**** BEGIN LOGGING AT Fri Feb 16 02:59:58 2007 Feb 16 04:04:35 03Tom 07org.oe.dev * r6f55241b... 10/ (1 packages/libxml/libxml2-native_2.6.26.bb): libxml: update libxml2-native to 2.6.26. Closes 1837. Thanks. Feb 16 04:04:41 03Tom 07org.oe.dev * rff4a9eec... 10/ (1 packages/libxslt/libxslt-native_1.1.17.bb): libxslt: add native package. Closes 1839. Feb 16 07:01:39 03koen 07org.oe.dev * rc7b1f1c0... 10/ (1 conf/distro/include/sane-srcdates.inc): sane-srcdates: bump srcdate for EDS-dbus at Ross' recommendation Feb 16 07:01:46 03koen 07org.oe.dev * r0d6f4236... 10/ (1 packages/openmoko-pim/openmoko-dates_svn.bb): openmoko-dates: catch up with eds and autotools changes Feb 16 07:01:55 03koen 07org.oe.dev * r0c0838b4... 10/ (9 files in 3 dirs): linux-gta01: update to 2.6.20 Feb 16 07:05:56 morning Feb 16 07:07:11 hrw: good morning Feb 16 07:08:29 koen: ross bumped esd-dbus in poky yesterday Feb 16 07:08:54 [08:34] CIA-20: koen org.oe.dev * rc7b1f1c0... / (1 conf/distro/include/sane-srcdates.inc): sane-srcdates: bump srcdate for EDS-dbus at Ross' recommendation Feb 16 07:08:54 like that ? Feb 16 07:08:57 probably Feb 16 07:09:08 koen: poky has also granulated eds packaging Feb 16 07:09:37 I need to run poky build Feb 16 07:09:47 hrw: heh Feb 16 07:09:58 ~lart amd64 for not working with 3 dimms Feb 16 07:09:58 * ibot lowers amd64's priority for not working with 3 dimms Feb 16 07:10:01 hrw: poky got the granulated packaging for OE, which I did :) Feb 16 07:10:16 * hrw has 1.5GB ram instead of 2 :( Feb 16 07:10:26 good morning guys. Feb 16 07:10:48 koen: iirc guys manipulated it last days Feb 16 07:10:51 Laibsch: hi Feb 16 07:15:51 hrw: I had a good laugh yesterday reading OESF Feb 16 07:16:06 hrw: pdaX wants more devels, but not more than 3 Feb 16 07:16:11 ;DD Feb 16 07:16:23 looks like time to look there Feb 16 07:16:30 "sorry, all our dev spots are taken now, please get lost" Feb 16 07:16:42 (if you read between the lines) Feb 16 07:18:37 remember thread? Feb 16 07:19:23 hrw: http://www.oesf.org/forums/index.php?showtopic=23242&view=findpost&p=154184 Feb 16 07:21:35 thx Feb 16 07:21:56 * koen builds a small oabi toolchain to build uboot Feb 16 07:32:38 03koen 07org.oe.dev * r85ec50cc... 10/ (1 packages/eds/eds-dbus_svn.bb): eds-dbus: sync packaging with poky Feb 16 07:37:17 zecke: morning Feb 16 07:37:26 koen: so more changes for eds? Feb 16 07:42:54 How come that I can build task-base, but when I build opie-image, all of the sudden not all dependencies are satisfied: http://rafb.net/p/I4wOB238.html Feb 16 07:48:20 Laibsch: avoid bitbake shell? Feb 16 07:48:35 if that helps, OK Feb 16 07:48:44 Do you think this might have caused it? Feb 16 07:51:39 I had problem with it Feb 16 07:55:23 hrw: we need to release bitbake-1.6.6 (now) Feb 16 07:56:16 zecke: noticed. Feb 16 08:00:27 zecke: http://www.openembedded.org/bonsai/view/scm/6/ doesn't seem to get updated anymore Feb 16 08:00:47 zecke: neither does http://www.openembedded.org/bonsai/view/scm/3/ Feb 16 08:00:52 hmm Feb 16 08:00:56 fsck.. second hang today.. Feb 16 08:00:59 broken cron? Feb 16 08:02:48 security update, be back later Feb 16 08:08:27 shit. dns failed Feb 16 08:11:11 greetings ! Feb 16 08:48:04 re Feb 16 08:54:07 howdy y'all Feb 16 08:59:17 cu Feb 16 08:59:28 time to test memory Feb 16 09:07:11 hi all ... Feb 16 09:07:47 i have completelly off-topic question ... how i can print the fosdem grid ??? Feb 16 09:09:35 hi psokolovsky_ Feb 16 09:10:50 gremlin[it], Hi! So, your bug was fixed, just do new build. Feb 16 09:11:16 ok il'' do this evening ... !!! Feb 16 09:17:06 good morning Feb 16 09:35:04 03koen 07org.oe.dev * r069cb75c... 10/ (1 site/arm-linux): site/arm-linux: remove forced unwind, glibc/oabi doesn't like it Feb 16 09:35:18 03koen 07org.oe.dev * r80417389... 10/ (11 files in 4 dirs): s3c2410 utils: merge from openmoko svn Feb 16 09:54:04 http://www.openembedded.org/bonsai/view/rev/24536/ that is clueless^3 Feb 16 09:55:13 http://www.openembedded.org/viewmtn/diff.psp?id1=762759b3b4d3a2e666f5c0ac9446fabdc7617929&id2=0a027f255ab4229460aa779443c611039530e4b3&fname=packages/linux/linux_2.6.20.bb Feb 16 09:55:21 IMAGE_TYPE = "bzImage" Feb 16 09:55:29 IMAGE_TYPE_simpad = "bzImage" Feb 16 09:55:32 zecke: I have fix for it Feb 16 09:55:39 had problem to push it yesterday Feb 16 09:56:16 looks like today monotone.oe.org is still too busy for me Feb 16 09:59:06 koen|away: how do I see who comitted this? Feb 16 10:00:15 zecke: it was me and bzImage was mistake Feb 16 10:00:24 fixed soon after Feb 16 10:00:37 hrw: did you look at the mq200 patch Feb 16 10:01:01 hrw: only 10% of these patch are mq200 stuff, the other stuff should not be part of this file... Feb 16 10:01:12 zecke: nope Feb 16 10:01:51 http://rafb.net/p/fXcDlA16.html Feb 16 10:02:06 am I the only one who thinks *~ files should not be part of patches? Feb 16 10:02:24 03hrw 07org.oe.dev * r760aa490... 10/ (13 files in 6 dirs): Feb 16 10:02:24 lighttpd: lot of bugfixes Feb 16 10:02:24 - forced PCRE detection so it is linked in Feb 16 10:02:24 - cleaned directory structure Feb 16 10:02:24 - added /www/ dirs needed Feb 16 10:02:25 - added empty homepage Feb 16 10:02:27 - fixed initscript Feb 16 10:02:28 we are not mvista Feb 16 10:02:33 cleaned Feb 16 10:03:46 mr_nice == bernhard? Feb 16 10:04:52 his switches driver could have 37 lines less without sacrifcing readability :) Feb 16 10:05:20 if ( key == 0 ) reportKey0(); else if( key == 1 ) reportKey1() else if ... screams for an array of keys :) Feb 16 10:09:21 zecke: atleast someone tries to get 2.6 on simpad again Feb 16 10:10:23 pushed fixes Feb 16 10:10:24 hehe, I appreciate that. this is why I who bernhard is to give comments Feb 16 10:43:20 03zecke123 07bitbake-1.6 * r759 10/ (ChangeLog MANIFEST): Add more files to our MANIFEST. It looks complete enoug now Feb 16 10:45:14 zecke: ah.. I never touched MANIFEST file.. Feb 16 10:45:23 ;) I know Feb 16 10:45:35 another leasson learned Feb 16 10:49:42 'learnt' Feb 16 10:50:24 learned, learnt is not really english Feb 16 10:50:41 fscking inglish grammar Feb 16 10:51:14 its like spelt and spelled Feb 16 10:51:33 its in common use but is really english Feb 16 10:51:52 grr missing " ' " in its Feb 16 10:51:59 it's == it is Feb 16 10:52:12 its == belongs too Feb 16 10:52:16 grr to Feb 16 10:52:18 even Feb 16 10:52:22 * ade|desk gives up Feb 16 10:52:36 hehe Feb 16 10:52:43 dontcha love the english language Feb 16 10:52:49 I made a word up the other day Feb 16 10:53:03 in the vain of employee/employer Feb 16 10:53:12 I assumed there was a word insuree Feb 16 10:53:15 but there isn't Feb 16 10:53:22 i like meld, no such word, it is weld and melt together Feb 16 10:54:10 irregardless is another made up word Feb 16 10:54:16 hehe Feb 16 10:54:30 where you from ade|desk ? Feb 16 10:54:37 regardless, ok , irrespective is ok too Feb 16 10:54:49 UK