**** BEGIN LOGGING AT Thu Feb 26 02:59:57 2009 Feb 26 03:08:59 why does opkg need gpgme? Feb 26 03:11:19 ah, i see Feb 26 03:17:40 yes. exactly. :( Feb 26 03:18:47 i can see it being useful, but it should definately be something for which there's a variable in OpenEmbedded to toggle the behavior, imo Feb 26 03:18:55 it pulls a lot of extra deps into the so called minimal-image Feb 26 03:19:13 opkg-nogpg addresses that. Feb 26 03:19:20 s/dd/d/ Feb 26 03:19:32 ahh, didn't see that Feb 26 03:19:33 cool Feb 26 03:21:04 It also pulls in libcurl, which pulls in a lot of crypto stuff isn't needed for minimal things either, but nobody was interested in trimming that out. So I had a go at it so it would run on the 32MB hardwares, and created the horrible hack called opkg-nogpg-nocurl. Feb 26 03:33:36 03Khem Raj  07org.openembedded.dev * r2c15084856 10openembedded.git/conf/checksums.ini: checksums.ini: Add checksums for fakeroot_1.12.1.tar.gz and netbase-4.21.tar.gz Feb 26 03:44:06 khem, thanks for cleaning up the checksums! Feb 26 03:53:33 Crofton, fixed whatever came in way of building console-image Feb 26 08:22:03 good morning Feb 26 09:49:42 good morning Feb 26 10:03:47 hi florian Feb 26 10:23:50 hi, in the linux-omap-pm kernel, does the message "SR1: VDD autocomp is not active" means that smartreflex is deactivated? Feb 26 11:00:20 http://www.linuxdevices.com/news/NS2276515578.html Feb 26 11:00:26 morning Feb 26 11:02:31 hi hrw Feb 26 11:03:51 hrw: That's much better: http://www.linuxdevices.com/news/NS8394566429.html ;) Feb 26 11:05:37 cool news ! :-D Feb 26 11:07:16 that one I already read Feb 26 11:07:16 florian: will GPE project mail LD to change handhelds.org entry? Feb 26 11:07:37 florian: Question 6 and 7 have entry "Handhelds.org (Familiar, GPE, Opie, Intimate, etc.)" which is wrong - GPE project moved from Handhelds.org to Linuxtogo.org hosting nearly two years ago.... Feb 26 11:08:08 hrw: oh, I didn't notice this yet Feb 26 11:08:24 hrw: everyoen should complain! Feb 26 11:10:02 "BTW - Familiar distribution which is part of handhelds.org died two years ago..." Feb 26 11:10:08 thats end of my mail to ld Feb 26 11:10:33 Intimate was pre-familiar thing? Feb 26 11:11:59 "The problem with handhelds.org is that most of their project are no longer developed. Intimate (which you list) was deprecated by Familiar which ended own life two years ago." Feb 26 11:12:43 hrw: yes right... Intimate is really ancient Feb 26 11:15:00 mail sent Feb 26 11:15:25 hrw: cool, many thanks! Feb 26 11:20:50 we need more peopl eto clcik the OE box Feb 26 11:25:11 morning Feb 26 11:26:12 good work, we are up to 8% Feb 26 11:26:26 Crofton|work: we can not mention this on ML Feb 26 11:26:36 otherwise it will be like 2-3 years ago Feb 26 11:26:59 LD wrote something like 'high position of openembedded is due to aggressive promotion of survey on their ML' Feb 26 11:28:12 hrw, you are correct Feb 26 11:28:32 but, all of these surveys have the same issue Feb 26 11:28:43 I would assume the debian guys do some promotion Feb 26 11:28:50 the .com crowd not so much Feb 26 11:29:21 Promotion? Debian? Errr... Feb 26 11:29:29 I would view the skewing of the results as a measure of community involvement :) Feb 26 11:41:13 8.4% Feb 26 12:02:19 03Khem Raj  07mickey/org.openembedded.dev * ra53bdf1edf 10openembedded.git/conf/checksums.ini: Feb 26 12:02:19 checksums: Remove spurious control characters in ++DFB-0.9.25.tar.gz url Feb 26 12:02:19 * This led to 'NOTE: Creating the CheckSum parser failed' but the build Feb 26 12:02:19 continues without checking for the checksums. Ideally it should have Feb 26 12:02:19 errored if OE_STRICT_CHECKSUMS was set. But I will leave that for Feb 26 12:02:21 other day. Feb 26 12:02:25 03Khem Raj  07mickey/org.openembedded.dev * rd9bd245d15 10openembedded.git/conf/distro/include/preferred-om-2008-versions.inc: preferred-om-2008-versions.inc: Use pidgin version 2.5.3 Feb 26 12:02:28 03Dmitry 'MAD' Artamonow  07mickey/org.openembedded.dev * rfa76888ead 10openembedded.git/ (5 files in 3 dirs): Feb 26 12:02:31 xdiskusage: xdiskusage shows disk space usage information, it can produce a PostScript version of the display. Feb 26 12:02:33 Signed-off-by: Khem Raj Feb 26 12:02:35 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r40da08b6db 10openembedded.git/ (3 files in 3 dirs): frameworkd-devel: use mickey/ogsmd-new-timeout branch for now Feb 26 12:11:39 03Jeremy Lainé  07org.openembedded.dev * r9db305be9d 10openembedded.git/packages/linux/ (4 files in 2 dirs): linux-2.6.27: improve boc01 EEPROM support Feb 26 12:23:42 how to get with bitbake shell a variable value used when parsing specific package (task)? Feb 26 12:24:09 i'd like to know what $DISTRO_INIT_MANAGER is when building task-boot Feb 26 12:30:21 and how to redirect bitbake shell output to a file? I'd like to get 'environment' output into file Feb 26 12:39:02 Could it be somebody forgot to update the openembedded/conf/checksum.ini when they updated base-passwd to 3.5.20 ... trying to build SlugOS firmware now generates the following error: ERROR: base-passwd-3.5.20: ftp://ftp.debian.org/debian/pool/main/b/base-passwd/base-passwd_3.5.20.tar.gz has no entry in conf/checksums.ini, not checking URI Feb 26 12:40:01 Caelian, most likely Feb 26 12:40:07 let me take a look Feb 26 12:42:27 booxter: bitbake task-boot -e|grep ^DISTRO_INIT_MANAGER Feb 26 12:48:08 hello all! Feb 26 12:48:42 yep Feb 26 12:48:46 I was wondering, is anyone still using qtopia-core 4.3.3 (as opposed to qt-embedded 4.4.3)? Feb 26 12:48:59 Would people stop commiting new recipes without checksums! Feb 26 12:49:09 heheh Feb 26 12:49:15 I know it happens sometimes Feb 26 12:49:29 Let me fix it Feb 26 12:49:49 Glad to have brought it to your attention :) Feb 26 12:51:42 Crofton, no reviews - no way to fix the problem Feb 26 12:51:58 automate it or die :) Feb 26 12:57:13 03Philip Balister  07org.openembedded.dev * r4fe53e316d 10openembedded.git/conf/checksums.ini: checksums.ini : Add checksum for base-passwd-3.5.20. Feb 26 12:57:15 Caelian, pushed, sorry for the trouble Feb 26 12:57:47 no trouble at all ... i know stuff like this happens :) Feb 26 12:59:38 was just about to do a fresh firmware image / package build-run to put my automated