**** BEGIN LOGGING AT Tue Oct 17 02:59:57 2006 Oct 17 06:23:03 * koen reverts some breakage Oct 17 06:23:16 03koen 07org.oe.dev * rff788cda... 10/ (1 conf/machine/h5000.conf conf/machine/h5xxx.conf): Oct 17 06:23:16 h5xxx: rename machine back to its original name Oct 17 06:23:16 * renaming machines breaks overrides. In this case the kernel couldn't find its defconfig Oct 17 06:23:16 * renaming machines breaks ipkg since existing installations don't know about the new arch Oct 17 06:23:16 * renaming machines breaks existing distributions Oct 17 06:23:16 So please don't rename machines without providing an upgrade path Oct 17 06:25:02 he koen Oct 17 06:25:38 03hrw 07org.oe.dev * rc8b6f226... 10/ (1 packages/fltk/fltk-1.2+cvs20040806): fltk: dropped empty dir - close #1490 Oct 17 06:28:21 03hrw 07org.oe.dev * r649af75f... 10/ (1 MAINTAINERS): MAINTAINERS: added Rolf Leggewie - close #1493 Oct 17 06:28:58 Good Morning Oct 17 06:29:51 hello blizzard Oct 17 06:30:31 Is opie-writer only supposed to read and write html? Oct 17 06:30:41 I thought it could do rtf Oct 17 06:30:46 wrong channel? Oct 17 06:31:00 I thought someone here had some experience Oct 17 06:31:14 I asked twice on #opie, and they won't answer Oct 17 06:31:55 No matter Oct 17 06:32:27 because they are not in front of theier computers? Oct 17 06:32:41 easiest is to check the source code Oct 17 06:33:03 Ah Oct 17 06:33:06 Will do Oct 17 06:33:27 The website and the name of the program say it does, but the program doesn't Oct 17 06:39:22 hi, I'm trying to find the version of libgaim-protocol-msn in oe. I'm looking in the "browse tree" option in the website, but I don't seem to find it under "packages", can anybody point me to the right direction? Oct 17 06:39:42 YoG: it's part of the gaim recipes Oct 17 06:40:12 so I should look inside the qpe-gaim package? Oct 17 06:40:48 get the ipkg and look at the Source: field in the control file Oct 17 06:41:22 ah ok, thanx Oct 17 06:43:39 03hrw 07org.oe.dev * r43f0c625... 10/ (3 files in 3 dirs): Oct 17 06:43:39 keymaps: fixed tosa/2.6 keymap - close #1500 and partially #1498 Oct 17 06:43:39 - Fn+Space = bar like it is on keyboard (part of #1498) Oct 17 06:43:39 - 'bullseye' (center of ring) = space (close #1500) Oct 17 06:43:39 - Fn+z is no longer bar Oct 17 06:44:41 hm frist bugday? Oct 17 06:44:45 first Oct 17 06:46:03 03hrw 07org.oe.oz354x * r2d8c7644... 10/ (3 files in 3 dirs): Oct 17 06:46:03 keymaps: fixed tosa/2.6 keymap - close #1500 and partially #1498 (from .dev) Oct 17 06:46:03 - Fn+Space = bar like it is on keyboard (part of #1498) Oct 17 06:46:03 - 'bullseye' (center of ring) = space (close #1500) Oct 17 06:46:03 - Fn+z is no longer bar Oct 17 06:46:03 applied changes from 649af75f6b16159321cf1119882aba851afe11be .dev Oct 17 06:46:05 through 43f0c625d45be3cacb24330821a0102dae836549 Oct 17 07:18:43 yop Oct 17 07:24:27 03koen 07org.oe.dev * r46e19d94... 10/ (12 files in 2 dirs): gtk+: add ttf-dejavu-sans to RRECOMMENDS, otherwise applications requiring fonts will crash Oct 17 07:25:13 IceWM should have some kind of font added Oct 17 07:25:31 Cause otherwise it displays empty boxes Oct 17 07:25:45 patches welcome Oct 17 07:25:50 Sure Oct 17 07:26:24 I like the fact that IceWM doesn't depend on X though :) Oct 17 07:30:02 Is RDEPENDS or RRECOMMENDS more appropriate? Oct 17 07:30:34 As far as I know, IceWM can't be usable without a font Oct 17 07:30:42 But it can be used Oct 17 07:32:26 How do I push changes upstream? Oct 17 07:34:01 morning Oct 17 07:34:20 Good Morning Oct 17 07:40:49 B_Lizzard: bugtracker Oct 17 07:40:51 hey XorA Oct 17 07:42:07 * XorA wishes linux still had fast delete Oct 17 07:42:29 use 2.6.17 and xfs Oct 17 07:42:30 hey florian_kc Oct 17 07:42:46 koen: isnt that the FS that hrw keeps losing files on? Oct 17 07:42:56 hi Oct 17 07:43:02 XorA: 2.6.17 has a huge xfs bug Oct 17 07:43:08 XorA: a very fast delete :) Oct 17 07:43:52 * XorA chuckles, so xfs has reached reiserfs levels of usablility Oct 17 07:45:46 XorA: on linux, yes :) Oct 17 07:46:03 or is that deathrowFS Oct 17 07:47:18 Not quite yet Oct 17 07:47:27 He is innocent, as of yet Oct 17 07:48:50 morning Oct 17 07:48:58 hey hrw|work Oct 17 07:49:00 morning Oct 17 07:50:22 B_Lizzard: thought youd been told, everyone is guilty until proved innocent Oct 17 07:50:57 Well, I didn't expect a kind of spanish inqusition! Oct 17 07:50:59 just say "terrorist" and you don't need any trials anymore in the US Oct 17 07:51:03 * B_Lizzard looks at the door Oct 17 07:51:25 Aye, koen is pretty much right about that Oct 17 07:52:03 Koen, do I submit the whole .bb file, or a patch containing the changes? Oct 17 07:52:19 patch, as attachment, not inline Oct 17 07:52:24 Aye Oct 17 07:52:59 * koen waves to likewise Oct 17 07:53:17 * likewise waves back Oct 17 07:53:49 good morning all Oct 17 07:54:04 hi likewise Oct 17 07:54:17 likewise Oct 17 07:54:27 florian_kc: hi florian, B_Lizzard Oct 17 07:55:34 koen: I had this crazy idea of bitbake adapting the "alldepends.dot" file to colour the nodes during the build, so that one can graphically see the build progress. Cool nah? Oct 17 07:55:56 quick someone add another .dd, we are less than 4000 Oct 17 07:56:01 s/dd/bb/ Oct 17 07:56:08 helloworld.bb Oct 17 07:56:13 byebyeworld.bb Oct 17 07:56:20 4000.bb? Oct 17 07:57:36 likewise: I had that idea as well :) Oct 17 07:57:59 likewise: but on a webpage with googlemaps like scrolling Oct 17 07:58:14 koen: you didn't apply for a patent did you? Oct 17 07:58:55 :) Oct 17 07:59:07 koen: this would probably be a bbclass, that adds some pre/post tasks right? Oct 17 07:59:24 koen: BitTube.com :-) Oct 17 07:59:52 already exists... nm Oct 17 07:59:59 likewise: I think it would be a bitbake module Oct 17 08:00:40 koen: a bitbake module plugs into bitbake? so, in the bitbake sources? Oct 17 08:00:52 yes Oct 17 08:01:23 likewise: pornobake.com Oct 17 08:01:57 XorA: probably Google won't buy it then :-) Oct 17 08:02:12 likewise: dont google own flickr? Oct 17 08:02:26 XorA: that's yahoo Oct 17 08:02:39 ah, thats why seaching for cats works so well Oct 17 08:04:42 morning all Oct 17 08:04:55 hey rp Oct 17 08:05:18 RP: what license is stage-manager? MIT? Oct 17 08:06:45 hey RP Oct 17 08:08:21 RP: should devmand be started with the current zaurusd code? Oct 17 08:09:21 RP: good morning Oct 17 08:12:36 03koen 07org.oe.dev * r3dae7e8b... 10/ (8 files in 5 dirs): packaged-staging: add a recipe for stage-manager and DEPENDS on it Oct 17 08:13:11 koen: nice Oct 17 08:13:18 hi [cc]smart RP Oct 17 08:15:20 03koen 07org.oe.dev * r5668d8e4... 10/ (1 classes/packaged-staging.bbclass): packaged-staging.bbclass: update instructions Oct 17 08:15:41 florian_kc: yes, that was the intention. I've just not had time to do it Oct 17 08:15:43 hrw|work: now it's 33% easier to use ;) Oct 17 08:16:19 koen: I should probably GPL(v2) it I guess Oct 17 08:16:51 RP: what a well known problem ;) Oct 17 08:16:53 RP: could you add that to the file itself? Oct 17 08:17:04 koen: will do Oct 17 08:17:23 thanks Oct 17 08:23:24 hrw|work: http://msevior.livejournal.com/12752.html Oct 17 08:27:12 I filed the "bug report" Oct 17 08:27:17 I hope it's allright Oct 17 08:27:33 I used to be a virgin Oct 17 08:27:35 BUT NO LONGER! Oct 17 08:35:20 koen: no inherit native? Oct 17 08:37:37 RP: I wanted as less dependencies as possible Oct 17 08:37:46 03rpurdie 07org.oe.dev * re6c1fd77... 10/ (3 files in 3 dirs): stage-manager: Add copyright + licence, improve .bb file Oct 17 08:38:04 koen: Without it, it will end up in the wrong bindir Oct 17 08:38:13 you'll also notice it's using do_install instead of do_stage Oct 17 08:38:22 RP: there is only one bindir... Oct 17 08:38:44 STAGING_BINDIR = "${STAGING_DIR}/${BUILD_SYS}/bin" Oct 17 08:40:12 koen: right, I probably just broke it... Oct 17 08:40:37 native.bbclass looks pretty harmless Oct 17 08:40:44 and it does strip of PACKAGE_DEPENDS :) Oct 17 08:40:54 s/of/off/ Oct 17 08:41:24 koen: At some point, I'd like to see OE using an arch independent bin dir for scripts like this Oct 17 08:41:36 yes Oct 17 08:43:33 hey zecke Oct 17 08:43:57 ho Oct 17 08:44:01 hi zecke Oct 17 08:44:49 RP: the problem now is how to handle our patched ipkg-build Oct 17 08:45:08 RP: since it will break the ipkg-collateral package iirc Oct 17 08:46:30 03koen 07org.oe.dev * r56d82178... 