Feb 16 10:54:58 whereabouts? Feb 16 10:55:10 i live 20 miles north of cambs Feb 16 10:55:24 work 25 miles north of that Feb 16 10:55:40 * Paperface tries to work out where that might be Feb 16 10:55:42 huntingdon? Feb 16 10:56:03 live near ely, work in king's lynn Feb 16 10:56:15 you? Feb 16 10:57:43 live in norwich, work in norwich ;) Feb 16 10:58:11 know ely/cambridge quite well as I went to uni in Cambridge Feb 16 11:01:56 i like norwich, always do christmas shopping there Feb 16 11:02:05 yeh, it is pretty nice Feb 16 11:02:33 and there is a maplins there too Feb 16 11:03:47 * ade|desk is a secret canary, but don't tell anyone Feb 16 11:03:49 FYI: bitbake git fetcher uses a '-n' switch which is not available on the git version shipped with RHEL and CentOS. Feb 16 11:04:16 * rwhitby continues building on Debian Etch instead Feb 16 11:07:43 03zecke123 * r760 10/ (6 files in 6 dirs): Another try: Release BitBake 1.6.6 with updated MANIFEST Feb 16 11:15:17 zecke: mtn automate certs iirc Feb 16 11:17:55 * koen just had 2 hours of FT -> DTFT -> DFT -> FFT joy Feb 16 11:18:05 now I'm finally allowed to use FFT in assignments :) Feb 16 11:20:50 03zecke123 * r761 10bitbake/MANIFEST: trunk/bitbake: Update the MANIFEST as well Feb 16 11:26:18 could someone have a look at this -> http://www.pastebin.ca/358780 Feb 16 11:26:59 when doing bitbake glibc and although glibc 2.5 is defined it say it can not find it and tries 2.3 Feb 16 11:27:34 but when trying to clear it with bitbake glibc -c clean it tries to clean the 2.5 .... Feb 16 11:27:37 missing PREFERRED_PROVIDER? Feb 16 11:27:51 1.6.6 of BitBake released Feb 16 11:28:50 koen : shouldn't in both case try to build and clean the same version ? Feb 16 11:29:02 I think it should Feb 16 11:29:53 zecke : do you think this is a bitbake issue ? Feb 16 11:31:06 Ifaistos: no Feb 16 11:31:39 do you use icecream? glibc might deserve a PARALLEL_MAKE="" Feb 16 11:33:31 zecke : no its on a single amd64 Feb 16 11:33:50 zecke : it does not use icecc Feb 16 11:42:28 koen: could you please add kuhlen's key? he is one sith folks sitting next to me and will do random bugfixing and testing Feb 16 11:43:07 zecke: I would if he had sent mickeyl and me his key Feb 16 11:44:04 koen: he is about to send it to you, I ruled out mickeyl as he seems to be asleep Feb 16 11:44:31 koen: he is probably checking your GnuPG key atm Feb 16 11:45:04 zecke: sending it to both of is a backup :) Feb 16 11:45:18 and it makes sure mickeyl and I know who has access Feb 16 11:58:54 hi all ... where i can find info about ipaq h63xx (h6340 exactly) ... i can buy one for 250Euo ... is a value price ? Feb 16 12:08:48 koen: bonsai stopped updating due a change of the mtn db Feb 16 12:09:45 zecke: even the svn ones? Feb 16 12:10:28 zecke: use a versioned binary Feb 16 12:10:31 e.g. /usr/bin/mtn-0.32 Feb 16 12:11:09 will do the upgrade and then use a versioned binary Feb 16 12:11:28 the svn ones are broken due a broken pysvn, but I will take a deeper look now Feb 16 12:13:08 drat Feb 16 12:13:25 zecke: sorry, I typo'd in a kill command Feb 16 12:13:53 lol Feb 16 12:13:58 again? Feb 16 12:15:10 no Feb 16 12:18:13 what version of monotone was used to create OE.mtn? Feb 16 12:19:08 0.32 Feb 16 12:19:16 as the GettingStarted wiki says Feb 16 12:19:53 03Bernhard 07org.oe.dev * re6225ed3... 10/ (3 files in 3 dirs): simpad: move USE_DEVFS to simpad-2.4 include Feb 16 12:19:57 03hrw 07org.oe.dev * rb5dacbd6... 10/ (2 files in 2 dirs): linux/simpad: cleaned mq200 patch from backup files, defconfigs - spotted by Zecke Feb 16 12:20:03 03koen 07org.oe.dev * r471af30b... 10/ (5 files in 4 dirs): eds stuff: bump PR to notice new EDS packaging Feb 16 12:20:07 03koen 07org.oe.dev * r63b307f0... 10/ (1 packages/eds/eds-dbus_svn.bb): eds-dbus: bump PR as well Feb 16 12:20:10 morning all Feb 16 12:20:12 03koen 07org.oe.dev * r487c97d1... 10/ (1 packages/glibc/glibc_2.5.bb): glibc: add a workaround for some strange shlib renaming problems, needs a proper fix Feb 16 12:21:22 hi RP Feb 16 12:21:27 hi RP Feb 16 12:21:31 hey RP Feb 16 12:22:12 RP: I'm a bit lost why glibc needs that PACKAGE_DYNAMIC Feb 16 12:23:06 koen: What happens without it? Feb 16 12:24:01 zecke: I can't compile bitbake, but I guess I've screwed up somehow: http://rafb.net/p/EvK2z372.html Feb 16 12:24:31 Other python programs install fine Feb 16 12:24:35 RP: RDEPENDS = "libc6-dev" doesn't work Feb 16 12:25:02 koen: Right, thats expected. You want glibc-dev Feb 16 12:25:02 koen: RDEPENDS = glibc-dev? Feb 16 12:25:20 last time I tried that, that didn't work as well Feb 16 12:25:23 * CM is afk Feb 16 12:26:02 koen: It should work. If not, I need to see the error Feb 16 12:26:20 CM: Looks like you have a previous bitbake installed you need to uninstall first? Feb 16 12:30:40 * rwhitby *never* installs bitbake Feb 16 12:38:10 re Feb 16 12:40:46 03koen 07org.oe.dev * r50793489... 10/ (1 packages/eds/eds-dbus_svn.bb): eds-dbus: rework packaging a bit to make it go from "not too broken" to "doesn't make ipkg freak out" Feb 16 12:50:06 CM: use 1.6.6 Feb 16 12:51:07 he zecke :( Feb 16 12:52:26 woglinde: yeah, I messed up the release :) Feb 16 12:53:45 zecke yeah I saw Feb 16 12:53:54 and I sucked in the exam Feb 16 12:53:59 :( Feb 16 12:56:21 zecke: I added the mtn key you requested Feb 16 13:00:07 thanks a lot Feb 16 13:05:20 zecke: That is is 1.6.6 :/ Feb 16 13:06:03 * koen shutsdown the router to change some pci cards Feb 16 13:07:23 CM: no it is not Feb 16 13:09:49 koen: your router has pci swappable cards? Feb 16 13:09:49 zecke: Ah, maybe I was too quick to download it from berlios or so. I'll re-fetch it Feb 16 13:14:26 cu Feb 16 13:15:54 zecke: I was in the untared 1.6.6 dir, but I still had 1.6.4 installed. Now I get almost the same error, but with PIL instead of bb: http://rafb.net/p/pxyNIZ25.html Feb 16 13:18:07 CM: never heard of PIL Feb 16 13:18:08 no matter how I set IMAGE_FSTYPES in my local.conf I still get ext2.gz, jffs2, and rootfs.tar.gz files in tmp/deploy/images, am I missing something? Feb 16 13:18:39 xumbi8: Is it set anywhere else? Feb 16 13:19:07 NAbyss: let me see... Feb 16 13:19:15 CM: this must be a local issue. In contrast to 1.6.4, I did not just package it, but tested the execution Feb 16 13:19:20 zecke: Me neither, but that's what I get running setup. I don't have any "site-packages" folder either. I guess gentoo is messing it up somehow. I'll see if I can figure it out Feb 16 13:19:52 would be appreciated, python -vvv setup.py can help Feb 16 13:20:03 I'll try that Feb 16 13:20:13 1.6.4 installed fine this morning Feb 16 13:20:19 NAbyss: I checked task-base.bb, and my machine .conf file, where else could it be set? Feb 16 13:20:36 The only thing I've done since that is to run rwithbys moko makefile Feb 16 13:21:11 NAbyss: whoops, I just answered my own question, I found it set in distro/generic.conf Feb 16 13:21:20 NAbyss: thanks though Feb 16 13:23:29 xumbi8: np :) grep is usually your friend in cases like that.. there's quite a few places things can be set Feb 16 13:24:05 NAbyss: definitely, I don't know why it didn't occur to me to look elsewhere Feb 16 13:36:03 xumbi8: consider the order of conf files in conf/bitbake.conf Feb 16 13:37:18 xumbi8: it seems like IMAGE_FSTYPES should be set with ?