flashing/software-installation script through another test run when i noticed the build breakage Feb 26 13:03:52 hi, is there a way to tell the fetcher not to try the mirror, especially when I try to fetch a local svn server? Feb 26 13:29:16 Crofton|work: spotted another one ... ERROR: tzcode-native-2007k: http://bent.latency.net/bent/darcs/zoneinfo-2007k/src/tzcode2007k.tar.gz has no entry in conf/checksums.ini, not checking URI Feb 26 13:29:54 though i am wondering if that one should not be updated to something a bit more recent :) Feb 26 13:30:03 frak Feb 26 13:30:09 let me check Feb 26 13:35:22 its origin is openembedded/packages/tzcode/tzcode-native_2007k.bb (in case you need it) Feb 26 13:35:40 yeah, I think this is a mirror issue .... Feb 26 13:37:12 ~curse lack of syncing tools Feb 26 13:37:13 May the fleas of a thousand camels infest your most sensitive regions, lack of syncing tools ! Feb 26 13:41:07 SRC_URI = "\ Feb 26 13:41:07 http://bent.latency.net/bent/darcs/zoneinfo-2007k/src/tzcode${PV}.tar.gz \ Feb 26 13:41:07 http://bent.latency.net/bent/darcs/zoneinfo-2007k/src/tzdata${PV}.tar.gz \ Feb 26 13:41:07 " Feb 26 13:41:09 ? Feb 26 13:41:12 why twice? Feb 26 13:41:25 tzcode/tzdata Feb 26 13:42:12 doh Feb 26 13:42:17 thanks :) Feb 26 13:42:27 * Crofton|work is an idiot Feb 26 13:42:37 ;) Feb 26 13:43:15 ;D Feb 26 13:43:29 Crofton|work: you probably forgot contact lenses Feb 26 13:43:53 yeah Feb 26 13:44:11 I have my glasses on atm, and the screen is a bit blurrry Feb 26 13:44:33 Crofton|work: cleaning screen and glasses helps a lot ;) Feb 26 13:46:02 bbl Feb 26 13:47:19 * hrw is doing firmware update on nokia phone.. Feb 26 13:50:05 I just found another way to defeat checksums Feb 26 13:50:10 remove checksums.ini Feb 26 13:52:16 genius! Feb 26 13:52:45 Crofton|work: cd build; touch conf/checksums.ini Feb 26 13:52:51 thats even better Feb 26 13:53:26 heh Feb 26 13:53:36 I am not encouraging this behavior Feb 26 13:53:43 bitbake use first found in BBPATH Feb 26 13:53:54 kinda defeats the original purpose ... but it would remove the headache :) Feb 26 13:54:50 heh.. win32 is funny... nokia software updater do flashing new firmware to my phone. this require installing few drivers (which is done automatically by nsu). during flashing windows asks for reboot due to new drivers installed Feb 26 13:56:15 03Philip Balister  07org.openembedded.dev * r834db40851 10openembedded.git/conf/checksums.ini: tzcode-native : Add checksums for txcode and tzdata. Feb 26 14:00:57 ok ... works now ... (able to successfully build openntpd again) Feb 26 14:01:47 thanks for being patient and not disabling checksums Feb 26 14:02:21 hehe ... the more i can use a stock build system .. the less maintenance hassle falls upon me :) Feb 26 14:03:35 there's a reason i opted fro writing scripted SSH-sessions do to additional software installation/removal ... instead of trying to somehow modify the actual firmware image build Feb 26 14:03:48 s/do to/to do/ Feb 26 14:12:18 how to force OE rebuild packages for older versions if packages with larger PR are not present in packages tree Feb 26 14:13:08 bitbake -c rebuild foo will rebuild foo. Feb 26 14:16:06 kergoth, the problem is that I messed my BBPATH with actual upstream ones, and when I realized that I don't need them (only packages that I "forked" at some time in the past) I already had lots of packages with newer versions in my build/tmp built. I'd like to downgrade all the packages for which I have only older versions Feb 26 14:16:51 but when I try to build my images I have newer versions supplied :( Feb 26 14:17:15 I don't want to repeat build task from scratch - it costs too much time Feb 26 14:17:54 rm packages from tmp/deploy/*/ipk/* Feb 26 14:19:31 hrw: wouldn't one need to also remove the tmp/stamps/*do_package* files as well for that to work ? Feb 26 14:20:08 no Feb 26 14:20:17 touch: cannot touch `/oe/build/tmp/deploy/glibc/ipk/Packages': No such file or directory Feb 26 14:20:47 ow I understand, I removed the directory too... Feb 26 14:21:46 ok ... i stand corrected ... still not too intimately familiar with the oe-internals :) Feb 26 14:24:45 hrw, http://pastebin.com/m3e1d667c Feb 26 14:25:13 http://marcin.juszkiewicz.com.pl/2009/02/26/syncing-mobile-devices/ Feb 26 14:25:45 booxter: I love how opkg suxx with messaging Feb 26 14:26:50 hrw, me too. but wtf?:) Feb 26 14:27:08 it occured after I removed all the ipk/ contents Feb 26 14:44:18 03Koen Kooi  07org.openembedded.dev * rf8e5442bb3 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 26 14:44:18 03Koen Kooi  07org.openembedded.dev * rbe8f0f5b8f 10openembedded.git/packages/dsplink/gstreamer-ti_svn.bb: gstreamer-ti: bump to r127 Feb 26 14:54:15 what are samba and samba-essential for? Feb 26 14:58:57 hi there Feb 26 14:59:12 i have problems with the latest gstreamer releases Feb 26 14:59:25 seems they upgraded to a new libtool version Feb 26 14:59:52 and since then it tries using the host machine's libs in /usr/lib instead of the staging target specific ones Feb 26 15:00:10 it also doesn't build a mipsel-linux-libtool any more, just a normal one Feb 26 15:00:26 morning Feb 26 15:00:47 fraxinas: well, OE is running autoreconf... Feb 26 15:01:34 i noticed that the main libtool configuration used to be in aclocal.m4 and now moved to a new file m4/libtool.m4 of the working dir Feb 26 15:02:36 here's what i posted on their bugzilla: http://bugzilla.gnome.org/show_bug.cgi?id=572532 Feb 26 15:03:14 fraxinas: what happens if you do do_configure_prepend() \n{ \nrm ${S}/m4/libtool.m4 \n} ? Feb 26 15:03:46 i shall try. just a moment. Feb 26 15:06:06 * * OE Bug 5044 has been created by di(AT)fh-wedel.de Feb 26 15:06:08 * * The source archive for dash 0.5.2 is no longer downloadable Feb 26 15:06:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5044 Feb 26 15:07:31 cool that's new Feb 26 15:07:41 RP: ping Feb 26 15:08:38 zecke: http://pastebin.com/d57b41849 Feb 26 15:09:14 fraxinas: I'm on a GPRS line and have some downloads going, use words :) Feb 26 15:09:56 ah it loaded here... okay... hmm Feb 26 15:10:02 oh okay so the main problem is this i guess: checking host system type... mipsel-unknown-linux-gnu Feb 26 15:10:36 fraxinas: this is not instantiating