**** BEGIN LOGGING AT Sat Mar 29 02:59:56 2008 Mar 29 03:16:19 03rwhitby 07org.oe.dev * r641cb531... 10/ (3 files in 3 dirs): linux-ixp4xx: fsg3 patch updated to match upstream submission Mar 29 03:24:19 NOTE: :name 'map_uboot_arch' is not defined while evaluating: Mar 29 03:24:20 ${@map_uboot_arch(bb.data.getVar('ARCH', d, 1), d)} Mar 29 03:24:27 something's broken Mar 29 03:25:30 configure: error: GMP 4.1 and MPFR 2.2.1 or newer versions required by fortran Mar 29 03:26:17 I'm trying to run bitbake nano as test and it's not able to configure gcc-cross Mar 29 03:39:43 hi mithro Mar 29 03:43:20 I'll try to rebuild gmp-native and mpfr-native Mar 29 03:48:27 hmm that did not work still getting configure: error: GMP 4.1 and MPFR 2.2.1 or newer versions required by fortran Mar 29 03:48:42 gmp-native and mpfr-natvie built fine Mar 29 03:50:19 ing for correct version of mpfr.h... yes Mar 29 03:50:29 checking for any version of mpfr.h... no Mar 29 03:50:42 hmm so correct version yes, any version no Mar 29 03:50:48 this may be the answer Mar 29 05:30:59 RP: Your assistance/insight with a meta-toolchain build issue would be greatly appreciated, when you find a spare moment: http://pastebin.com/m4b33c2ba Mar 29 07:28:05 how frustrating, I removed /stuff/build/tmp dir and started again and came up with same error sight Mar 29 07:28:19 looks like I need to wait for someone to fix it oh well Mar 29 07:28:33 configure: error: GMP 4.1 and MPFR 2.2.1 or newer versions required by fortran Mar 29 07:28:46 trying to build gcc-cross_4.2.2.bb Mar 29 07:29:07 I'll put in a bug report Mar 29 07:38:04 * * OE Bug 4130 has been created by dmoulder(AT)qtome.com Mar 29 07:38:06 * * gcc-cross 4.2. 2 does not configure due to problems with gmp and mpfr Mar 29 07:38:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4130 Mar 29 10:12:04 * * OE Bug 4131 has been created by andrea.adami(AT)gmail.com Mar 29 10:12:06 * * gtk+-directfb broken depends? Mar 29 10:12:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4131 Mar 29 10:33:11 03woglinde2 07org.oe.dev * rcbe4fc0e... 10/ (3 files in 3 dirs): (log message trimmed) Mar 29 10:33:11 linux 2.6.24: fix simpad rtc Mar 29 10:33:11 Revision: ead8fe4eb9f4a05e10d6ad072d127ba6aa6820f7 Mar 29 10:33:11 Ancestor: 557257ed260a028b08cdb8f9a859b3ea652317ee Mar 29 10:33:11 Author: woglinde2@openembedded.org Mar 29 10:33:12 Date: 2008-03-25T17:30:23 Mar 29 10:33:14 Branch: org.openembedded.dev Mar 29 10:33:16 03koen 07org.oe.stable * ra0a3cc00... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom 2007: bump to r14 Mar 29 11:17:34 'morning Mar 29 11:19:37 right now do_deploy fails in linux-handhelds-2.6.inc here because ${KERNEL_FILENAME} isn't defined. Mar 29 11:19:54 5042c98815b59089f6a52790f0dba3c831b5590d contains its removal, replacing it with KERNEL_IMAGE_BASE_NAME Mar 29 11:19:59 is this just a merge gone wrong? Mar 29 11:21:30 hm, no. looks like a missed occurence. Mar 29 11:28:07 mwester: You've put your finger on the problem, it needs the sysroot variable set in those flags Mar 29 11:29:25 03pH5 07org.oe.dev * r17cf1462... 