= in distro.conf so we can override it in local.conf. Feb 16 13:38:21 cbrake: hello Cliff, yes that would make more sense Feb 16 13:38:32 cbrake: thanks for the tip on bitbake.conf Feb 16 13:43:31 mickeyl: good morning! Feb 16 13:43:46 cheers Feb 16 13:47:20 03freyther 07org.oe.dev * rb866b827... 10/ (6 files in 5 dirs): Feb 16 13:47:20 Add the DNARD/Shark machine configuration and kernel Feb 16 13:47:20 The DNARD/Shark was the Network Computer constructed Feb 16 13:47:20 by DEC with spare StrongARM processors. Here we use them Feb 16 13:47:20 as music basestations. Feb 16 13:55:22 zecke: Thanks, found the problem. I had old files from 1.6.4 in /usr/lib64/python2.4/site-packages/bb/ messing up the import bb stuff because it looked there first. Feb 16 14:10:26 zecke: Hallo ! Feb 16 14:10:38 mickeyl: Hallo 2u2! Feb 16 14:11:17 schurig: Hi! Wanted to catch you for some time and ask what's the state of mnci machine in OE.dev? Feb 16 14:11:40 psokolovsky_: "Neglected" would be a good name :-) Feb 16 14:11:48 schurig: hi holger Feb 16 14:12:13 psokolovsky_: I once used OE to build the image and SDK for the MNCI-RX, but I didn't continue this Feb 16 14:12:23 psokolovsky_: Our customers still use the old SDK for their work Feb 16 14:12:41 schurig, well, would you like to update/maintain it? because I'd like to cleanup at least opie's machine deps, and mnci os one which has adhoc scripts. Feb 16 14:12:57 psokolovsky_: MNCI also runs still Kernel 2.4, another device of us is here newer, it runs with 2.6 (but for this device there is no SDK, just a CD with the GPL code) Feb 16 14:13:30 psokolovsky_: feel free to kill MNCI our of Opie. It turned out that no one really needed Opie on MNCI. I just used it for a short time as a demo. Feb 16 14:14:00 psokolovsky_: later we wrote our own demo for Logistics stuff, in Qt3/Embedded, then I never used Opie as long as it stayed Qt2/Embedded (not sure about it's current state) Feb 16 14:14:16 schurig, ok, thanks. I won't "kill" it, but will make as much of opie packages as possibel mach-independent. mnci config will be left for reference. Feb 16 14:14:28 psokolovsky_: we and customers need Qt's SQL stuff with an sqlite database, so Qt2 and therefore Opie was no option Feb 16 14:14:28 ok, thanks! Feb 16 14:15:37 psokolovsky_: Is Opie now Qt3 or Qt4 ? Feb 16 14:16:01 schurig, old good opie1 is still on qte2. Feb 16 14:16:15 schurig, there're opieII in works, on qte4 Feb 16 14:16:36 I kinda maintain old good opie1 in OE ;-) Feb 16 14:16:40 psokolovsky_: sound's better, althought I've never adapted Qte4 to MNCI or RT4000 Feb 16 14:16:49 psokolovsky_: did lorn ever lay out what he plans for OpieII? Feb 16 14:17:23 zecke: did TT ever lay out their plans with XYZ? Feb 16 14:17:33 zecke: does the shark really need a new kernel? Feb 16 14:17:42 zecke: couldn't it be added to linux_2.6.20? Feb 16 14:17:43 zecke, I didn't hear anything since initial qtopia4 gpl release was made. I know some folks build it, but I'm still waiting while it is added to OE ;-) Feb 16 14:17:47 schurig: yes, WCE in 2008 Feb 16 14:18:11 zecke: Wow, that would be interesting. We would instantly buy a license :-) Feb 16 14:18:17 koen_: yes, sadly the network driver needs a bad hack which will never be in mainline (I could extract that hack) Feb 16 14:18:30 schurig: I ported a Qt2/Qt3 subset to Windows Mobile last month ;) Feb 16 14:18:40 zecke: SRC_URI_shark += "file://nethack.patch"? Feb 16 14:18:46 zecke: if "WCE" is QtX-on-Windows-CE-dotNet-whatever-it's-currently-called Feb 16 14:19:19 zecke: yeah, but that would then be GPL by definition? Feb 16 14:19:46 schurig: yes, at TT DevDays they said Windows CE port will be released in 2008 and they just started with it (November 2006) Feb 16 14:20:08 zecke: re-started maybe. I once saw Qt on CE. Some years ago. Feb 16 14:20:12 schurig: I started with KWQ from Apple which is BSD (even if it contains verbatim copies of Qt... bad apple Inc) Feb 16 14:20:24 schurig: check for Q_OS_TEMP in you win32 copy of Qt Feb 16 14:20:54 koen_: I will consider just extracting this evil hack for the network driver Feb 16 14:20:55 zecke: Hu? Apple made a GPL/proprietary thing BSD-licensed? Feb 16 14:21:37 schurig: Off Topic here: They created WebKit. Extracted the SubSet of Qt used by KHTML, called their Qt KWQ and licensed it under a BSD license Feb 16 14:22:10 hi all Feb 16 14:22:17 schurig: somehow this code contains copies of functions that look 1:1 as the one in the original Qt. Well I started with that KWQ and created QWidget/QApplication.. stuff for Windows Mobile Feb 16 14:24:17 zecke: hi, do you have some time for me? Feb 16 14:25:09 little, caught a cold and try to catch up with the neglected things Feb 16 14:26:55 zecke: ah, ok. just one little thing. what was the reason for taking the 2.6 driver for the mq200 from uclinux and not trying to port the old 2.4 one? Feb 16 14:27:23 licensing Feb 16 14:27:41 the 2.4er one is illegal, it is property of NVidia Corporation Feb 16 14:27:54 zecke: ah, ok Feb 16 14:30:18 zecke: do you know if it is illegal to use the mq200 manual to write a driver? Feb 16 14:30:56 mr_nice: back then I believed that, but as google can find the manual nowadays I doubt there are any tradesecrets left... Feb 16 14:31:32 mr_nice: this is why I used the uclibc skeleton + memory dumps Feb 16 14:32:00 zecke: so you think I don't get into trouble if I will use the manual? Feb 16 14:32:50 mr_nice: If I would have time, I would use this manual Feb 16 14:33:00 zecke: ok Feb 16 14:33:27 mr_nice: just document where you have downloaded the manual from Feb 16 14:33:47 zecke: yes good idea Feb 16 14:35:34 does anyone know of a howto or a walkthrough of creating a bitbake recipe? Feb 16 14:35:56 I've found plenty of references, but nothing step-by-step Feb 16 14:36:03 is it easy to specify a *revision* from svn rather than a SRCDATE? Feb 16 14:36:46 xumbi8: did you check the manual? Feb 16 14:36:53 mccarthy: sure? Feb 16 14:37:04 hrw|gone: did you update the manual after changing the svn fetcher? Feb 16 14:37:31 zecke, google couldn't find it for me ... is there a good example around? Feb 16 14:37:44 mccarthy: as easy as adding ;rev=123 Feb 16 14:38:07 zecke: yes but I was looking for something "higher level" if possible Feb 16 14:38:34 zecke: I find looking at existing .bb files more helpful than reading the bitbake manual, so far Feb 16 14:39:22 xumbi8: conf/documentation.conf is also fairly