libtool at all now... the basic idea is: We have our own libtool and apps should use it Feb 26 15:10:46 i think it doesn't like haveing that m4 deleted Feb 26 15:10:56 RP: with bitbake 1.8.13, rebuild is .. not useful. it refuses to clean, it wants to get a successful execution of the default task first. if you're getting an error in a task, then rebuild is useless as is Feb 26 15:11:28 fraxinas: well, what is inside the libtool.m4? just a copy of the original or some additional things? Feb 26 15:11:38 yeah i've read that nice line before... # Always use our own libtool. Feb 26 15:12:09 i haven't gotten around to comparing it line-by-line to the old code... it's over 7000 lines :/ Feb 26 15:12:47 it pretty much contains stuff that looks like what used to be in aclocal.m4 in the previous release Feb 26 15:15:05 sorry i'm such an autotools n00b :/ Feb 26 15:15:08 kergoth: pong Feb 26 15:15:29 kergoth: rebuild is probably broken :( Feb 26 15:15:54 kergoth: I remember trying to fix it and finding no nice way to do it Feb 26 15:15:58 okay, just making sure someone knew about it. not that difficult to manually clean and rerun like we used to Feb 26 15:16:12 what changed to make it not work? Feb 26 15:16:47 kergoth: I switched build.py not to have its own task graphs and use the main one from runqueue/taskdata Feb 26 15:16:58 ah Feb 26 15:17:29 It was a worthwhile code cleanup and rebuild breaking was the main fallout Feb 26 15:17:44 RP: hey! Feb 26 15:17:50 hi zecke Feb 26 15:18:25 kergoth: We had to do that to get packaged-staging stamp handling to work iirc Feb 26 15:26:44 fyi, package_stagefile_shell explodes if you move DEPLOY_DIR out of TMPDIR :) Feb 26 15:31:25 RP: thought about reworking the bitbake versioning? if oe tasks get broken in a micro version, maybe you should consider bumping the minor or major instead. Feb 26 15:32:25 kergoth: I don't quite follow? Feb 26 15:32:59 micro versions should be for bugfixes, not cleanup that breaks things Feb 26 15:33:16 kergoth: Ah, that version number :) Feb 26 15:33:26 i'd argue against staying on the 1.8 branch with changes like that Feb 26 15:33:27 * kergoth nods Feb 26 15:33:50 kergoth: I know what you mean Feb 26 15:34:22 People are kind of comfortable with the 1.8 series though and I'm kind of ok leaving it at that untill trunk is ready... Feb 26 15:34:26 seems like a lot of projects fall into this mode, where they are so nervous to bump the major, they end up on a bugfix branch indefinately :) Feb 26 15:34:56 kergoth: The plan is to have 1.9 -> 2.0 with the UI split code Feb 26 15:35:01 well, i could easily see someone following the 1.8 branch in their production use, to pick up bugfixes, and suddenly finding the tools that used to work in their daily operation no longer do. Feb 26 15:35:04 and that's problematic, imo Feb 26 15:35:21 Yes, it wasn't intentional to break rebuild :/ Feb 26 15:35:23 kergoth: agreed Feb 26 15:35:32 I use 1.9 anyway Feb 26 15:36:49 kergoth: Basically I agree but I'd prefer not to rock the boat and continue with the 1.9 -> 2.0 plan Feb 26 15:37:13 makes sense. please consider a hard versioning policy for post 2.0 though :) Feb 26 15:38:25 Random bitbake question, will 1.9 use git archive ? Feb 26 15:39:41 RP: are the bitbake plans for going forward written down anywhere? Feb 26 15:41:25 Tartarus: Depends if someone sends patches ;-) Feb 26 15:41:25 kergoth: parts probably in poky/trunk changelog ;D Feb 26 15:41:40 kergoth: various things have been mentioned on various mailing lists Feb 26 15:41:43 nothing official Feb 26 15:41:43 heh :) Feb 26 15:42:25 * kergoth would like to get back into the swing of things.. mostly back up to speed with the changes in OE, though packaged staging is still voodoo, but want to get up to speed with what needs doing in bitbake so he can contribute there Feb 26 15:43:19 kergoth: It really needs the UI split and xmlrpc stuff stress testing Feb 26 15:45:10 not my area of expertise, but what the heck, need something to do Feb 26 15:45:13 heh Feb 26 15:45:16 sounds good Feb 26 15:45:19 * kergoth looks over that code Feb 26 15:45:52 I'd be tempted to rewrite performance critical stuff in Vala Feb 26 15:46:06 but i need some months off Feb 26 15:46:08 for that Feb 26 15:47:22 i think itd be kind of neat to have a version where the classes are code, holding to a specific api and versioning scheme... whether they're python classes or .net assemblies or whatever.. just one of those oddball possibilities thats been floating around my head Feb 26 15:47:40 yeah Feb 26 15:48:53 ironpython could be fun to play around with, could write the classes in f# or something. hehe Feb 26 15:50:04 oh, i wanted to ask, what's the current required version of python for bitbake? Feb 26 15:50:17 so i know what i can use in it Feb 26 15:50:37 any of you guys played with the Nix package manager? Feb 26 15:50:40 kergoth: 2.4 still iirc Feb 26 15:50:41 it's an interesting concept Feb 26 15:50:51 2.4 as I understand it for 1.8 Feb 26 15:50:56 k Feb 26 15:50:59 1.0 is 2.5 Feb 26 15:51:06 1.9 is 2.5 even Feb 26 16:03:59 feel free to bump 1.9 to 2.6, but don't even think about 3.0 Feb 26 16:04:45 mickey|cafe: I don't have 2.6 so that won't happen quite yet ;-) Feb 26 16:04:55 righto Feb 26 16:05:20 although 2.6 is a straightforward update Feb 26 16:05:24 2.5 is in most of distros but 2.6 not yet Feb 26 16:07:08 :) Feb 26 16:07:43 is this a right place to ask about iPAQ 36** ? Feb 26 16:10:18 flameman: personally I would just put them into drawer to make them gather dust Feb 26 16:10:55 :P Feb 26 16:12:06 flameman: unless you want to play with developing drivers for device which has 15 users on whole glob ;D Feb 26 16:12:33 ahaha Feb 26 16:12:52 http://www.elinux.org/Ipaq-Flameman Feb 26 16:13:05 flameman: I saw that blob of misc text yesterday Feb 26 16:13:17 ok Feb 26 16:13:21 flameman: split it into separate pages to make it readable Feb 26 16:13:44 good suggestion :) i will split in the next rewrite ! Feb 26 16:14:17 more progress today Feb 26 16:14:29 the stage4 is at his 40% :P Feb 26 16:14:43 issues are still in the kernel side Feb 26 16:15:18 flameman: I do not care about how gentoo works on any device. for h36xx the thing which needs work is kernel. userspace just works