**** BEGIN LOGGING AT Mon Mar 06 10:59:56 2006 Mar 06 11:02:20 hrw|work: hi! Mar 06 11:02:40 hrw|work: trying to keep some of my social contacts intact :-) Mar 06 11:03:10 schurig: good thinking Mar 06 11:03:29 schurig: long time no see - how things? Mar 06 11:03:46 * Bernardo wonders what are social contacts... Oh, those people I haven't seen in ages, could that be it? Mar 06 11:04:05 Bernardo: rather yes Mar 06 11:04:12 lol Mar 06 11:04:23 my english is not perfect when it comes outside of IT Mar 06 11:04:30 It feels like I haven't seen other people than my coworkers dfor ages Mar 06 11:04:38 hrw|work: it was perfect Mar 06 11:05:01 it's just that I do need to start seeing people again one of these days Mar 06 11:05:47 yep Mar 06 11:06:23 time to spend some cash and drive to some long-time-no-see friends Mar 06 11:06:48 (as it is cheaper then drive to not-yet-seen-face-to-face friends) Mar 06 11:12:28 * france is away: Away Mar 06 11:20:28 * france is back (gone 00:08:00) Mar 06 11:21:32 * france is away: Away Mar 06 11:23:22 hrw|work: I rebuild subversion for the collie, but now my friend gets a "Unrecognized URL scheme" error when he tries to connect to a repo, looks like subversion isn't shipping with the https plugins? Mar 06 11:23:35 dont know Mar 06 11:28:03 hrw|work: looks like a ld.so.conf problem, he says his only has /usr/local/lib, and subversion's shared libs are in /usr/lib Mar 06 11:30:49 you don't (generally) need to put /usr/lib in ld.so.conf; ld.so has that path built in. Mar 06 11:32:29 pb__: yep, it didn't fix his problem Mar 06 11:35:46 looks like libsvn_ra_dav didn't get built, as ldconfig -v | grep ra_svn gives libsvn_ra_svn.so.0 -> libsvn_ra_svn-1.so.0.0.0 Mar 06 11:36:03 I'll have to check build deps tonight, as it seems it needs libneon Mar 06 11:41:28 morning all Mar 06 11:42:02 hi do13 Mar 06 11:42:26 hi dirk Mar 06 11:42:32 hi do13_ Mar 06 11:42:52 hey pb__ , Marcin, Bernardo Mar 06 11:57:46 ~lart isa servers Mar 06 11:57:46 * ibot beats isa servers into protomatter with the andromeda galaxy Mar 06 12:24:10 hi dirk Mar 06 12:24:43 hi all Mar 06 12:24:58 hey greentux Mar 06 12:28:50 hi florian_kc alan|xchat greentux Mar 06 12:29:17 florian_kc, hi Mar 06 12:32:48 RP: Do you see any reason why we shouldn't have your updated SDK compiler stuff in the stable branch? Mar 06 12:34:09 florian_kc: its rather 'why noone updated it yet' Mar 06 12:34:37 florian_kc: did you tested it and it works ok? Mar 06 12:35:10 hrw|work: that's why i ask... i was thinking about doing it - we would have a much more consistent SDK then. Mar 06 12:35:20 hrw|work: do you talk about gcc 4.1 + proper glibc? Mar 06 12:35:44 schurig: no - 3.4.4 Mar 06 12:35:50 florian_kc: I will push it in a minute then Mar 06 12:36:06 schurig: 4.1 etc is koen toy Mar 06 12:36:10 hrw|work: i ran some basic tests compiling things, seems to work fine... Mar 06 12:36:19 hrw|work: okay, great... Mar 06 12:37:22 RRA [cc]smart Mar 06 12:42:04 florian_kc: Having it in the branch would be very beneficial, if it all works - I didn't test it as throughly as I'd like but given you're also found it works... Mar 06 12:42:22 CIA-1: get ready Mar 06 12:42:30 florian_kc: It also means openzaurus can have an sdk as with the current sdk files, softfloat is broken... Mar 06 12:44:18 RP: even better Mar 06 12:46:33 03hrw 07org.oe.oz354fam083 * r3fcb60ed... 10/packages/gcc/ (3 files): Mar 06 12:46:37 added gcc-cross-sdk 3.4.4 - thx to Richard Purdie Mar 06 12:46:39 - taken from .dev as gcc-cross-sdk 3.4.3 Mar 06 12:46:41 - updated to 3.4.4 by me Mar 06 12:46:43 - tested by Florian Boor Mar 06 12:46:48 03xora 07org.oe.dev * rc46e4bd0... 10/packages/sylpheed/ (3 files in 2 dirs): sylpheed_2.2.2.bb : new version courtesy of John Bloom Mar 06 12:50:19 03hrw 07org.oe.oz354fam083 * r3f8937c1... 10/contrib/buildscripts/openzaurus/build-oz.sh: Mar 06 12:50:21 OpenZaurus: added build script for easy image building Mar 06 12:50:23 - script clean all packages which has $MACHINE related parts Mar 06 12:50:25 - script clean all packages related to kernel Mar 06 12:50:27 - build standard images: bootstrap/gpe/opie Mar 06 12:50:29 - also build experimental images: e-image/e-image-core Mar 06 12:50:36 Script is based on Familiar buildscript. Mar 06 12:51:04 it also build meta-sdk Mar 06 13:10:12 http://www.oesf.org/forums/index.php?showtopic=18042 Mar 06 13:32:53 hmm, something's wrong with quilt? Mar 06 13:32:55 can't find file to patch at input line 8 Mar 06 13:32:56 | Perhaps you used the wrong -p or --strip option? Mar 06 13:32:56 | The text leading up to this was: Mar 06 13:32:56 | -------------------------- Mar 06 13:32:58 | | Mar 06 13:33:00 | |# Mar 06 13:33:03 | |# Patch managed by http://www.holgerschurig.de/patcher.html Mar 06 13:33:05 | |# Mar 06 13:33:08 | | Mar 06 13:33:10 | |--- quilt-0.32/Makefile.in~install Mar 06 13:33:13 | |+++ quilt-0.32/Makefile.in Mar 06 13:33:15 | -------------------------- Mar 06 13:33:18 | File to patch: Mar 06 13:33:20 | Skip this patch? [y] Mar 06 13:33:23 | Skipping patch. Mar 06 13:33:26 | 1 out of 1 hunk ignored Mar 06 13:33:28 NOTE: package quilt-native-0.42: failed Mar 06 13:33:31 ERROR: Build of opie-image failed Mar 06 13:33:39 ~pastebin Mar 06 13:33:47 it has been said that pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Mar 06 13:34:18 @_@ Mar 06 13:34:24 I'm sorry about that. Mar 06 13:35:08 why quilt cann't build? Mar 06 13:35:11 did i miss something? Mar 06 13:36:37 douglee: he doesn't find quilt-0.32/Makefile.in Mar 06 13:37:15 douglee: which package do you want to create, quilt_0.42.bb ? Mar 06 13:38:05 yes Mar 06 13:40:08 douglee: hmm, for me this worked Mar 06 13:40:55 douglee: look at the end of your "tmp/work/quilt-0.42-r0/temp/log.do_patchcmd.*" file, what does it say? Mar 06 13:42:26 it said: Mar 06 13:42:28 can't find file to patch at input line 8 Mar 06 13:42:29 Perhaps you used the wrong -p or --strip option? Mar 06 13:42:29 The text leading up to this was: Mar 06 13:42:29 -------------------------- Mar 06 13:42:31 | Mar 06 13:42:33 |# Mar 06 13:42:36 |# Patch managed by http://www.holgerschurig.de/patcher.html Mar 06 13:42:38 |# Mar 06 13:42:41 | Mar 06 13:42:43 |--- quilt-0.32/Makefile.in~install Mar 06 13:42:46 |+++ quilt-0.32/Makefile.in Mar 06 13:42:48 -------------------------- Mar 06 13:42:51 File to patch: Mar 06 13:42:54 Skip this patch? [y] Mar 06 13:42:56 Skipping patch. Mar 06 13:42:59 1 out of 1 hunk ignored Mar 06 13:43:22 douglee: USE PASTEBIN! Mar 06 13:43:26 douglee: http://pastebin.ca/44634 Mar 06 13:44:10 douglee: so, please do me a favour and use pastebin and paste the last 20 lines (or so) of your tmp/work/quilt-0.42-r0/temp/log.do_patchcmd.