:59:11 polyonymous: wonderful, but it should be either 0.9.0+cvs, or 0.9.0+cvsXXXXYYZZ Jul 06 20:59:20 isn't it? Jul 06 20:59:26 or do you mean .bb name? Jul 06 20:59:35 polyonymous: .bb name Jul 06 20:59:38 ah sure Jul 06 21:03:59 woglinde: -> http://www.pastebin.ca/606432 , uff Jul 06 21:04:21 mrdata hm Jul 06 21:04:57 yes seems goods Jul 06 21:05:00 * * OE Bug 2610 has been created by  Jul 06 21:05:02 * * mtd-utils can not be build becuase GIT_AUTHOR_NAME and GIT_COMMITTER_NAME are not set Jul 06 21:05:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2610 Jul 06 21:05:06 tha damn __adjtimex problem Jul 06 21:05:30 woglinde: and the work is not from me -> http://blackfin.uclinux.org/gf/project/uclinux-dist/scmsvn/?action=browse&path=%2Ftrunk%2Fuser%2Fntp%2Futil%2Ftickadj.c&view=markup Jul 06 21:07:47 hm there was a good explanation to the problem Jul 06 21:07:51 let me find it Jul 06 21:08:24 are here it is Jul 06 21:08:26 https://lists.ntp.org/pipermail/bugs/2007-March/005578.html Jul 06 21:10:03 * mr_nice|afk happy Jul 06 21:10:27 mrdata: it realy was the swap stuff :). I was able to build angstrom-minimal-image Jul 06 21:10:38 mr_nice hehe good Jul 06 21:10:49 mrdata: now I will try to build opie :) Jul 06 21:10:59 woglinde: yes :) Jul 06 21:11:14 mr_nice I just pushed the missing patches Jul 06 21:11:30 woglinde: thx Jul 06 21:13:24 woglinde: also thx from me for pushing this patches Jul 06 21:14:34 okay Jul 06 21:14:50 I will integrate now the tickadj patch Jul 06 21:15:26 woglinde: it's quite cool to have one oe dev which owns a simpad :) Jul 06 21:15:37 there are 2 Jul 06 21:15:42 * woglinde eyes to zecke Jul 06 21:15:50 but zecke hasnt time for simpad Jul 06 21:16:45 and better, this oe dev have write access again ;-) Jul 06 21:16:48 woglinde: jepp. he already did enought for the simpad commuinty :) and should also be able to just use it Jul 06 21:18:04 woglinde: have your simpads any mods inside Jul 06 21:18:32 no Jul 06 21:19:29 ah, okay, but 128MB ram inside helps a lot ;-) Jul 06 21:20:00 I am not good at smd soldering Jul 06 21:20:25 this is one part of my job Jul 06 21:20:39 mrdata, solder it for woglinde :) Jul 06 21:20:41 mrdata you were in berlin? Jul 06 21:20:57 woglinde: no, in LE Jul 06 21:21:48 but 200km is not that problem, i think Jul 06 21:22:11 le == leipzig? Jul 06 21:22:19 polyonymous: yes ;-) Jul 06 21:22:28 mrdata: but be carefully else our oe connection is broken *g Jul 06 21:22:37 ah. I think Laibsch is your neighbour then :) Jul 06 21:22:56 hehe Jul 06 21:22:57 lol Jul 06 21:23:04 polyonymous: i have reading this, but his address gives brilon Jul 06 21:23:05 hm my dad lives in leipzig Jul 06 21:23:28 mrdata, I'm never sure of what I'm saying, anyway :) Jul 06 21:23:42 so many germans around... :) Jul 06 21:24:08 and, though not german, I live in HH :) Jul 06 21:24:11 polyonymous: no, no, you are right, he gives also leipzig as possibility Jul 06 21:24:46 i have read his home page 2 days ago Jul 06 21:24:49 mrdata, he doesn't have any particular location, just the probability :) Feels like a nuclear physics or something :) Jul 06 21:24:51 polyonymous germany is the countries with the most opensource developers Jul 06 21:24:58 mrdata: Where have you been looking? My arcor.de page? Jul 06 21:25:03 hm Jul 06 21:25:07 polyonymous is correct, I live in LE Jul 06 21:25:10 woglinde, I know. Jul 06 21:25:21 ntp dir gives me some headattach Jul 06 21:25:23 hey, Laibsch but I thought you're elsewhere now? Jul 06 21:25:36 Laibsch: hallo and yes, but your impressum gives brilon Jul 06 21:26:20 could someone please paste me what the packages/ntp/ dir contains? Jul 06 21:26:32 mrdata: impressum? That might be my mums page Jul 06 21:26:42 online web seems wrong Jul 06 21:26:50 Or the TOL impressum? Jul 