Feb 26 16:15:46 well the issue is in the sleeve 2xpcmcia device Feb 26 16:15:51 in the IrDA device Feb 26 16:16:00 other things are well working Feb 26 16:16:40 unfortunately without a 2xpcmcia sleeve you can't have any connectivity ! Feb 26 16:16:54 and space for rootfs Feb 26 16:17:02 exactly Feb 26 16:17:26 actually the rootfs is on a microdrive (4Gb), attached to a pcmcia to cf2 Feb 26 16:17:27 hm.. new fvwm is out.. Feb 26 16:17:46 looks that I should create new recipe.. Feb 26 16:17:59 anyway as the fact i can't contact anybody interested about h36** Feb 26 16:18:10 i ve planned to switch myself to h5550 Feb 26 16:18:33 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rb7f96978a7 10openembedded.git/packages/linux/linux-openmoko-2.6.28_git.bb: Feb 26 16:18:33 linux-openmoko-2.6.28: build out of the stable branch; bump to latest Feb 26 16:18:33 version Feb 26 16:18:36 so i will rewrite the ewiki page converting my work for the new platform Feb 26 16:18:37 :) Feb 26 16:18:44 flameman: h5xxx are better supported Feb 26 16:18:50 yes ! openmoko will be supported, too :D Feb 26 16:19:05 hrw yes, they are arm v4 pxa2*** Feb 26 16:19:08 * mickey|cafe working on HTC raphael these days Feb 26 16:19:17 so lcd framebuffer is attached to it Feb 26 16:19:23 i think the HTC are the most interesting non-free devices these days Feb 26 16:19:37 on h36** everything is attached to sa11** that suxs Feb 26 16:19:49 flameman: h5xxx is armv5te not armv4 Feb 26 16:20:14 is should be pax2***, isn't it ? Feb 26 16:20:36 pax2 is v4, but it may be i am wrong Feb 26 16:20:40 pxa250/255 is armv5te Feb 26 16:20:45 mickey|cafe: what sorts of things are you doing with the Raphael? Feb 26 16:20:55 all XScale are armv5te Feb 26 16:21:33 hrw ok :) in any case a stage4 arm v4 little endian should be fine the same Feb 26 16:21:35 CosmicPenguin: trying to port FSO to it Feb 26 16:21:42 so we have middleware Feb 26 16:21:52 and then cowork with SHR or what not to put a UI on top Feb 26 16:22:03 bbl, need to leave the cafe now Feb 26 16:24:33 i have to debug a mmc_over_gpio now, i wrote that driver but there is something wrong about the amount of the byte Feb 26 16:24:52 it says that a 2Gb mmc/sd card is "512Mb" Feb 26 16:25:01 that's funny :P Feb 26 16:25:27 sandisk handbook has been donwloaded and i am enjoying it :D Feb 26 16:27:08 it is a bit sad that 2.6 still doesn't work on h3600. those drivers have been 90% complete for years now. Feb 26 16:27:09 flameman: how old kernel? Feb 26 16:28:08 2.6.26 Feb 26 16:28:26 pb_ Feb 26 16:28:42 the sleeve 1xpcmcia is working, the 2xpcmcia is not Feb 26 16:29:07 you can check it out from the hh repositories (kernel 2.6.21.hh) Feb 26 16:29:48 here i am working with both the kernel series: vanilla 2.6.26 + patch, handhelds patched 2.6.21.hh Feb 26 16:30:20 the issues is related to the sa11**, the way it handles his pcmcia support for 2 pcmcia slots Feb 26 16:31:07 right. it's clearly not intractable though, it works in the 2.4 kernel. Feb 26 16:31:39 dunno, i suppose it was worked with 2.4, never checked Feb 26 16:32:22 i contacted a developers by email ... still waiting for a reply Feb 26 16:32:36 flameman: handhelds.org developers? Feb 26 16:32:43 it seems nobody is working / interested / active with 36** Feb 26 16:32:45 yes, it did work in 2.4. that's why I say it is a bit sad that, four or five years later, these things still don't work in 2.6. Feb 26 16:32:47 hrw yes Feb 26 16:33:13 all the people i can see in the email list Feb 26 16:33:16 still, it sounds like the h3600 support in 2.6 is much better now than it used to be. Feb 26 16:33:31 it is not Feb 26 16:33:40 also the LCD has an other issue Feb 26 16:34:09 so .. no 2xpcmcia, no Irda, no connectivity ... and the LCD that power off in 90 sec Feb 26 16:34:28 anyway the machine is well compiling with an uptime of 4 days Feb 26 16:34:45 (that is sad, cause it should be a PDA) Feb 26 16:35:38 well, if the single pcmcia sleeve works then presumably you can have connectivity that way. otherwise there's always usbnet, though obviously that doesn't work too well in a mobile environment. Feb 26 16:36:22 pb_: usbnet and sa11xx == problem Feb 26 16:36:29 it is a bit of a shame that the lcd is broken but presumably that also can't be too hard to fix. if all the h3600_micro bits are working: charger, touchpanel, that kind of thing, then everything else should be fairly straightforward. Feb 26 16:36:36 I do not remember is it supported in mainline Feb 26 16:36:39 oh, right, I forgot about the sa11xx suckage with usb Feb 26 16:36:44 usb is not working Feb 26 16:36:48 irda is not working Feb 26 16:37:00 the only pcmcia is used for the rootfs Feb 26 16:37:01 Thomas Kunze worked on sa11xx usb gadget driver some time ago Feb 26 16:37:16 well now it is not working Feb 26 16:37:22 you mean old kernel ? Feb 26 16:37:25 2.6.12 ? Feb 26 16:37:45 err, i think packaged staging just tried to install opkg-native into staging, but opkg from opkg-native isn't there to install it.. Feb 26 16:37:48 * kergoth scratches head Feb 26 16:37:49 usbned = issue ! Feb 26 16:38:04 (tested on 2.6.21) Feb 26 16:38:49 ahahah to be extremely sad Feb 26 16:39:04 i could use the /dev/ttySA0 as pppd dev Feb 26 16:39:20 just a serial slip net :D Feb 26 16:39:35 (but i have to remove and loose the console) Feb 26 16:39:42 (console = ttySA0) Feb 26 16:40:18 LinuxToGo is now present in LinuxDevices embedded survey ;) Feb 26 16:40:25 thanks for all which mailed LD Feb 26 16:41:39 kergoth: hmm it installs ok for me Feb 26 16:42:04 kergoth: I have opkg-cl in staging Feb 26 16:42:12 i just did a build of an image, wiped tmp, then built again (with pstage outside of tmp) Feb 26 16:42:17 hrw: seems that Henry (of LD) have had a lot of correction requests today :-) Feb 26 16:42:28 then it used pstage for everything, including opkg-native, and failed Feb 26 16:42:29 indeed Feb 26 16:42:36 i had to clean adand rebuild opkg-native, then it went fine Feb 26 16:42:44 kergoth: ah I did not try that Feb 26 16:42:52 kergoth: hmm Feb 26 16:45:06 * mckoan noticed a lot of liveliness with LinuxDevices poll, and this is a good thing for OE and for Linux! :-D Feb 26 16:50:04 mickeyl: good morning Feb 26 16:50:28 good morning pb_ Feb 26 17:09:41 