10/ (3 files in 2 dirs): stage-manager: correct PV Oct 17 08:55:30 koen: I'm not sure exactly what changes you had to make to it... Oct 17 08:57:32 zecke: hi zecke Oct 17 08:58:22 likewise: hey Oct 17 09:02:57 RP: ignore /usr/lib/ipkg Oct 17 09:04:53 no feedback on bonsai Oct 17 09:06:59 zecke: I think I have fixed icecc.bbclass to the point that it works. Could you care to commit? Oct 17 09:08:50 * likewise is pushing developers to commit - push push push! Oct 17 09:09:06 Push me: http://bugs.openembedded.org/show_bug.cgi?id=1477 Oct 17 09:09:38 likewise: attach recipes, not inline Oct 17 09:09:49 03hrw 07org.oe.dev * rf772c2cf... 10/ (1 packages/stage-manager/stagemanager-native_0.0.1.bb): stage-manager: fixed typo in inherit Oct 17 09:09:56 zecke: FWIW, I think its useful Oct 17 09:10:10 * hrw|work sends 0.5l cup of coffee to koen Oct 17 09:10:10 * koen mumbles something about inline attachments Oct 17 09:10:35 likewise: The requested URL /archives/cairo/attachments/20061013/5c1e044a/librsvg-clip.bin;patch=1;pnum=0 was not found on this server. Oct 17 09:10:44 ops.. Oct 17 09:10:47 * hrw|work need coffee too Oct 17 09:11:03 koen: tinderbox has now proper license headers Oct 17 09:11:10 zecke: awesome Oct 17 09:11:11 koen: LTG must send a postcard to me now Oct 17 09:11:50 zecke: The only thing that I couldn't get to work was looking at a list of files in the root directory of a given repository Oct 17 09:12:36 RP: that is tricky Oct 17 09:12:37 hrw|work: what is the recommended root directory for diffs? Should I be outside or inside the branch directory? Oct 17 09:12:58 likewise: 'mtn diff' ? Oct 17 09:13:14 likewise: for me you can just attach recipe file. Oct 17 09:13:57 RP: I will need to handle paths better, e.g. with svn they start with '/', with mtn not Oct 17 09:15:36 ERROR: Exception:bb.MalformedUrl Message:${GPE_SVN} Oct 17 09:15:36 ERROR: Printing the environment of the function Oct 17 09:15:36 ERROR: ${GPE_SVN} while parsing /home/hrw/devel/oe/org.openembedded/packages/gpesyncd/gpesyncd_svn.bb Oct 17 09:15:40 hmm.. Oct 17 09:17:05 hrw|work: update bitbake.conf? Oct 17 09:17:09 NOTE: package gpesyncd-0.0+svn20061017-r0: task do_build: completed Oct 17 09:17:09 oh GPE_SVN? lol Oct 17 09:17:41 zecke: updated repo to head Oct 17 09:17:56 koen: could you reload apache again Oct 17 09:18:28 done Oct 17 09:24:50 koen: which django did you use Oct 17 09:24:54 pb_: HEY!!! Oct 17 09:25:11 hi Phil Oct 17 09:25:19 zecke: the one from the tarball link you gave me Oct 17 09:25:29 hi hrw|work Oct 17 09:25:31 hail zecke Oct 17 09:25:41 hail mickeyl Oct 17 09:25:52 koen: +patch? Oct 17 09:26:25 zecke: I explained to you yesterday I can't morally justify applying that patch since it's tainted Oct 17 09:26:35 koen: why is it tainted? Oct 17 09:27:01 I explained that yesterday and I'm not going to repeat myself Oct 17 09:27:22 ERROR: name 'get_kernelversion' is not defined while parsing /home/hrw/devel/oe/org.openembedded.dev/packages/linux/xanadux-un-2.6_2.6.12.bb Oct 17 09:27:28 did not we get rid of it? Oct 17 09:27:59 pb_: howdee Oct 17 09:28:01 koen: no, you didn't Oct 17 09:28:17 pb_!!!!!!!!!!!!! Oct 17 09:28:33 pb_: good to see you! How has life been to you? Oct 17 09:28:37 morning btw. Oct 17 09:29:33 koen: ERROR: upwards chain is: dnsmasq (dnsmasq) -> ipkg-utils-native (ipkg-utils-native) -> quilt-native (quilt-native) -> stagemanager-native (stagemanager-native) Oct 17 09:29:52 que? Oct 17 09:29:59 it has inhibit_default_deps set Oct 17 09:30:11 checking for libiconv_open in -liconv... no Oct 17 09:30:12 | configure: error: *** No iconv() implementation found in C library or libiconv Oct 17 09:31:09 mickeyl: pretty good, thanks. how are you? Oct 17 09:32:53 koen: I use bitbake 1.6 Oct 17 09:33:04 hrw|work: I don't Oct 17 09:33:21 koen: p-s does not warn that it need trunk... Oct 17 09:33:24 pb_: busy, but pretty good as well. submitted my thesis (finally), waiting for the advisory board to come up with an appointment for the public defense now. started a freelancer career and I'm working on the first project atm. Oct 17 09:33:52 Super Secret Project (TM) Oct 17 09:33:54 hrw|work: it seems that bitbake trunk is broken, since INHIBIT_DEFAULT_DEPS = "1" should work Oct 17 09:33:55 hehe Oct 17 09:34:45 mickeyl: you scared him away Oct 17 09:34:50 darn Oct 17 09:34:53 mickeyl: or pb_ == lrg Oct 17 09:34:59 SSP has some known features but we would be forced to write NDA if I would tell which :D Oct 17 09:35:03 hi liam Oct 17 09:35:21 zecke: unlikely ;) Oct 17 09:35:43 morning all Oct 17 09:35:57 hey lrg Oct 17 09:35:57 hrw|work: Cell SSP? Oct 17 09:36:26 Super Secret Project Oct 17 09:36:39 hi lrg Oct 17 09:37:42 likewise: social engineering makes miracles... blah, blah, blah, it will have bt, blah,blah,blah,no irda, blah, blah.. Oct 17 09:37:54 * mickeyl chuckles Oct 17 09:38:42 zecke: now Im scared, we have submarines pb_s Oct 17 09:40:32 anyone interested in mmx2 support in ARM ? Oct 17 09:41:05 lrg: has intel finally bribed you? Oct 17 09:41:46 * XorA is even more scared as terrier turns itself on Oct 17 09:41:55 koen: no, there's a public patch for gcc now Oct 17 09:42:12 koen: building stagemanager-native with !empty staging fails Oct 17 09:42:37 koen: http://pastebin.ca/206236 Oct 17 09:43:22 koen: I've got a Zylonite board now Oct 17 09:44:20 zylonite? Oct 17 09:44:34 lrg: which SoC's support mmx2? Oct 17 09:44:51 hi lrg Oct 17 09:45:08 koen: Monahans, zylonite is the ref board Oct 17 09:45:12 hey pb_ Oct 17 09:45:27 lrg: ah, right Oct 17 09:45:38 lrg: I bet you ads is still faster Oct 17 09:47:01 koen: i think so, probably because it has VFP Oct 17 09:51:34 pb_: we have missed you! Oct 17 09:53:06 zecke: heh, I have missed you too Oct 17 09:55:16 :) Oct 17 09:55:37 dpkg-deb: building package `openembedded' in `../openembedded_1.0-1_i386.deb'. Oct 17 09:57:15 hrw|work: i386 as arch? not any/all? Oct 17 09:57:26 zecke: will be all Oct 17 09:57:39 dpkg-deb: building package `openembedded' in `../openembedded_1.0-1_all.deb'. Oct 17 10:02:29 hrw|work: what does that .deb do? Oct 17 10:03:11 koen: Depends: python, ccache, build-essential, patch, m4, sed, bison, wget | curl, bzip2, cvs, gawk, subversion, coreutils, quilt, unzip, texi2html, texinfo, git-core, monotone Recommends: python-psyco Oct 17 10:03:44 hrw|work: so it's more like openembedded-requirements_1.0-1_all.deb Oct 17 10:03:48 http://gcov.php.net/ Oct 17 10:03:52 koen: yep Oct 17 10:03:55 this looks interesting Oct 17 10:13:48 hrw|work: nice Oct 17 10:14:59 hrw: consider calling it openembedded-essential.deb Oct 17 10:15:17 hrw: conforming to "build-essential" in debian (alike) OSs Oct 17 10:15:42 gnu-openembedded-essential Oct 17 10:16:41 ok Oct 17 10:22:28 03mickeyl 07org.oe.dev * rc7455782... 10/ (1 packages/xoo/xoo_svn.bb): xoo svn add missing dependencies to gtk+ and libglade Oct 17 10:24:28 wb pb_ Oct 17 10:25:08 RSS works again :) Oct 17 10:26:56 RP,zecke: did you get my mail about bitbake versioning problem? Oct 17 10:27:08 mickeyl: Yes. It looks very broken :-( Oct 17 10:27:21 ok, thanks for confirming that. i started to worry about it being a local thing here Oct 17 10:27:33 hi rp Oct 17 10:27:41 mickeyl: I haven't reproduced but its certainly a bug Oct 17 10:28:40 zecke: what's the state of making inherit behave like require? Oct 17 10:28:53 zecke: are we still in discussion mode or should we just do it(tm)? :) Oct 17 10:29:15 mickeyl: My personal opinion is do it Oct 17 10:30:34 mickeyl: One question - when you say that gtk+ issue, what was the PREFERRED_VERSION of gtk+ ? Oct 17 10:30:42 s/say/saw/ Oct 17 10:31:05 RP: it wasn't set Oct 17 10:31:22 mickeyl: You're 100% sure about that? Oct 17 10:31:27 hmm Oct 17 10:31:31 need to check Oct 17 10:31:37 *booting workstation* Oct 17 10:31:38 mickeyl: If PREFERRED_VERSION was 2.8.x, bitbake would be doing the right thing in its mind Oct 17 10:31:59 Although we probably should error in that case, at least until we get packaged staging Oct 17 10:32:09 hmm, you _could_ be right Oct 17 10:32:37 it's possible that forced gtk+ 2.10.3 to be built initially by bitbake gtk+-2.10.3 Oct 17 10:32:43 and on the 2nd run it "upgraded" to 2.8.x then Oct 17 10:32:48 let me try to reproduce Oct 17 10:33:00 s/that/that I/ Oct 17 10:35:51 ehm Oct 17 10:35:54 hah Oct 17 10:36:01 gtk+2.10.3 contains DEFAULT_PREFERENCE=-1 Oct 17 10:36:12 that explains it Oct 17 10:36:59 mickeyl: right. Its a confusing message though. "upgrading" sould really be "replacing" Oct 17 10:38:28 openembedded-essential_1.0-1_all.deb Oct 17 10:49:33 Hi! Oct 17 10:49:52 hey psokolovsky Oct 17 10:50:11 koen: Sorry about h5xxx stuff. About overrides I didn't think at that time ;-(( Oct 17 10:50:48 koen, but looking at that it uses 2.6 kernel which is not yet production ready IIRC, I thought that at least no distro uses it currently. Oct 17 10:51:20 koen, I'll post RFC to the list and prepare more careful patch, and test it. Oct 17 10:51:31 hrw|work: master of web? how do I write a simple query (without form)? Oct 17 10:52:01 hrw|work: I want a user to enter a name in a textbox and then select query Oct 17 10:55:56 ~botmail for psokolovsky: http://angstrom-distribution.org/unstable/images/20061012/h5xxx/ Oct 17 10:56:39 ~lart rpm-howto for 1999 Oct 17 10:56:39 * ibot slaps rpm-howto around with a large trout for 1999 Oct 17 10:56:44 zecke: you need form Oct 17 10:56:53 hrw|work: or javascript? Oct 17 10:57:02 hrw|work: can I format an hrw|work: If I use form, I will handle post, then I can create a really complex form Oct 17 11:00:51 03koen 07org.oe.dev * r015d8066... 