useful. Feb 16 14:40:08 cbrake: ah cool, I will check that out Feb 16 14:40:14 xumbi8: http://www.openembedded.org/user-manual&dpage=commonuse_new_package like this? help us to create it Feb 16 14:40:56 xumbi8: http://www.openembedded.org/user-manual&dpage=recipes_examples is that not useful? a hint on why it is not helpful would be nice Feb 16 14:41:00 zecke: but that page just says "This section is a stub, help us by expanding it" Feb 16 14:41:28 xumbi8: yes, so help us? the other link is not a stub Feb 16 14:41:39 zecke: sorry, gotcha Feb 16 14:42:00 zecke: as I make progress I will see if I can add to that page Feb 16 14:44:17 re Feb 16 14:44:41 zecke: forgot... Feb 16 14:44:54 mccarthy: packages/tremor/ are good example Feb 16 14:47:12 hrw, zecke thanks Feb 16 14:50:31 hrw: thx for adding the patches :). I have uploaded the new patchfile to bugzilla Feb 16 14:50:58 mccarthy: about gpesyncd: libeventdb api changed in svn, so we need a gpesyncd in OE that works against the released versions of libeventdb, libtododb, etc Feb 16 14:51:27 koen_, I spoke with neal about that yesterday in #gpe Feb 16 14:51:28 * koen_ fixes the "no sound" problem by putting some firmware in /lib/firmware Feb 16 14:51:51 koen_: your device need firmware for sound? Feb 16 14:52:13 hrw: yes Feb 16 14:52:20 pvr150, hardware mpeg2 tuner Feb 16 14:52:31 koen_, he said that he would get on it today (I modified a local copy of gpesyncd so that it compile, but the intended error handling is probably broken since I made no effort to actually make it work) Feb 16 14:52:53 ironically I bought a pvr150 because my bt848 had a broken audio chip since last week Feb 16 15:08:07 what do people typically do with the graphviz *.dot files generated by bitbake? Is there a graphical format that is useful, or do you typically just look at the dot files? Feb 16 15:09:04 also, are there any tools that will generate a run-time dependency tree for ipk packages? Feb 16 15:13:11 is mtn.oe.org stuck somewhere? Feb 16 15:13:42 yeah Feb 16 15:13:43 we have a load of 5 Feb 16 15:13:50 guess it's mtn server restart time Feb 16 15:13:52 or something Feb 16 15:14:05 hmm, I regenerate rosters on serenity... Feb 16 15:14:11 oh uh Feb 16 15:14:12 * zecke looks innocent and offers cookies Feb 16 15:14:14 that's it Feb 16 15:14:18 * mickeyl takes one Feb 16 15:14:50 don't we have two cpu's in serenity? Feb 16 15:15:12 don't think so Feb 16 15:15:59 I think I'm guilty for the load Feb 16 15:16:12 morning Feb 16 15:16:35 man nice Feb 16 15:16:41 for the next time Feb 16 15:19:55 03koen 07org.oe.dev * raa2c7cca... 10/ (6 files in 4 dirs): libftdu: add libftdi from openmoko svn Feb 16 15:20:00 03koen 07org.oe.dev * r13f5cb98... 10/ (3 files in 2 dirs): openocd: merge openocd from openmoko svn Feb 16 15:20:06 03koen 07org.oe.dev * r9f203ec0... 10/ (4 files in 4 dirs): netbase: merge in fic-gta01 bits from openmoko svn Feb 16 15:20:11 03koen 07org.oe.dev * r1c587b28... 10/ (1 packages/lm_sensors/lmsensors-apps_2.10.1.bb): lmsensors: merge in changes from openmoko svn Feb 16 15:20:18 03koen 07org.oe.dev * r9ad1a49a... 10/ (6 files in 5 dirs): base-files: merge in fic-gta01 and qt2410 files from openmoko svn Feb 16 15:20:21 koen does not scream yet Feb 16 15:20:24 03freyther 07org.oe.dev * ra8bf0677... 10/ (1 classes/insane.bbclass): Feb 16 15:20:24 classes/insane.bbclass: Grep through config.log to check for CROSS Compile errors Feb 16 15:20:24 -Inspired by doku's work on mpd's buildsystem grep through Feb 16 15:20:24 the config.log and find broken autotools tests Feb 16 15:20:24 -Make errors fatal so we will notice these things more quickly. Feb 16 15:21:20 Can OE do a canadian cross and build my toolchain for cygwin? Feb 16 15:21:32 * koen wonders why oe@ltg now has 3 screens running Feb 16 15:21:35 cbrake: Typically I read the dot files. If you find a good viewer let me know :) Feb 16 15:21:50 RP: if you have OSX I know one Feb 16 15:22:18 RP: you should mallum to get you a mac Feb 16 15:22:25 you should ask... wow I'm tired Feb 16 15:23:44 zecke: I don't think that a dot visualiser will be quite enough to convince him ;-) Feb 16 15:24:29 you could run two instances of the dot visualiser (wow today I forget more verbs than ....)? Feb 16 15:25:12 zecke: That would make all the difference :) Feb 16 15:25:49 hmm Feb 16 15:25:55 shall i get a MacPro or a MacBookPro? Feb 16 15:25:58 * mickeyl undecided Feb 16 15:26:06 mickeyl: desktop or laptop? Feb 16 15:26:13 get flithy rich and buy both Feb 16 15:26:15 that's the question Feb 16 15:26:16 hah Feb 16 15:26:21 not with my current salary Feb 16 15:26:28 i know... Feb 16 15:26:44 forget about this OpenSource crap and get flithy rich :) Feb 16 15:27:02 mickeyl: I think your IBM will last another year or two (it is not as sexy though) Feb 16 15:27:11 emphasis on filthy Feb 16 15:27:29 mickeyl: so get a MacPro, it should make a good replacement for your dual xeon Feb 16 15:27:34 zecke: yeah. i'm really satisfied with the thinkpad Feb 16 15:27:38 zecke: i guess i will do that Feb 16 15:27:47 excellent Feb 16 15:27:58 you think i'll have more performance even through the parallels overhead? Feb 16 15:28:05 I never use my laptop, can't type on it's nastly little keyboard Feb 16 15:28:57 mickeyl: should be okay Feb 16 15:29:04 *nod* Feb 16 15:29:20 * mickeyl notes for April Feb 16 15:29:28 I must admit my thinkpad does its job well... Feb 16 15:29:42 koen: I want to make insanity the default (insane.bbclass) soon Feb 16 15:29:42 * mickeyl agrees Feb 16 15:29:49 RP: you have a r60? Feb 16 15:30:00 zecke: z60m Feb 16 15:30:14 zecke: don't do bb.fatal on .so files being in the main package Feb 16 15:30:32 zecke: think about plugin based things like perl, gnome-vfs, gtk, etc Feb 16 15:30:33 koen: it is only okay for plugins Feb 16 15:30:42 zecke: There are some issues with, we had to disable it for a handful of packages in poky Feb 16 15:31:44 RP: I want to see the issues now Feb 16 15:31:44 zecke: but yes, insane should be the default Feb 16 15:31:50 zecke: Grep a poky checkout for INSANE_SKIP_ Feb 16 15:32:19 RP: hehe, will do Feb 16 15:32:57 zecke: Looks like gcc, bg and elfutils are broken at least Feb 16 15:33:24 RP: I like unsound incomplete tools, SKIPPING is the obvious answer for false positives but dangerous as well (just because they are not broken now, does not mean they will not be broken soon) Feb 16 15:33:59 koen: do you know when pysvn was updated? Feb 16 15:34:10 zecke: yesterday Feb 16 15:34:16 zecke: I know. I was reluncant to let that happen but didn't have time to find ross an alternative Feb 16 15:34:19 zecke: pester florian about it Feb 16 15:34:29 RP: hehe Feb 16 15:34:44 RP: did it find some bugs though? Feb 16 15:34:45 MacPro is dual xeon Feb 16 15:34:53 zecke: a few, yes :) Feb 16 15:35:05 hrw: hmm yeah. so the question is will it really be an upgrade