anyone know why I would be getting 'Stale NFS file handle' when trying to read/write to a CF with OE built filesystem on target - I dont' have any nfs or otherwise network mounted filesystems Feb 26 17:10:08 or nfsd running... Feb 26 17:11:42 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r13ac0a793a 10openembedded.git/conf/distro/include/sane-srcrevs.inc: sane-srcrevs: bump linux-openmoko-2.6.28 Feb 26 17:11:52 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r4ca003cb04 10openembedded.git/packages/linux/ (16 files in 2 dirs): linux-openmoko-2.6.28: remove patches that have been applied upstream Feb 26 17:12:08 03zecke123 07bitbake-1.8 * r1151 10/lib/bb/fetch/git.py: Clean up by Otavio Salvador Feb 26 17:13:10 RP: in case you wonder, I will put that into trunk as well.. just fetching the sources Feb 26 17:16:47 zecke: cool, np :) Feb 26 17:18:27 anyone mind if i update all the fetchers to use runfetchcmd method, so they all obey the metadata PATH? Feb 26 17:19:31 tharvey: what filesystem are you using? Feb 26 17:20:29 various non-nfs filesystems do have the capability to generate ESTALE under certain circumstances. Feb 26 17:24:42 pb_, the particular fs is ext2 on a CompactFlash Feb 26 17:26:11 tharvey: best to check the ext2 code then, see what causes it to emit that error Feb 26 17:27:13 ok thx Feb 26 17:31:00 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r9b4c27d165 10openembedded.git/ (conf/distro/include/sane-srcrevs.inc packages/qi/qi.inc): qi: enable build debug output again and bump to latest version Feb 26 17:31:34 03Jan Lübbe  07fso/milestone5.5 * r6c8e82e903 10openembedded.git/ (3 files in 3 dirs): libexif: add 0.6.17 (security/bugfix release) and update it in preferred-om-2008-versions Feb 26 17:31:37 03Jan Lübbe  07fso/milestone5.5 * r22e9328e78 10openembedded.git/packages/images/fso-image.bb: fso-image: add cellhunter Feb 26 17:31:38 03Jan Lübbe  07fso/milestone5.5 * r2830c21b55 10openembedded.git/packages/freesmartphone/ (menu-freesmartphone.bb menu-freesmartphone/applications.menu): menu-freesmartphone: create a simple XDG menu for illume which shows all applications Feb 26 17:31:38 03Jan Lübbe  07fso/milestone5.5 * r6daf9b8afe 10openembedded.git/ (3 files in 2 dirs): tangogps: upgrade to 0.9.6 Feb 26 17:31:40 03Jan Lübbe  07fso/milestone5.5 * r5825c2dfa7 10openembedded.git/packages/e17/e-wm_svn.bb: e-wm: split out XDG menu definition into e-wm-menu Feb 26 17:31:43 03Jan Lübbe  07fso/milestone5.5 * rdf101dfac1 10openembedded.git/packages/images/fso-image.bb: fso-image: use menu-freesmartphone and drop the .desktop file hack Feb 26 17:46:42 kergoth: I thought we'd done that? Feb 26 17:46:52 kergoth: It was certainly the intention anyway... Feb 26 17:47:37 done what? i forgot what i said :) Feb 26 17:47:48 * kergoth trying to do too much at once, clearly Feb 26 17:48:00 kergoth: "anyone mind if i update all the fetchers to use runfetchcmd method, so they all obey the metadata PATH?" Feb 26 17:48:25 oh, right, i didn't check trunk, feel free to ignore me :) Feb 26 17:48:41 trunk may be bust as well, I really don't know Feb 26 17:48:50 I know what the intention was :) Feb 26 17:48:53 * kergoth just got bit by 1.8's wget fetcher doing its own fetch command execution Feb 26 17:49:09 overriding the FETCHCOMMAND is easy enough, but .. :) Feb 26 17:49:24 * RP had fun with that in a world of firewall pain Feb 26 17:54:39 bye all Feb 26 17:58:11 03zecke123 * r1152 10bitbake/lib/bb/fetch/git.py: Clean up by Otavio Salvador Feb 26 18:02:53 hello Feb 26 18:03:00 I am tying to bitbake libgpsmgr Feb 26 18:03:04 I get an error: http://pastebin.com/d34adad44 Feb 26 18:03:13 anybody got an idea on how to fix? Feb 26 18:09:24 udovdh: you can try to look at configure.ac and look at AC_INIT there Feb 26 18:10:00 RP: ever thought about BBFILES = "${@' '.join([path for path in d.getVar('BBPATH', 1).split(':')])}" ? Feb 26 18:10:23 RP: with some fixes so it doesn't try to parse directories, instead ignoring them, it does work Feb 26 18:10:35 (same for nonexistant paths) Feb 26 18:10:59 * kergoth 's doing all sorts of oddball things lately Feb 26 18:11:12 mr_nice, will have a look Feb 26 18:11:49 AC_INIT(Makefile.am) it says Feb 26 18:12:46 mr_nice, so what next? Feb 26 18:13:59 error: http://pastebin.com/d34adad44 Feb 26 18:14:44 udovdh: I am no autoconf expert the error message tell that ac_init needs package and version Feb 26 18:15:03 so the format becomes? Feb 26 18:15:14 AC_INIT(Makefile.am, x, y) ? Feb 26 18:15:18 where x is name Feb 26 18:15:21 and y is a number? Feb 26 18:16:54 udovdh: i don't know the right format Feb 26 18:17:02 AC_INIT(Makefile.am, x, y) ? Feb 26 18:17:11 AC_INIT(x, y) it is Feb 26 18:18:25 I also see Can't locate Date/Parse.pm in @INC Feb 26 18:18:30 isn't that fatal? Feb 26 18:24:50 morning all Feb 26 18:25:06 does openembedded work with the freescale i.mx processors? Feb 26 18:28:50 garren: I see conf/machine/mx31ads.conf conf/machine/mx31moboard.conf Feb 26 18:29:00 it should do Feb 26 18:30:38 03Frans Meulenbroeks  07org.openembedded.dev * r064adeb426 10openembedded.git/ (4 files in 3 dirs): pmount: updated to 0.9.18 Feb 26 18:32:43 garren: I use an OE userspace regularly on i.MX31. Feb 26 18:33:04 The ADS config is borked due to bitrot in the kernel recipie. Feb 26 18:33:16 but mx31moboard is actively developed. Feb 26 18:35:39 awesome thanks broonie and khem Feb 26 18:38:14 still looking at this error: http://pastebin.com/d34adad44 Feb 26 18:38:19 anyone got a tip? Feb 26 18:38:57 udovdh: what version are you compiling Feb 26 18:39:11 the one in OE.dev Feb 26 18:39:13 euhm... Feb 26 18:39:23 0.1 Feb 26 18:39:35 in the maemo4 directory Feb 26 18:40:05 I see perl Date module missing Feb 26 18:40:14 and autoconfig stuff Feb 26 18:40:16 wrong Feb 26 18:42:16 this build fine for me. so I guess something wrong with your env or host machine Feb 26 18:45:39 khem, any suggestions? dependencies ? Feb 26 18:46:06 see http://wiki.navit-project.org/index.php/Navit_on_OpenEmbedded_for_n810 for some documentation of my progress Feb 26 18:47:59 perl 5.8.8 is built here. right version? Feb 26 18:49:02 RP: ping Feb 26 18:58:15 zecke: ping Feb 26 19:00:08 pong Feb 26 19:01:54 zecke: bitbake has a variable called BBPKGS.. it