06 21:27:09 polyonymous: I've returned, back in LE now. Jul 06 21:27:16 Laibsch, welcome back then! Jul 06 21:27:21 thanks. Jul 06 21:28:21 mrdata: You see, Laibsch, laaaiibbbbssscchh? Rings a bell? ;-) Jul 06 21:28:36 Just speak it very softly ;-) Jul 06 21:28:45 * polyonymous also wants his bell to ring :) Jul 06 21:29:00 (bell, not ball!!!) Jul 06 21:29:09 damm, sächsch is a good language Jul 06 21:29:25 * Laibsch is not sure if he can agree Jul 06 21:30:21 Laibsch: welcome back. buncha stuff on opie side, needs testing ;-) Jul 06 21:30:30 I am already at it. Jul 06 21:30:38 psokolovsky: Thanks for the very good work Jul 06 21:30:51 Or at least a lot of work :)) Jul 06 21:30:52 Laibsch: -> http://home.arcor.de/leggewie/ gives brilon Jul 06 21:30:54 psokolovsky: Saw all those nice and numerous patches Jul 06 21:31:15 I didn't mean to say the work isn't good, just that I can't confirm it's good, yet :) Jul 06 21:31:47 Laibsch: so i was not sure what brilon <-> leipzig puts together Jul 06 21:32:01 * * OE Bug 2611 has been created by jpuhlman(AT)mvista.com Jul 06 21:32:04 * * Perl bitbake file uses wrong define to address the cross linker... Jul 06 21:32:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2611 Jul 06 21:33:17 Laibsch: sorry, i can not answer you, pleas join #opensimpad.org Jul 06 21:33:30 does libsdl has x disabled intentionally? Jul 06 21:33:47 mrdata, register with freenode already :) Jul 06 21:34:22 polyonymous: i know, but someone has register my nick before Jul 06 21:34:32 mrdata, be creative then :) Jul 06 21:35:01 don't be so plural, shrink a bit and register mrdatum :) Jul 06 21:35:24 after all mr is supposed to be singular. Jul 06 21:35:24 mrdatii? Jul 06 21:35:30 mwester, nah, data is plural Jul 06 21:35:47 no! i like this nick, i is a part from me Jul 06 21:36:14 mrdata, it would still resemble part of you. datum IS a part of data :) Jul 06 21:36:27 if the nick has not been used for some long time, the admins can free it for your use, mrdata Jul 06 21:36:49 and meanwhile you can register datum :) Jul 06 21:37:18 mwester: right, but i must contact such admin Jul 06 21:37:23 mrdata: which target did you use for opie? Jul 06 21:37:35 misters are talking :) Jul 06 21:38:28 mr_nice: opie-image Jul 06 21:38:33 thx Jul 06 21:39:03 and for contact a admin, i am a little bit shyly Jul 06 21:39:07 mrdata: I am building from scratch now Jul 06 21:39:10 03polyonymous 07org.oe.dev * r6bf4895c... 10/ (5 files in 2 dirs): qemu 0.9.0+cvs: polished .bbs a bit and gave them default preference of -1, anyway Jul 06 21:39:22 mrdata, not in person, it is the net! :) Jul 06 21:39:55 yes, but my second problem is my poor english Jul 06 21:40:31 mrdata, c'mon, it's still the Net, most of us aren't native speakers, your english is definitely not the worst I've seen on the net :) Jul 06 21:40:53 and the proud worstenglishspeakers have never been shy :) Jul 06 21:40:54 polyonymous: thx for that Jul 06 21:41:27 you're welcome, but I'm not native speaker either, so... :))) Jul 06 21:42:04 * polyonymous is just a humble german language student ... der die das :) Jul 06 21:42:23 polyonymous: great Jul 06 21:42:53 you see! :) Almost Göthe and Schiller in one :) Jul 06 21:44:23 polyonymous: naja, this were great poets Jul 06 21:44:50 mrdata, yes. Of course I can also compete with Heine and Hesse :) Jul 06 21:51:55 hm is there something against to upgrade ntp to 4.2.4p0? Jul 06 21:52:49 yes, actual is ntp-4.2.4p3 Jul 06 21:54:17 heh, russians in germany when speaking russian also say "actual" when they mean "current" :) Jul 06 21:54:37 not that "current" is "current" in russian, of course :) Jul 06 21:54:58 polyonymous: right, i mean current Jul 06 21:55:34 mrdata, I know. but it's much funnier when russians who are native speakers