10/ (1 packages/linux/linux-handhelds-2.6.inc): Mar 29 11:29:25 linux-handhelds-2.6.inc: replace stray ${KERNEL_FILENAME} Mar 29 11:29:25 with ${KERNEL_IMAGE_BASE_NAME}.bin Mar 29 11:29:25 * this change was missing from 5042c98815b59089f6a52790f0dba3c831b5590d, Mar 29 11:29:25 causing do_deploy to fail. Mar 29 12:04:05 * * OE Bug 4132 has been created by franco.bez(AT)web.de Mar 29 12:04:07 * * Missing Dependency to tzdata in x11-gpe-image Mar 29 12:04:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4132 Mar 29 12:06:50 03ifaistos 07org.oe.dev * r1f9b9f08... 10/ (1 packages/netpipes packages/netpipes/netpipes_4.2.bb): Mar 29 12:06:50 packages/netpipes/netpipes_4.2.bb : New package - netpipes Mar 29 12:06:50 A package to manipulate BSD TCP/IP stream sockets Mar 29 12:06:54 03ifaistos 07org.oe.dev * rc3368713... 10/ (1 packages/haserl/haserl_0.9.23.bb): packages/haserl/haserl_0.9.23.bb : Update haselr to latest version Mar 29 12:48:48 mwester: fix pushed Mar 29 12:49:02 hi rp Mar 29 12:49:05 morning folks Mar 29 12:49:09 hi woglinde, mickeyl Mar 29 12:49:25 03rpurdie 07org.oe.dev * r77c5b906... 10/ (38 files in 2 dirs): gcc: Include TARGET_CC_ARCH in ARCH_FLAGS_FOR_TARGET by default for everyone, not just as a slugos specific override. Mar 29 12:50:46 gm Mar 29 12:51:10 mickeyl, why hasn't your wife made you go sliing in the US? Mar 29 12:51:22 er skiing Mar 29 12:52:00 hi crofton Mar 29 12:52:10 hi woglinde Mar 29 12:52:40 hi all Mar 29 12:53:47 hi flo Mar 29 12:54:55 Crofton: some of group need to save money and we only have one week, so we always go to austria Mar 29 12:55:20 Crofton: actually, this year i will not ski, but continue writing that Linux on Smartphones book Mar 29 12:55:44 We have seen a number of European's skiing Mar 29 12:56:06 we stopped by an outlet store, and saw more stocking up Mar 29 12:56:38 apparently the exchange rate has boosted the number of people coming from Europe Mar 29 12:58:25 Crofton: stocking up sounds Dutch :-) Mar 29 13:03:58 crofton I dont know why so much people went through us-bullying bei the entry Mar 29 13:04:19 s/bei/in Mar 29 13:25:53 03mickeyl 07org.oe.dev * rd64ac77a... 10/ (3 files in 3 dirs): gsm0710muxd svn catch up version number with upstream, it's 0.9.0 now Mar 29 13:29:57 does anybody know a German distributor of the Fujitsu U810 UMPC by chance? Mar 29 14:08:46 hi mr_nice Mar 29 14:10:36 woglinde: hi Mar 29 14:17:21 Good Morning. Mar 29 14:19:36 hi mwester Mar 29 14:19:44 hi zecke Mar 29 14:19:45 hi Mar 29 14:24:38 moin Mar 29 14:24:39 zecke: I have no idea about the mesa error, looks like ipkg breaking somehow Mar 29 14:24:56 RP: was only once, yeah looked really weird Mar 29 14:38:57 ~lart people doing a mtn drop for resolving NCC's Mar 29 14:38:58 * ibot pulls out a ClueBat (tm) and thwaps people doing a mtn drop for resolving NCC's Mar 29 14:39:17 ncc? Mar 29 14:39:48 non content conflict Mar 29 14:41:32 woglinde: this is like a stupid CIA trick. You are the creator of a dir and a file, and next time you do mtn log your name is gone Mar 29 14:41:46 woglinde: and a broken copy of the file gets put there Mar 29 14:43:27 packages/openmoko-projects/illume_svn.bb Mar 29 15:09:59 RP: Re ARCH_FLAGS_FOR_TARGET -- I examined your fix, and enlightenment has walloped me over the head. It is the case, then that both ARCH_FLAGS_FOR_TARGET_slugos and ARCH_FLAGS_FOR_TARGET are set to different values, but the actual expansion of ARCH_FLAGS_FOR_TARGET occurs very late (after both are set), and selects only one of the two? Mar 29 15:13:05 so much for daggy fixes... Mar 29 15:13:31 * mwester googles "daggy" ;) Mar 29 15:13:42 mwester: ARCH_FLAGS_FOR_TARGET_slugos is an override and totally repaces the value of ARCH_FLAGS_FOR_TARGET Mar 29 15:14:01 mwester: It should really have been a TARGET_append_slugos Mar 29 15:14:13 Ahhh... Mar 29 15:14:33 mwester: However I've been looking at them for a while with a view to everyone using them Mar 29 15:14:40 Now to be sure I understand this, the assignment to XYZ_slugos does not replace the value XYZ at assignment time? Mar 29 15:14:54 rather at some later (expansion) time? Mar 29 15:15:10 mwester: Yes, when overrides are "expanded" Mar 29 15:15:20 which is typically when parsing completes Mar 29 15:15:44 Ok. Much now falls into place. Many thanks! :) Mar 29 15:16:53 BTW, testing is still required, but I now have an external toolchain for SlugOS. Also, packages staging builds completed without error as well. Mar 29 15:17:20 mwester: testing is good :) Mar 29 15:17:51 mwester: Like I said, I don't claim this stuff is bug free, just that we really need people to start playing with it :) Mar 29 15:33:53 hi mrdata Mar 29 15:34:00 woglinde: hi Mar 29 16:37:02 03mickeyl 07org.oe.dev * rcd2e3e62... 10/ (1 conf/distro/include/sane-srcdates.inc): sane-srcdates.inc: bump EFL_SRCDATE, pkgconfig detection works again Mar 29 16:37:06 03mickeyl 07org.oe.dev * r5ee427c7... 10/ (1 classes/efl.bbclass): efl.bbclass: ship .debug files in proper package Mar 29 16:37:11 03mickeyl 07org.oe.dev * rf818d337... 10/ (1 packages/efl1/etk_cvs.bb): etk cvs packaging fixes Mar 29 16:51:04 * * OE Bug 4133 has been created by bz-oe(AT)kdzbn.homelinux.net Mar 29 16:51:06 * * gtk (from python-pygtk) can't be imported from Python Mar 29 16:51:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4133 Mar 29 19:32:44 hi Mar 29 19:33:05 moin Mar 29 19:33:23 hey likewise Mar 29 19:36:15 hi jineld Mar 29 19:36:18 hi likewise Mar 29 19:42:19 Salutations! Mar 29 19:43:13 hi woglinde Mar 29 19:48:28 woglinde: how is the right workaround for gnutls-2.3.2-r0, first it wanted automake-1.10 but only automake-1.9.3 is currently available Mar 29 19:49:23 hu Mar 29 19:49:40 than some didnt checked Mar 29 19:49:40 also include files for gnutls are not complete, opencdk.h is missing there Mar 29 19:52:47 i patched configure.in to use automake-1.9.3 and copied opencdk.h to propper place Mar 29 19:54:29 mrdata you can modify the depends in the .bb Mar 29 19:55:03 so autotools < 1.10 ist used and opencdk is another package Mar 29 19:55:28 libopencdk10 - Open Crypto Development Kit (OpenCDK) (runtime) Mar 29 21:07:18 i use moko makefile and have two dirs with similar content: openembedded and org.openembedded.dev - is that normal? Mar 29 21:07:31 link? Mar 29 21:07:57 link what? no, it's no symlink.. Mar 29 21:08:14 hm Mar 29 21:13:34 emdete: hey, sounds like an old overlay (bb collections) Mar 29 21:13:57 zecke_: startet with mokomakefile yesterday. Mar 29 21:14:05 emdete: this allows to put fixed versions of OE into the openembedded until it is fixed in OE, but I have no idea if this is still used by the makefile :) Mar 29 21:14:09 emdete: ah then ;) Mar 29 21:14:19 what then? Mar 29 21:14:47 http://bitbake.berlios.de/manual/ch04s02.html#id870544 Mar 29 21:15:24 emdete: FYI: http://git.openmoko.org/?p=qtopia.git;a=blob;f=src/libraries/qtopiacomm/serial/gsm0710.c;h=e1a01cb86f00c51c9412e3744b3703613da3566f;hb=38bf8cfa6544114f603b6ee5d5cd5fb35a09d569 