* file Mar 06 13:45:37 schurig: got it, thanks for your kindly help Mar 06 13:45:39 douglee: also, it could help to restart from scratch, e.g. do "bitbake -b packages/quilt/quilt_0.4.2.bb -c clean" (this is now from my mind, correct typo's if there are any) Mar 06 13:47:07 anyway, kernel reboot time :-) Mar 06 14:02:55 CoreDump|afk: ping Mar 06 14:14:09 any objections on adding e2fs utils to base-images? Mar 06 14:16:07 hrw|work: as long as there is space... I usually add them myself Mar 06 14:21:06 I got another user with broken sd/mmc cards Mar 06 14:25:14 I didn't have the time to test 2.6 this weekend, I'll try today to build it again Mar 06 14:25:27 what was the problem and the size/brand/speed of the card? Mar 06 14:26:04 problem is i/o errors when you try to use card Mar 06 14:26:15 simple "md5sum /dev/mmcblk0" does not work Mar 06 14:26:27 size/brand/speed does not matter Mar 06 14:26:51 I have it with 64/256/1024M cards and does not have with other 256/1024 cards Mar 06 14:27:54 well, my 1Gb sd wouldn't work with the collie, with lots of i/o errors, but works well with the akita Mar 06 14:28:06 both with 2.4 and with the last 2.6 I tried Mar 06 14:34:12 CoreDump|afk: any public place to get your custom images for akita? Mar 06 14:38:35 katossi_uni: try http://www.hentges.net/tmp/do_not_use/ Mar 06 14:38:54 katossi_uni: coredump usually puts his images there Mar 06 14:39:05 I can't wait for when he has integrated all stuff into oe Mar 06 14:40:01 I hope that happens Mar 06 14:40:50 hrw|work: The important thing to tell Russell is whether his patch makes any card work that didn't before or vice versa. You should probably test with no mmc patches applied Mar 06 14:44:28 ok Mar 06 14:45:19 RP: building r11-no-mmc now Mar 06 14:58:57 03xora 07org.oe.dev * ra5072452... 10/packages/evince/ (3 files in 3 dirs): Mar 06 14:58:58 evince_0.5.1.bb : new version, seems to be significantly faster than Mar 06 14:59:00 previous ones. Mar 06 15:05:07 hi Mar 06 15:07:05 I can't wait for when he has integrated all stuff into oe Mar 06 15:07:14 Bernardo: lots of it already is in Mar 06 15:07:20 quite a bit will never be =) Mar 06 15:09:29 CoreDump|home: we were just talking about your roms Mar 06 15:10:59 yep, saw that ;) Mar 06 15:11:06 * florian_kc wonders why Z people always speak about ROMs Mar 06 15:11:13 Akita w/ kernel 2.6 rocks Mar 06 15:11:30 florian_kc: because sharp used cramfs ;( Mar 06 15:11:54 CoreDump|home: you have a image with 2.6 already? Mar 06 15:12:06 Bernardo: no sry Mar 06 15:12:30 hrw|work: heh Mar 06 15:13:26 :( Mar 06 15:13:42 * CoreDump|home 's build machine is sooooo down Mar 06 15:15:40 too bad your images aren't in oe yet Mar 06 15:16:12 * florian_kc builds a new one for the 770 Mar 06 15:17:09 Bernardo|away: the meta-data is available. you can roll your own images if you want to Mar 06 15:19:29 CoreDump|home: I've seen you have the data in your site Mar 06 15:20:33 my goal for now is to get colinux and a decent updated image so I can build oe at work Mar 06 15:21:49 I wonder if i could use the CATS machine here and build native... Mar 06 15:45:23 03hrw 07org.oe.oz354fam083 * ra0af0ce6... 10/packages/altboot/ (24 files in 11 dirs): removing old altboot Mar 06 15:45:42 03hrw 07org.oe.oz354fam083 * r67b76e29... 10/packages/altboot/ (35 files in 13 dirs): Mar 06 15:45:46 altboot: added 1.0.5 Mar 06 15:45:48 - It is 0.0.5 version from .dev bumped to 1.0.5 to provide upgrade path Mar 06 15:45:50 because previous version in OE did not have PV set (defaulted to 1.0) Mar 06 15:45:52 Change was discussed with maintainer. Mar 06 15:45:54 - this version has 2.4 and 2.6 kernels support Mar 06 15:45:56 - is able to create new rootfs from release tarballs (used ext2 images) Mar 06 15:46:04 TADA! Mar 06 15:47:08 * CoreDump|home cheers Mar 06 15:49:15 now e2fs tools for poodle/tosa/clamshells and TAG Mar 06 15:56:05 mickey|dinner: I hope that your build machine is working... Mar 06 15:56:42 03hrw 07org.oe.oz354fam083 * r1ef271ac... 10/conf/machine/ (poodle.conf tosa.conf zaurus-clamshell.conf): Zaurus poodle/tosa/clamshells: add e2fs tools (mkfs, fsck) into base system Mar 06 15:56:55 03hrw 07org.oe.oz354fam083 * rf3e9a55f... 10/contrib/buildscripts/openzaurus/feed-split/ (ipkg-split-feeds-koen-subfeeds ipkg.py split-world.sh): (log message trimmed) Mar 06 15:56:58 OpenZaurus: scripts used for splitting OZ 3.5.4-test5 feed into parts Mar 06 15:57:02 - I based on Koen Kooi version used to split Familiar 0.8.3-rc feed Mar 06 15:57:04 - all built packages need to be in 'world/' directory Mar 06 15:57:09 - script will move: Mar 06 15:57:11 - $MACHINE related stuff into machine/$MACHINE/ dir Mar 06 15:57:13 - OPIE/QPE stuff to opie/ Mar 06 16:05:35 hrw@bitbake:~/devel/oe/branch/contrib/buildscripts/openzaurus/feed-split$ mt list tags|grep zaurus Mar 06 16:05:39 openzaurus-3.5.4 f3e9a55fd47b949228a6bb81b0b349fd683df1a9 hrw@openembedded.org Mar 06 16:18:36 hrw|work: yay! Mar 06 16:22:45 koen|work: To which compiler does the stable feed default? Mar 06 17:12:38 03frederic 07org.oe.oz354fam083 * rbfffd966... 10/packages/linux/ (10 files in 4 dirs): Mar 06 17:12:40 opensimpad: MMC support + WE-18 Mar 06 17:12:42 backport MMC-SPI support from .dev Mar 06 17:12:44 Update to WE-18, taken from OZ (see http://bugs.openembedded.org/show_bug.cgi?id=595) Mar 06 17:14:00 03frederic 07org.oe.dev * r22576948... 10/packages/linux/ (4 files in 3 dirs): Mar 06 17:14:00 opensimpad: update to WE-18 Mar 06 17:14:02 WE-18 patches taken from OZ http://bugs.openembedded.org/show_bug.cgi?id=595 Mar 06 17:15:23 hrw|gone: Yay. I guess this means I might start to get involved :) Mar 06 17:16:29 koen|work: Any idea what familar is going to do before people start breaking, er, I mean adding things for 3.5.4.1? Mar 06 17:17:41 or do we start a seperate branch for 3.5.4.1 I wonder? Mar 06 17:22:43 RP: I'd suggest to use a new branch - even if it might work at the moment it is easier to handle this in fouture e.g. to create updates and so on. Mar 06 17:32:00 florian_kc: Agreed. The question therefore becomes do we branch from the 3.5.4 tag or from .dev?... Mar 06 17:33:19 RP: well... from the name i'd suggest to branch from 3.5.4... not sure id .dev is ready for this. Mar 06 17:33:56 are oz and dev branches going to be completely independent now? Mar 06 17:34:37 polyonymous: No, just a question of how to make this release... Mar 06 17:35:10 florian_kc: Yes, thinking about the changes to GPE, that would make sense... Mar 06 17:35:15 Doesn't it sound naturally then to merge with .dev and then branch? Actually, I'm only speaking common sense, since I'm no oz developer. At least yet :) Mar 06 17:36:21 polyonymous: In theory, yes and in theory everything in the branch is in .dev. In practise, I'm not sure it'll be that easy Mar 06 17:37:13 I don't think it's going to be easy, but isn't it inevitable, anyway, and the sooner you go that path the easier it is? Mar 06 17:37:18 RP: various changes in .oz354fam083 are not in .dev Mar 06 17:38:19 reenoo: I know, hence the "in theory". Are you aware of anything in particular that needs merging back? Mar 06 17:40:22 RP: would need to check all of koen's commits after a certain date I think Mar 06 17:41:41 I guess for light entertainment, I could read through a diff of the branch vs. .dev and see how bad it is :-/ Mar 06 18:23:57 I'm trying to finish up a custom package, my first, and i'm getting this message: Mar 06 18:24:20 NOTE: Multiple libraries found and LEAD_SONAME not defined Mar 06 18:24:47 i looked at the debian.bbclass file, and i can't quite make out why this is important. anyone? Mar 06 18:45:42 HopsNBarley: it's not all that important, which is why it's just a NOTE and not an error. Mar 06 18:45:59 okay, thanks. Mar 06 18:46:00 it means that the library auto-namer couldn't figure out what name to give your package. Mar 06 18:46:18 ah, but if i've already given it a name (foo-dev) then it's a non-issue? Mar 06 18:46:39 if you've already given it a name that's based on the SONAME, yes, it's a non-issue. Mar 06 18:46:51 if your name isn't SONAME related then you ought ideally to do something about it. Mar 06 18:49:07 pb_: if you've got a sec, there is a larger question. runtime libs should be in the base package, and not in a -dev package? Mar 06 18:49:17 yes Mar 06 18:50:25 so that's why the base package FILES uses /lib/*.so*, while the -dev package FILES is /lib/*.so Mar 06 18:50:58 indeed Mar 06 18:51:58 and the packaging moves files out of image/ so subsequent packages built in same don't pick up duplicate files, yes? Mar 06 18:52:33 yeah Mar 06 18:52:46 thanks! i just about have hold of all this... Mar 06 19:07:02 re Mar 06 19:14:59 hey florian Mar 06 19:15:06 hi zecke Mar 06 20:19:48 zecke: busy? Mar 06 20:20:23 <_law_> n8 all Mar 06 20:21:24 somehow :) Mar 06 20:21:31 ashak: Konqueror Error? Mar 06 20:21:49 ashak: I'm trying to get konqe 3.5.1 compiling (for Qt2) Mar 06 20:22:19 zecke: yeah... I tried to compile with an earlier gcc, but I think i'm doing something wrong as it's still failing with the same errors. Mar 06 20:22:31 zecke: Although if you're working on it, I can quite happily wait ;) Mar 06 20:22:52 ashak: are you sure it rebuild the gcc? Mar 06 20:23:34 zecke: Ah... no it didn't.. can I clean gcc or something, or should I clean everything and rebuild it all again? Mar 06 20:23:58 ashak: I do not know if the C++ ABI changed from gcc3.4 to 4.0 Mar 06 20:24:19 ashak: you could just build a gcc 3.4 using 'bitbake -b' Mar 06 20:24:28 * zecke needs to fix the Ubuntu of his mother... Mar 06 20:24:52 zecke: ashak: NOTE: package konqueror-embedded-svn-r0: task do_package: completed Mar 06 20:25:31 zecke: it did change Mar 06 20:30:54 zecke: heh, do you really imagine the c++ developers would miss an opportunity like that? Mar 06 20:31:40 ashak: with 3.4.4 it compiles fine Mar 06 20:32:57 do13_: thanks, that's what zecke told me. I was just trying to work out how to get it to compile with 3.4.4 when i'd already built half of opie-image with 4.0.2 or whatever it was. Mar 06 21:09:42 03rw 07org.oe.dev * r9996b113... 10/ (3 files in 2 dirs): Mar 06 21:09:45 glibc, package.bbclass: generate old-style binary locales at build time using qemu. based on a patch originally written by Phil Blundell. fixes hh.org Bug #1385 (localedef OOM issue). Mar 06 21:09:48 - running localedef (in qemu...) for every locale takes a while. set Mar 06 21:09:49 BINARY_LOCALE_ARCHES = "" to avoid the new behaviour. Mar 06 21:13:04 yay Mar 06 21:18:08 :) Mar 06 21:19:42 pb_: We c++ fix bugs whenever we can (changing the output format - abi change ) Mar 06 21:19:54 do13_: oeh? Mar 06 21:20:05 do13_: what version of konqueror did you build? Mar 06 21:20:17 zecke: latest svn Mar 06 21:20:21 re Mar 06 21:20:27 do13: Qt2? Mar 06 21:20:36 zecke: yep. But it won't work Mar 06 21:20:44 zecke: it seems. we c programmers, by contrast, fixed all our bugs long ago and can now just lounge around. Mar 06 21:20:46 do13: atleast it builds Mar 06 21:20:51 do13: do you have patches? Mar 06 21:20:59 pb_: your ABI is too easy for us Mar 06 21:21:05 I need some cleanup Mar 06 21:21:12 pb_: it is unattractive Mar 06 21:21:29 zecke: right, it must seem primitive Mar 06 21:21:32 pb_: it is like drinking beer when you can have Leffe Mar 06 21:21:45 quite so Mar 06 21:22:06 * pb_ drinking californian wine at present, however Mar 06 21:22:42 do13: how did you make it build? Mar 06 21:22:50 <[lala]> do13: is that the konq4 one? Mar 06 21:23:07 [lala]: no, it is a paid backport of konq3.5.1 to Qt2 Mar 06 21:23:16 [lala]: basysKom.de is paid for it Mar 06 21:23:43 <[lala]> i see Mar 06 21:24:01 zecke: I took a few hours :) Mar 06 21:24:18 do13: I in my scattered five minutes I didn't achieve much Mar 06 21:24:24 do13: I will post some hints for you Mar 06 21:25:39 mickey|dinner: long dinners are not good for stomach.. Mar 06 21:25:54 do13: you will need a Qt2 with exceptions and rtti Mar 06 21:26:01 do13: and threads Mar 06 21:26:27 <[lala]> zecke: is there any problem with enabling threads in opie? Mar 06 21:26:52 [lala]: none, besides incompatibilities with Qtopia :) Mar 06 21:27:26 <[lala]> zecke: does qtopia doesn't work with threads enabled or is it just turned off there by default? Mar 06 21:27:39 zecke: exceptions? Mar 06 21:27:47 do13: wait a second :) Mar 06 21:28:05 hrw: congrats on oz354 Mar 06 21:28:10 koen: thx Mar 06 21:28:16 zecke: ok. start with http://www.do13.de/oz/temp/dcopidl-native.bb Mar 06 21:28:16 hrw: CONGRATS! Mar 06 21:28:38 do13: okay (I cheated in installing kde dev packages) Mar 06 21:29:19 ~hail reenoo for fixing localegen Mar 06 21:29:21 * ibot bows down to reenoo for fixing localegen and chants, "I'M NOT WORTHY!!" Mar 06 21:30:42 koen: pb_ did the intial work. I just happened to have the same idea (the same day) and finished his patch Mar 06 21:31:02 do13: http://oe.pastebin.com/587708 <- that is ROAD's konq.bb Mar 06 21:31:10 do13: it has a couple of flaws I just saw Mar 06 21:31:18 do13: the configure method is exciting Mar 06 21:31:26 do13: but the compile flags are right Mar 06 21:32:42 <[lala]> zecke: is that the same version do13 is about to compile or a different one? Mar 06 21:32:57 [lala]: it is the same Mar 06 21:33:36 [lala]: with some brain power applied you will find out who is paying for that port easily Mar 06 21:34:12 zecke: phone manufacturer which people does not call? Mar 06 21:34:43 zecke: My version: http://www.do13.de/oz/temp/konqueror-embedded_svn.bb (It needs some cleanups) Mar 06 21:34:50 hrw: no Mar 06 21:34:53 hrw: ROAD Mar 06 21:34:59 hoho Mar 06 21:35:03 ~lart the other manufacturer Mar 06 21:35:04 * ibot beats the other manufacturer to within 2.54cm of his life Mar 06 21:35:39 <[lala]> --with-gui=road Mar 06 21:35:41 <[lala]> ;-) Mar 06 21:36:13 http://websvn.kde.org/trunk/kdenox/konq-embed/src/mainwindow_road.cc?rev=515076&view=log Mar 06 21:38:32 do13: doesn't look too broken Mar 06 21:38:59 zecke: why doesn't http://www.kernel.org/pub/linux/kernel/v2.6/testing/patch-2.6.16-rc5.bz2;patch=1 work? Mar 06 21:39:11 zecke: There is one prob with prce-config Mar 06 21:41:36 zecke: configure: Mar 06 21:41:38 if test -n "$PCRE_CONFIG" && $PCRE_CONFIG --libs >/dev/null 2>&1; then Mar 06 21:41:38 LIBPCRE=`$PCRE_CONFIG --libs-posix | sed -e "s,-L/usr/lib ,,"` Mar 06 21:41:38 PCRECFLAGS=`$PCRE_CONFIG --cflags` Mar 06 21:42:34 koen: due the .bz2? Mar 06 21:42:38 sed -e "s,-L/usr/lib is wrong, because prce-config returns -L/..STAGING_LIBDIR... Mar 06 21:42:49 koen: check oe/classes/base.bbclass Mar 06 21:43:38 do13: which pcre config got picked? Mar 06 21:44:23 zecke: STAGING_BINDIR/prce-config Mar 06 21:45:58 zecke: After manual removal of the dir, it compiles Mar 06 21:46:31 zecke: .bz2 should not matter - ${RPSRC}/alsa/asoc-0.10rc2.diff.bz2;patch=1 \ Mar 06 21:46:43 zecke: I don't want to put a 27mb patch in monotone or force users to download it Mar 06 21:46:43 zecke: this one is bzipped and works Mar 06 21:48:05 weird Mar 06 21:49:50 step by step pals Mar 06 21:50:06 do13: Where is it pulling -L/usr/lib from? Mar 06 21:50:13 koen: NOTE: Applying patch 'patch-2.6.16-rc5' Mar 06 21:50:25 koen: fetched bz2 patch and applied Mar 06 21:51:06 * koen tries again Mar 06 21:51:13 maybe bitbake r384 is too old Mar 06 21:52:53 hmm, maybe we could use qemu to preinstall packages etc so first boot won't be so slow Mar 06 21:53:29 mreimer: that's wath the patch does :) Mar 06 21:53:37 mreimer, why hurry with the first boot? Mar 06 21:53:44 koen: also update-modules stuff? Mar 06 21:53:49 polyonymous: it freaks people out because they think they've bricked Mar 06 21:53:58 hrw: no, only localegen Mar 06 21:54:03 koen: really? great! Mar 06 21:54:23 mreimer, bricked? It's way to colorful for a brick :) Mar 06 21:54:23 oh, that's what I was referring to (update-modules and ipkg-cl) Mar 06 21:54:23 polyonymous: you can tell them 'it takes 10 minutes' as much as you want, but users just won't get it Mar 06 21:54:36 does it take 10 minutes? Mar 06 21:54:39 yes Mar 06 21:54:42 mreimer: we could run prelink as well :) Mar 06 21:54:44 hmm. Mar 06 21:54:52 mreimer: we coul allow users to customize their image Mar 06 21:55:01 mreimer: and generate a prelinked one for download :} Mar 06 21:55:04 mreimer, are you talking about the /etc/rcS.d/S98configure step? Mar 06 21:55:10 zecke: that's what I'm thinking of Mar 06 21:55:10 mreimer: the ipkg update-modules patch should be in ipkg .159 Mar 06 21:55:13 polyonymous: yes Mar 06 21:55:15 zecke: ???. My debian pcre-config --libs shows: -L/usr/lib -lpcre Mar 06 21:55:19 koen: what does that do? Mar 06 21:55:35 Doesn't take 10 minutes for me... Mar 06 21:55:37 mreimer: save up all update-modules call and execute it once Mar 06 21:55:37 do13: ah so it picked the pcre-config from /usr/bin? Mar 06 21:55:41 the biggest gain in the postinst executions would be to avoid doing the same things repeatedly. how many times is ldconfig and depmod run? Mar 06 21:55:43 instead of n times Mar 06 21:55:45 heh Mar 06 21:55:47 do13: could you upload your config.log please? Mar 06 21:55:48 ah good Mar 06 21:55:51 in hdd-based loop-ed filesystem... Mar 06 21:56:27 polyonymous: lucky you :) Mar 06 21:56:37 kergoth: right now? once each Mar 06 21:56:38 heh Mar 06 21:56:53 koen ;-) Mar 06 21:57:36 zecke: ok. I'll do a new configure run with a fresh pcre-config Mar 06 21:58:01 do13: why fresh pcre-config? Mar 06 21:58:18 do13: e.g. I patched a lot of sdl libs to not prepend /usr/bin/*-config Mar 06 21:59:51 zecke: I removed the -L...STAGING_DIR :) Mar 06 22:00:04 zecke: it works now, must be some weird bug with caching Mar 06 22:00:24 koen: any word on glibc + EABI? Mar 06 22:00:52 mreimer: not yet Mar 06 22:00:55 koen: so we have a bitbake-pxa250 version Mar 06 22:01:11 mreimer: I'm waiting for csl to address the armv4 problems Mar 06 22:01:17 ok Mar 06 22:01:21 mreimer: right now eabi only works on armv5 Mar 06 22:01:25 ah Mar 06 22:01:37 due to the thumb-interworking Mar 06 22:01:47 koen: do we have armv4 machines with working 2.6 kernels? Mar 06 22:02:06 hrw: h36xx is getting pretty close Mar 06 22:02:13 hrw: and my ep93xx board :) Mar 06 22:02:29 koen: hi, could you please help me review this http://pastebin.com/587777 ? Mar 06 22:02:32 koen: your ep93xx is weird board ;) Mar 06 22:02:34 heh. cirrus logic should get with the programme. Mar 06 22:03:02 it's a bit lame to be producing arm920 designs in this day and age. Mar 06 22:03:15 koen: and what will cause " checking for suffix of object files... configure: error: cannot compute suffix of object files: cannot compile Mar 06 22:03:16 " ? Mar 06 22:03:55 * hrw goes to bed - cu Mar 06 22:04:12 <_guillermo> qemu-native fails: ERROR: QEMU requires SDL or Cocoa for graphical output Mar 06 22:04:15 minipanda: did gcc build ok? Mar 06 22:06:32 koen: well, you mean in tmp/staging/i686 or arm ? Mar 06 22:06:36 zecke: http://www.do13.de/oz/temp/config.log Mar 06 22:07:00 minipanda: in tmp/cross Mar 06 22:07:34 do13: it selected the right pcre-config now? Mar 06 22:07:49 all: how can I see what packages are available in OE without installing it? Mar 06 22:07:59 koen: well, seems a gcc 4.0.2 is generated Mar 06 22:08:00 zecke: yeah. But the output is wrong. Mar 06 22:08:03 03koen 07org.oe.dev * red43fd4a... 10/packages/linux/ (ep93xx-kernel_2.6.15.bb ep93xx-kernel_2.6.15+2.6.16-rc5.bb): ep93xx: update to 2.6.16-rc5 and derevo14 Mar 06 22:08:10 nite Mar 06 22:08:10 includedir='/usr/include' Mar 06 22:08:13 ouch Mar 06 22:08:24 minipanda: in that case take a look at /conf/machine/include/tune-arm926* Mar 06 22:08:33 ah okay it is' right' Mar 06 22:08:56 zecke: pcre-config returns : -L/prj/zaurus/dev/build/tmp/staging/arm-linux/lib -lpcreposix -lpcre Mar 06 22:09:02 configure: 22410: /usr/include/jpeglib.h Mar 06 22:09:39 thanks. Mar 06 22:10:11 maybe I should patch cpp to never ever include /usr/include Mar 06 22:10:36 hmm need to get some other stuff done Mar 06 22:10:37 cya! Mar 06 22:10:48 koen: ok, they said "#if gcc breaks change arm926ejs to arm926ej-s Mar 06 22:10:49 " Mar 06 22:11:09 koen: but seems a cross gcc 4.0.2 is already build Mar 06 22:13:17 <_guillermo> if someone could