do it :) Jul 06 21:56:08 morning Jul 06 21:56:20 morning xora Jul 06 21:56:41 hi xora Jul 06 21:56:54 qemu's -append root= doesn't seem to do anything for spitz kernel :( Jul 06 21:56:55 Sandisk Connect+ fscks the kernel so badly it doesnt even oops :-( Jul 06 22:00:57 is there a way to pass root to spitz kernel with qemu? Jul 06 22:02:02 Crofton__: pong Jul 06 22:02:59 polyonymous: probably it's funky z kernel which doesn't parse kernel cmdline, eh? Jul 06 22:03:23 psokolovsky, dunno maybe... Jul 06 22:04:11 so there's no way around? Jul 06 22:04:47 polyonymous: I remember discussing sth in that regard with balrog-kun. either it was that z kernel doesn't parse atags properly, or its bootloader cannot grok them, so they're not implemented for it, sth like that... Jul 06 22:04:55 hmm... I remember like 10 years ago there were binaries to manipulate kernel images, but I don't even remember the name :) Jul 06 22:05:06 polyonymous: get a real PDA ;-F Jul 06 22:05:16 psokolovsky, I think it does parse, after all what's the use of root=/dev/hda1 in CMDLINE? Jul 06 22:05:22 psokolovsky, I like Z very much :) Jul 06 22:06:15 polyonymous: in cmdline compiled into kernel? yep it does parse in the end. all that stuff was about ATAGs. Jul 06 22:06:36 what's ATAG? :) Jul 06 22:07:25 polyonymous: ARM-Linux standard to pass boot params to kernel. supersedes some x86 adhoc crap. Jul 06 22:07:42 ah yeah, that's about passing cmdline to kernel, I guess. Jul 06 22:12:02 * * OE Bug 2612 has been created by xyzaid(AT)yahoo.fr Jul 06 22:12:04 * * kazekahase segmentation fault Jul 06 22:12:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2612 Jul 06 22:13:07 hmm... I recompiled kernel and now it says hda: blah blah blah, and unable to mount root on unknown-block(3,0)... wtf? Jul 06 22:13:22 ah, actually it does oops, altboot seems to disable console on serial :-( Jul 06 22:13:32 why is 3,0 unknown-block if it has hda? Jul 06 22:14:43 psokolovsky, oh well, do you think playing with MACHINE=qemuarm is about as good as spitz? Jul 06 22:15:28 re Jul 06 22:15:36 the only reason why I wanted spitz is because qemuarm image's opie LOOKS way different, but it's not necessarily related to machine. Jul 06 22:15:40 polyonymous: for what purpose? I debug userpace on integrator Jul 06 22:16:04 psokolovsky, for the purpose of playing with this new opie stuff without pulling it into Z. Jul 06 22:16:20 polyonymous: qemuarm is perfect fit for that Jul 06 22:16:47 polyonymous: and unknown-block is static string in kernel printk Jul 06 22:16:52 psokolovsky, great thanks. So, the lack of nice background is likely related to all that moves around? Jul 06 22:17:11 psokolovsky, ah, so it may just be that it can't find filesystem for some reason? Jul 06 22:17:49 nevermind though, I'll play with qemuarm then. Jul 06 22:18:37 polyonymous: yeah, it detected device (otherwise it would be (0,0)), but still can't use it for some reason Jul 06 22:18:52 gotta love those leenooks messages ;-\ Jul 06 22:18:59 ah ok. thanks for feeding my curiosity :) Jul 06 22:21:36 polyonymous: I also found that kernel simply hangs when it is fed too big initramfs (one it can't uncompress w/o getting OOM). And it doesn't take too big an initramfs for that. So, I lost faith in initramfs as complete replacement for initrd. Jul 06 22:22:30 psokolovsky, but does spitz use any of these? :) Jul 06 22:23:17 polyonymous: it's not about spitz, it's about stoopid leenooks! ;-) Jul 06 22:24:08 Ah. Well, nobody's perfect :) Jul 06 22:32:46 so, psokolovsky, am I right then that current opie-image is about totally broken? :) Jul 06 22:35:32 polyonymous: of course no. it's almost perfect by now ;-) Jul 06 22:36:02 psokolovsky, hmm.. when you run it in qemu do you have background? or is it not supposed to