**** BEGIN LOGGING AT Fri Apr 14 09:59:57 2006 Apr 14 09:59:58 All my boxes are gentoo-powered Apr 14 10:00:10 mine too :) Apr 14 10:00:15 And gentoo uses some very "special" packages Apr 14 10:00:18 Ecco_: and gentoo does not track where does file comes? Apr 14 10:00:21 (i.e. smalleer ones) Apr 14 10:00:23 ~lart gentoo then Apr 14 10:00:24 * ibot does a little 'renice 20 -u gentoo then' Apr 14 10:00:27 I don't think so Apr 14 10:00:37 it should Apr 14 10:00:38 equery belongs file Apr 14 10:00:40 RP: under 2.6.16 you tested that 'dd' thing? Apr 14 10:00:43 it does Apr 14 10:00:44 otherwise emerge --purge wouldn't work Apr 14 10:00:48 ha, thanks Genesis Apr 14 10:02:07 hrw|work: Yes, it doesn't work the same under 2.6.16. Apr 14 10:02:09 look equery documentation , it's very usefull Apr 14 10:02:20 Genesis: In that case, the arm system emulator doesn't support it Apr 14 10:02:20 Thanks, I wasn't aware of that tool Apr 14 10:09:47 for gentoo install emerge epm, then epm -ql to list files in a package Apr 14 10:10:50 XorA, hrw|work: abiword-plugins + deps are in the branch, all it needs is the updated enchant and abiword 2.4.4 Apr 14 10:11:16 koen: I think ive cracked spelling problem, just about to test, its a one line bug in configure.in Apr 14 10:11:21 equery files packet Apr 14 10:13:02 Genesis: ok, im outta date then :-) Apr 14 10:13:32 XorA: http://www.abisource.com/changelogs/2.4.4.phtml Apr 14 10:13:38 I'll have a look at poppler Apr 14 10:13:50 XorA, : epm seems to be for rpm Apr 14 10:14:06 ah no Apr 14 10:14:09 Description: rpm workalike for Gentoo Linux Apr 14 10:14:14 Genesis: epm was meant to mirror rpm commands to make stuff easier to go to gentoo Apr 14 10:14:21 yeap Apr 14 10:14:26 but the gentoo way is equery Apr 14 10:14:38 oki Apr 14 10:14:40 Genesis: back in bad old days it was only relistic way to query what was in a package Apr 14 10:21:26 * koen has found the poppler problem Apr 14 10:21:31 * koen add --enable-xpdf-headers to EXTRA_OECONF Apr 14 10:26:51 koen: erp, our aspell.bb doesnt do any wordlists Apr 14 10:29:15 XorA: http://oe.pastebin.com/659265 Apr 14 10:29:23 that's the latest progress on the plugins Apr 14 10:30:50 koen: cool Apr 14 10:35:43 03koen 07org.oe.oz354x * r80983972... 10/packages/mml-widget/gtkmathview_0.7.6.bb: gtkmathview: add 0.7.6 from .dev Apr 14 10:35:47 03koen 07org.oe.oz354x * r2b9ac288... 10/packages/abiword/abiword-plugins_2.4.4.bb: abiword-plugins: add 2.4.4 from .dev Apr 14 10:35:51 03koen 07org.oe.oz354x * r5020cd96... 10/packages/poppler/poppler_0.5.1.bb: poppler: add 0.5.1 from .dev Apr 14 10:35:55 03koen 07org.oe.dev * rca71e0d2... 10/packages/mml-widget/gtkmathview_0.7.6.bb: gtkmathview: add 0.7.6 Apr 14 10:35:57 XorA: does abiword need a fix or is it an enchant problem? Apr 14 10:35:59 03koen 07org.oe.dev * r95eda253... 10/packages/abiword/abiword-plugins_2.4.4.bb: abiword-plugins: enable mathml plugin Apr 14 10:36:03 03koen 07org.oe.dev * r1b045e26... 10/packages/poppler/poppler_0.5.1.bb: poppler: install xpdf headers for apps that need those Apr 14 10:37:21 koen: I have enabled aspell in enchant, so I dont think abiword needs to change Apr 14 10:38:42 XorA: ok, I've pushed abi 2.4.4 to the branch Apr 14 10:38:54 koen: I think problem now is the lack of wordlists Apr 14 10:38:58 so we only need the fixed aspell and enchant Apr 14 10:39:12 koen: I will push enchant now Apr 14 10:40:15 03koen 07org.oe.oz354x * re593df6f... 10/packages/abiword/abiword_2.4.4.bb: abiword: add 2.4.4 from .dev Apr 14 10:40:18 great Apr 14 10:41:01 koen: you familiar with aspell? Apr 14 10:41:06 nope Apr 14 10:41:15 koen: I think we need aspell-en.bb aspell-fr.bb etc etc Apr 14 10:42:50 with aspell.inc Apr 14 10:43:16 hrw|work: and aspell-native Apr 14 10:43:44 03xora 07org.oe.dev * rd2140203... 10/packages/enchant/ (enchant_1.2.5.bb files/configure.patch): Apr 14 10:43:44 enchant_1.2.5.bb : add a new version and fix a problem with configure.in Apr 14 10:43:44 that was preventing aspell support building. Apr 14 10:44:00 Just a general question here, but is it possible to "brick" a Zaurus ? Apr 14 10:44:07 Ecco_: only collie Apr 14 10:44:11 ok, great Apr 14 10:44:22 hrw|work: each country code aspell.bb would just be the dictionary for that language, aspell binary still in seperate package Apr 14 10:44:25 On the Akita for instance the flashing program cannot be overwriten, right ? Apr 14 10:45:05 Ecco_: it can, but the nand-restore is in a seperate ROM chip Apr 14 10:45:24 Ahem Apr 14 10:45:35 Which means ? (sorry, I'm a bit new to that) Apr 14 10:45:49 XorA: yep Apr 14 10:46:12 it means that you can overwrite the bootloader and stuff, but you can fix that by using the nand-restore in the emergency menu Apr 14 10:46:22 oh, ok, that's awesome :-D Apr 14 10:46:29 Ecco_: each Z>collie has nand-backup/restore in system diagnostics which dump/restore whole nand flash Apr 14 10:47:07 By the way, I have another question : g_ether (Ethernet over gadget usb) doesn't work on my C1000 Apr 14 10:47:13 Where do I report the problem ? Apr 14 10:47:50 oe bugtracker? Apr 14 10:47:54 ok Apr 14 10:53:29 hrw|work: http://www.oesf.org/forums/index.php?showtopic=18305&st=60&p=123159&#entry123159 Apr 14 10:54:07 gracias Apr 14 10:59:57 building Apr 14 11:04:23 ~lart kernel modules for lacking deps Apr 14 11:04:23 * ibot breaks out the Hoover and sucks up kernel modules for lacking deps Apr 14 11:04:59 03koen 07org.oe.oz354x * rf794c529... 10/packages/enchant/ (enchant_1.2.5.bb files/configure.patch): enchant: add 1.2.5 from .dev Apr 14 11:05:01 ieee80211-crypt-wep depend on ieee80211-crypt but also use arc4 Apr 14 11:09:55 How do you guys feel about that : Apr 14 11:09:59 http://bugs.openembedded.org/show_bug.cgi?id=841 Apr 14 11:15:55 Good day Apr 14 11:17:09 heh... my builddir is full of tmp/work in .la ;( Apr 14 11:22:35 hrw@bitbake:~/devel/build/oz354x$ grep 'tmp/work' tmp/staging/arm-linux/lib/*la -l|wc 59 59 2511 Apr 14 11:22:38 ;( Apr 14 11:22:45 hrw|work: hmm, shouldnt the sed in base.bbclass have a g on the end of it? Apr 14 11:23:09 XorA: that builddir is older then this change Apr 14 11:23:34 hrw|work: just checked mine and the change only seems to change the first entry Apr 14 11:23:50 hrw|work: though Im a bit crosseyed from aspell junk so I might be misreading Apr 14 11:27:49 koen: hrw|work spelling is now working in abiword Apr 14 11:28:19 koen: hrw|work give me ten mins to clean up my files and Ill check in aspell-lang stuff then everyone can add their own language Apr 14 11:28:30 amusingly fuck-wit is understood by aspell Apr 14 11:28:34 ok Apr 14 11:35:06 Help me out here: How do i patch the 'hciattch' with the file 5168-001.bin so i can test if the HTC universal can manage to run bluetooth at speed 921600.? Apr 14 11:35:24 The fil 5168-001.bin is form SDG Apr 14 11:37:17 patch hciattach.c < 5168-001.bin Apr 14 11:38:15 morning guys Apr 14 11:38:21 hail mickey Apr 14 11:38:41 do i have the *.bin file in the same directory as hciattach? Apr 14 11:38:53 03mickeyl 07org.oe.dev * radbb0fa0... 