Mar 29 21:15:42 emdete: http://git.openmoko.org/?p=qtopia.git;a=blob;f=src/libraries/qtopiacomm/serial/gsm0710_p.h;h=81cd0fdf450d034a5a746160dec0f5944816bf2f;hb=38bf8cfa6544114f603b6ee5d5cd5fb35a09d569 Mar 29 21:16:53 zecke_: nice links. so what? Mar 29 21:17:14 does a muxer help building? Mar 29 21:17:33 emdete: no, two unrelated threads/things Mar 29 21:17:56 zecke_: comment the links if you change topic ;) Mar 29 21:18:37 and: the first link does not help.. i dont understand oe enough. no time for that. thats the cause i use mokomakefile Mar 29 21:18:41 emdete: you didn't say you have something related to building. You asked why you have two similar named directories. I assume it is bb collection related, you can find out by grepping through your configs for the strings from the bitbake manual Mar 29 21:19:24 zecke_: i don't even know what "my config files" are ;) Mar 29 21:20:44 emdete: MUXER: I'm a fan of good separations. As an example I want to show you Qtopia's design. framing/checksumming/channel allocation is the lowest layer, implemented in the files above, then the next layer is flow control, handling the logic for ack's, timeouts, requests for allocating and closing channels, the next one would be on top and be dbus or such Mar 29 21:21:06 emdete: CONFIG: you will have some directories called conf, they contain config files :) Mar 29 21:21:22 and: yes a also have a build problem. step by step: i cant build and get an error edbus missing. mickey said he fixed that yesterday, i need a ml3 in a bb. i have two of those bb files in those similar dirs. i don't know why those exist Mar 29 21:21:26 emdete: CONFIG: But yeah, I'm not helpful as I don't use moko makefile and don't know it too well (5 minutes exposure) Mar 29 21:21:55 zecke_: go ahead with qt. i dont care. no time to discuss. Mar 29 21:23:29 emdete: hehe, the two above links are plain c, but you got no time, :) Mar 29 21:23:36 zecke_: if you get the qt muxer running as a unix daemon using dbus i will throw away my code. no problem. Mar 29 21:23:49 emdete: you want to wait for rwhitby to wake up and help :) Mar 29 21:24:18 emdete: this was not my intention with posting the links. Mar 29 21:24:20 zecke_: for the muxer? Mar 29 21:24:37 zecke_: does qt muxer support PSC correctly? Mar 29 21:25:36 zecke_: dont understand relation of time and links to c code ;) Mar 29 21:26:30 emdete: hehe Mar 29 21:27:01 zecke_: dont understad me wrong. i am a pragmatic person and use what is usable. Mar 29 21:27:04 anybody using tailor to sync org.openembedded.dev with git? Mar 29 21:27:25 emdete: anyway. regarding your build issue: I have no idea, I don't use the moko makefile Mar 29 21:28:23 emdete: your two dirs are likely to be bitbake collections, I posted the link to bitbake collections, otherwise GettingStarted on openembedded.org explains basic config files (e.g. conf/local.conf) Mar 29 21:28:59 Esben: what is wrong with the OE git mirror on repo.or.cz? Mar 29 21:29:49 zecke_: prolly nothing :-) Mar 29 21:30:07 how often is it being updated? Mar 29 21:30:53 Esben: whenever I feel like Mar 29 21:31:07 zecke_: i startet using oe without mokomakefile which did not work out. so i went back to it. i made everything like the dog says but got bb not to compile something. the doc may be wrong... Mar 29 