be there? Jul 06 22:36:24 not that it would justify the statement of it's brokenness :) Jul 06 22:37:15 polyonymous: it should have 240x320 background. and yep, that's not criterion of brokenness ;-) Jul 06 22:37:43 it has white background... well, I'll play more now. Jul 06 22:38:36 polyonymous: I yes, right, it has no b/g. I left sth for you to fix ;-)) Jul 06 22:38:55 psokolovsky, ah, that was nice of you, thanks :))) Jul 06 23:25:42 night all Jul 06 23:27:22 polyonymous: FYI: qemu-native-0.9.0+cvs20070701-r1/qemu/linux-user/mmap.c:237: error: `MAP_32BIT' undeclared (first use in this function) Jul 06 23:27:50 psokolovsky, how did you build it? Jul 06 23:28:17 polyonymous: that's older checkout, np. just regarding "if it works for everyone" Jul 06 23:28:38 psokolovsky, well, I said I'm sure, not SURE : Jul 06 23:28:39 :) Jul 06 23:28:44 ;-) Jul 06 23:28:48 but that's strange, anyway. Jul 06 23:29:28 psokolovsky, or do you mean it doesn't work on 32bit x86? Jul 06 23:29:46 * polyonymous pokes Marex into ribs. Jul 06 23:29:48 yep Jul 06 23:30:16 Well, I'm glad I defprefed it, thanks for the hint. Jul 06 23:30:17 ouch :E Jul 06 23:30:22 wassup? Jul 06 23:30:27 I was hoping it would hurt :) Jul 06 23:30:37 marek.wassup :) Jul 06 23:30:56 Marex, patches against current mtn to make it work for 32bit platform as well are welcome :) Jul 06 23:32:02 and yes, I admit I am responsible for any breakage this qemu might introduce. Jul 06 23:33:15 MAP_32BIT is only supported on amd64 ... damn :-T I told you rather merge poky qemu :-T Jul 06 23:33:35 polyonymous: actually, last I heard on that issue is that Marex was pointed out that poky has some kind of solution for amd64 problem too, and was asked to have a look at it. of course, he didn't note that fact in the original bug... Jul 06 23:33:37 Marex, okay, no patches from you, I will then. Jul 06 23:33:48 he did. Jul 06 23:33:55 ;-) that's too conservative ;-) Jul 06 23:34:04 And I was going to take a look at it as well. Jul 06 23:34:08 just watch his patches thoroughly ;-) Jul 06 23:34:24 well :) Jul 06 23:34:50 the default_prefference = -1 is there, isnt it? Jul 06 23:34:51 anyway, in this case, he did refer to poky in the bug, it's just that I've had it in my tree by that time and postponed looking at poky version. Jul 06 23:34:55 it is. Jul 06 23:35:03 and if it wasn't it would be my fault, not yours. Jul 06 23:35:13 or at least not only yours :) Jul 06 23:35:31 actually, no "if" here it WAS my fault until psokolovsky pointed it out to me :) Jul 06 23:35:34 is there some way to apply certain patches only if host=AMD64 ? Jul 06 23:35:50 Marex, put #ifdef into patch? :)) Jul 06 23:36:05 Marex: you should just code patch in such way that it works everywhere Jul 06 23:36:29 But heck, I think I should just take poky version and that's it. Jul 06 23:36:32 psokolovsky, I didnt noticed that MAP_32BIT works only on amd64 ... sorry Jul 06 23:36:42 polyonymous, right Jul 06 23:36:55 Marex, did all right, he indeed point out that poky version exists. Jul 06 23:37:25 polyonymous, I said in that bugreport it will be better than mine, didnt I ? :T Jul 06 23:37:26 Marex: well, np, that happens. we all just should be prepared to test patches to stuff on critical build path thoroughly. Jul 06 23:37:38 Marex, didn't I just tell you you did? :) Jul 06 23:38:00 polyonymous, ok ... sorry ... I feel braindamaged from hunting opie bugs :T Jul 06 23:38:05 heh Jul 06 23:38:22 yeah, I remember, facioplastic on the rotting corpse :) Jul 06 23:39:00 polyonymous, yeah ... though I have to reconstruct deeper stuff first ... and not it seems more like brain surgery :T Jul 06 23:39:07 *now Jul 06 23:39:26 on the rotting corpse? Sounds appealing :) Jul 06 23:40:16 yeah ... its melting under my hands :T Jul 06 23:40:48 