10/packages/qt/qt4-x11-free_4.1.1.bb: qt4: set platform to ${TARGET_OS}-g++ to prevent host based assumptions Apr 14 11:39:33 doesn't matter Apr 14 11:39:54 mickeyl: hey, and cool Apr 14 11:40:01 hey mickeyl Apr 14 11:40:10 hi mickeyl Apr 14 11:43:56 mmh Apr 14 11:44:01 python bindings to libqanava, anyone? Apr 14 11:44:02 http://www.libqanava.org/ Apr 14 11:44:15 * mickeyl doesn't need yet another project but would love to have bindings... Apr 14 11:45:59 mickeyl: qanava looks a lot like http://www.kdedevelopers.org/node/1927 Apr 14 11:46:43 although qanava makes graphing a lot easier Apr 14 11:48:43 03xora 07org.oe.dev * r33dbaffa... 10/packages/aspell/ (aspell-lang.inc aspell5-en_6.0-0.bb): Apr 14 11:48:43 aspell5-en_6.0-0.bb, aspell-lang.inc : Add English dictionary for aspell Apr 14 11:48:43 everyone go ahead and copy/alter for languages you can read/test. I am Apr 14 11:48:43 unfortuneately unilingal Apr 14 11:49:01 heh Apr 14 11:49:01 koen: spelling is now complete Apr 14 11:49:15 XorA: cool, thanks Apr 14 11:53:14 * pb_ stabs apmd Apr 14 12:03:45 koen: I think you broke perl for non gnueabi case :-( Apr 14 12:03:47 * mickeyl waits patiently for another incarnation of qt4 Apr 14 12:04:00 XorA: I did? Apr 14 12:04:13 koen: or forgot to check a file in Apr 14 12:04:59 hi Apr 14 12:05:19 koen: you forgot config.sh-arm-linux-gnueabi Apr 14 12:05:25 D'oh Apr 14 12:05:33 I forgot to remove it from SRC_URI Apr 14 12:06:35 XorA: fixed Apr 14 12:06:45 hi CoreDump Apr 14 12:08:20 CoreDump|home: scroll back, you've broken openssh/dropbear Apr 14 12:09:01 second Apr 14 12:09:07 03koen 07org.oe.dev * rd93900aa... 10/packages/perl/perl_5.8.7.bb: perl 5.8.7: remove extra copy of config.sh from SRC_URI Apr 14 12:10:12 * CoreDump|home yawns Apr 14 12:10:26 now what did I break? Apr 14 12:10:42 CoreDump|home: you added RCONFLICTS/RREPLACES? Apr 14 12:10:42 ipkg upgrade will remove dropbear and install openssh-sshd Apr 14 12:11:20 XorA: correct, RCONFLICTS / RREPLACES for openssh-sshd. Apr 14 12:11:53 which is wrong Apr 14 12:12:21 CoreDump|home: openssh-sshd neither CONFLICTS or REPLACES dropbear Apr 14 12:12:24 huh? It is supposed to remove dropbear when you manually install openssh-sshd, whats so wrong with that? Apr 14 12:12:52 CoreDump|home: I can have openssh-sshd as sshd + dropbear as ssh Apr 14 12:13:48 the wasn't the behaviour last time I checked. It would try to run both deamons Apr 14 12:15:00 however, I didn't intend openssh to automatically replace dropbear on a simple ipkg upgrade heh Apr 14 12:28:49 koen: looks like abiword update will have to wait - fscked my base 354 builddir ;( Apr 14 12:34:36 hrw|work: :-( Apr 14 12:35:50 hmm nmap depends on libstdc++6 Apr 14 12:40:01 XorA: I can provide armv5te binaries today Apr 14 12:49:33 03rwhitby 07org.oe.dev * r3392085b... 10/packages/meta/slugos-packages.bb: slugos-packages: Demoted icecast (fails to build). Apr 14 12:51:30 Hello guys Apr 14 12:51:50 What's the best way to modify ony the kernel of a distro in OE ? Apr 14 12:53:05 depends what you want to do Apr 14 12:53:06 Ecco: Add the patch to the list in the bitbake recipe Apr 14 12:53:26 and edit defconfig for your machine Apr 14 12:57:49 ok Apr 14 12:58:02 but let's say I'd like to use "make menuconfig" ? Apr 14 12:58:25 use make menuconfig, save the resulting defconfig, overwrite the one in OE, rebuild the kernel Apr 14 12:58:33 then look how bitbake/OE works and then you will see how to use make meuconfig during build Apr 14 12:59:18 Ok, thanks Apr 14 12:59:32 hi zecke Apr 14 13:00:02 *home* Apr 14 13:00:13 mickeyl: hey Apr 14 13:00:30 welcome back! Apr 14 13:01:24 hail zecke Apr 14 13:01:55 hi zecke & mickeyl Apr 14 13:03:21 yo CoreDump|home Apr 14 13:04:28 someone familiar with kernel module dependencies (in kernel source)? Apr 14 13:04:44 interpackage dependencies are generated from module infos - but for example ieee80211_crypt need arc4 and crc32 to be present but does not depend on them (in KERNEL/net/ieee80211/KConfig is SELECT them automatically but not DEPEND) - is that normal or a bug? Apr 14 13:05:06 hrw|work: what happens if they are not present? Apr 14 13:05:37 the package dependencies are generated from depmod output, not from Kconfig; the dependencies in the configuration system are fairly irrelevant Apr 14 13:06:03 wifi0: registered netdevice wlan0 Apr 14 13:06:03 ieee80211_crypt_wep: could not allocate crypto API arc4 Apr 14 13:06:03 wlan0: could not initialize WEP: load module hostap_crypt_wep.o Apr 14 13:06:43 thats happen Apr 14 13:06:52 ah, right, that is unfortunate Apr 14 13:07:09 I would consider that a bug in the kernel, but I suspect the kernel people might disagree. Apr 14 13:07:26 I can workaround this by adding more modules into machine configs but its not a solution Apr 14 13:07:40 the best workaround is probably to put RRECOMMENDS_kernel-module-ieee80211-crypt = "..." in kernel.bbclass. Apr 14 13:08:21 and sooner or later we could get long kernel.bbclass... Apr 14 13:08:35 indeed, that's the risk you run Apr 14 13:09:12 03mickeyl 07org.oe.dev * r7dc4f399... 10/ (classes/qt4x11.bbclass packages/qt/qt4-x11-free_4.1.1.bb): qt4: embrace STL Apr 14 13:09:12 03mickeyl 07org.oe.dev * r044a5885... 10/packages/zope/zope_2.9.2.bb: zope: add 2.9.2 Apr 14 13:09:15 03mickeyl 07org.oe.dev * rdcf8e3a2... 10/packages/libqanava/ (libqanava_0.0.3.bb libqanava_0.0.7.bb): libqanava: upgrade to 0.0.7, needs qt4x11 now Apr 14 13:09:19 kernel.bbclass is already nearly 400 lines long, though, so adding that one statement would only be an increase of approximately 0.25% Apr 14 13:09:20 03mickeyl 07org.oe.dev * r2ff27dc2... 10/packages/plone/ (plone_1.0.5.bb plone_2.0.0.bb plone_2.1.2.bb): plone: upgrade to 2.1.2, remove 1.0.5 Apr 14 13:09:25 I think that's probably tolerable. Apr 14 13:10:20 alternatively, you could send hate-mail to the kernel h4x0rs until they fix their code. Apr 14 13:10:35 or I suppose yet another option would be to fix the code yourself Apr 14 13:11:54 * mickeyl agrees with pb_ Apr 14 13:12:21 mickeyl: http://openzaurus.berlios.de/ - oz wiki Apr 14 13:12:37 mickeyl: taking an easter holiday? Apr 14 13:13:20 XorA: kind of that, yeah. i need a bit of OE maintaining work to relax :) Apr 14 13:14:03 ~lart collie for its too smal flash reat estate Apr 14 13:14:03 * ibot breaks out the Hoover and sucks up collie for its too smal flash reat estate Apr 14 13:14:28 CoreDump|home: opie-image created for collie from feeds booted Apr 14 13:14:54 hrw|work: I know, found my problem. My build were indeed too large :\ Apr 14 13:15:03 had to remove the lovely bootsplash heh Apr 14 13:18:54 ;( Apr 14 13:19:45 ~lart dansawyer, even after explaining him _3_ times what an .ext2 file is, he is _still_ complaining that he can't use / flash it Apr 14 13:19:45 * ibot send killer squirrels to attack dansawyer, even after explaining him _3_ times what an .ext2 file is, he is _still_ complaining that he can't use / flash it Apr 14 13:21:43 mickeyl: http://openzaurus.org/wordpress/2006/04/14/openzaurus-wiki/ Apr 14 13:27:28 hrw|work: good. eventually we may completely move from sf.net Apr 14 13:28:40 koen, XorA: something for you Apr 14 13:28:41 | checking for /proc/self/maps... configure: error: cannot check for file existence when cross compiling Apr 14 13:28:44 | FATAL: oe_runconf failed Apr 14 13:28:46 NOTE: Task failed: /home/hrw/devel/build/3541/tmp/work/armv5te-linux/enchant-1.2.5-r0/temp/log.do_configure.25935 Apr 14 13:28:53 http://pastebin.com/659498 Apr 14 13:29:06 --disable-binreloc Apr 14 13:29:09 hrw|work: crappit, forgot to check in change to sitefile Apr 14 13:29:33 koen: ah, does that do the same thing? Apr 14 13:31:08 XorA: it should Apr 14 13:31:23 testing.. Apr 14 13:31:28 NOTE: package enchant-1.2.5-r0: task do_configure: completed Apr 14 13:31:34 no idea what it does, but it works Apr 14 13:31:44 and the resulting packages still work as well Apr 14 13:31:47 hrw|work: can you update packages? Im in mioddle of a gpe-image build Apr 14 13:31:48 heh Apr 14 13:32:45 XorA: adding it to the site file might be a good solution as well Apr 14 13:33:22 koen: looks like it means you can install the ipkg anywhere and it will figure stuff out itself Apr 14 13:34:14 NOTE: package enchant-1.2.5-r0: task do_distribute_sources: started Apr 14 13:34:14 NOTE: Licenses (['unknown']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Apr 14 13:34:17 hmm.. Apr 14 13:34:53 * hrw|work votes for adding src_distribute_local with 'symlink' method to default set on INHERIT Apr 14 13:34:58 XorA: iirc binreloc was about fancier stuff (scratchbox depends on it), but I haven't looked into it Apr 14 13:35:10 ipkg install -d still works Apr 14 13:35:50 koen: looks bloody useful for installing stuff on SD/CF cards without needing ipkg-link Apr 14 13:36:02 yeah Apr 14 13:36:23 pb_ has some assembly magic for that in gpe-appmgr Apr 14 13:37:17 so we just need to decide which solution is the best Apr 14 13:38:23 I remember the scratchbox people saying "any non-ancient glibc supports it" Apr 14 13:40:57 03rwhitby 07org.oe.dev * ra9390448... 10/ (39 files in 6 dirs): unslung-kernel, unslung-image: Now that 6.8-beta is released, moved mainline to R63 and dropped R29 (people can get it from SVN if they want to build old versions) Apr 14 13:41:02 03rwhitby 07org.oe.dev * r57ac6022... 10/conf/distro/unslung.conf: unslung.conf: Moved to 6.9-alpha in preparation for any upcoming bug fixes Apr 14 13:44:16 koen: how much stuff uses binreloc in OE Apr 14 13:45:24 ~lart libtool badly Apr 14 13:45:25 * ibot squeezes libtool badly till libtool badly turns blue like papa smurf Apr 14 13:45:29 http://pastebin.com/659514 Apr 14 13:45:34 XorA: http://oe.pastebin.com/659516 Apr 14 13:45:59 mickeyl: could you change bbshell to use oe.pastebin.com? Apr 14 13:46:00 hi zecke Apr 14 13:46:23 hey RP & zecke Apr 14 13:46:38 oe.pastebin.com? Apr 14 13:46:39 mickeyl: I sent an email to the hostap list about the suspend/resume problem Apr 14 13:46:42 hi koen Apr 14 13:46:43 hrw|work: thats what I get as well :-( Apr 14 13:46:48 RP: good, thank you. Apr 14 13:46:49 hrw|work: hence why I clean /tmp Apr 14 13:47:11 koen: ah, cool. will do Apr 14 13:48:45 mickeyl: opie need update for tskeys Apr 14 13:49:31 opie needs update for many things . . . Apr 14 13:49:54 CoreDump|home: can we move tskeys to other Fxx? under opie 'blanks' are like some off apps keys Apr 14 13:50:52 hrw|work: sure, but remember that we can not go above F20.... and since we are assigning 15 keys, that's going to be a problem =) Apr 14 13:52:29 CoreDump|home: or we need to patch keylaunch to get F2x Apr 14 13:53:03 I'm not even so sure that it is keylaunchs fault. Might just as well be a kbd driver limitation Apr 14 13:53:32 or X or whatever Apr 14 13:53:35 * CoreDump|home dunnos Apr 14 13:55:00 I doubt keylaunch would care. You'd have to check with xev to see what X is doing. Apr 14 13:56:25 keylaunch is evil anyway (when connected with matchbox) Apr 14 13:56:52 keylaunch says "unknown symbol F21" or something like that. But F21 is of course mapped Apr 14 13:58:59 uh huh | make[5]: *** No rule to make target `-lpcre', needed by `konqueror'. Stop. Apr 14 13:59:28 polyonymous: get angry against autotools Apr 14 13:59:32 polyonymous: you need the lpcre dev files on your host Apr 14 13:59:40 polyonymous: and install libpcre-dev tools on your bost, rebuild konqueror Apr 14 13:59:42 zecke|food, I don't like getting angry Apr 14 14:00:00 I sure do have libpcre on the host, but what does it have to do with oe? Apr 14 14:00:07 -dev Apr 14 14:00:16 zecke|food, what -dev on gentoo host? :) Apr 14 14:00:28 And I sure do have all headers, etc. Apr 14 14:00:47 time to get angry ;) Apr 14 14:03:13 CoreDump|home: that sounds like F21 is missing from the x keysym table. does xmodmap know about it? Apr 14 14:03:50 btw, target '-lpcre' is not exactly the same like can't find libpcre... Apr 14 14:04:42 pb_: probably not then. yay for X11 keyboard handling Apr 14 14:04:46 03hrw 07org.oe.oz354x * r9ef23afe... 10/packages/enchant/enchant_1.2.5.bb: enchant: added --disable-binreloc to get it building Apr 14 14:04:50 03hrw 07org.oe.oz354x * r86fe7fbd... 10/packages/opie-stumbler/ (4 files in 2 dirs): added opie-stumbler from .dev Apr 14 14:04:51 polyonymous: right Apr 14 14:04:55 03hrw 07org.oe.oz354x * r154f24f8... 10/packages/opie-stumbler/opie-stumbler_cvs.bb: opie-stumbler: fixed cvs ver Apr 14 14:05:04 polyonymous: some weird autoconf behaviour Apr 14 14:05:31 autoconf works fine when handled properly, in fact. Well, I'm running off, I'll look into the problem when I'm back. Apr 14 14:08:20 hrw|work: gtkmathview is an optional dependency for abi-plugins, so if it doesn't built, ignore it Apr 14 14:08:53 polyonymous: a nuclear chain reaction works fine as well - if handled properly Apr 14 14:09:00 you'll lose mathml support, but we didn't have it before either Apr 14 14:09:45 zecke|food: r423 is 33% faster as r412 on cpu bound tasks Apr 14 14:09:52 ~lart knetwalk for being so small Apr 14 14:09:52 * ibot beats knetwalk to within 2.54cm of his life for being so small Apr 14 14:10:05 zecke|food: I tested it with tmp/cache/ in a ramdisk Apr 14 14:11:21 koen: hey Apr 14 14:11:36 and iirc mithro wanted you to ping him Apr 14 14:11:48 * zecke|food goes upstairs to install SuSE in VMware server, hack on bitbake or Qephyr Apr 14 14:12:00 koen: we wasn't online when I read your botmail Apr 14 14:13:17 mickeyl: I'm in need for python vodoo Apr 14 14:15:21 zecke: for what in particular? Apr 14 14:15:50 mickeyl: world domination, I want to 'shadow' the builtin file implementation Apr 14 14:16:01 mickeyl: but need to call the real file from within my impl Apr 14 14:16:21 mickeyl: I think I need to play with __builtins__ here Apr 14 14:19:06 i don't think that's possible Apr 14 14:19:23 why can't you derive from File and use MyFile everywhere? Apr 14 14:19:25 koen: I'm sure you have read the blog about Nautilius mem usage? Apr 14 14:19:33 zecke|baking: yes Apr 14 14:19:36 mickeyl: I want to 'monitor' each use of file Apr 14 14:19:50 koen: they should use a more modern class library Qt, .NET with better string classes Apr 14 14:19:58 :) Apr 14 14:21:01 koen: other performance blogs are interesting, but this was 'yesterday'... Apr 14 14:21:36 zecke|baking: hmm, go ahead and try changing like that Apr 14 14:21:42 __builtins__.