Feb 16 15:35:07 RP: so it was no complete waste of time Feb 16 15:35:17 zecke: I've agreed to enable it on our autobuilder although thats not happened quite yet Feb 16 15:35:28 zecke: No, I think its a great idea Feb 16 15:35:57 mickeyl: it has server xeons Feb 16 15:36:44 mickeyl: and add tons of ram to the macpro, you won't regret it Feb 16 15:37:31 hmm Feb 16 15:37:36 tons = Feb 16 15:37:37 2G? Feb 16 15:37:38 4G? Feb 16 15:37:56 mickeyl: ever heard of highmem support? Feb 16 15:37:59 mickeyl: at least 2G Feb 16 15:39:21 morning Feb 16 15:39:31 hi kergoth Feb 16 15:40:03 hey kergoth Feb 16 15:40:11 mickeyl: iirc macpro support 16GB Feb 16 15:40:19 heh Feb 16 15:40:28 i'm not a billionaire Feb 16 15:40:29 ah and obviously don't buy ram from apple Feb 16 15:40:55 then you *would* have been a billionare. Feb 16 15:40:57 right, use the kingston online configurator and write down the type Feb 16 15:49:19 RP: OE has an updated http://www.openembedded.org/repo/org.openembedded.dev/packages/xserver-kdrive-common/xserver-kdrive-common/etc/X11/Xserver , could be usefull for poky as well Feb 16 15:51:46 koen: You mean the Xw100 bits? Feb 16 15:51:53 or the dpi? Feb 16 15:52:33 Xw100, Ximageon, dpi bits, extra devices Feb 16 15:52:58 it merges in all updates since the fork Feb 16 15:53:00 koen: w100 will be soon tested on e740 Feb 16 15:53:08 hrw: yes, I heard Feb 16 15:53:12 koen: somehow a bogus space was in the path, bonsai updates work again Feb 16 15:53:23 zecke: cool, thanks for looking at it Feb 16 15:53:30 koen: I'll have to run the dpi changes past mallum and he's rather busy atm. I'll keep it in mind though, thanks Feb 16 15:53:53 RP: I'll send a patch to florian to merge in the o-hand changes into the upstream gpe one Feb 16 15:54:05 actually working on it is on my agenda now, well before my boss sells my soul for further windows work ;) Feb 16 15:54:19 koen: ok, thanks Feb 16 15:54:38 RP: in OE we set fontconfig DPI to 160, so the change affect that Feb 16 15:54:50 but a test would be good :) Feb 16 15:55:29 koen: Where is the fontconfig dpi set? Feb 16 15:55:49 * koen thinks Feb 16 15:56:04 /etc/fonts/local.conf iirc Feb 16 15:56:04 hrw: hi, did you send the invoice? to where? essen or dresden? Feb 16 15:56:09 I don't really understand this dpi business/problems... Feb 16 15:56:13 coredump wrote a script for it Feb 16 15:58:38 greentux: did had a time to send it. can we /query? Feb 16 15:58:43 koen: Which package is that in? Feb 16 16:00:22 ./packages/xserver-common/files/setDPI.sh Feb 16 16:02:03 koen: thanks. I think that might solve some of my dpi concerns :) Feb 16 16:03:06 koen, hrw: I have a problem now. Can you help me shed some light to the two questions I raised in Feb 16 16:03:10 !oebug 1877 Feb 16 16:03:12 * * Bug 1877, Status: NEW, Created: 2007-02-15 18:17 Feb 16 16:03:13 * * bugs.openembedded.org(AT)rolf.leggewie.biz: linux-rp-2.6.20 will not build for collie (do_patch fails) Feb 16 16:03:14 * * http://bugs.openembedded.org/show_bug.cgi?id=1877 Feb 16 16:04:04 Laibsch: I don't have any helpfull remarks on both questions Feb 16 16:04:09 helpful* Feb 16 16:04:09 Laibsch: collie/2.6 is not maintained rather. Feb 16 16:04:27 Laibsch: catch pavel machek on fosdem and kick his ass until he will cooperate ;D Feb 16 16:04:47 hrw: Did you not see the message a while back - he's no longer working on collie Feb 16 16:05:06 RP: ah.. not tracing lkml too often Feb 16 16:05:10 RP: thx for info Feb 16 16:05:30 Laibsch: you can be official collie/2.x maintainer if you want Feb 16 16:05:44 I already volunteered to do that. Feb 16 16:05:52 But of course I will need a lot of help. Feb 16 16:06:02 I can be mainly useful in catching the problems. Feb 16 16:06:30 hrw: But I guess you should be able to shed some light as to why oz is building a 2.6 kernel after all? Feb 16 16:06:44 This must be a misconfigured dependency somewhere, I guess. Feb 16 16:06:51 Well, I was hoping to fix this in the next 24 hours. Feb 16 16:06:59 svn suxx Feb 16 16:07:21 I will be meeting with YamaKen, the head developper of uim, here in Tokyo. Feb 16 16:07:54 Laibsch: in .dev all machines are 2.6 iirc Feb 16 16:08:03 Since I was having so many difficulties with the official, old collie images I was thinking about recompiling a new one to talk with him about uim on it. Feb 16 16:08:34 hrw: OK. Is there a way to change this quickly and easily? Feb 16 16:09:08 I guess I could just try and fix the patch as well. Feb 16 16:09:14 Laibsch: vi conf/machine/collie.conf Feb 16 16:09:23 It seems that the lines are off by about 10 lines, that is all. Feb 16 16:09:28 hrw: OK, thanks. Feb 16 16:10:33 hrw: That file says MACHINE_KERNEL_VERSION ?= "2.4" Feb 16 16:11:06 Should I just s/?=/=/ and see what happens? Feb 16 16:11:14 Where else could it be set? Feb 16 16:11:18 Laibsch: local.conf Feb 16 16:14:20 hrw: Nothing in there like it: http://rafb.net/p/6M0IHI45.html Anything else? Feb 16 16:16:05 Laibsch: you have gcc-2.95? Feb 16 16:21:09 hrw: yes, it is int the path Feb 16 16:21:15 as last entry Feb 16 16:49:10 hrw: Maybe heeding the advice and defining a PREFERRED_PROVIDER would help: http://rafb.net/p/SPlwZd94.html ? Feb 16 16:58:19 Laibsch: PREFERRED_PROVIDER_hostap-modules = "hostap-modules"? Feb 16 17:00:25 hrw: for example. The stuff that is suggested. I see that linux-rp is always an option. I wonder if bitbake picks the wrong one (linux-rp when hostap-modules would have been better, same for the other cases) Feb 16 17:00:40 Making explicit what to pick might fix this. Feb 16 17:01:20 PREFERRED_PROVIDER_kernel is of course gonna be tricky. Feb 16 17:01:54 Is it possible to lower the chance of linux-rp being chosen for that without making an explicit choice for example for the 64-0 image? Feb 16 17:02:22 It seems collie and linux-rp don't match well. Feb 16 17:02:38 Can I conflict them? Feb 16 17:04:43 Laibsch: PREFERRED_PROVIDER_hostap-modules = "hostap-modules"? Feb 16 17:05:54 hrw, yes. For example. Feb 16 17:08:41 hrw: http://rafb.net/p/zJWFjd32.html is what I hope fixes this. And in that case I would consider commiting it. Feb 16 17:12:40 RP: is linux-rp from you? Feb 16 17:12:54 Laibsch: yes Feb 16 17:13:12 OK. I guess linux-rp is always 2.6? Feb 16 17:13:19 Laibsch: very much so ;-) Feb 16 17:13:41 Would it be possible to conflict linux-rp with anything that is always built for the collie? Feb 16 17:13:52 Thus making sure it will not get picked. Feb 16 17:14:08 I guess that would be a better solution than http://rafb.net/p/zJWFjd32.html Feb 16 17:14:13 Laibsch: linux-rp has collie 2.6 support Feb 16 17:14:27 Laibsch: I'd guess the key is PREFERRED_PROVIDER_hostap-modules = "hostap-modules", the other two look uneeded Feb 16 17:14:49 You might want a PREFERRED_PROVIDER_virtual/kernel = though Feb 16 17:14:55 RP: Well, but it breaks, see bug 1877. Until it does it would make