21:31:18 s/dog/doc/ Mar 29 21:31:44 emdete: I have no idea. I just know XorA and John are quite sensitive to not break things in org.openmoko.dev Mar 29 21:31:57 emdete: you, went back? Mar 29 21:32:02 emdete: wait a second :) Mar 29 21:32:33 zecke_: the doc seems to be wrong. probably oe is not broken, only the doc :D Mar 29 21:32:56 Esben: the script/tool to convert is inside the contrib directory of OE, if you want to run it yourself Mar 29 21:33:58 Esben: computer programs are never broken, they work really well and deterministic. Sometimes they do things we don't like and blame them :) Mar 29 21:34:19 emdete: computer programs are never broken, they work really well and deterministic. Sometimes they do things we don't like and blame them :) Mar 29 21:34:22 emdete: http://lists.openmoko.org/pipermail/distro-devel/2008-March/000280.html Mar 29 21:34:35 emdete: did you read the above mail from rod? does it apply for you? Mar 29 21:34:47 zecke_: should I just run mtn2git.py within an empty git repository? Mar 29 21:35:12 zecke_: yes, after going back to mokomakefile i did that Mar 29 21:37:14 emdete: no idea then :) Mar 29 21:37:32 zecke_: okay Mar 29 21:37:52 zecke_: thnx. so when can we see a working muxer from u? Mar 29 21:38:50 emdete: the Qtopia muxer does work (tm), it is disabled due suspend/resume issues and the issue you see yourself. That the modem goes to deep sleep after five seconds and wakes up and loses the first/more bytes :) Mar 29 21:39:18 zecke_: so we can use a arbritrary gsmd on top of that? Mar 29 21:39:42 zecke_: as you said above it is be nice, modular code.. Mar 29 21:40:01 zecke_: or can qtopia's gsmd used by a different gui? Mar 29 21:40:37 emdete: right now. If you don't mind linking to non gui Qt libs and have event loop integration, sure :) Mar 29 21:40:54 Esben: I want to paste the script, but my GPRS link is... Mar 29 21:41:22 zecke_: that's what i would name modular nor find unix-like... ;) Mar 29 21:42:09 emdete: this is fact twisting. Check the header file. you can hook anything on top of the core GSM 07.10 muxer :) Mar 29 21:42:18 s/would/wouldnt/ Mar 29 21:42:22 emdete: in contrast to everything in one blob including dbus handling :) Mar 29 21:42:43 zecke_: would you drop dbus from linux? hahaha Mar 29 21:42:48 zecke_: on the neo, multiplexing is disabled because we cannot get the modem to work correctly, or at all, in that mode. Mar 29 21:43:05 emdete: has it been a usecase for Trolltech to not use their GSM stack, no, have they proper abstraction to allow such things, yes Mar 29 21:43:08 ljp: with qt muxer? Mar 29 21:43:24 zecke_: qtopia is a non-unix. yes. Mar 29 21:43:26 ljp: I have seen the comment in the fic multiplexer plugin Mar 29 21:43:43 zecke_: another point for another muxer Mar 29 21:43:59 emdete: lol, fact twisting... I don't mind. If you think it is unix to put three things + dead code in one file, you are right :) Mar 29 21:44:07 if multiplexing worked, we could then enable GPRS Mar 29 21:44:25 ljp: if you /could/ do gprs, yes.. :D Mar 29 21:44:49 zecke_: what dead code do you talking about? Mar 29 21:44:53 gprs works fine Mar 29 21:45:04 ljp: in muxer on neo? Mar 29 21:45:12 not on neo, Mar 29 21:45:21 ljp: you are funny :D Mar 29 21:45:47 i mean tthat gprs is dependent on a working multiplexer Mar 29 21:45:47 emdete: the one in comments. I