10/ (1 conf/machine/hx4700.conf): hx4700.conf: hx4700acx, not hx4700acx Oct 17 11:02:42 zecke: form can be get Oct 17 11:02:48
Oct 17 11:05:47 I wonder how people are able to use rpm to build sourceless packages Oct 17 11:06:10 sourceless? Oct 17 11:06:11 hrw|work: probably by using cpio directly Oct 17 11:06:19 mickeyl: pure metapackages Oct 17 11:06:24 ah Oct 17 11:07:39 you need to provide Source0: Oct 17 11:07:41 you need to provide Patch0: Oct 17 11:08:21 etc.. Oct 17 11:10:40 stupid udhcpc. DoSes me because ISP's router gone mad Oct 17 11:10:47 koen: ok Oct 17 11:10:54 Hmm - OE is no longer buildable on Debian Sarge. Oct 17 11:11:02 eno: Couldn't find package git-core Oct 17 11:11:28 why is OE sanity requiring git now? Oct 17 11:11:43 koen: Btw, could you please still post/refresh summary of EABI support on non-armv5 devices? http://lists.linuxtogo.org/pipermail/openembedded-devel/2006-October/000584.html Oct 17 11:12:29 psokolovsky: gcc is still broken, since we are still using the same version Oct 17 11:12:53 koen: updating gcc would fix stuff? Oct 17 11:13:02 * rwhitby uses http://backports.org/debian/pool/main/g/git-core/git-core_1.4.1.1-1~bpo1_i386.deb Oct 17 11:13:05 zecke: no, fixing gcc would fix it Oct 17 11:13:12 koen, But will armv4 (thumb-less in general) devices support EABI? Oct 17 11:13:28 psokolovsky: only in theory Oct 17 11:13:43 psokolovsky: armv4t is as low as EABI goes Oct 17 11:13:54 psokolovsky: http://wiki.debian.org/ArmEabiPort Oct 17 11:14:46 koen: Thanks, rereading it. Oct 17 11:16:07 ~lart rpm Oct 17 11:16:07 * ibot crushes rpm with a full height scsi disk Oct 17 11:16:24 looks like newest rpm manuals are from 2002 Oct 17 11:17:48 Anyway, has anyone used OE to build for Samsung S3Cxxx CPUs? I bet yes, as there's h1940 for example. But no tune-*.inc for it. Anyone has an idea? Oct 17 11:18:19 hrw|work, last I heard, RPM is unmaintained, lol Oct 17 11:18:42 psokolovsky: http://www.openembedded.org/bonsai/query/description/s3c/ Oct 17 11:19:27 yeah, too little Oct 17 11:19:42 psokolovsky: http://www.openembedded.org/bonsai/query/description/smdk244/ Oct 17 11:20:02 koen: hmm why is RSS broken again :} Oct 17 11:20:21 zecke: no idea Oct 17 11:20:38 zecke: I didn't change anything Oct 17 11:20:51 koen: hehe ;) Oct 17 11:20:59 koen: RSS for desc is not implemented Oct 17 11:21:00 psokolovsky: me Oct 17 11:21:08 has anyone worked with U-Boot before ? Oct 17 11:21:10 psokolovsky: linux-smdk2440.bb is in there Oct 17 11:21:10 XorA: Hello, any hints regarding gcc -m flags for S3Cxxx CPUs? Oct 17 11:21:26 XorA: I see, nice! Oct 17 11:21:52 psokolovsky: -march=armv4t -mtune=arm920t Oct 17 11:22:08 like tune-arm920t does? Oct 17 11:22:17 koen: tune-arm920t is wrong Oct 17 11:22:23 i am trying to develop loader.bin and U-boot.bin for AT91rm9200 from a linux-x86 PC. Can anyone guide me ? Oct 17 11:22:31 can we fix tune-arm920t then? Oct 17 11:22:33 Failed to start message bus: Failed to read directory "/space/kde/install//etc/dbus-1/${prefix}/share/dbus-1/services": Oct 17 11:22:36 XorA: fix it :) Oct 17 11:22:37 dbus is so stupid Oct 17 11:22:47 XorA, ok, thanks! ok to create tune-s3cXXX.conf or something? Oct 17 11:23:06 psokolovsky: I shall commit fixed tune-arm920t I had forgotten I fixed it locally Oct 17 11:23:25 XorA, ok. So, use it then, right? Oct 17 11:25:43 psokolovsky: you on the smdk2440 board? Oct 17 11:26:09 XorA, nope, my whole interest is in consumer devices, like pocketpc's ;-) Oct 17 11:26:42 XorA, I'd like to annotate h1940.conf with needed optmizations and add rx3000.conf. both are ipaq models on S3Cxxx. Oct 17 11:26:43 psokolovsky: just pushed that fix Oct 17 11:26:50 XorA, thanks Oct 17 11:31:00 whhich Ipaq would Zaurus 3200 by most like in terms of physical size Oct 17 11:32:16 ~lart yum Oct 17 11:32:17 * ibot gives yum a good seeing to Oct 17 11:39:42 XorA, if I kept in my cache what Z 3200 is... ;-) lemme look up. Oct 17 11:40:48 XorA, well, no ipaq like that at all. no modern clamshell models with kbd from known US or EU based brands. Oct 17 11:41:21 psokolovsky: No, I just need to know will my Z fit in a Ipaq sized hole when closed :-) Oct 17 11:41:27 XorA, of VGA ipaqs and pocketpc, only hx4700 has good kernel support (and yes, it's one of most active pocketpc ports at all) Oct 17 11:42:11 XorA, gotcha! Any ipaq would make Z-brick in terms of size and weight ;-P Oct 17 11:42:40 * XorA lives in scotland and wants waterproof case Oct 17 11:43:36 <[lala]> hi guys Oct 17 11:43:36 XorA, coolest pocketpc device now (gsm-less) is acer n311, which packs VGA in h1900/h4000 case (that's small and cute). it's S3 too, btw Oct 17 11:44:18 <[lala]> how can i debug what exactly the fetcher is doing? i have the problem, that the zlib_1.2.3.bb doesn't use one of the mirrors for downloading :-( Oct 17 11:44:45 <[lala]> that's pretty bad for me since i don't have any outside network connection on my build machine Oct 17 11:45:57 psokolovsky: Acer not good enough, also i has FS C550 - it is good, but hx4700 best. Oct 17 11:46:31 XorA: 'wants waterproof case' - like this http://www.teleology.ru/files/hx4700/cover01.png :) Oct 17 11:47:03 den-ros: not quite as cyberpunk/subsea looking as that Oct 17 11:47:10 den-ros: :-D Oct 17 11:47:21 den-ros, lol. acer should be good, but not with wince ;-). It has GoForce chip at least ;-). FS expensive, hx4700 is brick. anyway, OT here. Oct 17 11:48:00 XorA: But if you falldown from bicycle - hx4700 did not broke :P Oct 17 11:48:13 den-ros, I bet, that's some palm T inside, and for sure not hx ;-) Oct 17 11:49:04 psokolovsky: I hear, that this casemodel support hx4700. Oct 17 11:50:19 psokolovsky: waite a little, i find an article, where explain, why Acer not good enough even without WinCE. Oct 17 11:51:07 den-ros, yeah, just picking on ;-). Well-known cases anyway. XorA: maybe "aquapac" cases will be good for you, ask google. Oct 17 11:53:45 XorA: iirc the 3200 is roughly the size of 2 stacked hx4700s Oct 17 11:55:22 03xora 07org.oe.dev * r2acba65c... 10/ (1 conf/machine/include/tune-arm920t.conf): Oct 17 11:55:22 conf/machine/include/tune-arm920t.conf : fix so that -march and -mcpu arent Oct 17 11:55:22 confilivting. THis caused all -Werror compiled programs to fail. Oct 17 12:06:04 <[lala]> can noone comment on my mirror question above? Oct 17 12:09:30 [lala]: zlib fetch from zlib.net which does not has mirrors defined iirc Oct 17 12:13:56 <[lala]> hrw|work: shouldn't the fetcher always try at least the fallback mirrors, e.g. oesources.org? Oct 17 12:14:38 <[lala]> it works perfectly for instance for qemu Oct 17 12:14:42 [lala]: oesources.org is defunct :) Oct 17 12:14:57 [lala]: but yes, see the last two lines in MIRRORS in base.bbclass Oct 17 12:15:16 Koen: Can you comment: "Thumb interworking requires every return and indirect function call execute BX instruction (or LDR/LDM on armv5t) to set the core to the correct state." I guess, ity's typo there, should be "LDR/LDM on armv4t", right? Oct 17 12:15:44 <[lala]> zecke: well, it doesn't even TRY to use the last two lines - that's my point actually! Oct 17 12:16:13 oh well they wouldn't work anyway :) Oct 17 12:16:19 psokolovsky: ehm, I'm an arm asm noob Oct 17 12:16:27 ok Oct 17 12:16:31 feel free to take a look at lib/bb/fetch/wget.py Oct 17 12:16:34 <[lala]> zecke: i have a local mirror for the stuff and i want bitbake to use it Oct 17 12:16:44 <[lala]> but it doesnt in all cases as it seems Oct 17 12:17:05 <[lala]> i have, but i don't speak python very well and i don't understand the code entirely Oct 17 12:17:28 [lala]: add 'print's until you know what is going on :) Oct 17 12:17:45 <[lala]> i give it a shot Oct 17 12:19:13 :) Oct 17 12:19:26 I'm still busy flashing devices Oct 17 12:27:28 re Oct 17 12:28:19 wb woglinde Oct 17 12:29:53 hey woglinde Oct 17 12:30:18 he Oct 17 12:41:45 03koen 07org.oe.dev * r8b3a67cc... 10/ (9 files in 7 dirs): gpe-scap: replace gpe-screenshot and update metadata accordingly Oct 17 12:50:00 koen: you removed gpe-screenshot from metadata? Oct 17 12:50:24 yes Oct 17 12:50:33 why? maybe I like it... Oct 17 12:50:44 gpe-scap does exactly the same Oct 17 12:50:58 but with less bugs Oct 17 12:52:27 gpe-screenshot isn't allowed to be part of gpe anymore, so the gpe people forked it and fixed some bugs Oct 17 13:03:04 http://www.hrw.one.pl/2006/10/17/rpm-is-really-poorly-maintained/ Oct 17 13:04:59 Did anyone get mail from one of the hh.org lists lately? Oct 17 13:06:25 H Oct 17 13:06:31 Hello everyone Oct 17 13:06:38 hi scx Oct 17 13:06:51 scx: were you invited by Liam? Oct 17 13:07:10 Yes Oct 17 13:07:17 scx: cool, I am Graeme, I work with him Oct 17 13:07:20 lrg: wake up Oct 17 13:07:46 Hello Graeme Oct 17 13:07:47 scx: he is talking to someone at moment, Ill give him a nudge Oct 17 13:09:27 florian_kc, seems, hh.org lists are stuck again. couldn't catch france on irc too. Oct 17 13:09:52 hey wookey Oct 17 13:10:19 hey Stanley Oct 17 13:10:42 hey scx Oct 17 13:10:54 morning Oct 17 13:11:03 scx: is the mmx2 patch now in gcc ? Oct 17 13:11:49 lrg: O...yes. Actually it is in binutils. Oct 17 13:12:02 scx: stanley from pxa-linux.sf.net? Oct 17 13:12:11 yes. Oct 17 13:12:25 I and my team are building the project Oct 17 13:12:38 scx: ah, is it in an official release yet ? Oct 17 13:13:07 psokolovsky: nasty... Oct 17 13:13:21 I can double check it. A codesourcery guy helped us to do it Oct 17 13:14:49 what is the mmx2 patch? Oct 17 13:14:56 scx: have you tried building MPlayer with it ? Oct 17 13:15:32 Crofton: new mmx