take a look at this http://pastebin.com/587802, it's qemu-native error Mar 06 22:13:24 minipanda: I can't think of anything else going wrong Mar 06 22:13:53 koen: ok, i've changed that tune options and let me try again Mar 06 22:14:11 _guillermo: 1.) file a bug report that qemu-native should depend on sdl-native... Mar 06 22:14:19 <_guillermo> ok Mar 06 22:14:23 _guillermo: we will put sdl-native to assum_provided Mar 06 22:14:27 koen: BTW: we've already rolling to use gcc 4.0.x in our default release? Mar 06 22:14:31 _guillermo: you will need to install SDLnative :) Mar 06 22:14:42 <_guillermo> zecke: thx Mar 06 22:15:04 actually that issue is tougher :} Mar 06 22:15:16 for the locale generation we just need a simple qemu Mar 06 22:15:31 for running the image with X11 we will need it with SDL Mar 06 22:16:19 minipanda: I've been using images built with gcc4 with quite some success Mar 06 22:18:02 koen: well, i just worried about this point especially for build image against OSK board ... and seems the same issue didn't happened last month with gcc 3.x Mar 06 22:19:44 zecke: yah, you should have a separate package for the sdl-enabled ueberqemu. Mar 06 22:20:19 it would suck to require you to build native versions of sdl, x11 and all that stuff just to compile glibc. Mar 06 22:20:58 koen: and could you let me know the last stable gcc 3.x version we used ? Mar 06 22:21:19 koen: and where can i override gcc/ glibc version ? Mar 06 22:22:03 pb_: virtual/windowing-system-native Mar 06 22:22:07 minipanda: see conf/distro/angstrom-2006.9.conf Mar 06 22:22:45 * minipanda checking angstrom .... Mar 06 22:23:42 pb_: heh, okay, so depmod and ldconfig were bad examples :) Mar 06 22:23:56 so qemu is benig used now eh Mar 06 22:25:06 kergoth: heh, yeah, ipkg knows how to eliminate duplicate invocations of depmod, ldconfig, and update-modules. Mar 06 22:25:35 yeah, glibc uses qemu (all hail reenoo) to generate the binary locale packages nowadays Mar 06 22:26:26 slick Mar 06 22:26:30 kergoth: qemu also runs arm OE images quite happily Mar 06 22:26:33 I think RP and zecke also have some kind of system-emulation thing going on that lets them boot a whole opie system under qemu Mar 06 22:27:03 would be cool to get a bbclass going that does the scratchbox binfmt_misc bits to do the initial site file population Mar 06 22:27:06 someday Mar 06 22:27:15 that would be cute Mar 06 22:27:58 pb_: xserver based images work happily (although I didn't specifically try gpe). opie only gets as far as calibrating its imaginary touchscreen Mar 06 22:28:29 RP: ah Mar 06 22:29:26 kergoth: We're getting ever closer :) Mar 06 22:30:39 RP: tko expresses interest in building maemo using OE (with some blessing from the big N) Mar 06 22:30:56 zecke: Now it compiles fine. There was a prob with my mixed system include dirs. Mar 06 22:30:57 RP: but I had to catch my train before I could ask for details Mar 06 22:31:19 koen: Really? Mar 06 22:31:40 RP: yes Mar 06 22:31:51 koen: Is tko Tommi? Mar 06 22:32:01 RP: also yes Mar 06 22:32:20 koen: Can you ask him to mail me/mallum as I think I might be able to help a lot here... Mar 06 22:32:54 RP: sure, I have to talk to a lot of N people soon about eabi, OE, etc Mar 06 22:33:46 koen: I've mentioned OE a fair bit to them and have also raised the eabi issue in the past... Mar 06 22:34:12 kergoth: do you happen to know how I can undelete a file using svn? Mar 06 22:34:20 kergoth: using merge? Mar 06 22:34:50 yeah Mar 06 22:34:54 zecke, update to the certain revision? Mar 06 22:34:58 did you look at the svn book? Mar 06 22:35:03 no :) Mar 06 22:35:13 well more than once, but I did not read it Mar 06 22:35:50 http://svnbook.red-bean.com/en/1.1/ch04s04.html#svn-ch-4-sect-4.2 Mar 06 22:35:55 and the next section :) Mar 06 22:36:08 I had to resurrect a branch yesterday Mar 06 22:36:11 Crofton: that sounds like too much Mar 06 22:36:19 but I think I ended up making another copy Mar 06 22:36:25 koen: I'm having trouble believing this... Mar 06 22:36:27 Crofton: I wish websvn would have a 'Attic' directory Mar 06 22:36:29 or copying from the previous rev Mar 06 22:36:30 RP: it may be a good idea to have a variable containing a list of the test result variables which that app requires to crosscompile successfully, and make it warn/error if any of those are missing in teh current site file. Mar 06 22:36:47 would then use that to do the population with qemu once that bbclass is written Mar 06 22:36:49 trac makes it much easier to find the right revs Mar 06 22:37:01 trac is awesome Mar 06 22:37:09 yeah Mar 06 22:37:25 i love the crosslinking between svn/wiki/bts Mar 06 22:37:27 good stuff Mar 06 22:37:55 kergoth: That sounds good. We're heading into a sort of similar position with debian.bbclass as we could have that cache the previous renaming to check to changes and also removing its need to build all rdepends Mar 06 22:38:31 zecke , is this in the oe svn you need to do this? Mar 06 22:38:35 RP: it surprised me as well Mar 06 22:38:40 s/to chan/for chan/ Mar 06 22:39:07 any volunteers for writing a package_deb.bbclass? Mar 06 22:39:17 Crofton: no Mar 06 22:39:33 koen: cp package_ipk.bbclass package_deb.bbclass Mar 06 22:39:41 koen: I dread to think how many of my reports read "using openembedded" :). I know the one I have ready for tomorrow does... Mar 06 22:39:45 Crofton: good idea, I could sync again :) Mar 06 22:40:09 reenoo: that was my first plan Mar 06 22:40:34 if I get some time and I can do it without buggering my work trac setup, I'll get it going against oe Mar 06 22:43:49 reenoo: I can't find a clean way to get OE to give the resulting package a .deb suffix Mar 06 22:56:35 yeah, there's no really good answer to that. Mar 06 22:56:57 one option would be to add the packaging code directly to package_ipk.bbclass, and discontinue use of ipkg-build. Mar 06 22:57:18 that'd also remove the requirement for the ipkg-utils-native dependency, which I guess would be a general improvement in road safety Mar 06 22:57:29 hehe Mar 06 22:57:31 indeed Mar 06 22:58:44 ipkg-build is just 'copy in control stuff, tar, tar, ar', right? Mar 06 22:59:05 more or less, yeah Mar 06 22:59:28 it's a shell script, you can easily inspect it. Mar 06 23:00:06 * koen adds that to his todo Mar 06 23:00:15 'night all Mar 06 23:00:30 'night koen|sleep Mar 06 23:02:45 'night all Mar 06 23:03:06 cya! Mar 06 23:03:09 night RP, koen|sleep Mar 06 23:04:53 night RP, koen|sleep Mar 07 00:59:19 'night all Mar 07 02:30:55 who maintains the gcc bb's in OE? Mar 07 03:25:41 night Mar 07 03:40:37 hi, folks . I got a proble while bitbake manufacturers . Mar 07 03:41:51 here is the message : http://rafb.net/paste/results/iqzLnp31.html Mar 07 03:42:06 any idea ? Mar 07 06:12:48 Hi. Does anyone here work with embedded non-Linux wireless? Mar 07 06:13:16 what type of wireless? Mar 07 06:13:45 802.11b Mar 07 06:14:08 ideally, a compact flash card Mar 07 06:14:19 there are a few people here who did work on the specific devices Mar 07 06:14:47 not sure if they can release info tho Mar 07 06:14:47 I ordered a book on it, should be here soon, bur from what I have heard, it's slim on options Mar 07 06:14:57 wow Mar 07 06:15:22 I certainly wouldn't be looking for confidential info... I'm building a one-ff project thats a wireless musical instrument Mar 07 06:15:35 you trying to make a .11b device? Mar 07 06:15:36 I was hoping to use 802.11 to transfer data Mar 07 06:15:52 yes, which would talk to a WAP router. Mar 07 06:15:56 why not use a prebuilt module? Mar 07 06:15:56 sending UDP Mar 07 06:16:17 If I could find one that would do throughput over 230 Kbps, I'd buy it in a heartbeat Mar 07 06:16:20 saves the expense of testing to pass federal regs Mar 07 06:16:54 I was hoping the Compact Flash wireless card would be my wireless interface... that or an off-the-shelf wireless module Mar 07 06:16:57 .11b doesnt allow that i do not belive Mar 07 06:17:14 heh you're joking Mar 07 06:17:23 802.11b is 11 Mbps throughput Mar 07 06:17:24 i am trying to remember Mar 07 06:17:37 802.15.4 is the 230 Kbps one Mar 07 06:17:51 whoops bad me, read wrong B Mar 07 06:17:53 that's being championed by ZigBee Mar 07 06:18:06 i thought you ment 230KBps Mar 07 06:18:40 yeah I'll be in the payload range of 3 Mbps if I were to stream all channels constantly... Mar 07 06:18:56 that's 384 KB/s Mar 07 06:19:08 i am pretty sure i've a link to some ... Mar 07 06:19:20 I've seen dpactech, airdrop, and rabbit Mar 07 06:19:32 rabbit = $$$$ and the company has pulled their devkit Mar 07 06:19:53 DPACtech said "yeah um well er maybe, but you will have to write your own stack" Mar 07 06:20:22 Airdrop - only has SPI interface, and one guy said he spent months getting it up to around 1 Mbps Mar 07 06:20:32 and said it involved a LOT of re-engineering Mar 07 06:20:48 so I'm feeling pretty crushed at the moment Mar 07 06:21:16 I wonder how those portable windows devices use 802.11b CF cards? Maybe they are limited to 230 Kbps data rate? Mar 07 06:25:28 hrmm ... cant find it in my bookmarks Mar 07 06:25:42 sec i'll grab a couple mags Mar 07 06:27:53 basically, I need someone to tell me what the chip requirements would need to be, in order to collect the 3 Mbps from my sensor array, and control a CF wireless card. Mar 07 06:28:06 cool, thanks for your help Mar 07 06:29:25 yay, wpa_supplicant working Mar 07 06:33:42 Has anyone recently tried a version of firefox from .dev ? Mar 07 06:34:42 Iron_Monk, you sure you dont want to just use a pcmcia/cf wireless card? Mar 07 06:35:15 i can find all sorts of articals using those .... then you can skip the fed RF tests :) Mar 07 06:35:54 btw i havent found the link yet ... it was on the same site as the GPS modules ... Mar 07 06:37:03 emte - uhh Mar 07 06:37:04 [01:20] I was hoping the Compact Flash wireless card would be my wireless interface... that or an off-the-shelf wireless module Mar 07 06:37:19 CF stands for Compact Flash Mar 07 06:37:35 yeah i missed that Mar 07 06:37:44 the problem is, using those embedded Mar 07 06:37:58 there are quite a few articals at www.circuitcellar.com Mar 07 06:38:00 I can plug them into a PCMCIA slot computer just fine Mar 07 06:38:25 using them for data aquisition Mar 07 06:39:24 I'll bet they all use SPI at 230 Kbps max. Mar 07 06:39:50 need to find out whom maintains GCC bb's in OE Mar 07 06:40:05 they're in need of a big overhaul for gcc 4.2 Mar 07 06:41:54 Enfora CompactFlash CF GSM GPRS Modem GSM0110 Mar 07 06:43:55 Iron_Monk, the "Embedded WiFi with TRENDnet" artical has a screen capture that says "2 Megabits per second" Mar 07 06:44:28 from the Netasyst Sniffer tool Mar 07 06:46:09 yeah I have Fred Eady's book on order, actually Mar 07 06:46:21 * emte guesses jnc is incapable of reading the MAINTAINER tag in bb Mar 07 06:48:06 pfft, since when has that been accurate Mar 07 06:48:17 the mysterious Jenna arrives. Mar 07 06:48:34 mysterious ? Mar 07 06:48:36 if its not acurate it needs to be removed Mar 07 06:49:38 anyway gb2 is the maintainer Mar 07 06:51:08 well thanks emte, I actually hadn't seen this particular article. I saw the one where he was schilling for rabbit Mar 07 06:51:53 the other ones i browsed over did not mention throughput Mar 07 06:52:13 and thats only a screencapture, ... Mar 07 06:53:47 sure Mar 07 06:54:38 but I guess if the CF bus is 16 bit, I could maybe build my own dev board and try generating dummy data, on a ramp up to see how high it can go Mar 07 06:59:30 * Iron_Monk kicks the bleeding edge and watches it fall over Mar 07 07:02:20 heh Mar 07 07:02:22 "Again, after the Wi-Fi design concept is validated, it doesn’t matter which microcontroller or microprocessor you choose as your host controller. Your design has to do with how you adapt of the Wi-Fi driver code as well as your ability to construct, debug, and program a custom host microcontroller or microprocessor plug-in hardware module." Mar 07 07:03:52 Fred made the Airdrop. Mar 07 07:03:57 now it all makes sense Mar 07 07:04:05 this article is his genesis of the Airdrop Mar 07 07:36:49 morning Mar 07 07:57:40 good morning all Mar 07 07:57:50 morning Mar 07 08:01:12 heh Mar 07 08:01:24 sharp has really innovated this time with the c3200 Mar 07 08:01:32 lots of new stuff Mar 07 08:01:35 I am blown away by the innovation Mar 07 08:01:39 it's incredible Mar 07 08:01:45 * johnX 's head explodes Mar 07 08:02:37 on the other hand, it's less work for RP :) Mar 07 08:03:20 do we know the dog name of it already? Mar 07 08:08:02 what? Mar 07 08:08:08 so is it official? Mar 07 08:08:09 the 3200? Mar 07 08:08:17 the rumors are true? Mar 07 08:08:22 yeah Mar 07 08:08:25 it's incredible Mar 07 08:08:28 link please Mar 07 08:08:30 !!!!!!!!!!!!!! Mar 07 08:08:31 champ: Don't know, don't care Mar 07 08:08:31 I bet you're glad you waited Mar 07 08:09:16 champ: let me dig up the link again Mar 07 08:09:20 ok Mar 07 08:09:21 thanks Mar 07 08:09:29 c3100, but with a 6gb microdrive Mar 07 08:09:56 http://www.oesf.org/forums/index.php?showtopic=18086&hl= Mar 07 08:10:24 upgraded ram? Mar 07 08:11:05 nope Mar 07 08:11:21 pschh Mar 07 08:11:22 same rom, same ram, new apps to learn english, and 6GB MD Mar 07 08:11:25 just 2 more gigs? Mar 07 08:12:17 well, if you're not interested in having it teach you english, then yes the 6GB drive is the only difference Mar 07 08:12:30 read the Japanese page though and tell us if we're wrong :) Mar 07 08:13:03 if it could teach me japanese