blame Trolltech a lot for the quality of Qtopia, the muxer is one of the better things. you could learn something from it (design and abstraction wise). If you don't want to, I don't care :) Mar 29 21:46:50 zecke_: i know the weak points of my muxer but it works. thats the point. if i wouldnt loose my time in oe i would polish it - shure :D Mar 29 21:47:44 emdete: and regarding PSC. This GSM 07.10 standard is pretty weak, e.g. check the multiplexer guidelines from Siemens and how they 'derived' the PSC command :) Mar 29 21:48:13 emdete: it is their replacement for AT+CFUN in muxer mode Mar 29 21:48:16 zecke_: url? and: what does siemens help if the modem does nothing here? Mar 29 21:52:07 emdete: the point is: You can always go and say: "My interpretation of the standard is XYZ and modem FOO is violating it". But so such life. E.g. for the Calypso and the software we run on it, PSC is useles (to my knowledge), it goes to deep sleep whenever it wants to and is dropping the first byte on the uart to wakeup Mar 29 21:52:47 (the latter is a hardware limitation), or even expensive modules like the ones from Siemens don't implement PSC the way you and me read the standard :) Mar 29 21:52:59 zecke_: if that is true my muxer is fine. Mar 29 21:53:26 (this is my pragmatic side), the only use of PSC and the TEST command I have seen so far (well I only know three modules) has been on Siemens and their abuse of PSC Mar 29 21:54:23 my pragmatic side says: the muxer works, i go with it. Mar 29 21:54:36 Esben: http://rafb.net/p/NLEIzJ95.html this is the script I use to update the mirror Mar 29 21:54:48 my other side says: oe build does not work an steals my time :D Mar 29 21:55:43 emdete: hehe, I go with ugly code does not work, I bash myself for it, I bash Trolltech (see their qtopia-interest list), I _inform_ you :) Mar 29 21:56:44 emdete: http://mc75.ru/lib/mux_drv_dev_guide_v05.pdf I think this is the link regarding the muxer on Siemens modules (yes we do not have a siemens one, it gives some insight though) Mar 29 21:58:02 emdete: and I agree, OE is stealing your time (if it doesn't build on random things) Mar 29 21:59:12 (eek wasn't the right manual) Mar 29 22:01:22 zecke: you dont need to inform me about qtopia code. i know that code quite well. i played around with a greenhone for a while Mar 29 22:01:33 http://www.mobiltim.com/images/modul/tc45modul/mux_guide_v05.pdf this sounds better Mar 29 22:01:50 zecke: compared to oe a greenphone&qtopia is a time after burner! :D Mar 29 22:02:36 emdete: I agree, if one would ask me if I like my Qtopia/X11 task, the answer is not at all. Mar 29 22:02:43 anyway more suspend/resume work todo Mar 29 22:03:38 zecke: and dont get me wrong - i like to see qtopia running well on a neo. it will raise numbers :) Mar 29 22:05:05 emdete: this is the only reason I do it Mar 29 22:07:22 zecke: thnx alot btw :-) Mar 29 23:02:21 bye Mar 29 23:08:35 hi all Mar 29 23:12:00 hi darmou Mar 29 23:23:47 bye all Mar 29 23:32:07 good nite Mar 30 00:50:47 its great that bug reports are answered so quickly I put in a report yesterday and got a reply only a few hours later with a suggestion Mar 30 00:52:00 fingers crossed it will work, I removed the tmp folder and am rebuilding from scratch so I'll see what happens Mar 30 00:52:36 hi greentux_ **** ENDING LOGGING AT Sun Mar 30 02:59:56 2008