goes in the list of packages to be built.. but those packages are _always_ built, even if the user specifies what they want to build on the cmdline. I can't think of a use case for that behavior. If we changed it to only build that if the user doesn't specify on the cmdline, then you could have a sane default provider to be built, letting someone "cd foo; . ./env; bitbake" assuming the right setup Feb 26 19:01:59 opinion? Feb 26 19:08:23 sure Feb 26 19:09:29 still looking at this error: http://pastebin.com/d34adad44 Feb 26 19:09:42 is the error the cause of a host-side autoconf problem? Feb 26 19:09:50 (Fedora 10 x86_64) Feb 26 19:12:11 hi all Feb 26 19:13:59 is qemu-native-0.9.1+svnr4242-r14 broken? bitbake barfed while trying to compile it. Feb 26 19:18:31 ~pastebin Feb 26 19:18:32 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Feb 26 19:18:34 show us the problem :) Feb 26 19:18:45 something about :| /home/bmyers/OE_Palm/stuff/tmp/work/i686-linux/qemu-native-0.9.1+svnr4242-r14/trunk/target-i386/ops_sse.h:597: error: unable to find a register to spill in class `GENERAL_REGS' Feb 26 19:19:09 ah Feb 26 19:19:14 ah ha ha Feb 26 19:19:16 So, what distro? Feb 26 19:19:25 the problem is qemu-native needs gcc-3.4 Feb 26 19:19:36 and in some cases, the host one might be found, used, but be too old still Feb 26 19:19:39 ah ah ha, 3.4 is not my default Feb 26 19:19:44 4 is Feb 26 19:19:53 yes Feb 26 19:20:01 but if you have a host gcc-3.4 as well, it calls it by name Feb 26 19:20:04 but that might still be "wrong" Feb 26 19:20:11 and the gcc-3.4 OE builds not get used Feb 26 19:20:48 is there something very special about this qemu? I have the ubuntu installed and the hackndev and they work Feb 26 19:25:19 Lets say yes :) Feb 26 19:25:25 i get the point.. Feb 26 19:25:26 So, what distro? Feb 26 19:25:37 And if you do 'gcc-3.4 what do you get? Feb 26 19:25:59 ubuntu Feb 26 19:26:31 i have 3.4 install but i have to call it that way... gcc-3.4 not just gcc Feb 26 19:26:55 OK Feb 26 19:27:11 Do you have 'ASSUME_PROVIDED += "gcc-3.4-native" or so in your local.conf? Feb 26 19:27:31 hmm. in the build dir you mean? Feb 26 19:27:44 in your local.conf file Feb 26 19:29:26 # ASSUME_PROVIDED += "virtual/${TARGET_PREFIX}gcc virtual/libc" Feb 26 19:29:40 OK Feb 26 19:29:44 03woglinde  07org.openembedded.dev * rcf7e4c7dbd 10openembedded.git/packages/update-alternatives/ (3 files in 2 dirs): update-alternatives: update to newer version of dpkg Feb 26 19:29:52 What branch of OE are you using, the .dev branch? Feb 26 19:31:04 This should be working in .dev, but if you were on stable perhaps that would exlpain it Feb 26 19:31:04 should be this one Feb 26 19:31:08 $ svn co svn://svn.berlios.de/bitbake/branches/bitbake-1.8/ bitbake Feb 26 19:31:16 branch of OE, not bitbake, sorry Feb 26 19:32:07 i iwas working from this Feb 26 19:32:11 http://wiki.openembedded.net/index.php/Getting_Started Feb 26 19:32:28 $ git clone git://git.openembedded.net/openembedded Feb 26 19:32:34 should be that right? Feb 26 19:32:43 Yeah, hm Feb 26 19:32:53 Not sure why, but it looks like the wrong gcc-3.4 is being called Feb 26 19:33:02 Do you use gcc-3.4 on your machine for other stuff? Feb 26 19:33:21 only quirky old shit that will compile only with 3.4 Feb 26 19:33:43 hmm, ok Feb 26 19:33:48 but it is there. Feb 26 19:33:58 for now can you apt-get remove gcc-3.4 and bitbake -c rebuild qemu-native Feb 26 19:33:58 ? Feb 26 19:34:26 it doesnt like 3.4 or it needs 3.4? Feb 26 19:34:35 It needs certain versions of 3.4 Feb 26 19:34:40 And will build the one it needs Feb 26 19:34:47 My theory is that it's calling the wrong one, your host one Feb 26 19:34:59 because qemu tries to be overly smart and outsmarts OE Feb 26 19:35:42 just remove 3.4? Feb 26 19:38:58 yes Feb 26 19:39:16 done. it's cooking on task do_compile: started Feb 26 19:39:29 task #115 Feb 26 19:40:07 gonna go get a bit bucket incase it pukes again... Feb 26 19:48:25 woglinde hi Feb 26 19:48:31 hi Feb 26 19:59:03 mr_nice did you work on the simpad lately? Feb 26 20:09:26 woglinde: a bit but i am currently more working on userland/opie Feb 26 20:09:49 ah okay Feb 26 20:11:04 woglinde: can you give me a hint how to add PREFERRED_PROVIDER_virtual/libsdl ?= "libsdl-qpe" for opie? Feb 26 20:11:56 hm Feb 26 20:12:09 do we have a class for opie? Feb 26 20:12:39 jupp Feb 26 20:12:43 I would set it there Feb 26 20:12:55 but better ask on the mailing list too Feb 26 20:13:06 woglinde: ok, thanks Feb 26 20:13:16 hm Feb 26 20:13:17 args Feb 26 20:13:18 nope Feb 26 20:14:27 hm nope Feb 26 20:14:33 its a candiate for local.conf Feb 26 20:14:35 or distro Feb 26 20:14:57 or for opie-image Feb 26 20:15:15 hm, the image may be the best compromise Feb 26 20:16:27 if the libsdl-qpe has the same version as libsld-x11 is it possible to use packages which where build with libsdl-x11 with libsdl-qpe? Feb 26 20:16:51 hm dont know Feb 26 20:16:54 I guess no Feb 26 20:16:58 but not 100% sure Feb 26 20:17:11 ok thanks i will try it out Feb 26 20:21:59 Tartarus-- thanks. that did the trick.. Feb 26 20:32:55 * mwester notes that his emails to the oe list are not going through, and wonders if someone has decided to moderate the list. Feb 26 20:34:34 woglinde, hi,I don't know what to do about emacs/emacs-x11:I saw emacs(non x11) in the FSO-unstable reposiroty but me and other people were unable to compile(there was somebody that reported a build failure in one of the lists I follow) it so I fixed it but I'm not sure if the fix are ok,moreover I used a qemu that was highly modified(and I don't know which patches fix the compilation faillure of emacs)and compiles only fo Feb 26 20:34:34 r arm Feb 26 20:34:43 i'll pastebin the recipe Feb 26 20:34:57 dorry Feb 26 20:35:06 I dont know about emacs compilation Feb 26 20:35:48 http://rafb.net/p/V7RyWu66.html Feb 26 20:35:50 ok Feb 26 20:35:57 who should I talk to then? Feb 26 20:36:41 mailnglist Feb 26 20:36:48 ok thanks Feb 26 20:36:58 just that irc was more convenient and faster... Feb 26 20:37:31 and that I fear the reactions about the modified qemu on the mailing lists Feb 26 20:46:51 if I install a file say libmaniac.a in usr/lib, where would it get packaged? -dev? Feb 26 20:49:26 Crofton yes Feb 26 20:49:44 