sense to conflict it. Feb 16 17:15:14 PREFERRED_PROVIDER_virtual/kernel is in collie.conf Feb 16 17:15:42 Laibsch: It might well break but that doesn't mean we should make it conflict. Earlier versions of linux-rp might work better btw Feb 16 17:16:41 RP: http://rafb.net/p/SPlwZd94.html Feb 16 17:17:00 That is why I think the other two are needed as well to make sure linux-rp is not being built. Feb 16 17:17:35 RP: I just need an opie-image within the next 14 hours ;-) Feb 16 17:17:53 Laibsch: The list shows the order it will choose them in. Note linux-rp is not first in two of them, only hostap Feb 16 17:18:20 With a recent bitbake even that is the other way around Feb 16 17:18:24 OK, I did not know the order is already for preference. Feb 16 17:18:35 Well, the recent bitbake ;-) Feb 16 17:18:48 That breaks other things for me. And regularly. Feb 16 17:18:57 But I cannot remember exactly what it is. Feb 16 17:19:11 Its in development ;-) Feb 16 17:19:19 But I can switch easily between them now by updating a symlink. Feb 16 17:19:35 I usually stay with stable, though. Feb 16 17:20:54 So we agree that adding PREFERRED_PROVIDER_hostap-modules = "hostap-modules" to collie-2.4.conf is a good idea? Feb 16 17:21:08 yes Feb 16 17:21:16 OK, I will commit that later, then. Feb 16 17:22:24 Since this is collie-*2.4*.conf, I guess it still would make sense to explicitly forbid linux-rp which as you say is 2.6. Feb 16 17:22:37 has anyone tried to build php lately ? Feb 16 17:22:39 How to do that is another question. Feb 16 17:22:55 Laibsch: I was just about to ask you how you proposed doing that :) Feb 16 17:23:14 RP: At first I thought "impossible" Feb 16 17:23:25 But now I think there might be a way, even two. Feb 16 17:24:34 a) if there is something that is always being built for 2.4 collie, make it conflict with linux-rp Feb 16 17:24:53 This sounds a bit hacky and my wording is not 100% clear, but I think you get the idea. Feb 16 17:25:36 Laibsch: You can't make two bb files conflict though, only resulting packages Feb 16 17:25:44 b) make a meta package, DEPEND on it and thus create what I was looking for in a) Feb 16 17:26:34 Laibsch: I suspect both approaches will result in confusion :/ Feb 16 17:26:48 yes, they are a bit hacky. I admit. Feb 16 17:27:38 RP: I guess we don't have "DISPREFERRED_VERSION"? Is that even and english word, disprefer? Feb 16 17:27:40 ;-) Feb 16 17:28:39 LOATHED_VERSION? Feb 16 17:28:51 Well, that is too strong ;-) Feb 16 17:29:22 I have no personal issue with linux-rp, I just don't like it for collie Feb 16 17:30:01 Laibsch: No, its not English :) Feb 16 17:30:04 * koen reads http://butterfeet.org/?p=8 Feb 16 17:30:22 RP: I was afraid it wasn't :-D Feb 16 17:31:08 Well, let's see what that hostap stuff brings us Feb 16 17:31:12 I'm building openmoko for the first time. bitbake has stopped at package gta01-kernel-2.6, says it fails to fetch gta01-kernel-2.6-moko7-r1 Feb 16 17:31:40 it's been building for hours... :-) Feb 16 17:33:36 PsychoChicken: That is normal Feb 16 17:33:54 I suppose you mean the "building for hours" Feb 16 17:33:59 yes Feb 16 17:34:09 Compiling instills patience ;-) Feb 16 17:34:15 How about the fetching of the kernel? ;-) Feb 16 17:34:30 * koen ponders about adding the openmoko images to the pslash bb Feb 16 17:34:30 The svn location seems to work (http proto) Feb 16 17:35:11 PsychoChicken: http://rafb.net and paste the console error Feb 16 17:36:21 http://rafb.net/p/2Vb0NA93.html Feb 16 17:37:17 line 71 is where it goes bad Feb 16 17:38:46 cu Feb 16 17:40:35 PsychoChicken: yes. Strange indeed since browsing to the URL shows a couple of nice little patches. Koen, what about line 71 in http://rafb.net/p/2Vb0NA93.html ? Feb 16 17:41:11 I'm running bitbake with -v now, to get some more detail... Feb 16 17:41:36 Laibsch: line 71 is a bug in bitbakes svn fetcher combined with SRCDATEs iirc Feb 16 17:42:07 workaround? Feb 16 17:42:46 What do I need to do to get the Kernel modules builded and included into the resulting image? Feb 16 17:43:40 Ah - I just saw that they are builded but just not included into the image Feb 16 17:44:12 How can I change this? Someone hinted me to change something at distro? Feb 16 17:45:57 mr_nice: modules *needed* for the machine to operate properly should be in the MACHINE.conf, other ones in the distro or image Feb 16 17:46:22 mr_nice: see hx4700.conf Feb 16 17:47:08 koen: ok I will have a look there - thx Feb 16 17:50:40 koen: is there something possible like kernel-module* Feb 16 17:52:08 I'm struggling trying to create my first .bb recipe, could someone point me to a .bb file that just takes a file or two and puts it in ${libdir} or something similar? I need something simple Feb 16 17:55:21 xumbi8: generally the population of what goes where is handled by autotools. Feb 16 17:55:24 mr_nice: no, you *must* specify the seperate modules, since you might want to build optional kernel-modules at a later point Feb 16 17:55:53 xumbi8: so the best way to build an app with OE, is use autotools, and then in your bb, "inherit autotools" Feb 16 17:56:38 xumbi8: it works _really_ well Feb 16 18:00:06 cbrake: that's cool, autotools sounds good for a full app, but right now I'm just trying to create a package that puts one library in /usr/lib Feb 16 18:00:22 no compiling at all Feb 16 18:01:26 I created my folder under packages, then a "files" folder under that with my library Feb 16 18:02:38 I point SRC_URI to the name of the library, then I have a do_install() function with 'install -m 0644 ${WORKDIR}/ ${D}${libdir}' Feb 16 18:04:07 it seems to make sense based on the recipe example in the wiki, and after looking at some other .bb files, but it fails with "cannot create regular file `/oe/build/tmp/work/armv5te-linux/-r0/image/usr/lib': No such file or directory" Feb 16 18:04:33 clearly I don't understand how .bb files work :) Feb 16 18:04:54 xumbi8: can you pastbin your bb, and I'll take a quick look? Feb 16 18:05:12 cbrake: that would be great, if you don't mind Feb 16 18:07:26 xumbi8: I'm guessing you need to create the directory first Feb 16 18:08:02 xumbi8: here is an example when I'm creating a file that has a few udev rules: http://pastebin.ca/359304 Feb 16 18:08:26 cbrake: here it is http://pastebin.ca/359306 Feb 16 18:09:55 cbrake: ah, I think I understand... Feb 16 18:10:18 xumbi8: :-) Feb 16 18:11:16 RP: the order of preference for kernel looks suspiciously like the one in packages/linux/collie-kernels-2.4-embedix.bb. I guess the 24-8 kernel is not the best choice in most cases. Would it make to move up the 64-0 image to the first position in that file? Feb 16 18:11:53 cbrake: wow, no idea how I missed that, I guess I had it in my mind that there was an existing directory structure containing all packaged files or something Feb 16 18:12:49 cbrake: thanks once again for the help