**** BEGIN LOGGING AT Tue Dec 11 02:59:57 2007 Dec 11 03:46:05 * * OE Bug 3448 has been created by sonicvanajr(AT)gmail.com Dec 11 03:46:07 * * MPlayer lacks Real Video support Dec 11 03:46:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3448 Dec 11 04:00:24 03mickeyl 07org.oe.dev * r54e46ae8... 10/ (1 packages/linux/linux-openmoko.inc): Dec 11 04:00:24 linux-openmoko.inc: catch up with naming changes in linux.inc Dec 11 04:00:24 (hey, you could as well have adjusted linux-openmoko.inc while you were doing it...) Dec 11 04:00:31 03mickeyl 07org.oe.dev * r5480bec3... 10/ (4 files in 2 dirs): directfb: refactor commonalities into directfb.inc, bring back 1.0.0 Dec 11 04:01:13 * mwester-road scratches his head whilst reading the commit messages... Dec 11 07:04:04 * * OE Bug 3449 has been created by xydarcher(AT)hotmail.com Dec 11 07:04:06 * * genext2fs file size error Dec 11 07:04:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3449 Dec 11 07:32:12 !logs Dec 11 07:32:13 Channel logs for #oe are archived at: Dec 11 07:32:14 http://hentges.net/tmp/logs/irc/%23oe Dec 11 07:32:15 Live-logs are available at Dec 11 07:32:16 http://hentges.net/tmp/logs/irc/livelogs/%23oe.livelog Dec 11 07:32:18 See ?? help-logs for usage instructions Dec 11 08:24:00 hi ! if i want to use my own kernel config for an at91sam9263ek board and a 2.6.23 kernel, as well as my own patchset, what is the proper way to do it ? Dec 11 08:24:10 i guess packages/linux/linux-2.6.23/at91sam9263ek/defconfig for the config, but for the patches ? Dec 11 08:26:31 hmm, maybe i've found for the patches ... Dec 11 08:35:12 geekounet: I'm a bitbake newbie, but I guess you can put the patches in the files folder and update the .bb file to include them? Dec 11 08:37:08 hm, in fact i've found that in linux_2.6.22+2.6.23-rc3.bb : SRC_URI_append_kb9202 = " http://maxim.org.za/AT91RM9200/2.6/2.6.23-rc3-at91.patch.gz;patch=1 " Dec 11 08:37:20 i just have to do the same in the 2.6.23 :) Dec 11 08:37:36 Ah, ok :) Dec 11 09:18:23 03hrw 07org.oe.dev * r1a297661... 10/ (3 files in 3 dirs): zaurus-updater: do not run shell at the end - reboot instead Dec 11 09:50:48 hi Dec 11 09:52:48 hi all Dec 11 10:33:29 03hrw 07org.oe.dev * ra42329cb... 10/ (1 classes/rootfs_deb.bbclass): rootfs_deb.bbclass: few fixes to get it working Dec 11 10:33:34 03hrw 07org.oe.dev * rf9729b78... 10/ (2 files in 2 dirs): matchbox-keyboard-inputmethod: fix RCONFLICTS to get openmoko-image built Dec 11 10:43:33 weird, i cannot compile gcc-cross because of arm-angstrom-linux-gnueabi-gfortran missing ... :/ Dec 11 10:46:57 is there a way to have it ? :P Dec 11 10:49:50 morning Dec 11 10:52:48 morning hrw Dec 11 10:56:54 'morning all Dec 11 10:58:05 morning koen Dec 11 11:13:59 koen: #3884 time to get those binary codecs workin on arm :-) Dec 11 11:14:40 XorA|gone: win32 ones? Dec 11 11:14:56 hrw: yes, they work fine :-D Dec 11 11:15:37 XorA|gone: my c760 plays h.264 fullHD just fine with m$ codecs installed Dec 11 11:15:59 hrw: cool, blog that one :-) Dec 11 11:20:53 http://o-hand.com/ - new website Dec 11 11:28:47 mickey|zzZZzz: http://bugs.openembedded.org/show_bug.cgi?id=3288 is your toy... Dec 11 11:30:51 geekounet: some days ago I have the same problemwith the fortran stuff... make an update from oe Dec 11 11:39:17 hi all Dec 11 11:41:33 i'm trying to compile psplash ... and got error Dec 11 11:42:08 as happened to me other time bitbache try to retrieve via svn the revision 1 ... instead of the HEAD or the last one ... Dec 11 11:42:28 hi koen ! Dec 11 11:42:56 gremlin[it]: DISTRO=generic is broken Dec 11 11:43:02 gremlin[it]: use angstrom-2007.1 Dec 11 11:44:10 mhhhhh :( .... Dec 11 11:44:30 i really cannot at now 'generic' is not so bad for me ... Dec 11 11:46:38 which files i have to compare and check to make generic behave as angstrom-2007.1 regards svn ? Dec 11 11:47:08 you're on your own if you want to use generic Dec 11 11:48:42 re Dec 11 11:48:45 03hrw 07org.oe.dev * r91d5e4c4... 