instruction support for ARM Oct 17 13:15:45 This interests me Oct 17 13:15:58 scx: did you test lennert's iwmmxt patch yet? Oct 17 13:16:03 mmx2 is Intel's Wireless Multimedia accelerator coprocessor Oct 17 13:16:19 You may hear of MMX in IA world Oct 17 13:16:28 yeah Oct 17 13:16:45 iwmmxt kernel patch? Oct 17 13:16:48 My real job involves software defined radio Oct 17 13:16:53 scx: yes Oct 17 13:17:24 is there a web page describing the mmx2 work? Oct 17 13:18:05 koen: that patch on lkaml do we have that? Oct 17 13:18:38 XorA: If I read it right, that patch should be part of glibc 2.4 Oct 17 13:19:02 koen: not yet. Fengwei is doing it. Oct 17 13:19:28 We published the mmx2 spec. You can find the link in pxa-linux.sf.net Oct 17 13:19:33 scx: any idea when xscale3 documentation will become public? Oct 17 13:19:47 koen, 2.5 you mean? Oct 17 13:20:00 psokolovsky: right, sorry Oct 17 13:20:07 I am working hard to make it happen Oct 17 13:20:22 scx: thanks Oct 17 13:20:56 scx: we could probably do some OE recipies for Monahans when it's public :) Oct 17 13:20:57 No problem. I am going to work for Marvell, which acquired our department. ;( Oct 17 13:21:15 lrg:Yes, absolutely. Oct 17 13:21:54 scx: I can make a start to some in the near future, as soon as my bsp arrives :) Oct 17 13:22:07 lrg: I am working to make our team more friendly and open to the open source community. Oct 17 13:22:31 scx: your doing a good job. :) Oct 17 13:22:44 Monahans, you know, now PRQed. the product name is PXA320. Oct 17 13:23:26 scx: ah, it's been called several things in the past, i.e. pxa280, pxa290, pxa300, Oct 17 13:24:11 yes. :D. a lot of changes there. Oct 17 13:24:13 A320, the Airbus that falls out of the sky :-) Oct 17 13:24:25 heh Oct 17 13:24:31 hehe Oct 17 13:25:21 koen: btw, there are some mmx2 docs on the pxa linux page Oct 17 13:26:26 scx: did you speak with RP about the new MMC drivers yet ? Oct 17 13:27:11 koen: when we upgraded our tool chain from GCC 3.4.3 to GCC 4.x, we found some calling conventions are changed. Oct 17 13:27:21 lrg: no. Oct 17 13:27:31 RP: hello Oct 17 13:31:50 XorA: http://sources.redhat.com/cgi-bin/cvsweb.cgi/ports/ChangeLog.arm?rev=1.53&content-type=text/x-cvsweb-markup&cvsroot=glibc Oct 17 13:35:14 scx: I have EABI + iwmmxt running quite nice on a pxa270: http://www.openembedded.org/~koen/hx4700.JPG Oct 17 13:35:30 XorA: we could add that patch for glibc 2.4, make glibc 2.4 use the ports/ from 2.5, or use 2.5 :) Oct 17 13:36:03 scx: do you or Fengwei have any plans to attend FOSDEM next year in Brussels ? Oct 17 13:36:24 lrg: what's FOSDEM? Oct 17 13:37:14 scx: http://www.fosdem.org/ - there should be a lot of pxa hackers in attendance Oct 17 13:37:36 although the site is not being very helpfull right now :( Oct 17 13:37:44 when is it? Oct 17 13:37:53 februari Oct 17 13:38:24 the fosdem embedded room dude is in #oe as well: Philippe Oct 17 13:38:36 any idea what dates? Oct 17 13:39:33 iirc 24 and 25 februari Oct 17 13:39:35 Crofton: iirc it was late Feb, and over a weekend Oct 17 13:40:02 scx: hi! Oct 17 13:40:09 RP: hi Oct 17 13:41:22 lrg: I can try to get a VISA to Swiss. Tomorrow I have to go to AMchant for US VISA. :( Oct 17 13:42:37 RP: we made a little bit enhancment on MMC infrastructure. I am talking with Pierre Ossman. Oct 17 13:42:44 Crofton: FOSDEM is listed in http://openembedded.org/ical/oe-events.ics :) Oct 17 13:43:29 scx: For reference, I've done a fair bit of work on the pxa2xx support in mainline and maintain the Sharp Zaurus (pxa) kernels in mainline linux. I've been watching the new chip with interest Oct 17 13:43:38 RP: I would like to invite you (of course, all guys in the channel, in the community) to review the design and code. and give us some feedbacks. We try to push it the kernel. Oct 17 13:45:02 scx: I assume it should also work on pxa25x and pxa27x ? Oct 17 13:45:52 lrg: I ported to pxa27x. Oct 17 13:46:13 pxa27x hardware is capable of 4bit sd? Oct 17 13:46:57 scx: I was part of the group that got SD support into the mainline kernel. I presume the link you mentioned on LAKML is still the most up to date reference? Oct 17 13:47:02 for FOSDEM, looks like I would have to wait until last minute to see if airfare is sane Oct 17 13:48:17 hrw|work: yes, the pxa27x hardware can do 4 bit Oct 17 13:48:55 hi koen Oct 17 13:49:28 hey Professor mccarthy Oct 17 13:50:09 RP: I will update it this week.We got better block device driver. ;) Oct 17 13:51:52 koen, not in #gpe anymore these days? Oct 17 13:52:14 nope Oct 17 13:52:24 hmm, how hard to get from Amsterdam to Brussels? Oct 17 13:52:25 although florian keeps pushing me to join it :) Oct 17 13:52:49 Crofton: train to rotterdam and then straight to brussels with another Oct 17 13:53:12 * lrg and Xora will fly to Brussels Oct 17 13:53:13 about how much? Oct 17 13:53:44 anyone know what MACHINE I should be using to build familiar-unstable for an h5xxx? I tried, surprisingly, "h5xxx", but it seems to insist on building a 2.6 kernel; then I remembered that familiar always releases jffs2 images for the h5xxx that are *named* h3900 (so I presume I need to use that MACHINE, but I would guess I need to change the necessary modules) Oct 17 13:54:00 RP: About SD spec, the SDA says “any implementatioon may require a license from SDA" Oct 17 13:54:15 Crofton: actually it's only one train :) Oct 17 13:54:28 seems pb has disappeared as well; related incidents? or coincidence? Oct 17 13:54:45 RP: English is not my mother Language. I can not understand their disclaimer and notice very well. Oct 17 13:55:13 pb ate koen and fell ill Oct 17 13:55:19 scx: I can't see how they could enforce that without patenting it and its not patented (otherwise we'd actually know how it works!) Oct 17 13:55:33 Crofton: €66.70 for amsterdam -> brussels Oct 17 13:55:57 ~change 234 pln to eur Oct 17 13:56:02 234.00 Zloty (PLN) makes 60.2111 Euro (EUR) (from http://www.xe.com/) Oct 17 13:56:18 Crofton: including the trip back Oct 17 13:56:27 hmm Oct 17 13:56:38 scx: Also, at least some of the specifications were opened up Oct 17 13:56:50 that is about $700, wait, the dollar isn't that bad, yet Oct 17 13:57:07 60 EUR sounds nice... Oct 17 13:57:09 must finish thesis first Oct 17 13:57:18 66 eur to usd Oct 17 13:57:20 ~change 67 eur to usd Oct 17 13:57:24 67.00 Euro (EUR) makes 84.0634 United States Dollar (USD) (from http://www.xe.com/) Oct 17 13:57:44 how about changing eur to RMB Oct 17 13:57:50 brussels also has international airport iirc Oct 17 13:57:51 last time I was in brussels 67 eur was about 67 usd :( Oct 17 13:58:00 yeah, but flights are bad for me Oct 17 13:58:55 scx: 1 pound = 12 RMB iirc, so 1 eur is probably about 18 RMB Oct 17 13:59:02 poznan -train-> warsaw -flight-> brusells and back for me Oct 17 13:59:07 ~change 67 eur to rmb Oct 17 13:59:12 67.00 Euro (EUR) makes 21,061.81 Dollar (ZWN) (from http://www.xe.com/) Oct 17 13:59:43 * koen tries to remember the trick to convert to rmb Oct 17 13:59:48 rmb? what currency it is? Oct 17 13:59:58 China currency Oct 17 14:00:08 ~change 67 eur to CNY Oct 17 14:00:13 67.00 Euro (EUR) makes 664.502 China Yuan Renminbi (CNY) (from http://www.xe.com/) Oct 17 14:00:30 well, not too bad. Oct 17 14:00:44 * den-ros :) (ibot great financial) Oct 17 14:01:41 RP: Things may be easier after we move to Marvell... Oct 17 14:01:57 scx: let's hope so. Oct 17 14:02:19 is there a call for papers for FOSDEM? Oct 17 14:03:20 Crofton: should be Oct 17 14:03:34 Crofton: OE plans to have atleast two presentations Oct 17 14:04:50 Are you interested in suspend-to-disk implementation on ARM? Oct 17 14:05:22 scx: some zaurus have disks Oct 17 14:05:23 I has someone who asked about it recently Oct 17 14:06:58 scx: suspend-to-ram has been much more common on handhelds although I can see the attraction of suspend-to-disk Oct 17 14:07:10 A guy from Sony did some work on suspend-to-disk and boot-from-snapshot. Oct 17 14:07:42 I hope it could be helpful to speed up the boot up process. Oct 17 14:08:26 suspend-to-disk let the kernel store and restore the memory. But in handheld, we could let boot loader do it. Oct 17 14:09:56 hrw|work and scx: That someone was me. And I just posted a message to oesf this morning: http://www.oesf.org/forums/index.php?showtopic=21752 Oct 17 14:10:03 more interesting would be a pre-booted OZ image Oct 17 14:10:15 first flash takes seconds to boot Oct 17 14:10:28 www.linuxsymposium.org/2006/linuxsymposium_procv2.pdf Oct 17 14:12:26 scx: nihao ma? Oct 17 14:12:34 scx: I think you are Chines, right? Oct 17 14:12:40 XorA: that only leaves small problem with dropbear, but we can work around that Oct 17 14:12:45 Laibsch: yes. Wo hen hao. Oct 17 14:16:32 RP: The attraction of suspend-to-disk is being able to change battery for example. Oct 17 14:17:46 It could boot to the GUI in a step. very quick Oct 17 14:17:55 scx: Are you considering to work on suspend stuff? Oct 17 14:18:14 As you can see from the posted link, I would be very interested in this. Oct 17 14:18:57 Laibsch: yes, I do. Oct 17 14:19:03 scx: Just to introduce myself. Wo she Rolf (but that is about as much Chinese now as I have learned and