__dict__["open"] = 40 Apr 14 14:22:14 and file Apr 14 14:22:18 and see what happens Apr 14 14:22:21 hehe Apr 14 14:22:28 mickeyl: have you looked at CMake? Apr 14 14:22:34 no, what's that? Apr 14 14:22:42 mickeyl: the build system used by KDE4 Apr 14 14:22:53 ah. no, didn't see that yet Apr 14 14:22:54 mickeyl: after bksys did not suceed (take off) Apr 14 14:23:09 http://aseigo.blogspot.com/2006/04/cmake-sadness.html Apr 14 14:23:20 " i understand the need for something better than autotools, but in my book the baseline for "better" is "actually builds reliably". meh." Apr 14 14:23:24 mickeyl: it got a horrible syntax, is inferior to bksys (which used python) Apr 14 14:23:51 mickeyl: but it can generate Makefiles, .pro, Visual Studio projects Apr 14 14:24:01 mickeyl: and it looks like being useful for cross compilation Apr 14 14:27:27 XorA: http://uwog.net/blog/?p=33 Apr 14 14:27:54 XorA: so we actually got it in OE before the official anouncement :) Apr 14 14:30:08 koen: is font rendering on OSX fixed? Apr 14 14:30:27 zecke|baking: no idea, haven't tried yet Apr 14 14:31:20 koen: I wonder if I managed to file that bug report before the announcement :-) Apr 14 14:33:47 lol Apr 14 14:35:04 koen: it is not yet fixed Apr 14 14:38:15 morning Apr 14 14:42:16 mickeyl: qt4x11.bbclass needs a merge Im not qualified for Apr 14 14:49:06 happy Easter/weekend for everyone Apr 14 14:49:15 will be back in Thursday Apr 14 14:49:18 cu Apr 14 14:49:22 hrw|work: cya, have a good time Apr 14 14:59:17 XorA: ok, handling it... Apr 14 15:11:31 darn Apr 14 15:11:35 nslu merges Apr 14 15:11:39 which I don't want to mess with Apr 14 15:12:17 can anyone merge that? feel free to override qt4x11 I'll correct that afterwards Apr 14 15:13:14 heck, konq's configure looks for pcre-config in ${prefix}/bin/ /usr/local/bin/ /opt/local/bin... Apr 14 15:13:47 where ${prefix} is /opt/QtPalmtop (or /usr? it's passed to configure twice) Apr 14 15:18:25 * mickeyl merge done Apr 14 15:18:35 pushing.... Apr 14 15:21:27 mickeyl: cheers, sorry that wasnt a simple one for you Apr 14 15:26:32 heh, nm Apr 14 15:27:59 polyonymous: it is busted, but this doesn't explain the expansion to -lpcre Apr 14 15:28:42 zecke|baking, doesn't. well, why it happens is easy to see by the makefile, but haven't tracked down why it is the way it is in makefile yet. Keep getting distracted. Apr 14 15:35:01 03xora 07org.oe.dev * reee8b1a2... 10/packages/enchant/enchant_1.2.5.bb: enchant_1.2.5.bb : Add a linense and --disable-binreloc Apr 14 15:35:34 koen: know why gnome-vfs-dbus_svn.bb is DEFAULT_PREFERENCE=-1? Apr 14 15:38:44 btw, is it possible to bitbake ....bb -f -c compile from bitbake -i ? Apr 14 15:40:42 mickeyl: is that possible? Apr 14 15:41:21 I see ;-) Eloquence of silence :) Apr 14 15:55:32 XorA: no idea, but gnome-vfs-dbus doesn't compile with eabi Apr 14 15:56:15 koen: wierd on the build problem Apr 14 15:56:15 polyonymous,zecke|baking: yeah, iirc you can toggle FORCE in bitbake -i Apr 14 15:56:27 koen: should I re-enable it and put in a proper PV? Apr 14 15:56:35 XorA: gnome-vfs-dbus doesn't exist in my OE tree Apr 14 15:56:44 mickeyl, yeah, done that. but the build fails then. I think with something related to quilt coming out of nowhere... Apr 14 15:57:12 So, what in theory should be the same as -b -f -c, seems to be not. Apr 14 15:57:33 koen: you must have rm it then Apr 14 15:57:47 yep. there are unfortunately still some more or less hidden assumptions about the order of things executed in the bitbake core. we need to find them in order to make (G)UIs more sane... Apr 14 15:57:52 if only i had more time... Apr 14 15:57:53 anyone know where i can get the patch to hciattach (the one wich support speed:921600)? The patch i allready had was messed up. Apr 14 15:58:17 bye Apr 14 15:58:52 XorA: I did Apr 14 15:59:00 koen: need to get train and probably not be able to look at it again until sunday/monday Apr 14 15:59:12 bye people Apr 14 15:59:19 XorA: have fun Apr 14 16:01:24 how's the openslug status nowadays? Apr 14 16:01:35 should i flash a recent release or better try unslung 6.8 for a "productive" system ? Apr 14 16:01:40 cu xora Apr 14 16:02:13 mickeyl: updating to QtE2.3.12+ would give us IPv6 support :} Apr 14 16:02:25 zecke|baking: and open a can of other worms, right? Apr 14 16:02:37 mickeyl: my openslug system works quite good Apr 14 16:02:45 mickeyl: but I do not use much beside ssh and rsync Apr 14 16:02:58 mickeyl: the latest openslug is really nicely engineered, but unslung might save you some time Apr 14 16:02:59 zecke|baking: i will consider this when we switch from qte to qte-mt Apr 14 16:03:25 koen,zecke|baking: basically i want to use it as a samba/usb device for now Apr 14 16:03:42 and usnlung it is ;) Apr 14 16:03:49 *nod* Apr 14 16:04:01 guess i just flash the latest unslung for now Apr 14 16:04:06 :} Apr 14 16:04:13 * mickeyl not in experimentation-mood Apr 14 16:04:23 koen: how should I know how it is named, it just works Apr 14 16:05:01 :) Apr 14 16:08:16 zecke|baking,mickeyl: what's your take on updating to monotone 0.26? Apr 14 16:08:33 0.27 would get my vote Apr 14 16:08:43 we've waited so long... doesn't hurt to wait a bit more Apr 14 16:08:45 koen: I'm retrying the rosterify (I interrupted many times...) Apr 14 16:09:31 mickeyl: what shoud monotone 0.27 fix? (is that release actually planned?) Apr 14 16:09:59 mickeyl: what do we expect to be broken? Apr 14 16:10:07 it would fix my feeling. would turn it into the good feeling that 0.26 has been used for some time. Apr 14 16:10:30 CP|PMSummit: heh, I read that as PMS summit Apr 14 16:10:46 mickeyl: I think we are the biggest user anyway, we will uncover... Apr 14 16:10:56 mickeyl: we could use gcov and let their test suite run Apr 14 16:11:07 mickeyl: and see how much of the new merge code and roster stuff is covered Apr 14 16:11:09 zecke|baking: or coverrty Apr 14 16:11:12 coverity* Apr 14 16:11:58 njs: did you apply for coverity scans? Apr 14 16:13:27 mickeyl: OT: do you know if coverity found some true issues in python? Apr 14 16:24:21 zecke|baking: nothing serious from a quick glance @ python-dev Apr 14 16:24:24 hmm Apr 14 16:24:34 woot! Apr 14 16:24:43 a new UPnP device has been found Apr 14 16:24:53 first time i've seen this message on Windows Apr 14 16:31:27 mickeyl: but there is a report, I will look at py-dev then Apr 14 16:31:35 Who owns the omap kernels? Apr 14 16:31:52 mickeyl: my cousin started more intensive python hacking Apr 14 16:32:09 mickeyl: I think I will tell him more about threads... Apr 14 16:32:22 mickeyl: he has used pygame for image manipulation ;) Apr 14 16:32:42 mickeyl: loaded an image, had