10/ (1 packages/curl/curl_7.16.4.bb): curl: really disable libidn (from Poky) Dec 11 11:49:04 * * OE Bug 3450 has been created by  Dec 11 11:49:06 * * python-pygtk-2.10.4-ml5-do_configure Dec 11 11:49:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3450 Dec 11 11:58:33 koen, angstrom have too much stuff to be a base for a custom distro :( ... Dec 11 11:59:08 gremlin[it]: what do you mean? Dec 11 11:59:33 there are tons of custom distros based on angstrom Dec 11 12:00:35 i got huge touble switching from kdrive to xorg with angstrom :( ... Dec 11 12:00:40 gremlin[it]: you are possibly confusing distro vs. images Dec 11 12:01:16 rschuster: indeed Dec 11 12:02:04 * * OE Bug 3451 has been created by  Dec 11 12:02:06 * * eet-native-0.9.10.041+cvs20071206-r0-do_fetch Dec 11 12:02:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3451 Dec 11 12:02:16 * * OE Bug 3452 has been created by  Dec 11 12:02:18 * * eet-0.9.10.041+cvs20071206-r0-do_fetch Dec 11 12:02:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3452 Dec 11 12:02:48 ok ... trying to do x11-image with angstrom (switching from kdrive to xorg) i got lot of trouble ... first of all .. kdrive still was builded Dec 11 12:02:51 selecting the xserver is in 99% of the case just a matter of doing XSERVER="xserver-xorg" in local.conf or in your custom image Dec 11 12:07:35 koen, what i did ... but not so straight ;( Dec 11 12:07:58 go to lunch Dec 11 12:08:20 rschuster: what's the ETA on merging all your java work? Dec 11 12:12:45 hey mickeyl Dec 11 12:13:01 koen: well. from my side I could start immediately. I hoped to get some input from henning but he hasn't much time these days Dec 11 12:13:39 I'd say: merge it and have henning fix his pet peeves afterwards Dec 11 12:13:39 koen: what do you suggest to get things started? Dec 11 12:13:45 :) Dec 11 12:14:01 koen: can I do manual routes in maemo mapper? Dec 11 12:14:07 mickeyl: hi Dec 11 12:14:11 morning Dec 11 12:14:18 mickeyl: http://bugs.openembedded.org/show_bug.cgi?id=3288 is your toy... can you look? Dec 11 12:14:22 hrw: create a .gpx file Dec 11 12:14:36 koen: can not make it from mapper itself? Dec 11 12:14:42 hrw: never tried it Dec 11 12:14:51 ok Dec 11 12:14:58 hrw: I'd rather have zecke doing this. He's the qmake dude nowadays Dec 11 12:15:10 ok Dec 11 12:17:31 mickeyl: all the E stuff is broken, see #3453 Dec 11 12:18:04 * * OE Bug 3453 has been created by  Dec 11 12:18:06 * * edje-native-0.5.0.41+cvs20071206-r0-do_compile Dec 11 12:18:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3453 Dec 11 12:18:37 courtesy of the autobuilder Dec 11 12:18:41 fun Dec 11 12:18:52 strange though, especially since it works fine here Dec 11 12:18:58 read the bug Dec 11 12:19:33 as soon as you start using rm_work or packaged-staging, it won't work fine Dec 11 12:19:48 mickeyl: for x86+arm builds? Dec 11 12:20:15 hrw: yo Dec 11 12:20:20 so what needs fixing? Dec 11 12:20:32 fast way to catch multimachine bugs is build for qemux86 and then for arm and again x86 Dec 11 12:20:34 efl's usage of autotools? Dec 11 12:20:38 or libtool itself? Dec 11 12:21:32 * hrw -> afk Dec 11 12:22:13 mickeyl: I'd put in an extra sed statement, but your best bet is to ask RP Dec 11 12:22:43 mickeyl: since yesterday we finally have sane .pc files in -dev packages Dec 11 12:22:53 * koen hugs sed Dec 11 12:25:38 ERROR: QA issue: libecore_fb.la failed sanity test (reference to workdir or installed) Dec 11 12:25:47 insane.bbclass is your friend Dec 11 12:26:37 i don't know anything about libtool. if libtool generates the .la files, it sounds to me that libtool needs to be fixed. Dec 11 12:27:00 you could try libtool-1.5.24 .... Dec 11 