remembered in the ten days intensive course) Oct 17 14:19:14 scx: Great. I would be very interested. Oct 17 14:19:23 But apparently for different reasons. Oct 17 14:19:30 * Laibsch continues introduction. Oct 17 14:19:41 I am sort of "the Japanese guy" here. Oct 17 14:19:46 My nationality is German. Oct 17 14:22:45 But my interest stems from learning the Japanese language -> http://gakusei.sf.net and the equivalent for China http://xuesheng.sf.net Oct 17 14:22:45 Nice to meet you, Laibsch. My introduction could be short. I am Stanley. You can reach me at stanley.w.cai@gmail.com. My nationality is China. Oct 17 14:23:11 I tried over some time to get Japanese input methods into OE. Oct 17 14:23:21 So I guess this would be interesting for you as well. Oct 17 14:23:39 uim is already in OE Oct 17 14:24:10 CJKV, right? Oct 17 14:24:15 right Oct 17 14:25:31 But for me this is so far only related to Japanese. Just starting with Chinese. But I wanted to make sure scx know about the state of Chinese support in OE (close but not yet there). Oct 17 14:26:00 Wow, OE is really making inroads in Asia now ;-) Oct 17 14:26:39 hehe Oct 17 14:28:03 Laibsch: yes, and to prepare use zecke choose a real nice vietnamese restaurant on saterday :) Oct 17 14:28:45 Hi guys. I'd better have some sleep. It is 11pm here. Oct 17 14:28:48 Have a good day! Oct 17 14:29:21 scx: cu Oct 17 14:39:04 hrw|work: is makedevs included in your openembedded-essentials? I just ran into a missing makedevs for some reason. Does OE expect this to be available on the host? Oct 17 14:43:07 Crofton: http://lists.openembedded.org/pipermail/openembedded-devel/2006-October/000622.html Oct 17 14:44:20 hmm Oct 17 14:44:32 Looks like I need to adjust my list subscriptions Oct 17 14:49:12 hmm Oct 17 14:49:23 maybe not, why didn't I see this ... Oct 17 14:50:41 just as a nte, it looks like he tried to compile the kernel without the dsp gw enabled Oct 17 14:51:04 but that is enabled in the defconfig in the 2.6.x+git bb file Oct 17 14:54:05 re Oct 17 14:54:08 likewise: will add Oct 17 14:55:32 likewise: I do not have makedevs on my debian box - OE one is used Oct 17 14:55:55 hrw|work, one of us did commit the "stable" kernel version, right? Oct 17 14:56:41 koen: Too bad I missed the restaurant ;-) Any vietnamese around? Oct 17 14:57:00 not that I know off Oct 17 14:57:28 anyone got time for some really dumb questions? Oct 17 14:57:37 Crofton|home: you even booted it iirc Oct 17 14:57:42 yeah Oct 17 14:57:49 and mounted the cf drive Oct 17 14:58:05 I dropped him an email, hopefully he shows up here and I can check his config Oct 17 14:58:36 I will be happier when I see some people reproduce my results :) Oct 17 14:59:14 Crofton|home: rebuilding it now.. Oct 17 15:00:00 NOTE: package linux-omap1-2.6.18+git-r2: task do_unpack: started Oct 17 15:00:00 NOTE: Unpacking /home/hrw/devel/sources/dl_dir/git_source.mvista.com.git.linux-omap-2.6.git_c6051183c597b6a0fa73cdb59aac852c6148c5b6.tar.gz to /home/hrw/devel/build/celinux-test/tmp/work/omap5912osk-linux/linux-omap1-2.6.18+git-r2/ Oct 17 15:00:16 ah, the mailinglist post had 2.6.x-git Oct 17 15:00:22 yeah Oct 17 15:00:26 I am wondering why Oct 17 15:00:35 stale repo probably Oct 17 15:00:35 mind you the git one should have built Oct 17 15:00:38 yeah Oct 17 15:00:53 he said he updated, but I suspect a problem Oct 17 15:01:13 Crofton|home: 2.6.x+git should be DEF_PREF -1 Oct 17 15:01:25 ok Oct 17 15:01:29 in the bb file? Oct 17 15:01:33 yes Oct 17 15:01:41 DEFAULT_PREFERENCE = "-1" Oct 17 15:01:48 to not be selected Oct 17 15:01:48 I'll look up hoe to do that after I cross the street Oct 17 15:02:05 I'll wait 3 minutes before rebuild Oct 17 15:02:30 the openomap distro file calls for the "stable" git Oct 17 15:02:55 bbiab Oct 17 15:04:38 should someimage DEPEND on "mtd-utils" if someimage.bb uses jffs2? Oct 17 15:05:24 likewise: we have something like that Oct 17 15:05:56 conf/bitbake.conf:IMAGE_DEPENDS_jffs2 = "mtd-utils-native" Oct 17 15:09:53 parsing time kills me Oct 17 15:10:40 * chouimat decides to clean his office ... Oct 17 15:10:53 if I don't do it now I will never do it :) Oct 17 15:11:31 would someone mind helping me set up OE I am having some problems with monotone not liking the OE.mtn file. Oct 17 15:12:13 TheFool: monotone 0.30? Oct 17 15:12:21 hrw|work: yep Oct 17 15:12:22 TheFool: try reading the 'error' message mtn gives you Oct 17 15:12:26 they are quite clear Oct 17 15:12:41 TheFool: and it tells you to migrate database? Oct 17 15:12:52 literacy skills seem to be absent in the embedded community Oct 17 15:12:57 koen: yea I know it tells me to try and use migrate, saddly I haven't been able to figure out the syntax for that, pathetic I know Oct 17 15:13:16 TheFool: grab 0.30 snapshot Oct 17 15:13:54 hrw|work: thanks, I over-wrote DEPENDS right after inherit image_ipk (= instead of +=). Oct 17 15:15:07 hrw|work: as far as I know I am using the latest snapshot Oct 17 15:16:06 TheFool: default snapshot is for <0.30 Oct 17 15:16:23 TheFool: there is also >=0.30 snapshot on /snapshots/ dir Oct 17 15:17:03 can I ask bb to build a specific version of a b file? Oct 17 15:17:17 Crofton|home: bitbake package-version Oct 17 15:17:33 -version as opposed to _version? Oct 17 15:17:48 Crofton|home: yes Oct 17 15:17:53 NOTE: package linux-omap1-2.6.18+git-r2: task do_build: completed Oct 17 15:21:19 cu Oct 17 15:22:49 koen: I guess you were right, my litteracy skills must have been lacking I kept forgeting to add the -d flag when I tried to migrate the db. Oct 17 15:23:03 maybe just typing ... Oct 17 15:24:29 NOTE: package linux-omap1-2.6.x+git: completed Oct 17 15:25:11 Is there a reason why we can't simple only publish one snapshot for mtm 0.30? Or is that buggier than previous releases? Oct 17 15:25:44 that would force everyone to upgrade Oct 17 15:26:54 upgrade their mtn binaries Oct 17 15:27:04 I ahve ine machine at 29 and the other at 30 Oct 17 15:27:24 > Military intelligence is a contradiction in terms. Oct 17 15:27:24 > -- Groucho Marx Oct 17 15:27:24 So what? So's George Bush! Oct 17 15:27:28 largely because I am too lazy to work around a boost version problem on the older machine Oct 17 15:27:41 do not get me started on Bush Oct 17 15:27:52 Crofton hahaha Oct 17 15:28:11 I had to go to Canada to find "Impeachmints" Oct 17 15:28:14 Crofton: yeah, that would it. Is there a reason not to upgrade to mtm 0.30 ? Oct 17 15:28:22 schurig, not that I know off Oct 17 15:28:33 other than it needs a new boost library Oct 17 15:28:56 Crofton: didn't noticed this with boost, apt-get handled that for me :-) Oct 17 15:28:58 I think people are reluctant to force an upgrade on people at the moment Oct 17 15:29:55 the other was is to name all snapshots according to the MTM version, e.g. not just "snapshot.bz2" but "snapshot-for-mtn-0.29.bz2" (or similar, don't know the proper name be heart) Oct 17 15:30:14 it's just that in the last two days that I'm in #oe I have seen one person each day using the wrong snapshot Oct 17 15:30:33 djeez Oct 17 15:30:39 schurig joins the illiterate Oct 17 15:30:42 and the error message "migrate db" might be "right", but a better way would be to use the correct snapshot in the first place (less time consuming) Oct 17 15:30:43 * koen points to http://www.openembedded.org/snapshots/ Oct 17 15:31:04 koen: yeah, get rid of OE.mtn.bz2 Oct 17 15:31:20 koen: several people in the last two days downloaded the wrong one Oct 17 15:31:32 there is no 'wrong one' Oct 17 15:31:35 koen: actually, not "get rid", but "rename" Oct 17 15:32:01 koen: there is a "wrong one". If you download an older and need to migrate db and re-rosterify, you need more time compared to the correct download Oct 17 15:32:10 "wrong" in quotation marks Oct 17 15:32:15 * koen points to http://www.openembedded.org/user-manual&dpage=ch03s02 Oct 17 15:32:33 koen: but you're king and you always see red whenever i type something Oct 17 15:32:36 It sounds like the getting started instructions need some clarification? Oct 17 15:32:57 OE.mtn.bz2 will always contain the latest OE metadata, regardless of monotone version Oct 17 15:33:45 * schurig bows down before koen and chants "hail the king" Oct 17 15:33:46 :-) Oct 17 15:34:11 complainers-not-doers Oct 17 15:34:17 so rename this to OE-this-is-for-mtn-0.28.mtn.bz2 Oct 17 15:34:35 it isn't for 0.28 Oct 17 15:34:39 it takes less time to "mtn pull" compared to "mtm db migrate" and to re-roster Oct 17 15:34:48 so? Oct 17 15:35:21 It takes less time to /ignore you, but I still try to argue Oct 17 15:35:51 koen: actually I would prefer that you /ignore me Oct 17 15:35:53 by your logic the OE website should be in german since "I say at least one german person in #oe" Oct 17 15:36:15 no, it's about preventing errors in the first case Oct 17 15:36:38 Like I said http://www.openembedded.org/user-manual&dpage=ch03s02 Oct 17 15:37:14 if you are not willing to do anything, just shut up Oct 17 15:37:27 koen: I'm not saying that the doc is WRONG, I'm saying that it can easily made more fool-proof Oct 17 15:37:54 the spec is correct, but some newbies made the same non-optimal "mistake" (in quotation marks