a selected color in a config file, and removed all pixels with this color Apr 14 16:37:19 CP|PMSummit: crofton did some work on the omaps and I think cbrake did as well Apr 14 16:37:34 CP|PMSummit: but afaik noone has touched them in ages Apr 14 16:37:46 they omap one needs updating Apr 14 16:37:59 I need to do this in the next couple of months ... Apr 14 16:38:01 zecke|baking: pretty cool :D Apr 14 16:38:11 but if someone else does this I would be happy Apr 14 16:43:23 Has anyone considered adding DPM support to any of the ARM/OMAP kernels? Apr 14 16:44:25 And if they haven't - would they appreciate help from the DPM guys to do that? Apr 14 16:44:49 That is a better question for the omap-linux list Apr 14 16:47:27 Crofton: ok - another question then Apr 14 16:47:55 Crofton: if somebody did help people play with DPM, would that increase its chances to be included into the ARM kernel? Apr 14 16:48:45 zecke|baking: you can use /dev/shm on ewi to store the bitbake cache if you suspect IO bound bottleneck Apr 14 16:48:57 zecke|baking: but don't forget to wipe it after you are done :) Apr 14 16:49:35 koen: thanks, currently I'm lost in thoughts and idling Apr 14 16:49:51 CP|PMSummit, I don't know the answer to that Apr 14 16:49:54 zecke|baking: can you remind me what version of bitbake is a good one for building thing with? Apr 14 16:50:03 Crofton: fair enough' Apr 14 16:50:21 pb_: .dev or any of the branches? Apr 14 16:50:29 zecke|baking: .oz254fam083 Apr 14 16:50:40 pb_: a released one or r329 Apr 14 16:50:41 er, 354, obviously Apr 14 16:50:46 zecke|baking: thanks Apr 14 16:51:08 pb_: or that should be fine as well Apr 14 16:51:26 does it look like a good workaround? Apr 14 16:51:28 -konqueror_DEPENDENCIES = $(LIB_KJSHTML) $(LIB_KHTML) $(lib_addons) Apr 14 16:51:28 +konqueror_DEPENDENCIES = $(shell echo ${LIB_KJSHTML}|sed s/\s*-l\\w\\+//g) $(LIB_KHTML) $(lib_addons) Apr 14 16:51:36 zecke|baking: I seem to have r390 at the moment. I guess this one is not good. Apr 14 16:51:50 it certainly seems rather verbose Apr 14 16:52:40 * pb_ downgrade to r329 and try again Apr 14 16:52:58 polyonymous: can we get a proper fix as well? Apr 14 16:53:28 zecke|baking, the bug is really in Makefile -- it's not a good idea to put LIBS (which are -l) in DEPENDENCIES. Apr 14 16:53:37 or better -- which MAY be -l Apr 14 16:54:22 polyonymous: ah well, it gets generated ;) Apr 14 16:54:37 what does? Apr 14 16:54:54 polyonymous: Makefile, can't you fix the generation of the Makefile.in? Apr 14 16:55:05 zecke|baking, it's a patch for Makefile.am Apr 14 16:55:06 polyonymous: or is the Makefile.in correct and Makefile gets 'broken' Apr 14 16:55:13 So it gets propagated to .in and Makefile Apr 14 16:55:26 konq-embed/src/Makefile.am even Apr 14 16:56:02 is there a way to find what files get into staging by which package? Apr 14 16:56:12 polyonymous: sadly not Apr 14 16:56:15 I see. Apr 14 16:56:23 polyonymous: it is on the wishlist though Apr 14 16:56:32 Will look up package's log then Apr 14 16:57:03 wanted to see what pcre puts in there/ Apr 14 16:57:21 But I think this fix is the easiest. and it should work too. Apr 14 16:57:31 if we can rely on sed Apr 14 16:57:43 it just strips off -l*. Apr 14 16:57:55 (I know that regexp and sql statements are easier to write than to read:)) Apr 14 16:58:17 polyonymous: true, but please dig some more ;) Apr 14 16:58:25 is the -lpcre coming from pcre-config? Apr 14 16:58:45 No, I think it's a fallback in case pcre-config is inavailable Apr 14 16:59:07 polyonymous: okay, konqe should look into ${STAGING_BINDIR} for pcre-config Apr 14 16:59:23 The problem is that pcre-config supposedly will give this same line... Apr 14 17:00:10 There is no pcre-config in arm staging... only i686 -- isn't it for pcre-native? Apr 14 17:00:35 ah, no, it isn't. pcre-config-native is Apr 14 17:00:36 polyonymous: no STAGING_BINDIR -> i686 Apr 14 17:00:55 hacker@hippo ~/OE/build/tmp/staging $ ./i686-linux/bin/pcre-config --libs Apr 14 17:00:55 -L/home/hacker/OE/build/tmp/staging/i686-linux/lib -lpcre Apr 14 17:01:04 won't help much Apr 14 17:01:23 polyonymous: well, it tries to call it Apr 14 17:01:35 polyonymous: let us call it first and then see what is happening Apr 14 17:01:58 the same thing would happen. Only I'm afaid it will add -L.... to DEPs which will make things worse :) Apr 14 17:04:14 the more I look at it the more I think this solution isn't the worst. The other one is to substitute ${LIB_KJSHTML} withoutBPCRE manually, but it's no better even if it's more readable Apr 14 17:04:58 brb phone Apr 14 17:08:57 yay! Apr 14 17:09:12 hx4700 internal wlan can ping stuff Apr 14 17:09:24 neat Apr 14 17:09:27 koen: wow ;) Apr 14 17:09:58 koen: wow, HP sold a device with wlan, where wlan didn't work until now? Apr 14 17:10:31 zecke|baking: correct Apr 14 17:10:33 in linux that is Apr 14 17:11:13 ah linux Apr 14 17:11:18 that is obsolete anyway Apr 14 17:11:37 well, not even Linus has recognized what Xen is all about Apr 14 17:11:46 It slowly turns Linux into a micro kernel Apr 14 17:12:20 * koen wonders when hurd will be usuable Apr 14 17:12:55 koen: 2018 Apr 14 17:13:17 * koen marks it on his calendar Apr 14 17:13:23 koen: I have heard they will switch the micro kernel again Apr 14 17:13:32 koen: they don't like the code formatting of the current one Apr 14 17:13:33 I hope it will run Duke Nukem Forever... Apr 14 17:15:00 I'm not really impressed with microkernels and threading Apr 14 17:15:09 but maybe that's just osx sucking Apr 14 17:16:18 mickeyl: any idea on the gcc sysroot thing? Apr 14 17:16:47 http://bugs.openembedded.org/show_bug.cgi?id=812 Apr 14 17:20:58 sounds like a proper patch to me Apr 14 17:21:09 i'd like pb_ to comment though before you'd check that in Apr 14 17:21:43 yeah Apr 14 17:23:38 I think that's probably fine. It seems a bit weird to say "--with-sysroot=/", though: can you not just remove the option altogether? Apr 14 17:24:21 I have no idea Apr 14 17:24:24 ya, should be default Apr 14 17:24:26 toolchains are voodoo to me Apr 14 17:25:46 * RP returns Apr 14 17:26:18 zecke|baking: Have you looked at the secondary cache in bitbake yet or not? If not, do you want me to look at it? Apr 14 17:27:03 RP: wb Apr 14 17:27:49 What files do i miss? I am trying to build the modified hciattach, but the configure script complains about this : checking for hci_open_dev in -lbluetooth... no Apr 14 17:27:49 configure: error: Bluetooth library not found Apr 14 17:32:17 RP: please do so, I'm a bit 'disturbed' ATM Apr 14 17:32:58 zecke|baking: ok. I also filed a bug about the issues with the current bitbake head. If you need more info, let me know as I can reproduce the problems Apr 14 17:33:18 koen: Nice news on the wifi. I wonder if this means I can get hx2750 working? :) Apr 14 17:33:54 back Apr 14 17:33:56 RP: I hope so Apr 14 17:34:16 koen: I know it uses a different interface to hx4700 so we'll see... Apr 14 17:35:37 RP: we still need a solution to the ipaq-pxa270 problem Apr 14 17:35:59 RP: I'm tempted