12:27:15 Crofton|home: last time i tried that no plugin could be loaded anymore. Dec 11 12:27:20 althoug I do ot it would solve that prolem Dec 11 12:27:24 mickeyl: it all depends on how libtool is called Dec 11 12:27:48 koen: ah. and the rules for that are defined by? Dec 11 12:28:23 the makefile Dec 11 12:29:23 so I'd say 'broken autotools usage' and stick some extra sed statements base.bbclass Dec 11 12:29:52 hmm, yeah Dec 11 12:30:28 * Crofton|home hates libtool Dec 11 12:31:12 one of the netfilter packages only build with .24 Dec 11 12:32:09 Crofton|home: http://thomas.apestaart.org/log/?p=312 Dec 11 12:34:14 Crofton|home: and pulseaudio nowadays _needs_ libltdl >= .24 otherwise it asserts() out on runtime :D Dec 11 12:34:51 mickeyl: so make libtool_1.2.24 the default, but not libtool-cross_1.2.24 Dec 11 12:35:01 and have pulse DEPENDS = "libtool" Dec 11 12:35:28 in angstrom 2007.1 ß Dec 11 12:35:29 ? Dec 11 12:35:42 or globally Dec 11 12:35:55 * Crofton|home goes inane Dec 11 12:35:59 insane Dec 11 12:36:15 alright, I hae to do revisions Dec 11 12:36:16 I don't think much people have an informed opinion on libtool non-cross Dec 11 12:36:40 maybe RP or kergoth Dec 11 12:36:47 but it ends there Dec 11 12:37:21 libtool drives people crazy, I think that is informed .... Dec 11 12:37:51 except for fact that it does a lot of stuff that people do not wnat to figure out Dec 11 12:37:51 Crofton|home: you really should read that apestaart blog post :) Dec 11 12:37:55 yeah Dec 11 12:38:01 I skimmed it :) Dec 11 12:38:08 it starts slowly Dec 11 12:38:15 and finsihes with a bang Dec 11 12:38:16 "And finally, I’ve learnt that I should not just blame libtool for my incompetence like so many other hackers." Dec 11 12:38:34 evening! Dec 11 12:38:41 hey lumag Dec 11 12:38:59 I think netfilter suffers from the something like that .... Dec 11 12:39:10 "We are powerful. But we can only be wielded by those who are worthy." :) Dec 11 12:55:25 koen, can you circulate the url for the CISDR workshop at your school? Dec 11 12:55:39 Crofton|home: trying that already Dec 11 12:55:44 thanks Dec 11 12:56:10 Basically, Chris got a NSF grant to go to Ireleand Dec 11 12:56:29 informally, the group meetings were called "The Ireleand Pub Crawl:" Dec 11 12:56:39 or, "Help Chris get tenure" Dec 11 12:59:00 03mickeyl 07org.oe.dev * rcb2356de... 10/ (3 files in 3 dirs): python-pydirectfb: commit missing patch Dec 11 13:03:11 all: has anyone ever tried sourcepkg.bbclass ? It's weird, if I inherit sourcepkg in foobar.bb , bitbake foobar -c rebuild will success, but bitbake foobar will fail Dec 11 13:05:23 ERROR: InvalidTask event exception, aborting Dec 11 13:05:24 build.py:256 if not task_graph.hasnode(task): Dec 11 13:05:24 raise EventException("Missing node in task graph", InvalidTask(task, d)) Dec 11 13:06:19 urmelZAUS|work, it's a fresh new build :/ Dec 11 13:13:16 johncylee: sourcepkg isn't intended to be inherited by recipes Dec 11 13:14:26 koen: INHERIT += ? Dec 11 13:15:02 koen: wow, that will generate all source packages alone the way... Dec 11 13:15:41 ahem. sourcepkg is broken Dec 11 13:16:08 http://bugs.openembedded.org/show_bug.cgi?id=3385 Dec 11 13:16:16 s/alone/along/ Dec 11 13:16:45 rschuster: yeah I read that. that's why I tried to inherit it in bb Dec 11 13:17:22 rschuster: how do I make it direct inheritable just like autotools, etc. ? Dec 11 13:19:16 johncylee: no idea, sorry. I had hoped the class could be fixed for general use :| Dec 11 13:19:58 not being able to provide sources is a serious crime in these copylefted times ;) Dec 11 13:19:58 hmm... I will to trace this. Dec 11 13:20:17 s/to// Dec 11 13:23:27 rschuster: put DL_DIR and a copy of the OE database online Dec 11 13:23:58 koen: :) Dec 11 13:26:51 * koen stabs mysql Dec 11 13:29:06 03koen 07org.oe.dev * rd83056ff... 