again) Oct 17 15:38:14 and it would be easy to make the process simpler with a better speaking file name Oct 17 15:38:14 I have a better idea Oct 17 15:38:30 we delete all docs and say "schurig knows best, talk to him" Oct 17 15:39:05 bah, knock it off Oct 17 15:39:08 s/schurig/koen/ Oct 17 15:39:24 the only good that will come of this is I downloaded a static mtn-0.30 Oct 17 15:43:58 a+ Oct 17 15:45:14 anyway, I'm going home Oct 17 15:45:16 bye Oct 17 15:45:21 bye Oct 17 15:46:09 well thanks for the guidance, everything seems to be working now. Oct 17 15:47:13 good, if schurig decides to leave, I guess there is nothing much to do around here . I am hitting the road. CU Oct 17 15:47:43 * raduga yay Oct 17 15:47:53 ~emulate schurig Oct 17 15:47:55 * ibot schurig says: everything sucks, you suck, I know best Oct 17 15:48:05 ~emulate rp Oct 17 15:48:36 is RP self-virtualized? and not emulatable? Oct 17 15:48:59 raduga: nobody found the time to code this :) Oct 17 15:49:21 no he crashed at boot-up Oct 17 15:49:29 cu guys Oct 17 15:49:37 likewise: bye Oct 17 15:49:49 anyone here mess with SDL or framebuffer internals? Oct 17 15:50:47 I've been tracking some odd behavior with SDL/touchscreen interactions Oct 17 15:51:18 i thought it was application-specific bugs, but its looking more like something is borked at the SDL level, or in the TS driver even Oct 17 15:52:41 when SDL is projecting a 320x200 in QVGA, its not stretched (rather, centered in middle of field) Oct 17 15:52:58 and mouse clicks to the top and bottom margin kill the TS calibration Oct 17 15:53:27 perhaps there's some kind of exception that's not being caught and handled properly Oct 17 15:53:36 but the more I look for it... the more places I find it :) Oct 17 15:55:02 oddly, mousing to the left and right margins also seems to bork the TS; but this never seems to occur in 640x480 Oct 17 16:07:51 * den-ros 1 Oct 17 16:12:06 france: ping Oct 17 16:21:49 koen: IN the angstrom-2007.1.conf you have defines the FEED_URIS with the ${TARGET_ARCH} in the links. Is there a way to change this. Because the FEEDS are setup as "armv5te" but the TARGET_ARCH for htcuniversal and HX4700 is "arm". So with every new image i test i have to change the FEED from arm to armv5te so i can contact and download from the repos. Oct 17 16:22:21 I triedt to change my MAchine ARCH in the htcuniversal.conf file to armv5te, but it failed compiling. Oct 17 16:22:42 args Oct 17 16:22:46 whats that? Oct 17 16:22:48 glibc (2.5-r1) already staged but upgrading to 2.3.5+cvs20050627-r12 Oct 17 16:24:06 hm Oct 17 16:24:26 found Oct 17 16:24:32 workaround Oct 17 16:25:16 woglinde: I think I had that, too. Compiling for openzaurus-unstable? Oct 17 16:25:23 laibsch yes Oct 17 16:26:08 so I comment out there Oct 17 16:26:17 read bug 1421. openzaurus-unstable.conf was fixed to glibc 2.3.5 to make sure it is binary compatible. Oct 17 16:26:29 yes Oct 17 16:26:34 But the word upgrading is indeed funny ;-) Oct 17 16:26:35 that sounds reasonable Oct 17 16:26:41 yes Oct 17 16:49:17 03koen 07org.oe.dev * re9b9b873... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: use a newer binutils Oct 17 16:51:20 hm newer? Oct 17 16:51:35 2.17.50.0.1 vs 2.17 Oct 17 16:52:36 hehe Oct 17 16:57:31 03koen 07org.oe.dev * r77261f96... 10/ (1 conf/machine/navman-icn330.conf): Oct 17 16:57:31 navman-icn330: add support for the navman iCN 300 gps receiver Oct 17 16:57:31 * see http://www.duff.dk/navman/ for details about the linux port Oct 17 17:03:49 koen: hilarious ;-) Oct 17 17:05:36 psokolovsky: the navman? Oct 17 17:05:45 yeah ;-) Oct 17 17:07:16 koen, I'm looking for port, or at least, inside photos of Nokia e60 thingy too ;-) Oct 17 17:08:03 hmmm Oct 17 17:08:03 uboot seems to want kernel-headers Oct 17 17:08:05 and the kernel want mk-image from uboot Oct 17 17:08:34 Crofton: what provides mkimage? Oct 17 17:08:51 Crofton: u(-)boot or u(-)boot-native? Oct 17 17:08:57 uboot Oct 17 17:09:17 hmmm Oct 17 17:09:37 u-boot for what machine? Oct 17 17:10:03 | ../include/configs/icn330.h:38:31: error: asm/arch/pxa-regs.h: No such file or directory Oct 17 17:10:15 Crofton: navman iCN330 Oct 17 17:10:34 which version of u-boot? Oct 17 17:10:48 LABEL-38294582389758927y8rufweifj iejgf Oct 17 17:10:56 hmmm Oct 17 17:11:08 sounds like an issue with the particular machine? Oct 17 17:11:26 for omap5912osk, I can nuke tmp Oct 17 17:11:36 rebuild with u-boot happening first Oct 17 17:13:25 pretty small a port, indeed Oct 17 17:13:57 which bb file are you trying to build? Oct 17 17:14:57 let me sync some generic improvements to uboot first Oct 17 17:16:10 I would think u-boot would have to avoid depending on much of anything .... Oct 17 17:16:22 right Oct 17 17:16:38 I am re-rosterifying my db Oct 17 17:16:52 why not download a new one? Oct 17 17:16:59 you should have local changes :) Oct 17 17:17:03 heh Oct 17 17:17:10 I am wndering that myself ..... Oct 17 17:17:45 03koen 07org.oe.dev * ra8958088... 10/ (1 packages/uboot/u-boot_LABEL.2006.06.30.2020.bb): Oct 17 17:17:45 u-boot_LABEL.2006.06.30.2020: fix compiling for non-standard target prefixes (like ones including target-vender) Oct 17 17:17:45 * machine machine specific as well Oct 17 17:17:52 s/shouldn/shouldn;t/ Oct 17 17:17:57 I just looked at log for git u-boot and don;t see anything about that machine Oct 17 17:20:03 external patch Oct 17 17:21:29 koen, does that thingy really has TS and usb? the page doesn't mention that, nor anything in patches. Oct 17 17:22:02 03koen 07org.oe.dev * r59a9ba47... 10/ (4 files in 3 dirs): Oct 17 17:22:02 linux-navman: add kernel for the navman-icn330 machine Oct 17 17:22:02 * due to some circular deps no uImage is created, but a zImage Oct 17 17:22:02 * the navman's uboot wants kernel headers in order to compile, and the kernel needs mk-image from uboot in order to build the uImage.... Time for an mkimage-native recipe? Oct 17 17:22:54 psokolovsky: it has at least usb according to the navman website Oct 17 17:23:32 ok. from page, i got guy boots it from the card Oct 17 17:23:47 right, he does Oct 17 17:24:09 that's why I removed the ROOTFLASHSIZE from the machine.conf Oct 17 17:27:51 koen: How come all my modules wouldnt install anymore. I allways have to add them with make_modules install manually after makeing a image. Oct 17 17:28:02 Even acx wouldnt install in to the image. Oct 17 17:28:15 use OE to compile your kernel Oct 17 17:30:01 I am using OE to compile my kernel, but there are several modules missing. Even though they are beeing created, but they dont get installed to it. Oct 17 17:30:55 goxboxlive, machines are being converted to use new task-base stuff. some breakage and repair should be expected. Oct 17 17:31:42 goxboxlive: Same here. I ask Koen today about this. Oct 17 17:31:54 psokolovsky: Thanks. Now clear. Oct 17 17:32:00 : I know that, and also expect that, but how do i force OE to install in example acx and htcuniversal_acx in to the image? Oct 17 17:32:14 psokolovsky: the conversion isn't the problem Oct 17 17:32:19 goxboxlive, I actually see a bit of problem. Oct 17 17:32:21 psokolovsky: me making zillions of typos is :) Oct 17 17:32:30 They are both defined in the htcuniversal.conf Oct 17 17:32:31 koen, yep ;-) Oct 17 17:32:38 goxboxlive: I add needed modules by hands. Oct 17 17:33:31 den-ros: I also do that, but the strange thing is that some modules are being installed, like irda, and usbnet, but not all i want, like sound and acx Oct 17 17:34:36 goxboxlive, try now Oct 17 17:34:48 psokolovsky: another common cause is underscores in module names Oct 17 17:35:08 ok, i will. hold on Oct 17 17:35:12 * koen kicks CIA-4 Oct 17 17:35:13 ow Oct 17 17:35:50 * koen nominates psokolovsky to merge the heads Oct 17 17:35:51 goxboxlive: Yes, we talk about same - there is ununderstandable thing yet. Oct 17 17:36:42 goxboxlive:Please repeat, what machine do you use? Oct 17 17:36:52 htc universal Oct 17 17:37:05 goxboxlive: Wow cool. Oct 17 17:37:08 03pfalcon 07org.oe.dev * r53d64b20... 10/ (1 conf/machine/htcuniversal.conf): htcuniversal: Fix typo. Oct 17 17:37:13 03koen 07org.oe.dev * ra28fabd7... 