that would be cool' Mar 07 08:14:14 that'd be nice actually Mar 07 08:28:50 morning Mar 07 08:29:52 I definitelly wont buy zaurus Mar 07 08:30:09 hey hrw|work Mar 07 08:32:13 anyone know vga portrait PDA with linux? (cheaper then hx4700) Mar 07 08:33:25 'night kergoth`zzz Mar 07 08:33:51 night Mar 07 09:02:30 morning Mar 07 09:14:21 morning folks Mar 07 09:15:44 morning all Mar 07 09:16:03 hi Mar 07 09:16:09 The C3200 is a bit of a joke. When anyone finds the kernel sources, let me know :) Mar 07 09:16:14 RP: hehe Mar 07 09:16:28 RP: not a very funny joke if you ask me Mar 07 09:16:28 RP: kind of c760->c860 situation Mar 07 09:17:10 hrw|work: Yes. Hopefully, the kernel won't even need any changes except perhaps a new PCMCIA ID... Mar 07 09:17:16 RP: i would like to see the suspend/resume issue fixed soon. my C3100 is in the mail, due to arrive tomorrow! i must sell my c3000 soon Mar 07 09:18:18 jnc: Porting the code in question isn't trivial :-/ Mar 07 09:18:33 RP: how about to the 2.4 kernel Mar 07 09:18:45 i want to see the bug fixed in OZ either way Mar 07 09:18:53 jnc: What about the 2.4 kernel? Mar 07 09:19:22 well, to fix suspend/resume. would it be easier to have that in the 2.4 OZ kernel? is it already there? Mar 07 09:19:42 jnc: It should already be there Mar 07 09:19:44 i don't know these things, since pure amd64 build host does not willingly build the 2.4 oz kernel Mar 07 09:19:47 okay Mar 07 09:20:00 if there is a test image, i will try this and report Mar 07 09:20:54 jnc: I have to stress the 2.4 kernels are dead as far as OZ is concerned Mar 07 09:21:00 oh okay Mar 07 09:21:43 i'm happy about that, the situation is i wish to bugtest and confirm fixed the suspend/resume issue which mysteriously only happens to my C3000, ... before i sell it :/ Mar 07 09:22:54 jnc: I'd really like to see that sorted out as well. I just can't say when I'm going to find several spare hours to look at the voltage/frequency scaling, even to hack it into a 2.6 kernel Mar 07 09:23:05 hmm Mar 07 09:23:10 do you think within 2 months? Mar 07 09:23:20 jnc: yes Mar 07 09:23:32 okay. i could even provide test hardware if really needed Mar 07 09:26:01 RP: did i show you the new test i ran? my device does wake up every 10 min or so and spit out info for a while, and eventually freezes up Mar 07 09:26:12 with power cord in Mar 07 09:26:30 jnc: That sounds like the correct bahaviour until it doesn't wake up Mar 07 09:27:30 jnc: We've ruled out a lot and have a good idea of wherehow its crashing. We just need to work Mar 07 09:27:38 why Mar 07 09:28:55 gcc devs have been receptive to suggestions about cross-compiling in a meta distro like OE. need to ping kergoth about fixing up the gcc builds to eliminate the build system patches Mar 07 09:29:59 jnc: One of our aims is to get patches upstream (or the issues the patches address solved as appropriate) Mar 07 09:31:29 yep. that is what i liked about the existance of Gentoo Mar 07 09:31:38 good morning Mar 07 09:31:43 it's users like myself i can't stand Mar 07 09:31:45 hi Bernardo Mar 07 09:31:47 Hey B Mar 07 09:31:53 jnc: http://wiki.debian.org/ArmEabiPort is also worth a read Mar 07 09:32:00 hey Bernardo Mar 07 09:32:02 * jnc groks Mar 07 09:32:22 ~lart autoconf Mar 07 09:32:23 * ibot executes killall -HUP autoconf Mar 07 09:32:31 ~lart subversion Mar 07 09:32:32 * ibot hauls subversion up by the scruff of the neck and spanks him until he waddles Mar 07 09:32:43 and http://sourceware.org/ml/libc-ports/2005-10/msg00031.html of course Mar 07 09:35:43 jnc: gcc 4.1.0 is out, wanna have a look at update the gcc-4.1 snapshot bits? Mar 07 09:36:21 i was going to jump into gcc 4.2.0 and co-ordinate with upstream to get our patches in Mar 07 09:36:37 that's also a nice plan :) Mar 07 09:36:40 plus look at reworking the whole gcc bb system to be simple Mar 07 09:37:02 it's way too complicated to deal with as-is. i have trouble submitting a bug report to gcc dev about anything Mar 07 09:37:26 now we just need to get the attention of the glibc people to complete our world dominination plans ;) Mar 07 09:38:15 is eabi considered a new arm arch? Mar 07 09:38:25 and what endianess is presumed Mar 07 09:38:47 jnc: it can be either LE or BE Mar 07 09:38:52 hmm Mar 07 09:38:56 the cpu itself doesn't care Mar 07 09:39:50 * koen joins #debian-arm where all the eabi action is supposed to happen Mar 07 09:43:50 koen: by the looks of it uclibc eabi bits are already good to go in gcc 4.1 :) Mar 07 09:56:18 ade|desk: cool Mar 07 09:57:22 pricejapan is already selling the 3200 Mar 07 09:57:31 http://www.pricejapan.com/PriceJapan_com.htm Mar 07 09:59:28 ~change 32000 jpy to eur Mar 07 09:59:32 32,000.00 Yen (JPY) makes 227.7450 Euro (EUR) (from http://www.xe.com/) Mar 07 09:59:36 ~change 65000 jpy to eur Mar 07 09:59:41 65,000.00 Yen (JPY) makes 462.5499 Euro (EUR) (from http://www.xe.com/) Mar 07 10:00:00 ~change 65000 jpy to pln Mar 07 10:00:06 65,000.00 Yen (JPY) makes 1,785.483 Zloty (PLN) (from http://www.xe.com/) Mar 07 10:02:19 3200? Mar 07 10:02:31 new zaurus Mar 07 10:02:38 extremely disappointing Mar 07 10:02:38 time for me to beg the wife to let me spend 400+ euros? :) Mar 07 10:03:19 what does it have on the 3100? Mar 07 10:03:31 Bernardo: 2G bigger uDrive Mar 07 10:03:48 and other case/keyboard colors (as usual) Mar 07 10:03:55 yeah 2 gigs Mar 07 10:03:58 thats about it Mar 07 10:04:01 not even more ram Mar 07 10:04:01 ;[ Mar 07 10:04:10 for me, it's nice to not be tempted to buy expensive things actually Mar 07 10:04:36 so it's 6GB instead of 4 Gb? Mar 07 10:04:40 yeah Mar 07 10:04:41 yes Mar 07 10:04:59 well, with 8GB cf at 150 US$ at amazon... Mar 07 10:05:33 too bad nobody is selling a "expansion kit" to add internal CF support for the akita Mar 07 10:05:56 Bernardo: it's missing too much important stuff Mar 07 10:06:03 believe me the thought has crossed my mind Mar 07 10:06:05 like bluetooth or wifi Mar 07 10:06:14 and/or Mar 07 10:06:22 I want tosa++ Mar 07 10:06:30 has anyone used the nokia 770? Mar 07 10:06:35 champ: yes Mar 07 10:06:51 koen: do you own one? Mar 07 10:06:52 hey zecke Mar 07 10:06:54 champ: yes Mar 07 10:07:12 koen: what do you think of it? Mar 07 10:07:31 it's a very cool device, but the software needs work Mar 07 10:07:38 johnX: yep, I know Mar 07 10:07:42 bluetooth and wifi may happen: http://www.oesf.org/forums/index.php?showtopic=13562&st=150 Mar 07 10:07:44 koen: hi, after modified tune-arm926ejs.conf, my build process can go further ... Mar 07 10:07:47 do you have a z as well? Mar 07 10:08:10 minipanda: ah, that's good to hear Mar 07 10:08:12 koen: but stopped at just 'build bootstrap-image' Mar 07 10:08:16 the 770 doesnt have cf or sd card slots, does it? Mar 