10/ (3 files in 2 dirs): Dec 11 13:29:06 mysql: workaround two problems: Dec 11 13:29:06 * mysql dislikes parallel compilation, so set PARALLEL_MAKE = "" Dec 11 13:29:06 * the mysql-native helper binaries reference ${WORKDIR}, so implement a custom, empty do_rm_work() method Dec 11 13:29:07 This will need further thought when packaged-staging(2) is used Dec 11 13:35:41 since pulseaudio was mentioned. anyone willing to commit this little patch: http://bugs.openembedded.org/show_bug.cgi?id=3297 ? Dec 11 13:48:04 * * OE Bug 3454 has been created by  Dec 11 13:48:06 * * mplayer-0.0+1.0rc2-r6-do_compile Dec 11 13:48:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3454 Dec 11 13:49:04 * * OE Bug 3454 has been RESOLVED (INVALID) by Dec 11 13:49:06 * *  mplayer-0.0+1.0rc2-r6-do_compile Dec 11 13:49:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3454 Dec 11 13:59:12 03thebohemian 07org.oe.dev * rb33c947e... 10/ (3 files in 2 dirs): (log message trimmed) Dec 11 13:59:12 pulse: build hal-detect. Closes #3297 Dec 11 13:59:12 Currently OE's pulseaudio recipe does not build pulseaudio's hal-detect module. Dec 11 13:59:12 The attached patch adds hal to the dependencies and activates the configure Dec 11 13:59:12 switch. Dec 11 13:59:12 For what hal-detect is neccessary? Dec 11 13:59:15 If you plug in a USB headset to your device a pulseaudio server with the Dec 11 13:59:19 03koen 07org.oe.dev * r2a2a73f1... 10/ (1 packages/angstrom/angstrom-feed-configs.bb): Dec 11 13:59:20 angstrom-feed-configs: add iwmmxt feeds for machine that have 'iwmmxt' as MACHINE_FEATURE Dec 11 13:59:23 * Currently mplayer is the only iwmmxt package, since that's the only package for which people have *proven* iwmmxt improved performance Dec 11 14:01:47 and OE does not generate binary locales for iwmmxt Dec 11 14:02:51 and AFAIK glibc still crashes at random when built with iwmmxt Dec 11 14:04:04 * * OE Bug 3455 has been created by  Dec 11 14:04:06 * * php-native-5.2.0-r1-do_compile Dec 11 14:04:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3455 Dec 11 14:04:39 I will not test it anyway :D Dec 11 14:11:39 koen: thanks for applying! Dec 11 14:13:58 is it possible to configure ipkg to replace config files during a package upgrade by default? Dec 11 14:14:48 for example, I want to replace /etc/syslog.conf when a new version is available, without being prompted Dec 11 14:15:17 with ipkg -force-defaults the file isn't updated Dec 11 14:15:44 you mean *altered* config files Dec 11 14:18:54 koen: yes, that's it Dec 11 14:20:11 koen: I thought I could customize busybox so it doesn't handle /etc/syslog.conf, then create another separate package to maintain that file, but I was hoping ipkg could be manipulated instead Dec 11 14:22:04 * * OE Bug 3297 has been RESOLVED (FIXED) by Dec 11 14:22:06 * *  pulseaudio: build hal-detect module Dec 11 14:22:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3297 Dec 11 14:24:02 err... so am I out of luck? Dec 11 14:46:29 hi, all! could anybody help me with jffs2 gc issues? Dec 11 14:49:21 which one? Dec 11 14:49:58 hi Dec 11 14:50:20 hrw, I'm in process of pasting debug log. It is about endless loop in case of open of log file (it seems so) Dec 11 14:50:32 mickeyl: ping Dec 11 14:51:40 koen: ping Dec 11 14:52:11 bitbake question: ASSUME PROVIDED Dec 11 14:54:15 hrw, http://ossfans.org/paste?id=309c8d Dec 11 14:54:23 slapin: let me guess - out of space? Dec 11 14:54:31 ant|work: yes/ Dec 11 14:55:14 hrw, how could I check this? it shows 22 Mb are used out of 32 Mb. Dec 11 14:55:27 no idea then Dec 11 14:55:56 hrw, how to check if it is out of space? Dec 11 14:56:18 hrw, I mean out of flash Dec 11 14:57:06 jffs2 has a real bad algo for gc and freeing up space Dec 11 14:57:23 logfs needs far less free space for gc and defragging Dec 11 14:58:14 koen, do you