just to make a hx4700.conf and a hx2000.conf Apr 14 17:36:16 hello all Apr 14 17:36:17 RP: I'm about to include your bb file in a test case Apr 14 17:36:22 hey mreimer Apr 14 17:36:23 does hx2750 have one of those TI chips too? Apr 14 17:36:26 hey koen Apr 14 17:36:32 koen: good news about hx4700, huh? Apr 14 17:36:34 won't be long Apr 14 17:36:37 mreimer: indeed Apr 14 17:36:56 now if we could just get power consumption in suspend under control, we'd be good to go Apr 14 17:36:57 mreimer: now we just need a bounty to get libaticore working on w3220 Apr 14 17:37:05 koen: That would be fine with me Apr 14 17:37:10 koen: what's that? acceleration? Apr 14 17:37:37 koen: I saw a patch on rpsys.net that enables acceleration of solid fills and something else, for w100fb Apr 14 17:37:38 mreimer: It has a similar wifi chip, yes Apr 14 17:37:44 well, zecke|baking I don't think that it makes sense to try to turn LIBCPRE into library pathnames, so the way to go is to remove it from _DEPENDENCIES and there are two ways -- substitute the variable (what was the name?) without LIBPCRE or sed it out. Apr 14 17:37:55 RP: similar interface? 16-bit slave Apr 14 17:38:08 mreimer: CF Apr 14 17:38:18 mreimer: yes accell for the ati framebuffer Apr 14 17:38:24 cool Apr 14 17:38:43 mreimer: w3220 is similar to w100, but I don't know how similar Apr 14 17:39:00 can someone look at this http://pastebin.com/658946 and let me know what might be wrong Apr 14 17:39:00 mreimer: That w100 accell patch went into mainline. I just hope it works on w3220 Apr 14 17:39:02 I googled once but couldn't find anything about w3220 Apr 14 17:39:13 RP: great. I'll give it a try sometime Apr 14 17:39:20 koen, if you remove the option altogether then you'll get no --with-sysroot support in ld. I'm not sure if it's really needed, but thought you should be aware of it. Apr 14 17:39:27 mreimer: w3220 = w100 + some extras Apr 14 17:39:27 RP: your LED patch is in -mm, on its way to mainline, right? Apr 14 17:39:35 (this is why I changed it to / instead of removing it) Apr 14 17:39:35 mreimer: No, mainline now Apr 14 17:39:48 RTC also made mainline Apr 14 17:40:06 RP: very good, on both counts. any pointers to where I can find out more about w3220 and its extras? Apr 14 17:40:20 also, what is the oe mailing list. How can I subscribe Apr 14 17:40:22 RP: you deserve an award for persisting so long to get that LED stuff in Apr 14 17:41:12 mreimer: Spyro has the most knowledge I know of. There is nothing online. It was myself and Spyro who told Todd how to get the hx4700 working ;-) Apr 14 17:41:29 :-) Apr 14 17:41:59 Spyro did the w3220 bits of w100fb.c Apr 14 17:42:35 ~seen spyro Apr 14 17:42:44 ~seen Spyro Apr 14 17:42:44 spyro was last seen on IRC in channel #handhelds.org, 33d 16h 48m 40s ago, saying: 'cya round!'. Apr 14 17:42:46 spyro was last seen on IRC in channel #handhelds.org, 33d 16h 48m 41s ago, saying: 'cya round!'. Apr 14 17:44:16 Its nice the LED class finally did make mainline I must admit. Hopefully hh.org will be able to convert to it now without too many problems :) Apr 14 17:46:13 mickey|dinner: could you have a look at building python with eabi? atm tcl fails in do_configure Apr 14 17:47:33 RP: I plan to convert to it as soon as I can. It'll be fun to try out your blinkenlights :-) Apr 14 17:49:41 The installed /usr/lib/libstdc++.la with pathnames pointing to host's directories prevents libtool from working. Removing file does the trick Apr 14 17:49:42 How can I subscribe to the OpenEmbedded mailing list Apr 14 17:53:24 https://www.handhelds.org/mailman/listinfo/oe Apr 14 18:07:15 RP: some comments: I would keep the Secondary Cache in one file first Apr 14 18:08:11 zecke|baking: I'll see how it goes, I'm undecided about that :) Apr 14 18:08:13 zecke|baking, may I ask you what do you think? Or should I just file the patch? Apr 14 18:08:25 polyonymous: file the patch :) Apr 14 18:08:31 Okay ;-) Apr 14 18:08:39 * polyonymous is heading to bugzilla Apr 14 18:08:44 zecke|baking: I was wondering if we'd get gains if we had the primary cache as one file through COW? Apr 14 18:09:14 RP: how would that work? Apr 14 18:09:32 RP: the common bits are reparsed anyway on start up Apr 14 18:10:11 zecke|baking: Ok, I'm confused then. Probably best to ignore me ;-) Apr 14 18:10:39 * zecke|baking adds the sucker to ever growing ignore list ;) Apr 14 18:11:46 zecke|baking: I can say things about you without you seeing them now? :) Apr 14 18:12:15 RP: right Apr 14 18:12:38 RP: but it shows up on my braille display Apr 14 18:18:01 lol Apr 14 18:18:09 :) Apr 14 18:18:36 I'm getting a bunch of errors like this when I try to bb http://pastebin.ca/49366 Apr 14 18:18:57 Crofton: rm tmp/cache -R Apr 14 18:19:06 hmmm Apr 14 18:19:10 I thought I had .... Apr 14 18:19:24 Crofton: or you suffer from a incompetent bitbake maintainer Apr 14 18:19:29 :( Apr 14 18:19:35 heh Apr 14 18:20:18 still there Apr 14 18:20:30 do you need a more complete log? Apr 14 18:20:43 Crofton: Which version of bitbake with which OE metadata? Apr 14 18:20:55 current/current Apr 14 18:20:58 I hope Apr 14 18:21:00 bb yes Apr 14 18:21:17 It should work :-/ Apr 14 18:21:18 OE data just pull, then prop'd to my provate branch Apr 14 18:21:41 * Crofton 's monotone skillz are not 3l33t Apr 14 18:22:12 Crofton: BitBake trunk is currently in motion Apr 14 18:22:18 heh Apr 14 18:22:40 Crofton: we work on speed and change major parts (the data structure) more often than our wives Apr 14 18:23:00 Crofton: this leads to major speed improvements but we have some issues as well :} Apr 14 18:23:14 any idea when it will work? Apr 14 18:23:33 r412 from svn works reasonably well Apr 14 18:23:34 Crofton: take a look in the OE or bitbake-dev mail archives Apr 14 18:23:49 Crofton: we have some 'known' ''good'' versions Apr 14 18:23:57 I haven't been paying attention Apr 14 18:24:06 and another guy asked me to make it work again Apr 14 18:24:13 I'll poke around Apr 14 18:24:17 Crofton: ah okay, well he is right ;) Apr 14 18:24:27 Crofton: we suck, we work on it, to suck less Apr 14 18:24:44 well, more get our braanch building and send him a copy :) Apr 14 18:25:15 can I fix this by uing current metadata and older bb? Apr 14 18:25:33 i found a mailing list post from 2002 where someone used jffs2 on a compact flash card using blkmtd Apr 14 18:25:55 TheMasterMind1: which is a hack Apr 14 18:26:12 TheMasterMind1: you can use jffs2 on your harddisk as well using that method Apr 14 18:26:36 twice the wear leveling is better =) Apr 14 18:26:44 heh Apr 14 18:26:55 people have actually stated that as a fact Apr 14 18:27:27 i can't find anything regarding journaling and sd/cf lifespan Apr 14 18:27:35 TheMasterMind1: store disk images on jffs2 and use ext3 as the filesystem Apr 14 18:27:44 TheMasterMind1: this way you have twice as much journaling Apr 14 18:28:11 Everything up to bitbake r419 works as far as I can tell Apr 14 18:28:13 double wear-leveling might not be good. graph the average of two random numbers and