07 10:08:20 koen: so could you please take a look at http://pastebin.com/588565 Mar 07 10:08:25 champ: only a rs-mmc slot Mar 07 10:08:36 koen: it's read like strange permission issue Mar 07 10:08:38 what exactly is a rs-mmc slot? Mar 07 10:08:51 champ: it takes rs-mmc cards Mar 07 10:09:03 minipanda: is your bitbake up to date? Mar 07 10:09:15 koen: i'm sure, i just svn up last night Mar 07 10:09:16 * koen is happy with his 1gb rs-mmc card Mar 07 10:09:23 do they make storage cards? gps cards? Mar 07 10:09:37 champ: no, that's what bluetooth is for Mar 07 10:09:54 it has the builtin bluetooth and wirless doesnt it Mar 07 10:10:30 integrated Mar 07 10:10:45 champ: yes Mar 07 10:10:53 johnX: There's a lot of physically impossible stuff being discussed in that thread :-/ Mar 07 10:11:12 RP: what specifically? Mar 07 10:11:20 RP: you mean in terms of space available? Mar 07 10:11:24 koen: there is no keyboard is there? is there a usb port or do you have to have a bluetooth kb? Mar 07 10:11:24 hi zecke Mar 07 10:11:53 johnX: also, for wifi I am thinking of getting my rt2570 based usb key working Mar 07 10:11:58 johnX: Like trying to trace the BTUART pins - if I remember rightly, they're used for other things so adding a BT module to them isn't going to work Mar 07 10:12:17 it is detected and powered by the akita, but the drivers aren't yet in the kernel Mar 07 10:12:19 champ: http://maemo.org/maemowiki/Nokia_770_Hardware_Specification?highlight=%28omap1710%29 Mar 07 10:12:29 it looks way thinner than the c760 Mar 07 10:12:41 Its clearest in the 2.6 kernel what the GPIO pins are connected to Mar 07 10:12:49 RP: they seem to have tapped USB wires that go to the USB client port, and have a wifi card detected Mar 07 10:12:49 when did the 770 come out/ Mar 07 10:13:02 card -> usb donge Mar 07 10:13:09 koen: thanks for the link Mar 07 10:13:23 RP: yes but 2.6 sucks because oz sucks beacuse opie sucks because it can't run opera Mar 07 10:13:43 RP: that's the thought chain of the average oesf forum user Mar 07 10:14:05 koen: can I quote this? Mar 07 10:14:14 sure Mar 07 10:14:20 hehe :) Mar 07 10:14:59 will add it into sig on oesf ;) Mar 07 10:14:59 The fact these people are going to open their devices and poke around inside without at least reading the documentation available scares me... Mar 07 10:15:14 koen: still far too positive towards OpieZar Mar 07 10:15:26 koen: so do you have a zaurus too? Mar 07 10:15:33 champ: yes, a c700 Mar 07 10:16:00 RP: at least two of them (albertr and "Da Blitz") seem to have at least a vague idea of what they're doing Mar 07 10:16:10 the rest is kind of noise Mar 07 10:16:21 koen: which do you prefer the 700 or 770? Mar 07 10:16:38 koen: this looks odd "| ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied Mar 07 10:16:42 |Installing task-bootstrap (1.0-r25) to root... Mar 07 10:16:42 " Mar 07 10:16:45 minipanda: that's harmless Mar 07 10:17:00 minipanda: unless you want ipkg to install itself on your host system Mar 07 10:17:10 koen: oh, i don't Mar 07 10:17:19 minipanda: it seems that the dependencies didn't get built Mar 07 10:17:19 johnX: "vauge" but also incorrect Mar 07 10:17:28 http://www.oesf.org/forums/index.php?showtopic=18042 Mar 07 10:17:53 koen: which do you prefer the 700 or 770? Mar 07 10:18:06 champ: neither, I like the ipaq h2200 best Mar 07 10:18:18 small, fast & long battery life Mar 07 10:18:31 the ipaq huh? got a link off the top of your head? Mar 07 10:19:13 I really need to improve the battery life on the hx2750. WinCE appears to run forever without charging... Mar 07 10:19:42 champ: that one: http://dominion.kabel.utwente.nl/koen/pda/h2210/images/h2210-wince.jpg, they don't sell them anymore Mar 07 10:19:51 guys, how should I solve the subversion/neon dependency? For subversion needs libneon to have http/https dav support, and subversion's autoconf does not check for the libs, but for neon-config (installed to /usr/bin, not present in staging) Mar 07 10:20:18 a missing 'inherit bin-config'? Mar 07 10:20:20 does the ipaq run linux? Mar 07 10:20:43 i like the look of that ipaq koen Mar 07 10:20:43 also, after hacking that, subversion looks for neon includes in staging/include, not staging/include/neon where they are Mar 07 10:20:55 koen: in neon.bb? Mar 07 10:21:10 RP: ah, well. I'll wait and see: Either they blow up their Zaurii and it's no skin off my back, or they get it working and internal bluetooth / wifi become a possibility Mar 07 10:22:50 Bernardo: that's the most likely suspect Mar 07 10:23:49 koen: ok, I'll check that tonight Mar 07 10:24:13 koen: thanks Mar 07 10:24:35 i really want to try out one of those 770's Mar 07 10:25:16 champ: I don't know where you are, but here in the US they have them out to play with at CompUSA Mar 07 10:26:47 hi guys Mar 07 10:27:02 hey mickeyl Mar 07 10:27:03 hi mickeyl Mar 07 10:29:21 hi mickeyl Mar 07 10:29:26 hail mickeyl Mar 07 10:32:30 really? i haven't been to compusa in like 7 years, i hate that place Mar 07 10:33:02 champ: me too, but AFAIK they're like the only retailer here to have it out Mar 07 10:37:10 crappit qemu-native is broken Mar 07 10:39:05 Note: I just sent a mail about my OpenEmbedded ISO A0 Poster + 6 Pages Flyer plans. Please give input :) Mar 07 10:39:27 (via oe@handhelds.org) Mar 07 10:39:34 * XorA now needs gcc-native Mar 07 10:41:47 hi mickeyl Mar 07 10:41:50 koen: NOTE: exceptions.Exception:variable BBINCLUDELOGS references itself! while evaluating: Mar 07 10:41:51 ${BBINCLUDELOGS} Mar 07 10:41:51 NOTE: exceptions.Exception:variable BBDEBUG references itself! while evaluating: Mar 07 10:41:51 ${BBDEBUG} Mar 07 10:41:59 mickeyl: started buildmachine? Mar 07 10:42:20 koen: seems that BBDEBUG is a 'int' .. .but i see it in some local.conf BBDEUBG="yes" ?? Mar 07 10:43:51 brb Mar 07 10:46:45 hrw|work: what do we need from me - the world build or also the images? Mar 07 10:46:59 (a world build has been started some hours ago) Mar 07 10:49:18 so has anyone got an easy solution to the fact qemu doesnt build or according to error work when built with gcc4 on how to make qemu-native work? Mar 07 10:55:08 XorA: could you rephrase that question? Mar 07 10:57:31 pb__: qemu refuses to build with gcc4 saying that gcc4 doesnt produce a working qemu, my host has gcc4, therefore the qemu-native in OE fails to build, hence glibc in OE fails to build Mar 07 10:58:26 that is unfortunate. I guess there isn't any easy solution to that problem: you'll have to build a different gcc, or fix qemu, or stop using qemu. Mar 07 10:58:48 ouch Mar 07 10:59:22 re Mar 07 10:59:38 we have now created the need for gcc-native in OE, but Im not clever enough for gcc builds **** ENDING LOGGING AT Tue Mar 07 10:59:56 2006