per default all files with .a ends up in the dev-package Feb 26 20:49:54 look at conf/bitbake.conf Feb 26 20:50:03 thanks, it made sense to me right as I hit return :) Feb 26 20:54:19 RP: i have a ghetto do_rebuild.. it runs a recursive function that traverses the deps flag :P Feb 26 20:54:28 RP: does work, though Feb 26 20:57:34 kergoth what you are working on? Feb 26 21:02:02 woglinde: lots of stuff, glad i'm getting back in the game, was getting rusty for a while there Feb 26 21:02:07 03woglinde  07org.openembedded.dev * rcf7e4c7dbd 10openembedded.git/packages/update-alternatives/ (3 files in 2 dirs): update-alternatives: update to newer version of dpkg Feb 26 21:02:18 03Frans Meulenbroeks  07org.openembedded.dev * r064adeb426 10openembedded.git/ (4 files in 3 dirs): pmount: updated to 0.9.18 Feb 26 21:02:18 03woglinde  07org.openembedded.dev * r4c6d99ece8 10openembedded.git/packages/libpng/ (libpng-native_1.2.35.bb libpng_1.2.35.bb): Feb 26 21:02:18 libpng: add version 1.2.35 Feb 26 21:02:19 * fixes security issues Feb 26 21:12:58 waha Feb 26 21:13:25 * woglinde scratches head about git Feb 26 21:14:41 OK after massive amount of research, I got gpe-image compiling. but half the apps crashes (Alignment Traps) :/ Feb 26 21:15:27 woglinde: what about git? Feb 26 21:15:40 kergoth I squshed my mislead commit Feb 26 21:15:50 and now its twice in the log Feb 26 21:16:31 hm Feb 26 21:16:37 its only twice in my log Feb 26 21:16:56 wahaa Feb 26 21:17:03 I dont get it Feb 26 21:18:06 hm Feb 26 21:18:08 03woglinde  07org.openembedded.dev * rf902914ccd 10openembedded.git/ (3 files in 2 dirs): Feb 26 21:18:08 libpng: add version 1.2.35 Feb 26 21:18:08 * fixes security issues Feb 26 21:18:08 03woglinde  07org.openembedded.dev * re94b0d5d17 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 26 21:19:05 can some please pull Feb 26 21:19:14 and say If there are problems Feb 26 21:28:58 back Feb 26 21:35:33 woglinde: libpng 1.2.35 compiled fine for me Feb 26 21:43:22 denix: ping Feb 26 21:49:29 kergoth: pong Feb 26 21:50:43 denix: enterprise.conf tries to remove certain questionably licensed stuff from the build via BBMASK. if you change its one BBMASK = to: Feb 26 21:50:46 LDBBMASK := "${BBMASK}" Feb 26 21:50:46 BBMASK = "${@base_conditional('ENTERPRISE_DISTRO', '1', '(mp3blaster|mpg123|mpg321|smpeg|liba52|lame|libmad|mpeg2dec|xmms-mad|madplay|python-mad|opie-mediaplayer1-libmadplugin|ffmpeg|mplayer)', '${OLDBBMASK}', d)}" Feb 26 21:50:50 (ugly, yes, but it works) Feb 26 21:50:53 s/LDBB/OLDBB/ Feb 26 21:51:27 course, it still drops your bbmask if ENTERPRISE_DISTRO is 1 in that case, should probably just rewrite the thing as an anonyous function Feb 26 21:51:29 anonymous, even Feb 26 21:51:31 heh Feb 26 21:51:35 ok, what if I want to use ENTERPRISE_DISTRO and BBMASK? Feb 26 21:51:47 right Feb 26 21:52:01 the issue is, you cant just += something onto bbmask, because its a regex, not a list Feb 26 21:52:11 khem okay Feb 26 21:52:14 thanks Feb 26 21:52:30 kergoth: that's the problem :) Feb 26 21:53:19 hmm Feb 26 21:53:23 * kergoth tries something Feb 26 21:55:51 kergoth: and btw, the question wasn't why it doesn't work, but rather how to fix it... :) as I explained here - http://thread.gmane.org/gmane.comp.handhelds.openembedded/21148, but you probably already knew that :) Feb 26 21:56:08 ah, i mustve missed that reply Feb 26 21:56:11 at any rate, working on a fix now :P Feb 26 21:57:21 kergoth: thanks a lot! I'm still very bad at Python :( Feb 26 22:00:25 there are DNS problems with "monotone.openembedded.org Feb 26 22:00:30 " Feb 26 22:00:56 we do not use monotone anymore Feb 26 22:00:58 Daga: monotone is long since gone... Feb 26 22:01:08 hmm, ok Feb 26 22:01:13 monogone Feb 26 22:01:28 * Daga wonders what he gets to do to fix this now ^_^ Feb 26 22:01:32 thanks Feb 26 22:01:39 Daga: use git :) Feb 26 22:02:00 this is a part of the build system for ThinStation Feb 26 22:02:19 I'll do some digging :) Feb 26 22:05:14 OLDBBMASK := "${BBMASK}" Feb 26 22:05:14 BBMASK = "${@base_conditional('ENTERPRISE_DISTRO', '1', '(${OLDBBMASK})|(mp3blaster|mpg123|mpg321|smpeg|liba52|lame|libmad|mpeg2dec|xmms-mad|madplay|python-mad|opie-mediaplayer1-libmadplugin|ffmpeg|mplayer)', '${OLDBBMASK}', d)}" Feb 26 22:05:22 denix: try changing your enterprise.conf to those two lines Feb 26 22:05:31 denix: see if that fixes it for you Feb 26 22:06:11 damnit, nevermind, that has another problem Feb 26 22:06:12 grr Feb 26 22:06:16 bbmask was always such a hack Feb 26 22:06:17 so was bbfiles Feb 26 22:06:21 this crap has got to go Feb 26 22:06:32 kergoth: I believe I tried this "(blah)|(blah|blah)" before and it didn't work, I'll double-check Feb 26 22:06:39 it worked here Feb 26 22:06:49 it just dies when BBMASK="" or unset before enterprise.conf loads Feb 26 22:06:54 heh Feb 26 22:07:56 hrmph Feb 26 22:15:08 03Stanislav Brabec  07org.openembedded.dev * rb084ea418e 10openembedded.git/ (5 files in 3 dirs): vte: Updated to 0.16.14. Added desktop file. Fixed sub-packages dependencies. Feb 26 22:15:24 denix: okay, i think i have something that might work... boy is it ugly as hell though.. Feb 26 22:15:29 denix: :\ Feb 26 22:15:42 :) Feb 26 22:17:51 denix: http://kergoth.pastey.net/109060 Feb 26 22:18:17 denix: that works here, tested to confirm that both the user specified mask and the enterprise bits function Feb 26 22:19:29 should pester zecke, isn't he the one that wrote it? i think he was Feb 26 22:19:44 kergoth enterprise yes Feb 26 22:19:54 but he is on the flight to taipeh Feb 26 22:20:08 ah Feb 26 22:20:11 well, no rush Feb 26 22:20:15 he just has a pestering coming is all Feb 26 22:20:16 :) Feb 26 22:20:17 I'd say he broke BBMASK... :) Feb 26 22:20:22 aye Feb 26 22:20:47 depends which feature is more important to someone :) Feb 26 22:22:00 no doubt enterprise stuff is useful, it's just I don't use it... Feb 26 22:29:07 hi ant Feb 26 22:31:46 03Theodore A. Roth  07org.openembedded.dev * r53d28637d0 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: getting_oe.xml: Update "Building Software" section. Feb 26 22:34:14 woglinde: hello there Feb 26 22:42:30 kergoth: sorry for the delay - had to try different combinations. works perfectly! thanks a lot! Feb 26 22:42:42 glad to hear