know of any external tools to diagnose jffs2 problems? Dec 11 14:58:28 slapin: nope, sorry Dec 11 14:58:55 rschuster: file a bug and alert the list, or send a patch to the list Dec 11 14:59:04 koen: and additionally logfs is reserving space for the gc as well Dec 11 14:59:32 koen: ok Dec 11 14:59:35 koen, hrw zecke: any idea to where to put debug output to check where stuff is messed up? Dec 11 15:00:40 koen: with the help of RP we sorted out http://ibot.rikers.org/%23oe/20071202.html.gz Dec 11 15:00:45 koen: http://www.pastebin.org/10222 Dec 11 15:01:29 Koen: the question is: are the two python modules to be installed on the host? Is it a requirement? Dec 11 15:01:41 koen: or is simply this broken Ark Linux... Dec 11 15:02:04 no idea on that Dec 11 15:03:02 koen: python-encodings and python-ctupes Dec 11 15:03:13 *ctypes Dec 11 15:03:30 zecke: welcome qmake master Dec 11 15:03:41 koen: are these instaled with python-dev on other distro? Dec 11 15:03:52 hrw: oh ha Dec 11 15:04:13 zecke: http://bugs.openembedded.org/show_bug.cgi?id=3288 Dec 11 15:04:41 zecke: according to my discussion with RP it should be cross not native Dec 11 15:04:52 koen: could'nt be the *native* version should be used ? Dec 11 15:05:32 koen: sorry for the horrible copy and paste english... Dec 11 15:05:53 I have no knowledge of python whatsoever Dec 11 15:05:57 koen, adding "require conf/distro/include/sane-srcrevs.inc" on generic.conf resolve the issues with svn and similar ... Dec 11 15:05:58 mickeyl is the python dude Dec 11 15:06:14 koen: he helped me tonight Dec 11 15:06:33 koen: anyway thx Dec 11 15:10:06 hrw: hey... how's the printer today??? Dec 11 15:10:45 ~works Dec 11 15:10:46 works is, like, something I do Dec 11 15:11:13 zecke: what do you think? Dec 11 15:12:03 morning Dec 11 15:12:15 hrw: hehe. mkspecs has nothing MACHINE specific. To admit I don't know the current difference between cross and native Dec 11 15:12:31 hrw: qmake2 the binary is native for sure, the data it is using is not dependant on any ARCH Dec 11 15:15:58 zecke: anyway if you build qmake2-native when building for qemux86 and then try to build for arm it will break Dec 11 15:16:01 and viceversa Dec 11 15:18:28 zecke: they are stored in tmp/staging/ARCH/ Dec 11 15:18:34 zecke: it installs one arch specific spec file (.e.g arm-eabi for c7x) Dec 11 15:19:15 zecke: but since it inherits native it won't install a new spec (.e.g powerpc for efika) when switchin machine, since OE thinks qmake2 is already built Dec 11 15:19:27 which it is, but it is lacking spec files Dec 11 15:20:31 koen: right, but in theory it would install all necessary mkspecs by default Dec 11 15:20:42 right Dec 11 15:20:45 but it doesn't Dec 11 15:22:16 what's up? Dec 11 15:25:50 when a package tries to change another package's config file during an upgrade, it prompts and says "The default action is to keep your current version", can that default action be changed to "The default action is to *replace* your current version"? Dec 11 15:26:43 I need to update a configuration file but I can't have it be interactive Dec 11 15:27:05 jsheldon: I'm affraid you have to check the ipkg source Dec 11 15:27:32 koen: nuts, I was afraid of that, ok thanks for the tip Dec 11 15:40:04 * * OE Bug 2568 has been RESOLVED (FIXED) by Dec 11 15:40:06 * *  Angstrom logo looks bad on 320x320 screen (psplash) Dec 11 15:40:08 * * http://bugs.openembedded.org/show_bug.cgi?id=2568 Dec 11 15:42:02 thanks for that one! Dec 11 15:46:17 zecke: ping Dec 11 15:47:04 koen: yes :) Dec 11 15:47:13 It is mail-day for mw Dec 11 15:49:46 zecke: see PM Dec 11 16:05:21 koen: cool Dec 11 16:07:57 ok I figured out what I want, ipkg needs an "--assume-yes" option Dec 11 16:08:01 the linux.com dude sounds like zoidberg Dec 11 16:08:08 now to figure out how to patch it... Dec 11 16:23:43 whic distro supports the asus eee pc? Dec 11 16:25:46 CoreDump: hallo Dec 11 16:26:15 CoreDump: I found out where I read about TAB = 4 spaces Dec 11 16:26:28 CoreDump: http://www.python.org/dev/peps/pep-0008/ Dec 11 16:26:48 CoreDump: but for C http://www.python.org/dev/peps/pep-0007/ Dec 11 16:26:55 bye Dec 11 16:27:29 CoreDump: .bb packages are python... Dec 11 16:28:21 do anybody knows of some framebuffer driver for spi-based LCD? Dec 11 16:30:45 bye! Dec 11 16:43:59 03slapin 07org.oe.dev * r1dd8a9a5... 