Feb 16 18:13:13 xumbi8: yes, that is understandable. There is a separate directory structure for each package, so you have to create it. Feb 16 18:13:33 xumbi8: welcome -- glad things are working. Feb 16 18:24:45 hi all Feb 16 18:25:11 hi pb_ Feb 16 18:44:07 I would like to push http://rafb.net/p/ZTZslS49.html (move 64-0 image to first position). Any objections? Feb 16 18:55:10 is the repo down? I have not been able to pull the latest 5 hours Feb 16 18:55:47 use of the mirrors, they are there for a reason Feb 16 19:05:34 ok Feb 16 19:13:04 http://code.google.com/soc/ Feb 16 19:13:10 summer of code is on! Feb 16 19:14:33 koen: Is bug 1178 fixed? Feb 16 19:14:37 !oebug 1178 Feb 16 19:14:39 * * Bug 1178, Status: NEW, Created: 2006-07-16 02:22 Feb 16 19:14:40 * * laforge(AT)gnumonks.org: librsvg-2.11.1-r0 mozilaa plugin depends on host include Feb 16 19:14:41 * * http://bugs.openembedded.org/show_bug.cgi?id=1178 Feb 16 19:25:42 Laibsch: can't say for sure Feb 16 19:27:11 wb mickeyl :) Feb 16 19:27:45 hi Feb 16 19:27:59 hey mickeyl Feb 16 19:28:01 mickeyl: so, one more week as a mere mortal :) Feb 16 19:28:10 *nod* ;) Feb 16 19:28:16 * mickeyl nervous as hell Feb 16 19:28:59 mickeyl: heh...well, on the 24th I'll be in Vegas, getting kicked around at the ATA nationals :) Feb 16 19:29:11 * hvontres|poodle also a bit nervous Feb 16 19:30:23 sounds cool Feb 16 19:31:32 mickeyl: yeah.. good exercise, and it can be a great way to reduce stress.. especailly when we get to kick inanimate objests..(pads etc) :) Feb 16 19:35:27 <_law_> why udev 1.00 is currently disabeld? Feb 16 19:37:27 _law_: untested and/or buggy Feb 16 19:37:42 <_law_> koen ok Feb 16 19:56:49 hmm Feb 16 19:56:58 do you guys remember how to add a ciabot client to a subversion repository? Feb 16 19:57:12 mickeyl: post commit hook? Feb 16 19:57:23 yeah, something like that Feb 16 19:57:28 how's that going exactly? Feb 16 19:58:01 mickeyl: http://lists.samba.org/archive/samba-cvs/2005-January/054380.html ? Feb 16 19:58:25 ah cool Feb 16 19:58:37 now i just need to find the appropriate location Feb 16 19:58:41 where is that hooks dir? Feb 16 19:58:54 on the server or in the repository? Feb 16 19:59:15 * koen is googling Feb 16 20:00:10 on the server it seems Feb 16 20:00:21 * mickeyl checks Feb 16 20:00:37 post commit hook Feb 16 20:00:38 ? Feb 16 20:00:44 ya Feb 16 20:00:50 server I would guess Feb 16 20:01:05 aah Feb 16 20:01:07 seen it Feb 16 20:01:31 mickeyl: http://svnbook.red-bean.com/nightly/en/svn-book.html#svn.reposadmin.create.hooks Feb 16 20:01:39 thanks Feb 16 20:03:45 Hello Feb 16 20:04:41 hm is the monotone server down? Feb 16 20:05:38 hey sirfred Feb 16 20:05:50 koen: Hi Feb 16 20:05:51 woglinde: no, it's overloaded, use one of the mirrors Feb 16 20:05:57 koen okay Feb 16 20:06:07 hm openmoko dos? Feb 16 20:06:08 *g* Feb 16 20:06:49 I have an strange effect with xmodmap. It doesn't work as expected when operating on the husky keyboard, it just ignores the changes. Feb 16 20:07:00 RP: do you have access to berlios where our bitbake sits? Feb 16 20:07:05 But, when I'm using a remote keyboard, it maps correctly Feb 16 20:07:07 i'd like to see how we customized the ciabot_svn for bitbake Feb 16 20:07:26 mickeyl: ssh in to berlios? Feb 16 20:08:11 hmm Feb 16 20:08:18 dunno whether i have access Feb 16 20:09:07 mickeyl: ping kergoth I guess Feb 16 20:09:20 hm or zecke Feb 16 20:10:31 BerliOS Shell Server Feb 16 20:10:31 Welcome to your BerliOS account Feb 16 20:10:35 guess i'm in Feb 16 20:11:34 although I suspect it's customized at the cia site Feb 16 20:21:03 re Feb 16 20:25:22 mickeyl: http://appleinsider.com/article.php?id=2500 Feb 16 20:25:52 * mickeyl reads Feb 16 20:26:16 apple's beefed up olpc meets umpc Feb 16 20:27:14 * koen posts some more OE propaganda to LAK Feb 16 20:28:02 * chouimat|busy reads about the Archos X04 ... now the question is which one I want :) Feb 16 20:30:18 Oh, I've just found that the /etc/X11/shepherd.xmodmap is only applied for 2.4 kernels. I was starting to get mad Feb 16 20:31:42 koen: lakml Feb 16 20:33:39 Laibsch: collie-2.4 should prefer 'openzaurus-sa' or 'collie-kernels'. I propose using first one and set COLLLIE_RAM/COLLIE_MEM (or how it was called) to 64/0 with =? Feb 16 20:34:35 koen: you was faster then me ;) Feb 16 20:34:51 :D Feb 16 20:35:18 koen: read my answer Feb 16 20:35:47 when it will arrive Feb 16 20:36:12 * koen reads Feb 16 20:38:39 koen: when angstrom will provide some kind of cross-sdk? Feb 16 20:38:57 hrw: you means a standalone one? Feb 16 20:38:59 yes Feb 16 20:39:02 no idea Feb 16 20:39:13 grab, unpack, use arm-angstrom-linux-gcc Feb 16 20:39:14 when someone fixes the cross-sdk recipes I guess Feb 16 20:39:35 I have a scratchbox toolchain that people can use in scratchbox Feb 16 20:40:09 hm hm Feb 16 20:40:10 I think that having cross-sdk ready before ELC would be nice Feb 16 20:40:36 re hrw Feb 16 20:40:56 koen: then blebleble about arm/eabi, oe/angstrom and point to angstrom/eabi toolchain Feb 16 20:40:59 hi woglinde Feb 16 20:41:13 hm what should I use for simpad now? Feb 16 20:41:19 in local.conf Feb 16 20:41:54 woglinde: machine=simpad, distro=angstrom-2007.1-oabi Feb 16 20:42:06 * hrw goes sleep Feb 16 20:42:07 cu Feb 16 20:42:16 hrw thanks Feb 16 20:42:19 woglinde: koen will tell the rest ;) Feb 16 20:42:37 woglinde: also check does simpad is 2.6 or 2.4 by default (I assume 2.6) Feb 16 20:43:29 hrw yes some one did some 2.6.20 work Feb 16 20:43:37 woglinde: it was mr_nice Feb 16 20:43:47 yes I thought this Feb 16 20:48:05 good morning Feb 16 20:48:11 he laibsch Feb 16 20:48:14 hrw|gone: I will make the proposed changes. Feb 16 20:48:26 hey woglinde Feb 16 20:48:28 how is life in japan? Feb 16 20:48:32 Nice. Feb 16 20:48:48 But badly jet-lagged. This has been going on all week :-/ Feb 16 20:51:22 so now I will drink a beer Feb 16 21:08:46 * cbrake boots from tiny cramfs partition :-) Feb 16 21:10:11 cbrake nice Feb 16 21:25:57 koen hm why angstroem need lbsdl native? Feb 16 21:26:16 woglinde: qemu-native Feb 16 21:26:43 hm but i dont want qemu native Feb 16 21:26:57 yes you do Feb 16 21:27:08 otherwise you won't have any glibc Feb 16 21:27:22 hrhrhrhrh Feb 16 21:27:31 QEMU is known to have problems when compiled with gcc 4.x Feb 16 21:27:32 | It is recommended that you use gcc 3.x to build QEMU Feb 16 21:27:38 apt-get install gcc-3.4 Feb 16 21:27:43 yes yes Feb 16 21:28:26 hm how is the compile path for glibc with qemu? Feb 16 21:28:50 slow :) Feb 16 21:28:59 hehe Feb 16 21:41:26 koen : is there an active mainter for the x86 kernels ? Feb 16 21:42:02 koen : seems they are a bit behind 2.6.17 Feb 16 22:08:03 I guess I cannot "svn up" from bitbake 1.6.2 to 1.6.6? I have to check out afresh? Feb 16 22:15:03 Laibsch: svn up should work Feb 16 22:20:20 koen: "svn up" tells me I am at revision 761. "bitbake --version" says