scalpel, you mean :) Jul 06 23:40:58 mr surgeon :) Jul 06 23:41:59 lol Jul 06 23:42:18 polyonymous, shovel would be just fine ;) Jul 06 23:42:22 nite Jul 06 23:42:28 heh Jul 06 23:42:33 night, woglinde Jul 06 23:42:45 * Marex thinks implementation of blouetooth visibility switching based on ioctls ripped from hciconfig isnt quite pretty :T Jul 06 23:43:04 Marex, I told you dbus would be just easier. Jul 06 23:43:08 marex why not Jul 06 23:43:16 hm Jul 06 23:43:19 polyonymous, I told you opie doesnt support dbus Jul 06 23:43:23 but opie dont preffer dbux Jul 06 23:43:30 *hides* Jul 06 23:43:57 Marex, but what prevents you from just linking in dbus and using dbus calls instead of ioctls? Jul 06 23:45:11 polyonymous, using libbluetooth + ioctls requieres just bluez + opie ... linking dbus in would add third dependency Jul 06 23:45:31 Marex, bluez depends on dbus. Jul 06 23:45:54 ahm .... right :T Jul 06 23:46:09 listen to what the wise man says :) Jul 06 23:46:28 ;-D Jul 06 23:46:28 polyonymous: so our next aim is to put bluez2 back into OE. Jul 06 23:46:40 psokolovsky, like why? Jul 06 23:46:55 polyonymous: as for dbus - it's easy, just port everything in opie from qcop to it, and we switch ;-D Jul 06 23:47:05 psokolovsky, I'm not excited about this idea. We can just as well simply revert the whole mtn tree :) Jul 06 23:47:15 Marex: to not bring in dbus dependency Jul 06 23:47:31 polyonymous: how revert relates here? Jul 06 23:47:38 bluez2 doesnt support bcm2035 ... Jul 06 23:47:50 psokolovsky, how about (a) having both qcop and dbus coexist (b) having a qcop over dbus? Jul 06 23:48:27 psokolovsky, going back to bluez2 is a lot like reverting, instead of pulling in the new stuff. Jul 06 23:48:49 anyway, I don't see how bt controlled via dbus mandates the whole opie switching to dbus. Jul 06 23:48:54 polyonymous: (a) easy, just provide images which fit into 8mb. (b) and all that over corba, please Jul 06 23:49:38 psokolovsky, (a) as long as bluez3 with opie fits, opie controlling bt over dbus should fit as well, (b) ;-))) Jul 06 23:49:52 polyonymous: new stuff is not always good. you're welcome to patch bluez3 to drop auxiliary dependencies not related to its core purpose - be an official Linux BT stack. Jul 06 23:50:53 psokolovsky, sometimes we just need to cope with these parts of new stuff... and don't expect these patches from me - it's far from being my own itch :) Jul 06 23:51:45 polyonymous: I'm glad that you agree that going back to bluez2 is the most sane way ;-) Jul 06 23:51:56 I don't :) Jul 06 23:52:14 Actually, I don't think we have any way back to the realm of sanity :) Jul 06 23:52:17 me neither ;-E Jul 07 00:01:21 u`re always can go and exhumate the corpse of affix bt stack :] Jul 07 00:02:38 фпрочем этот труп и так уже не воняет, так что бесполезно Jul 07 00:27:56 psokolovsky, what did you leave for me to fix? Implement runtime support in opie for loading background pixmap according to resolution? It suits Marex with his autopsy better :)) Jul 07 00:28:45 * Marex feels like pathologjst :T Jul 07 00:28:59 and well known for it :) Jul 07 00:29:24 polyonymous: well, you're likely right ;-). so, just update-alternatives for opie-taskbar-images-* would be enough ;-) Jul 07 00:29:54 yeah, I think that would make more sense :) Jul 07 00:30:17 but I'm off to bed now. once again, thanks for your torturing poor opie :) Jul 07 00:31:08 ;-) Jul 07 00:43:48 Hrm. pand was removed from bluez-utils... Doing an ipkg upgrade isolated me :) Jul 07 00:46:28 Necronom, ipkg install bluez-utils-compat Jul 07 00:53:02 btw, why such split happen? Jul 07 00:59:57 bluez folks are going madder and madder... Jul 07 01:00:40 psokolovsky, now I agree with you, but dont get too used to it ;-E **** ENDING LOGGING AT Sat Jul 07 02:59:56 2007