it, np Feb 26 22:42:45 itw as bugging me too :) Feb 26 22:43:29 kergoth: will you be commiting it or should we ask zecke? Feb 26 22:43:44 i'll commit it. ugly as it is, it's better than it was before :P Feb 26 22:44:56 ok, thanks. I was getting a feeling nobody cared about this bug - there were absolutely no replies on the ml... :) Feb 26 22:48:43 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r95d534d361 10openembedded.git/conf/distro/ (10 files in 2 dirs): Feb 26 22:48:43 minimal.conf: blatantly steal the collective wisdom (and I admit it must have Feb 26 22:48:43 been ton of work, thanks koen et. al.) people put in the angstrom distribution Feb 26 22:48:43 configuration and reorganize the sections a bit for more clarity. Most of this Feb 26 22:48:45 is really distribution-independent, so I'd welcome if angstrom would include one Feb 26 22:48:47 or the other of these sane-*.inc and maintain it there. Feb 26 22:53:06 ok, now we have a 'minimal' distro..let cross the finger and start a build Feb 26 22:56:11 I'm curious how minimal it ended up being, guess I need to pull Feb 26 23:00:07 I'd love a minimal.. angstrom with udev and all that takes a couple minutes to boot :/ Feb 26 23:03:05 BusError, minimal attempts to be a very simple featureless distro Feb 26 23:10:04 woglinde: is the uclibc build fixed now (was ARM_INSTRUCTION_SET = 'thumb')? Feb 26 23:17:56 ant_home: I would rather fix uclibc to be compilable in thumb mode Feb 26 23:18:19 We need to get rid of -fno-omit-frame-pointer for one file ldso.c Feb 26 23:18:29 and that will make gcc happy Feb 26 23:18:32 khem the ld.so bug is fixed with gcc-4.4 Feb 26 23:18:36 I think Feb 26 23:18:41 Nope Feb 26 23:18:47 nope? Feb 26 23:18:53 did you test it? Feb 26 23:19:00 looking at gcc 4.4 changes in this area Feb 26 23:19:11 it complains about not finding a register Feb 26 23:19:18 I did not test but I am fairly certain Feb 26 23:19:27 yes Feb 26 23:19:28 and there some changes on this topic in gcc Feb 26 23:19:41 yes there was a patch Feb 26 23:19:54 but that does not address the problem we have Feb 26 23:19:58 hm Feb 26 23:20:17 khem couldnt you wrote to the uclibc-ml on this topic Feb 26 23:20:28 hmm :) Feb 26 23:20:43 sure Feb 26 23:21:01 people want opts in ldso and at same time wants to keep frame info Feb 26 23:21:10 hm but -fno-omit-frame-pointer is a simpler solution yes Feb 26 23:21:29 for purists yes this should be fixed in gcc Feb 26 23:21:31 I still don't understand the difformity between various libc Feb 26 23:21:33 http://thread.gmane.org/gmane.comp.handhelds.openembedded/20414/focus=20416 Feb 26 23:21:49 okay I going to sleep now Feb 26 23:21:52 bye all Feb 26 23:22:02 'nite wog Feb 26 23:23:35 g'night Feb 26 23:40:10 kergoth: I'm 'happy' you confirm the issues with moving pstage and with -c rebuild...I've noticed it but then nobody had reproduced the issue...present since months ;-) Feb 26 23:40:55 hehe Feb 26 23:41:22 * kergoth hit like 16 bitbake and oe bugs in the past 48 hours.. just lucky, apparently.. Feb 26 23:41:28 ah, with vacuum-cleaned environment I had problems launching -c menuconfig (Xauth iirc) Feb 26 23:41:44 I'll retry in a few mins Feb 26 23:43:38 kergoth: prepend is same as += ? Feb 26 23:45:45 += is an append, =+ is a prepend Feb 26 23:45:49 but they're not the same Feb 26 23:45:54 append/prepend occur at the end of the parse Feb 26 23:45:58 += and =+ are immediate Feb 26 23:46:09 there's also no way to conditionally prepend/append with += and =+, you have to use the others Feb 26 23:46:32 right Feb 26 23:46:44 I confuse += and =+ Feb 26 23:46:49 ah Feb 26 23:47:09 of course, += and =+ use a space seperator Feb 26 23:47:21 you have to use .= or =. to append/prepend without seperator between the old and new content Feb 26 23:47:36 * kergoth mutters under his breath about annoyances with our file format Feb 26 23:47:43 hmm nice to know Feb 26 23:48:06 += and =+ are kind of intended to be word based, hence the spaces. but nobody ever wrote a wordwise -= :) Feb 26 23:48:32 heh Feb 27 00:06:31 ant_home: what confuses you about C libraries Feb 27 00:06:53 uclibc is machine specific? Feb 27 00:07:12 its configurable Feb 27 00:07:24 otherwise its like any other lib Feb 27 00:07:28 well, glibc too Feb 27 00:08:18 but ends up in armv5te and not in c7x0 Feb 27 00:08:28 ant_home: they use different mechanisms to be configurable Feb 27 00:08:59 oh thats OE specific thing, because they add configs based on machines to uclibc Feb 27 00:09:08 thats why it becomes machine specific Feb 27 00:09:23 it seems an historical legacy Feb 27 00:09:41 although I think it has no machine specific knobs used per say Feb 27 00:10:12 I can understand for klibc Feb 27 00:14:35 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rd88b40520a 10openembedded.git/packages/coreutils/ (19 files in 4 dirs): Feb 27 00:14:35 coreutils: make 6.0 the new default Feb 27 00:14:35 remove older ones Feb 27 00:14:41 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r43fa438235 10openembedded.git/packages/ncurses/ (ncurses.inc ncurses_5.4.bb): ncurses: modernize and fix packaging problems Feb 27 00:14:50 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r71e3268280 10openembedded.git/conf/distro/include/sane-toolchain.inc: sane-toolchain.inc: fix typo in binutils version Feb 27 00:14:57 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rd9bc47972d 10openembedded.git/packages/automake/ (17 files in 5 dirs): Feb 27 00:14:57 automake: update to 1.10.2 Feb 27 00:14:57 remove older ones Feb 27 00:15:06 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r0aad705da4 10openembedded.git/packages/coreutils/ (7 files in 2 dirs): coreutils: add 7.1. Not thoroughly tested yet, hence DP=-1 Feb 27 01:32:54 Vote for OE in Q8 of http://www.linuxdevices.com/cgi-bin/survey/survey.cgi?id=02252009112314 Feb 27 01:43:50 rwhitby, no stuffing the ballot box :) Feb 27 01:44:26 Crofton: one vote, one voice ;-) Feb 27 01:45:21 apparently one year OE was accused of over active promotion ... Feb 27 01:45:43 buyt, good on you mate Feb 27 01:46:23 I was pleased to see SlugOS specifically mentioned :-) Feb 27 01:46:46 yeah Feb 27 01:55:42 * scruggs just cast his vote for OE :) **** ENDING LOGGING AT Fri Feb 27 02:59:57 2009