you'll see what I mean Apr 14 18:28:44 mreimer: I would if my octave would build ;) Apr 14 18:28:48 :-) Apr 14 18:29:34 kergoth`: ping Apr 14 18:29:37 treke|home: ping Apr 14 18:30:14 Not at home Apr 14 18:30:16 but here Apr 14 18:30:34 ggilbert: hey Apr 14 18:30:43 ggilbert: how are you? Apr 14 18:30:51 ggilbert: and I have tons of questions :} Apr 14 18:30:51 hey Apr 14 18:30:52 Not bad Apr 14 18:31:11 I'm still getting hairballs on r419 Apr 14 18:31:15 hey kergoth` & ggilbert Apr 14 18:31:26 Crofton: try r412 (and wipe the cache) Apr 14 18:31:38 zecke|baking: What can I do for you? Apr 14 18:31:47 that was weird, was just googling for vim7 debian packages and one of the first links is hrw's blog Apr 14 18:32:15 well from what i have found sounds like either ext3 or reiserfs is the way to go Apr 14 18:32:20 Means it's popular? :) Apr 14 18:32:23 kergoth`: could we have a survey.oe.org and point it to Enter IP here Apr 14 18:32:24 koen|tv, still problems Apr 14 18:32:51 Crofton: did you base on .dev or oz... branch? Apr 14 18:32:57 dev Apr 14 18:33:15 kergoth`: 213.239.211.73 Apr 14 18:33:34 ggilbert: what is the state of oesources.org, what are the plans? Apr 14 18:33:38 Shouldn't we just remove these broken /usr/lib/*.la from ipks altogether? Apr 14 18:33:50 zecke|baking: The state is the hard drive is full Apr 14 18:33:51 ggilbert: and the other half of the tons is: What version if bugzilla do we run? Apr 14 18:34:08 The plan is to suck it up and try sorting through some of the cruft one weekend Apr 14 18:34:39 polyonymous: I proposed to stop shipping .la files a while ago, but I didn't get any reaction Apr 14 18:35:02 Be persistent. It doesn't do much good, unless you consider breaking libtool good :) Apr 14 18:35:17 zecke|baking: k, made the change over at easydns Apr 14 18:35:43 polyonymous: .la files should be in the -dev packages. if they're broken, they should be fixed. Apr 14 18:36:08 kergoth`, Well, libtool is normally used for dev, you know. Apr 14 18:36:08 trying to figure out what version this is :p Apr 14 18:36:27 And I think it's the only tool that cares about .la files and it does better without broken ones. Apr 14 18:36:27 "Reiser's Law of Software Engineering: Each order of magnitude increase in users finds more bugs, in a quantity equal to the previous order of magnitude increase in users. Success for software developers is measured by how long the frustration lasts." Apr 14 18:36:32 lol Apr 14 18:36:32 2.18.something Apr 14 18:36:51 poly, i know what libtool is, thanks, i'm the one that patched libtool for crosscompilation in oe :P Apr 14 18:36:52 They should be either sed-ed to usable state (I think that's possible) or dropped... Apr 14 18:36:53 ggilbert: okay, I want to get kbugbuster working with it Apr 14 18:36:56 kergoth`: thank you Apr 14 18:37:01 Want a more recent release? :p Apr 14 18:37:20 kergoth`, but you don't suppose I suspected that you don't? :) Apr 14 18:37:29 ggilbert: no, if it has no known security issues I'm happy to stay with it Apr 14 18:37:44 k. Apr 14 18:38:20 Is kbugbuster not working? Apr 14 18:38:38 poly, all i said was that they shoudl be fixed, not removed. Apr 14 18:38:59 spouting further comments isnt going to change my opinion on that Apr 14 18:39:28 kergoth`: How do we get you working actively on OE again? :) Apr 14 18:39:29 kergoth`, well, I do not object fixing them. Apr 14 18:39:53 I said just that it's better to remove them than to ship broken. Apr 14 18:40:17 i disagree. removing them just makes it less likely that itll be fixed properly in the future. Apr 14 18:40:20 And didn't mean to influence your point of view in favor of removing. Apr 14 18:40:22 itll be forgotten Apr 14 18:40:25 ggilbert: it fails to import most of the bugs Apr 14 18:40:34 zecke|baking: hrmm Apr 14 18:40:40 RP: excellent question. i'm thinking about picking up a 770 :) Apr 14 18:40:44 zecke|baking: Does it mention what versions it works with? Apr 14 18:41:05 kergoth`, right. I'd gladly take care of it, but I need to dig deeper into the process then... Apr 14 18:41:24 * ggilbert remembers the lovely broken/ directory that was like a black hole for packages Apr 14 18:41:39 kergoth`: The 770s are interesting bits of hardware. I still prefer spitz (Zaurus SL-C3000) over it but they both have their uses :) Apr 14 18:42:16 kergoth`: We have basic support for the 770 in OE already Apr 14 18:42:28 He left :p Apr 14 18:43:00 Hmm, yes. My irc client is playing games :-/ Apr 14 18:43:08 RP: did you get anywhere with the kernelsource? Apr 14 18:43:22 RP: you mentioned that people at OE didn't use the initfs anymore Apr 14 18:44:20 koen|tv: Those .bbs are on my laptop with the corrupt harddisk - I will fix it over the weekend at some point and hopefully recover them Apr 14 18:44:30 kergoth`: You aren't in Texas anymore, are you? Apr 14 18:44:43 RP: no hurry, just curious Apr 14 18:44:43 koen|tv: Correct, I don't use the initfs at all Apr 14 18:45:10 kergoth`: The 770s are interesting bits of hardware. I still prefer spitz (Zaurus SL-C3000) over it but they both have their uses :) Apr 14 18:45:14 ggilbert: kbugbuster can work with some bugzilla versions Apr 14 18:45:19 kergoth`: We have basic support for the 770 in OE already Apr 14 18:46:30 zecke|baking: http://bugs.kde.org/show_bug.cgi?id=102873 Apr 14 18:46:37 Not sure if that's the most recent state of things Apr 14 18:46:45 that's cool. i think i just need a new toy. all my zauruses are old and in minnesota, a couple thousand miles away from here Apr 14 18:46:49 koen|tv: I thought the .bb was elsewhere hence the holdup. I'll get it sorted as it will be useful to me if nobody else Apr 14 18:47:13 kergoth`: Get a 770 then :) Apr 14 18:47:27 kergoth`: Where are you at now? Apr 14 18:48:14 ggilbert: hmm Apr 14 18:48:20 kergoth`: FYI, OZ is now using 2.6 kernels on everything except collie, poodle and tosa. tosa support is beta state, poodle might improve now I'm getting a device - it can run GPE under 2.6 now Apr 14 18:48:38 kergoth`: and we see finally why bitbake-ng makes sense ;) Apr 14 18:48:57 zecke|baking: might just be sol Apr 14 18:49:07 we finally see... Apr 14 18:49:07 morning kergoth, ggilbert Apr 14 18:49:35 zecke|baking: I always did. I still think we can get a bit more efficient in bitbake though. I just wish my python was up to it :-/ Apr 14 18:49:54 Anyhow, I have to go. bbl Apr 14 18:50:01 RP: glad to hear it Apr 14 18:50:02 hey pb_ Apr 14 18:50:04 GPE still active? Apr 14 18:50:14 ggilbert: always Apr 14 18:50:18 sweet Apr 14 18:50:35 Last I checked in on opie, it looked like there was a bit of an explosion on the mailing list Apr 14 18:50:41 ggilbert: i'm in arizona now. left texas and went back to mn for 6 months or so workign a crappy embedded job there, then left that and moved out to phoenix for a job here (google for Land Warrior) Apr 14 18:50:47 working with Russ from #elinux Apr 14 18:50:49 kergoth`: ah Apr 14 18:51:43 intresting Apr 14 18:53:21 ooh. Someone backported Konq 3.5? Slik Apr 14 18:53:45 