10/ (1 packages/linux/linux-hackndev-2.6_git.bb): linux-hackndev: Since we support git:// protocol now, use it by default Dec 11 16:58:04 * * OE Bug 3456 has been created by andrea.adami(AT)gmail.com Dec 11 16:58:06 * * Zaurus won't resume if suspended during charge Dec 11 16:58:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3456 Dec 11 17:18:20 folks ... I'd like to buy one PDA phone and I'd like to know if someone has any hint about a neat device to buy (since neo isn't yet ready) Dec 11 17:19:05 RP: ping Dec 11 17:23:50 Can someone tell me how x11-image do_rootfs() works Dec 11 17:24:16 I am trying to fix an issue where it is trying to package glibc-localedata-i18n instead of eglibc-localedata-i18n Dec 11 17:24:16 Khem: what part of it? Dec 11 17:24:45 Khem: basicly the .bb file inherits rootfs and says what to install Dec 11 17:25:07 Khem: and then most of the magic is happening in classes/rootfs.bbclass and rootfs_{deb,ipk...}.bbclass Dec 11 17:25:19 Khem: e.g. tracing IMAGE_LINGUAS might be interesting for you Dec 11 17:25:26 zecke: http://pastebin.ca/812285 Dec 11 17:25:47 Khem: also checking the log generated by ipkg is interesting to see if that localedata got installed directly or if there is a wrong DEPENDS Dec 11 17:26:20 I think its the depends thingy Dec 11 17:26:31 Khem: grep for it in classes ;) Dec 11 17:26:41 what I am doing is that I am providing glibc* from eglibc* Dec 11 17:26:58 Khem: grep -rn glibc-localedata-i18n classes/* Dec 11 17:27:13 Khem: and provide a better fix, and make it share the stuff between deb and ipk Dec 11 17:27:19 zecke: found the problem with the cronjob? Dec 11 17:27:29 classes/rootfs_deb.bbclass:63 and 29 Dec 11 17:28:04 otavio: yes, still need to fix the conversion script... Dec 11 17:30:23 zecke: what kind of fix I did not get your point Dec 11 17:31:01 Khem: how do you want to fix it? Dec 11 17:31:19 all I need is eglibc awareness Dec 11 17:32:26 where is LINGUAS_INSTALL populated Dec 11 17:34:45 zecke: when do you intend to do that? Dec 11 17:37:03 zecke: I dont think we have a flag to note which libc variant we are using system wide Dec 11 17:37:21 zecke: its a bigger problem for glibc and eglibc as the triplets are same Dec 11 17:37:26 and also the os is same Dec 11 17:39:29 zecke: it needs something like ANGSTROM_MODE to be global Dec 11 17:39:50 otavio: once my paid work and uni assignment are done Dec 11 17:44:16 Crofton|home: do you have Davinci experience or am I thinking of someone else? Dec 11 17:49:35 a little Dec 11 17:49:41 I have run Linux on a DV Dec 11 17:49:44 but not much more Dec 11 17:50:52 cbrake, I am painting a room in the house at the moment, be back occasionally Dec 11 17:54:20 zecke: I am adding a new variable TARGET_LIBC and trying Dec 11 17:54:40 zecke: that should simplify things Dec 11 17:57:41 re Dec 11 18:00:00 koen: g'morning Dec 11 18:00:08 hey Khem Dec 11 18:00:25 hrw|gone, XorA|gone, mickey|lunch: http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2007-December/001088.html Dec 11 18:00:26 koen: I am hacking on a X11-image and found issues with eglibc config Dec 11 18:00:38 more references to glibc? Dec 11 18:00:43 koen: there are some hardcoded stuff that is meant for glibc only Dec 11 18:00:47 yep Dec 11 18:01:09 so I am planning to introduce a global variable