10/ (1 conf/machine/htcuniversal.conf): Oct 17 17:37:13 htc-universal: fix another batch of typos Oct 17 17:37:13 * I should get my typing license renewed :) Oct 17 17:37:29 conflicts sux most ;-) Oct 17 17:37:40 And for hx4700 please :) Oct 17 17:39:03 koen, may take me some time to merge due to conflict and me working on headless box now ;-( if you can, please fix ;-) Oct 17 17:39:27 I can't since I don't understand vimdiff Oct 17 17:39:56 and my buildbox is several hundred meters away Oct 17 17:40:49 koen: lol. me too ;-) ok, working on it ;-) at least we now know that mtn works ok with concurrent commits ;-E Oct 17 17:41:10 we already knew that Oct 17 17:41:26 that's why we choose it :) Oct 17 17:49:29 koen: this - vs _ issue: is it rooted in internal names used by kernel, or ipkg just uses '-' everywhere? Oct 17 17:49:51 psokolovsky: a bit of both Oct 17 17:52:00 koen, hm, then, let htcuniversal people try that with just __ fixed, because replacing each underscore with hyphen unlikely will give good results (at least, hx4700 still has some _) Oct 17 17:52:25 check the ipkg names, it will break Oct 17 17:52:32 that's what den-ros found out today Oct 17 17:52:40 XorA|gone you had mentioned you have directfb working on your system. What processor are you using? I keep getting glibc detected *** dfbg: double free or corruption when I run directfb. This error normally means there is a memory hole? Oct 17 17:53:14 Gerrath: http://dominion.kabel.utwente.nl/koen/cms/error-messages-that-arent-very-helpfull :) Oct 17 17:53:30 goxboxlive, still listening? now its "fixed" again, please restart your build to confirm it ;-) Oct 17 17:54:19 i will, but binutils where updated, and it trake som time to compile it. Oct 17 17:54:30 psokolovsky: all modules in RRECOMMENDS need to use hyphen, now we don't use kernel_modules.bbclass anymore Oct 17 17:55:50 psokolovsky: you can verify it bu checking the control files of task-base* Oct 17 17:58:41 koen, thanks, I think :-p, well I did an strace also but I'm not real sure that its any help either.. Oct 17 17:59:41 psokolovsky: I synced to my mac and merged it with FileMerge.app Oct 17 18:00:04 koen: Thanks , but it is Philippe(pH5) found '_' (i only ask) , if talk about it. Oct 17 18:00:07 koen, no! I already merged and pushed it ;-) Oct 17 18:00:19 hello. Any clue about this ipkg error: "Clearing state_want and state_flag for pkg=... (arch_priority=0 flag=0 want=2)" ? Oct 17 18:00:46 first time I see this error (just installed a fresh compiled image) Oct 17 18:00:58 nicolasfr: usually a wrong arch Oct 17 18:01:05 koen, and after all, only hyphens must be used, and '_' are -> '-' in linux-kernel-base.bbclass Oct 17 18:02:10 koen: where does ipkg looks for archtype? Cause uname gives armv5tel and packages are armv5te... Oct 17 18:02:24 nicolasfr: /etc/ipkg/arch.conf Oct 17 18:05:00 why do I inherit from autotools? Oct 17 18:05:44 03koen 07org.oe.dev * r1d7f3627... 10/ (1 conf/machine/hx4700.conf): hx4700: remove some more hyphens Oct 17 18:12:14 koen: :) Thanks again. Oct 17 18:12:27 psokolovsky_: I think I have caught most typos from the last few days :) Oct 17 18:13:09 still the same. Oct 17 18:13:26 goxboxlive: bitbake -c rebuild task-base Oct 17 18:13:36 i'll try Oct 17 18:13:55 i cleaed handhelds-pxa before i made the gpe-image thow Oct 17 18:18:18 CSI miami time Oct 17 18:18:21 koen: ok, I had removed arch.conf ;) Oct 17 18:18:49 btw the splash screen "a la poky" is nice Oct 17 18:22:08 goxboxlive: I will get an HTC Universal in a few weeks. I can't wait to boot linux on it Oct 17 18:22:20 03pfalcon 07org.oe.dev * rb042caec... 10/ (1 conf/machine/htcuniversal.conf): htcuniversal: A bit of merge funkyness due to vimdiff braindeadness. Oct 17 18:22:34 goxboxlive: any progress on the suspend problem? Oct 17 18:28:08 hi there. Yes it is great device. But actually, the phone part suck, it is not reliable, it sucks. But it will be a good linux machine thow Oct 17 18:28:31 suspend is not been working on yet. The RE man is working on the phone part now Oct 17 18:29:13 it means that you have to reboot to linux each time you shut the device off? Oct 17 18:34:44 psokolovsky_ & Koen : Now they where installed, Both sound and acx. thx guys. Oct 17 18:34:56 goxboxlive, cool! Oct 17 18:35:17 goxboxlive, just in case, can you tell filename of sound module ipk? Oct 17 18:36:54 ?? the patch to the sound module? Oct 17 18:37:27 goxboxlive, nope, filename of ipk built for the htcuniversal sound module. Oct 17 18:37:39 ok, hold on Oct 17 18:37:51 goxboxlive, that which contains snd & htcuniversal in it Oct 17 18:39:24 kernel-module-snd-htcuniversal-audio-2.6.17-hh1_cvs(date) Oct 17 18:40:28 | /home/balister/oe_home/tmp/staging/x86_64-linux/lib/libomniORB4.so: file not recognized: File format not recognized Oct 17 18:40:37 goxboxlive, nice, thanks! Oct 17 18:40:55 the x86_64-linux is the arch of the build machine Oct 17 18:40:57 pastebin.ca/207030 Oct 17 18:41:06 koen, just in case, I'm fixing lst typo in hx4700.conf now ;-) Oct 17 18:41:14 any idea how that gets into the libtool -L paths? Oct 17 18:41:51 I have also added snd-pcm-oss and snd-pcm-mixer. I dont know if i need them, but i used to have them with the 2.6.12 kernel Oct 17 18:42:03 this is in the ossiecf bb file? Oct 17 18:42:44 goxboxlive, I'm not sure either, but i assume, that should be controlled outside machine configs, as those modules are not machine-specific. Oct 17 18:44:45 psokolovsky_: Thanks for hx4700. Oct 17 18:45:38 03pfalcon 07org.oe.dev * r517f7efe... 10/ (1 conf/machine/hx4700.conf): Oct 17 18:45:38 hx4700: Fix last typo in kernel module package name (only hyphens are allowed there, Oct 17 18:45:38 underscores separate package name from version). Oct 17 18:49:05 python-pygtk needs a RDEPENDS on python-pycairo Oct 17 18:53:02 i havent added them to htcuniversal.conf, but i have added them to the defconfig for the linux kernel. Oct 17 18:53:24 Bye all. Oct 17 18:53:51 goxboxlive, I guess that's ok. I have no idea if ansgtrom is fully alsa-based. Oct 17 18:54:02 ok Oct 17 18:57:19 alsa + oss emu most likely Oct 17 18:57:43 I'd like to have ASoC for all machines, the giorgi is being a pain about that Oct 17 19:02:32 koen, I talked with Giorgio about that too. He told he couln't achieve concord communicating with eith either RP or Liam on it. He said also that he had issues with ASoC and DPM. Oct 17 19:02:57 ASoC is going to be in mainline kernel soon Oct 17 19:04:11 I have problems booting my new kernel with the new binutils-cross Oct 17 19:04:15 yep, so he said if anyone wants to play with it now, it's nice, but he'd rather wait and see how it goes Oct 17 19:04:15 so hh-pxa is going to end up with yet another unmergable diff from mainline Oct 17 19:04:25 it says: Oct 17 19:04:25 03crofton 07org.oe.dev * r9b4f1129... 10/ (1 packages/linux/linux-omap1_2.6.x+git.bb): linux-omap1_2.6.x+git.bb : Add DEFAULT_PREFERENCE = "1". Oct 17 19:04:41 Crofton: 1 or -1 ? Oct 17 19:04:50 hmmm Oct 17 19:05:05 I swear I cut and paste from openzaurus ,,,,, Oct 17 19:05:35 Crofton: you just typod in the commit message :) Oct 17 19:05:52 yeah :) Oct 17 19:05:55 bother Oct 17 19:06:08 /bin/sh /lib/libgcc_s.so1: version 'GCC_3.5' not found (reqiuerd by /bin/sh) and the same for libc.so.6 but for GLIBC _2.1.3 Oct 17 19:06:15 just checking to see if anyone was awake Oct 17 19:06:25 and GLIBC_2.0 Oct 17 19:07:39 goxboxlive, so, what exactly says that? Oct 17 19:08:25 init version 2.86 booting Oct 17 19:09:01 goxboxlive, ok, so at least it's not kernel Oct 17 19:09:14 It might be taht i have to rebuild a lot of stuff since there where an update in the cross compiler Oct 17 19:09:32 goxboxlive, yep, that would be first thought ;-I Oct 17 19:09:54 koen, do you think that rebuild would clear that ^^^ Oct 17 19:10:52 it think so Oct 17 19:10:58 gcc 3.5 seems a bit odd Oct 17 19:11:35 koen, that's likely, some ABI version, I guess Oct 17 19:11:40 so actually it the allmost every package needs to be rebuild? Oct 17 19:12:21 goxboxlive, gcc-cross & glibc to start with ;-) Oct 17 19:12:51 meh Oct 17 19:12:59 :-) and that is thoose who takes the longest time Oct 17 19:13:12 "get some bugfixes" turns in to "watch qemu grind trough 217 locales" Oct 17 19:13:48 goxboxlive, I guess, for you, it would be better to undo version bump locally and let other to catch nuts instead ;-) Oct 17 19:14:24 ok Oct 17 19:19:24 --with-mpfr=/home/balister/oe_home/tmp/staging/x86_64-linux Oct 17 19:19:43 this is in the arm-linux-gcc -v string Oct 17 19:19:55 what exactly is mpfr? Oct 17 19:20:22 some math thing iirc Oct 17 19:20:35 yay technical details Oct 17 19:20:53 why would my cross compiler point at a x86-64-linux directory ..... Oct 17 19:21:18 because its a x86_64 binary Oct 17 19:21:25 is there some command you can give gcc to make it cough up a list of link time include dirs? Oct 17 19:22:24 Koen: I have installed gtk+_2.10.0 under Angstrom but some gtk dependant libs are defined in their bb recipe to install under /usr/lib/gtk-2.0/2.4.0 instead of 2.10.0 (libgtkinput and libgtkstylus). Is there a "nice way" to fix the .bb? Oct 17 19:22:35 http://www.mpfr.org/ Oct 17 19:22:43 nicolasfr: actually there is :) Oct 17 19:22:58 why would my arm cross compiler care about mult precision floating point .... Oct 17 19:23:21 nicolasfr: those recipes need to use http://www.openembedded.org/viewmtn/revision.psp?id=593b9e6f5ea7ab1790ec54c7599baf883f22f12b Oct 17 19:23:31 * Crofton I have a bad feeling the cross bb file needs to disable this ... Oct 17 19:23:49 Crofton: for building ada and fortran compiler iirc Oct 17 19:24:07 koen: cool! Oct 17 19:27:30 damn, and I have only binutils-2.17.50.0.5.tar.bz2 locally. will try that. Oct 17 19:28:17 nicolasfr: and the bootsplash is actually the one from poky :) Oct 17 19:28:38 nicolasfr: I'm working with mallum to fix the white background issues Oct 17 19:28:54 http://ftp.kernel.org/pub/linux/devel/binutils/release.binutils-2.17.50.0.5: "It is purely for Linux. Oct 17 19:28:54 " Oct 17 19:29:11 ... kernel, I guess. should we take that warning? Oct 17 19:29:44 Changes from binutils 2.17.50.0.3: Oct 17 19:29:44 16. Various updates for