I have 1.6.2. What does "bitbake --version" say for you? Feb 16 22:20:39 Well, I guess you use bitbake unstable, though Feb 16 22:21:05 BitBake Build Tool Core version 1.7.4, bitbake version 1.7.4 Feb 16 22:21:19 I thought that 1.6.x were only tags, not branches Feb 16 22:21:23 seems that I'm wrong Feb 16 22:21:36 OK Feb 16 22:22:04 they are branches Feb 16 22:22:42 koen : who maintains the x86 kernels ? Feb 16 22:22:52 no idea Feb 16 22:22:55 BitBake Build Tool Core version 1.6.7, bitbake version 1.6.7 Feb 16 22:23:00 At revision 761. Feb 16 22:23:05 no entry in MAINTAINERS? Feb 16 22:23:08 that's the latest on 2.6 branch Feb 16 22:23:14 s/2.6/1.6/ Feb 16 22:24:40 yay: http://gitweb.freedesktop.org/?p=cairo;a=commit;h=2715f2098127d04d2f9e304580a26cd0842c0e64 Feb 16 22:25:18 hi Feb 16 22:28:16 rwhitby: How come the revision is the same but version is different? Feb 16 22:28:46 Laibsch: dunno - I just pull from svn Feb 16 22:29:21 me too ;-) Feb 16 22:31:03 Hey CoreDump|home :) Feb 16 22:31:16 koen: The change you made to SOURCEFORGE_MIRROR in 4c3e39344044dd60a48450fcc753d56c4739dad7 was that intentional or by accident? I suspect the latter and would like to reverse that change in that case. Feb 16 22:31:23 koen : no, there is no entry Feb 16 22:32:28 Laibsch: that was unintentional, sorry Feb 16 22:33:04 args Feb 16 22:33:12 gcc-4.1.1 crashes for me Feb 16 22:33:45 koen: OK. Feb 16 22:33:58 I will push a change later today. Feb 16 22:34:05 hm or I have corrupted memory Feb 16 22:37:15 I should test my memory Feb 16 22:37:51 but glibc for target is right vuilding now Feb 16 22:39:03 03koen 07org.oe.dev * rc8821bce... 10/ (1 packages/openmoko-apps/openmoko-dialer_svn.bb): openmoko-dialer: depend on eds-dbus Feb 16 22:39:37 hey guys I'm a fink developer and I'm working on getting openmoko devel stuff into fink Feb 16 22:40:04 I do have a major problem with the qmake build system in crrent svn in the openmoko-libs dir Feb 16 22:40:23 VERSION < 1 are not supported Feb 16 22:40:58 just wondering if that has been reported yet or not? Feb 16 22:41:18 TheSin: dont use qmake, use autotools Feb 16 22:42:11 the autotools version miss TONS of libs, mostly empty libs but since i'm new to openmoko I wasn't sure how important they where Feb 16 22:42:30 the autotools version works Feb 16 22:42:32 it also means I have to pkg each dir it would seem instead of all at once Feb 16 22:42:55 you have to package each package seperately Feb 16 22:43:03 unless you want to break fink even more Feb 16 22:43:19 I have successfully pkgs ipkg and just working out a few bugs in eds-dbus (mostly pkging stuff) so I'd like to start on openmoko on the weekend Feb 16 22:43:49 koen, I could have one build process and seperate them into seperate pkgs Feb 16 22:43:58 eeeeew Feb 16 22:43:59 make maintaining them much easier Feb 16 22:44:12 no wonder fink is such a steaming pile of breakage Feb 16 22:44:18 naw fink handles that part very well Feb 16 22:44:38 fink manages to break itself everytime I use it Feb 16 22:44:47 that sounds like user error Feb 16 22:44:54 I have been using it since the start Feb 16 22:44:59 and I have never had to reisntall once Feb 16 22:45:03 I only keep it because darwinports sucks even more Feb 16 22:45:10 that is true ;) Feb 16 22:45:23 is there a shortcut for apending to the compile-time include directory? In my case, I'm writing a recipie for logrotate which depends upon popt. I listed it in DEPENDS but how do I pass the proper include path to the compile step? Feb 16 22:45:33 I don't think fink is perfect but it's much better then doing everything by hand Feb 16 22:45:58 anyhow now that I know the autotools version works I'll use that route thanks koen Feb 16 22:46:29 hillct: e.g. -I${STAGING_INCDIR} Feb 16 22:46:30 but jsut and FYI the qmake way works too so long as you use QT4.1+ and the VERSION in teh pro files is 1+ Feb 16 22:46:51 anyhow I'll be back I'm sure ;) thanks for your help Feb 16 22:47:48 koen: thanks. I'll need some more detail, but lemme go browse the docs now that i know what to look for, first :) Feb 16 22:48:08 What is a mirror I can use for monotone.oe.org? Feb 16 22:48:24 the ones mentioned in GettingStarted? Feb 16 22:48:53 the ones after "If monotone.openembeddeded.org is down you can use the following backup monotone servers:" Feb 16 22:49:51 I guess they might do ;-) Feb 16 22:50:19 But of course they "backup monotone servers" and not mirrors so old fart google did not know about'em ;-) Feb 16 22:50:33 they are Feb 16 22:53:29 Oh shit 29 revisions in. That is going to take a while :-/ Feb 16 23:31:26 koen: having re-read the docs, I still don't understand how to specify the path to the header files I need to reference. In this case, I have a compile-time dependency to the popt package, but I don't see how to specify the path to those header files, in a flexible enough way to allow for popt version changes. I'm sure I'm missing someting simple Feb 16 23:37:47 03Laibsch 07org.oe.dev * ra04343c8... 10/ (1 conf/machine/include/collie-2.4.conf): collie-2.4.conf: Prefer hostap-modules over linux-rp as a workaround for bug 1877. Feb 16 23:37:52 03Laibsch 07org.oe.dev * r46952947... 10/ (1 packages/linux/collie-kernels-2.4-embedix.bb): collie-kernels-2.4-embedix.bb: the 64-0 image for collie is generally more appropriate than the 24-8 one. Feb 16 23:37:57 03Laibsch 07org.oe.dev * rdc26b71f... 10/ (1 conf/bitbake.conf): Feb 16 23:37:57 bitbake.conf: revert unintentional change of SOURCEFORGE_MIRROR introduced Feb 16 23:37:57 some time ago in 4c3e39344044dd60a48450fcc753d56c4739dad7 Feb 17 00:44:32 ok. so when I asked about this earlier, I didn't quite understand the structure of the staging directory, but that I understand now Feb 17 00:45:02 I still can't get the make to pickup the CFLSGS changes Feb 17 00:45:03 CFLAGS_prepend = "-I${STAGING_INCDIR} " Feb 17 00:45:23 I'm sure I'm still doing something stupid Feb 17 00:45:31 any guidance? Feb 17 00:48:53 hillct: One suggestion would be to change the = to a += to prevent you from clobbering other CFLAGS_prepend stuff Feb 17 00:49:22 ok Feb 17 00:49:57 I had figured prepend would do exactly that, on a repeated basis Feb 17 00:50:03 testing that adjustment Feb 17 00:50:21 hillct: unfortunately, that is pretty much the depth of my .bb file experience :( koen and the other guys should be back tomorrow morning, CET Feb 17 00:52:08 irc client problems... Feb 17 00:52:26 what was that last item? Feb 17 00:52:35 hillct: unfortunately, that is pretty much the depth of my .bb file experience :( koen and the other guys should be back tomorrow morning, CET Feb 17 00:52:56 what timezone are most of them in? Feb 17 00:53:23 hillct: Europe, GMT +1 Feb 17 00:53:30 ah Feb 17 00:53:32 thanks Feb 17 00:53:38 hillct: np. Feb 17 00:53:40 I'll keep beating on it **** ENDING LOGGING AT Sat Feb 17 02:59:57 2007