TARGET_LIBC Dec 11 18:01:24 which is same semantics as ANGSTROM_MODE but global Dec 11 18:01:30 is that really needed? Dec 11 18:01:49 there is no way right now to differentiate eglibc from glibc othertwise Dec 11 18:01:57 * XorA|gone waits an hour to receive that RFC Dec 11 18:02:16 * hvontres|work too - +1 from me early Dec 11 18:02:52 * Khem ipkg-cl core dumped Dec 11 18:03:46 koen: right now I am defining it as same scope as DISTRO etc. Dec 11 18:04:21 Khem: another option would be to make eglibc-localdata-* to RPROVIDES glibc-localdata-* Dec 11 18:04:41 koen: it does Dec 11 18:04:46 I have made that change Dec 11 18:05:11 but look at classes/rootfs_deb.bbclass:63 Dec 11 18:06:53 if it RPROVIDES glibc-localedata-i18n that bit in rootfs_deb.bbclass just works (TM) Dec 11 18:07:07 if it fails, it means that eglibc-localedata-i18n doesn't RPROVIDES glibc-localedata-i18n Dec 11 18:07:22 hmmmm Dec 11 18:07:31 * Khem looks Dec 11 18:09:13 I have added it to PACKAGES_DYNAMIC Dec 11 18:09:35 so that it has both glibc-localedata-* and eglibc-localedata-* Dec 11 18:11:26 will RDEPEND generate physcally two separate ipks Dec 11 18:12:00 because at classes/rootfs_deb.bbclass:63 it is looking for apt-get install glibc-localedata-i18n Dec 11 18:12:51 with eglibc we generate eglibc-localedata-i18n Dec 11 18:13:38 PACKAGES_DYNAMIC only informs OE that that recipe with generate packages matching that wildcard, since do_split_packages has no a priori knowledge of that Dec 11 18:14:04 Khem: you want something like http://rafb.net/p/3U4vo015.html Dec 11 18:14:17 ask RP or mickeyl, since my python is weak Dec 11 18:16:21 koen: I added that as parameter to do_split_packages call Dec 11 18:16:51 extra_depends parameter Dec 11 18:18:16 extra_depends works on RDEPENDS, not on RPROVIDES Dec 11 18:19:38 hmm ok Dec 11 18:19:54 I am verifying hard fix Dec 11 18:19:57 for now Dec 11 18:35:32 * XorA|gone wonders how people are managing to inject URLs into his drupal pages Dec 11 18:37:23 hrw|gone: btw, your blog is not showing up on planet.linuxtogo.org anymore :( Dec 11 18:49:40 huhu Dec 11 18:50:27 bonne soirée Dec 11 18:53:04 * * OE Bug 3457 has been created by thebohemian(AT)gmx.net Dec 11 18:53:06 * * provide a java5-capable compiler to compile native classpath and JVM Dec 11 18:53:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3457 Dec 11 19:09:05 * * OE Bug 3458 has been created by  Dec 11 19:09:07 * * Issues discovering BT device names with bluez 3.22 Dec 11 19:09:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3458 Dec 11 20:26:52 03koen 07org.oe.dev * r39b27e31... 10/ (1 packages/images/minimal-gpe-image.bb): Dec 11 20:26:52 minimal-gpe-image: add images that only offers a basic 'desktop' Dec 11 20:26:52 compared to other images: Dec 11 20:26:52 3.3M Angstrom-minimalist-image-glibc-ipk-2007.9-test-20071211-collie.rootfs.tar.gz Dec 11 20:26:52 8.5M Angstrom-console-image-glibc-ipk-2007.9-test-20071211-collie.rootfs.tar.gz Dec 11 20:26:53 11M Angstrom-minimalist-gpe-image-glibc-ipk-2007.9-test-20071211-collie.rootfs.tar.gz Dec 11 20:26:57 17M Angstrom-x11-image-glibc-ipk-2007.9-test-20071211-collie.rootfs.tar.gz Dec 11 20:26:58 03koen 07org.oe.dev * r43dea013... 10/ (1 conf/machine/include/collie-2.6.inc): collie-2.6.inc: copy installkit code from zaurus-2.6.inc Dec 11 20:58:38 good evening all Dec 11 21:11:31 hi Dec 11 21:18:32 * Khem verified koen was right with RDEPNDS thing in eglibc Dec 11 21:36:18 Khem: how much smaller is an eglibc build vs glibc? Dec 11 21:44:28 Hello. Anyone up for a question? -- I'm trying to create a new image. I've created my image file and assigned PACKAGE_INSTALL but my packages aren't getting into the image produced. Dec 11 22:04:35 morning all Dec 11 22:31:04 * * OE Bug 3458 has been RESOLVED (FIXED) by Dec 11 22:31:06 * *  Issues discovering