PPC, ARM, MIPS, SH, Xtensa. Oct 17 19:30:00 Changes from binutils 2.17.50.0.2: Oct 17 19:30:01 10. Improve MIPS, M68K and ARM support. Oct 17 19:31:38 psokolovsky_: we want the version with the fixes from tf, so you could add 2.17.50.0.latest Oct 17 19:31:44 how come hx4700 has dpi=200? Where is that setting? Htc universal comes out with dpi=100, and that is to small. It can be adjusted in the Xserver file, but how can i do that in OE? Oct 17 19:32:37 goxboxlive: add a patch to http://www.openembedded.org/repo/org.openembedded.dev/packages/xserver-common/xserver-common_1.10.bb Oct 17 19:32:46 thx Oct 17 19:32:53 goxboxlive: and send it to florian as well, so I can get upstream Oct 17 19:32:59 i will Oct 17 19:33:35 goxboxlive: you can get the latest version at http://projects.linuxtogo.org/plugins/scmsvn/viewcvs.php/trunk/base/xserver-common/?root=gpe Oct 17 19:39:44 koen: what would you think about making an alternative angstrom-matchbox task, which would provide a simple matchox desktop (I know I like poky, but having access to full OE feeds is much easier), without the full blown gpe? Oct 17 19:40:11 nicolasfr: sure Oct 17 19:40:26 nicolasfr: that's been the plan, just haven't had time to implement it Oct 17 19:41:15 oh nice! Oct 17 19:42:15 nicolasfr: the 'base' would be mb-wm+mb-desktop+(mbpanel or fbpanel) + gpe-conf Oct 17 19:43:18 what's gpe-conf? the default tools in settings folder? Oct 17 19:43:31 yes Oct 17 19:43:52 poky only has a theme selector last time I tried it Oct 17 19:44:00 also adding nano would save some users (those would don't like much vi for example!) Oct 17 19:44:11 koen: true, poky has no setting tools Oct 17 19:44:40 nicolasfr: my end goal is to not ship a xterminal Oct 17 19:45:39 thanks to mickeyl I can make some tests with latest pygtk and hopefully make some applets easily (wifi with wpa, card applets, ...). maybe that would be usefull (but of course would depend on python) Oct 17 19:46:08 If I can help with testing, let me know. Oct 17 19:46:34 gathering user experiences is very usefull Oct 17 19:46:46 especially now the software doesn't crash after 5 seconds :) Oct 17 19:47:40 hehe ;) Oct 17 19:48:17 a new icon theme would be nice as well Oct 17 19:48:21 tango for example Oct 17 19:49:09 exactly, tango icons are nice (much more than the gpe contact icon for eg;) Oct 17 19:49:38 I usually change to the industrial icon theme Oct 17 19:49:39 a nice background and logo would help. Any artists there? Oct 17 19:50:10 nicolasfr: see http://www.openembedded.org/~koen/hx4700.JPG Oct 17 19:50:24 that's how I usually configure my devices Oct 17 19:51:25 nice look Oct 17 19:51:51 I have a suspicion the arm compiler hosted on a x86_64 is including the x84_64 native lib from staging because of the --with-mfpr Oct 17 19:51:56 yes it looks nice, it should be like that at first boot. Oct 17 19:52:22 in any case mfpr path should be in staging/arm-linux Oct 17 19:52:49 Crofton: I'm not sure if that's true for the crosscompiler Oct 17 19:53:04 Crofton: conventional wisdoms don't work on gcc :( Oct 17 19:53:37 well, I am having a problem where it includes that directory in some link lines Oct 17 19:53:55 and this blows up the link because I have some libs there from native packages Oct 17 19:58:57 w00t, my PPCZone project was approved! Oct 17 20:00:10 kone: I saw your screenshor. Why cant you make the angstrom look like that at first boot. Oct 17 20:02:19 goxboxlive: I can, just haven't had time to sort out the patches needed Oct 17 20:02:31 Crofton: welcome to the efika club :) Oct 17 20:02:53 koen: Ok looking forward to that. Oct 17 20:03:09 Hopefully the OE support appears quickly Oct 17 20:03:27 my project is to build what I am making work on the OSK on that :) Oct 17 20:03:43 this mpfr stuff is pissing me off Oct 17 20:03:48 I think :) Oct 17 20:04:10 hi Oct 17 20:04:16 hey hrw Oct 17 20:04:29 hi Oct 17 20:04:48 koen: when I'm reading your discussions with schurig I feel that you should be in PR group. Oct 17 20:05:15 should not Oct 17 20:05:30 that would be like putting me in the technical writing group Oct 17 20:05:34 :) Oct 17 20:05:56 ~emulate schurig Oct 17 20:06:06 * ibot schurig says: everything sucks, you suck, I know best Oct 17 20:06:14 hrw: don't put me into a PR group either :-) Oct 17 20:06:26 schurig: ok;) Oct 17 20:07:05 koen: I often feel s/schurig/koen in this 'emulate' Oct 17 20:07:33 ofcourse I'm not perfect too Oct 17 20:07:52 ~oe.mtn Oct 17 20:07:52 rumour has it, oe.mtn is http://www.openembedded.org/snapshots/OE.mtn.bz2 Oct 17 20:08:25 hrw: at least we actually *do* something besides complaining Oct 17 20:08:25 ibot: no, oe.mtn is snapshot of OpenEmbedded metadata available on http://www.openembedded.org/snapshots/ as files for monotone <0.30 and for >=0.30 Oct 17 20:08:25 okay, hrw Oct 17 20:09:04 koen: then 'fix' gettingstarted with content of usermanual page? Oct 17 20:09:21 its a wiki, fix it yourself Oct 17 20:09:33 :D Oct 17 20:09:59 somehow there are people that understand "complaining" if someone merely throws in an idea ("maybe we should rename ...") Oct 17 20:10:21 anyway - time to shutdown to check how Ania's laptop works with external monitor Oct 17 20:10:36 but that is what most people understand as suggestion, to begin a discussion (not flame) about how to improve something Oct 17 20:11:23 13 slides for tomorrow - I hope that some people will come Oct 17 20:11:25 cu Oct 17 20:11:45 anyway, time to test a new kernel :-/ Oct 17 20:13:14 koen: With your hx4700 are you able to control your backlight in GPE? Oct 17 20:13:31 goxboxlive: yes Oct 17 20:13:50 thanks to one of the patches I wrote for minilite :) Oct 17 20:13:51 koen: Wich kernel version are you using? Oct 17 20:14:17 patch for the kernel or GPE? Oct 17 20:15:10 that patch has been in minilite for ~1.5 years now Oct 17 20:15:33 if your kernel uses the standard 2.6 interface for backlights, gpe can use it Oct 17 20:16:17 koen: Yes it worked in GPE until hh.org moved over to 2.6.17 Oct 17 20:16:51 goxboxlive, you should convert to corgi-bl for 2.6.17, as was RFCed on kernel-discuss@. Oct 17 20:17:48 goxboxlive, It works for me on h4000/2.6.17 and people submitted working patching for hx4700, h2200. Oct 17 20:17:56 join #htc-linux and explain for cr2. Oct 17 20:18:41 goxboxlive, ok, but tomorrow, good? ;-) Oct 17 20:18:50 ok Oct 17 20:19:35 Gerrath: its a zaurus c7x0 so arm Oct 17 20:19:46 Gerrath: under OZ 3.5.* something or other Oct 17 20:20:04 koen: any known bugs fixed by that new binutils? Oct 17 20:22:26 XorA|gone: it should fix some iwmmxt bugs Oct 17 20:22:49 XorA|gone: but I think I misread the changelogs a bit and we need 2.17.50.0.3 instead of .1 Oct 17 20:41:56 re Oct 17 21:23:02 * v8jlene is back Oct 17 21:27:13 we should add 'cu' in the OE.dev. Oct 17 21:35:42 goxboxlive: cu is a cli serial comms thingy right? Can't kermit do the same as what you'd use cu for? Oct 17 21:36:17 to communicate with the phone on htc universal Oct 17 21:49:15 florian: Will http://www.kernelconcepts.de/products/usb-cf.shtml ever become available again? (for us collie users) Any alternative suppliers of similar hardware? Oct 17 21:52:42 Laibsch: no :-( If you are really lucky there is still one somewhere... i'd need to check. Nut a'm not sure about the current driver situation... Oct 17 21:53:41 florian: I'd appreciate if you let me know. Oct 17 21:55:38 Laibsch: don't hestitate to remind me.. i really need to check tomorrow at the office. Oct 17 21:55:50 OK Oct 17 21:56:18 Can I just send you a mail to florain at kernelconcept.de? Oct 17 21:56:23 florian Oct 17 21:57:15 Laibsch: you can if you don't forget the last 's' :-) Oct 17 21:57:31 OK Oct 17 21:58:21 Laibsch: would this work: http://www.rexpccard.co.jp/english/products/subpages/cfu1u.html Oct 17 22:01:57 hvontres|poodle: I guess so. The dealer in Germany is only selling bulk it appears. Oct 17 22:04:48 :( Oct 17 22:35:50 boy that was priceless Oct 17 22:37:21 raduga: almost bash.org quality hehe Oct 17 22:43:04 ugh. Oct 17 22:43:19 i feel like trolling Cyorxamp. :/ Oct 17 22:43:25 probably a good sign that I should leave Oct 17 22:43:27 ... Oct 17 22:44:04 thanks for your help, developers. Oct 17 22:44:07 thanks *much* Oct 17 23:12:48 is it safe to use a newer mtn (0.30) with the OE db? Oct 17 23:14:21 bugs.oe.org is down Oct 17 23:19:14 mreimer, there's a DB download specifically for 0.30. or, youc an convert older DB yourself (long!!) Oct 17 23:19:31 hi psokolovsky_ Oct 17 23:19:40 mreimer, Hi! ;-) Oct 17 23:19:41 is there a recommended version to use atm? Oct 17 23:19:59 mreimer, 0.30 works very good ;-) Oct 17 23:20:34 ok, I'll upgrade. thanks psokolovsky_! Oct 17 23:20:37 mreimer, Matt, btw, wanted you to ask if you could try my common-buttonmap refactor on h2200. patches were posted to kernel-discuss. Oct 17 23:20:58 psokolovsky_: ok, I'll start a build. Does it include xmodmap patches? Oct 17 23:22:11 mreimer, yes, in the form of familiar tree patches. but it should be pretty easy to extract them to apply on already installed system. Oct 17 23:22:15 mreimer, http://www.openembedded.org/snapshots/ **** ENDING LOGGING AT Wed Oct 18 02:59:56 2006