BT device names with bluez 3.22 Dec 11 22:31:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3458 Dec 11 23:21:04 * * OE Bug 3459 has been created by andrea.adami(AT)gmail.com Dec 11 23:21:06 * * altboot-image for c7x0 stuck on altboot prompt Dec 11 23:21:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3459 Dec 11 23:24:26 hvontres|work: eglibc is configurable like uclibc Dec 11 23:25:49 hvontres|work: I guess in best case it could be half but i have not measured myself Dec 11 23:27:48 Khem: thantk Dec 11 23:27:59 Khem: err, thanks Dec 11 23:29:38 Good night Dec 12 00:03:41 RP: Hi Dec 12 00:04:19 RP: hit today http://bugs.openembedded.org/show_bug.cgi?id=3010 Dec 12 00:08:16 RP: do we need shared-mime-info and desktop-file-utils on host? Dec 12 00:12:31 RP: installed shared-mime-info and now bluez-gnome installs ok! Dec 12 00:12:54 RP: but smthg wrong NOTE: the following files were installed but not shipped in any package: Dec 12 00:13:05 RP: NOTE: /usr/share/mime/globs Dec 12 00:13:17 RP: NOTE: /usr/share/mime/magic Dec 12 00:13:28 RP: etc etc Dec 12 00:30:04 * * OE Bug 3445 has been RESOLVED (FIXED) by philippedeswert(AT)scarlet.be Dec 12 00:30:06 * *  gpe-mini-browser-0.2 QA issues Dec 12 00:30:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3445 Dec 12 00:50:09 Khem: So how far away are you from being able to build an eglibc-console image in Angstrom? Dec 12 00:58:30 hvontres|work: I am able to run console-image for angstrom 2007 and 2008 on omap boards Dec 12 00:58:44 hvontres|work: for couple of months now Dec 12 00:58:52 I am trying on x11-image now Dec 12 01:06:00 Khem: way cool :) Dec 12 01:29:25 hvontres|work: There is some more work that will make it a real gain Dec 12 01:29:58 hvontres|work: in eglibc you can generate locales in cross env Dec 12 01:30:08 so no need of qemu'ing Dec 12 01:30:42 then I would like to use the config options thats new in eglibc so we can chose features Dec 12 01:33:36 i'm targeting a mipsel machine - can anybody suggest me a good binutils/gcc/glibc combination? So far we have used glibc-2.3.5+cvs20051107, gcc-4.1.1, binutils-2.16.1 Dec 12 01:34:26 tmbinc_: any reason you are using glibc 2.3.5 Dec 12 01:38:51 no, it was the only working version back then in 2005 Dec 12 01:39:07 there are some scary nptl patches applied which i like to get rid of Dec 12 01:39:28 tmbinc_: I think 2.6.1 would be good option Dec 12 01:39:39 tmbinc_: you can also dare 2.7 Dec 12 01:39:55 tmbinc_: you will need to update binutils too Dec 12 01:40:06 2.18 would be good I think Dec 12 01:41:03 tmbinc_: I have successfully used gcc 4.2.1 glibc 2.6.1 binutils-2.18 combo on mips/mips64 machines Dec 12 01:51:43 Khem: inside OE? Dec 12 01:57:03 tmbinc_: not OE Dec 12 01:57:18 tmbinc_: on OE I have not tried mips at all Dec 12 01:57:41 but I think it should just work. If not let us know Dec 12 02:06:05 re Dec 12 02:46:05 * * OE Bug 3460 has been created by raj.khem(AT)gmail.com Dec 12 02:46:06 * * x11-image does not build with ANGSTROM_MODE=eglibc Dec 12 02:46:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3460 Dec 12 02:59:44 03pfalcon 07org.oe.dev * r320cd49f... 10/ (2 files in 2 dirs): Dec 12 02:59:44 matchbox-keyboard-inputmethod svn: Apply multiple-layouts patches. Dec 12 02:59:44 * Just the same as matchbox-keyboard. Dec 12 02:59:44 * Actually, this is adhox workaround for non-determinism in image building Dec 12 02:59:44 (when ipkg, having matchbox-keyboard and matchbox-keyboard-inputmethod, both Dec 12 02:59:45 Provide'ing each other, selects some non-deterministically). Dec 12 02:59:52 03pfalcon 07org.oe.dev * rfffc11db... 10/ (3 files in 3 dirs): Dec 12 02:59:52 angstrom-2007.1.conf: Use bluez 3.23. Dec 12 02:59:53 * Fixes #3458, timeout while getting bluetooth device names. **** ENDING LOGGING AT Wed Dec 12 02:59:57 2007