ggilbert: someone == a company with a Qtopia based product ;) Apr 14 18:53:54 ggilbert: where I had an internship last year Apr 14 18:54:25 hehe Apr 14 18:54:32 so someone == you? :p Apr 14 18:56:14 ggilbert: no, actually this company paid another company to do the port ;) Apr 14 18:56:20 ggilbert: and it is a work in progress Apr 14 18:56:39 ggilbert: www.road-gmbh.de *shameless advertising* Apr 14 18:59:11 sweet Apr 14 18:59:59 is advertising on IRC illegal? :) Apr 14 19:00:13 only if you're baking at the same time Apr 14 19:00:26 *cough* Buy AMD! *cough* Apr 14 19:00:33 I just did :p Apr 14 19:01:08 nice new athlon64 X2 4200 Apr 14 19:01:18 Aha. Apr 14 19:09:54 CP|PMSummit: Are Dual Core Athlons with Virtualisation Built-In available? Apr 14 19:12:22 zecke|baking: I don't think so - but I really haven't been keeping up Apr 14 19:12:46 A one core version? Apr 14 19:13:04 it would be really awesome to have lvm2, xen and 'Windoze' working together Apr 14 19:14:34 well. reiserfs definitely doesn't like its block device disappearing Apr 14 19:17:13 zecke|baking: Its out for the Opteron, but I don't know if the Athlons have it yet Apr 14 19:17:33 The only press release I saw said "Available in 2006" Apr 14 19:18:07 zecke|baking, going back in bb versions isn't helping me Apr 14 19:18:19 Crofton: hmm Apr 14 19:18:31 Crofton: going way back to r329 is doing something? Apr 14 19:18:37 Crofton: if not stay tuned Apr 14 19:19:14 Let me try that Apr 14 19:19:32 is there an equivalent of range in shell scripting? i.e. for i in range(0,100) Apr 14 19:19:36 Crofton: don't forget to cleat tmp/cache every change Apr 14 19:19:39 TheMasterMind1: seq Apr 14 19:19:45 that's it! thanks Apr 14 19:21:10 grr, I wonder why stupid localegen isn't working Apr 14 19:21:18 NOTE: generating locale es_NI (UTF-8) Apr 14 19:21:18 NOTE: Task failed: localedef returned an error. Apr 14 19:21:24 shame it doesn't say what the error actually was Apr 14 19:22:07 329 gives unparsed line 'conf/sanity.conf' Apr 14 19:22:17 Crofton: delete that line Apr 14 19:22:19 Crofton: due the 'require' keyword Apr 14 19:22:33 Crofton: hmm, then you are truely using .dev :} Apr 14 19:22:48 heh Apr 14 19:24:46 zecke|baking: do you have any idea what happens to stderr for commands run by os.system() inside do_package()? Apr 14 19:29:05 pb_: same as standard output Apr 14 19:29:31 zecke|baking: okay. can you tell me what happens to standard output? Apr 14 19:29:38 has anyone seen an error building gnome-vfs-dbus where DBUS_TYPE_DICT is undeclared? I'm trying to build abiword-plugins Apr 14 19:30:15 mreimer: what version of dbus do you have? Apr 14 19:31:07 pb_: I did a 'rebuild dbus' and it built 0.61-r0 Apr 14 19:31:23 pb_: In build.py is some magic to dup2 the Standard Output and Standard Error to the output of popen('tee') Apr 14 19:32:05 pb_: let me guess, popen only reads from one fd Apr 14 19:32:50 sorry Apr 14 19:32:53 I'm confused Apr 14 19:32:57 zecke|baking: I think that stuff only happens for shell functions, which do_package isn't. Apr 14 19:33:20 pb_: ah right, I'm even more confused than I thought I was Apr 14 19:34:21 pb_: hmm, we do not have any special handler for IO in bitbake Apr 14 19:34:27 pb_: stderr should just be printed :} Apr 14 19:34:36 drat Apr 14 19:34:41 I wonder why I don't see any error messages Apr 14 19:35:11 maybe localedef really is just exiting without printing anything Apr 14 19:35:13 that would be annoying Apr 14 19:35:51 I'll add some debugging to try to isolate the problem Apr 14 19:35:55 thanks Apr 14 19:36:11 meantime I will return to hating apmd Apr 14 19:38:54 pb_: meantime I wonder why I don't have a portinstall... Apr 14 19:39:52 ah Apr 14 19:39:54 -bash-2.05b$ PATH="/nexus/home/pb/oe/build-h3900/tmp/staging/i686-linux/bin/arm-linux:/nexus/home/pb/oe/build-h3900/tmp/staging/i686-linux/bin:/nexus/home/pb/oe/build-h3900/tmp/cross/bin:/usr/kerberos/bin:/usr/local/lib/ccache:/usr/local/arm/3.3.1/bin:/usr/local/bin:/bin:/usr/bin:/usr/X11R6/bin:/home/pb/oe/bitbake/bin" I18NPATH="/usr/share/i18n" qemu-arm -L /nexus/home/pb/oe/build-h3900/tmp/work/arm-linux/glibc-2.3.5+cvs20050627-r3/locale- Apr 14 19:39:54 tree /nexus/home/pb/oe/build-h3900/tmp/work/arm-linux/glibc-2.3.5+cvs20050627-r3/locale-tree/bin/localedef --force --old-style --no-archive --prefix=/nexus/home/pb/oe/build-h3900/tmp/work/arm-linux/glibc-2.3.5+cvs20050627-r3/locale-tree --inputfile=/usr/share/i18n/locales/es_NI --charmap=UTF-8 es_NI Apr 14 19:39:55 Segmentation fault Apr 14 19:39:57 -bash-2.05b$ Apr 14 19:40:02 that can't be good Apr 14 19:40:14 I guess my qemu is broken Apr 14 19:42:33 * pb_ tries on a different host Apr 14 19:43:54 drat, reparsing, I guess this will take a few minutes Apr 14 19:53:14 cbrake: Apr 14 19:59:51 Hmm... before upgrading to pcmciautils I've been setting up my BT card in /etc/pcmcia/bluetooth... I seem to be unable to figure out at the first sight how to go about it now... Apr 14 20:07:51 ah, here we go, localedef works fine on my other machine Apr 14 20:07:57 I guess qemu just hates my red hat host Apr 14 20:11:16 Red Hat... Apr 14 20:11:26 pb_: is that using gcc 2.96? Apr 14 20:11:59 Is handhelds.org mentoring for the Summer of Code again this year? Apr 14 20:12:47 CP|PMSummit: we should :) Apr 14 20:13:30 its open for mentors to apply Apr 14 20:13:39 http://code.google.com/soc/ Apr 14 20:14:01 move that to #hh.org Apr 14 20:14:42 hmm Apr 14 20:15:33 zecke|baking: no, 3.2.2 Apr 14 20:15:46 er, 3.3.2 Apr 14 20:16:46 Is there a way to override a certain variable from "packageX.bb" in local.conf, when the variable is defined in every .bb? I would need to override SRC_URI from libgpg-error_1.0.bb because ftp.gnupg.org didn't seem to work earlier and there are mirrors available Apr 14 20:18:50 Crofton: any luck yet? Apr 14 20:33:09 polyonymous: hehe Apr 14 20:33:19 polyonymous: it is autotools forcing you to fixup crappy usage... Apr 14 20:33:28 hmm Apr 14 20:33:32 polyonymous: Nuclear Power Plants are bad Apr 14 20:33:40 polyonymous: not because Atoms are evil Apr 14 20:33:50 zecke|baking, nah Apr 14 20:33:51 seg fault in ipk-build Apr 14 20:33:52 polyonymous: Atoms are really appreciated and important Apr 14 20:34:03 and the analogy is correct. Apr 14 20:34:06 polyonymous: but I do not like humans that want to control a chain reaction Apr 14 20:34:14 you can't blame atoms for what people do. Apr 14 20:34:34 polyonymous: same applies to autotools, The two people who fully understand autootols should use them Apr 14 20:34:35 zecke|baking, well, they may have to. But if they do, they should do it properly Apr 14 20:34:41 Right. Apr 14 20:34:50 polyonymous: they deserve it, the others should stay away - far away Apr 14 20:34:58 Many autotools users just use it because they think they have to. Apr 14 20:35:08 zecke|baking, I am going to try again on Monday ..... Apr 14 20:35:22 Crofton: we might have it fixed then ;) Apr 14 20:35:28 I know :) Apr 14 20:35:42 That is what I am hoping Apr 14 20:36:01 hehe Apr 14 20:36:04 zecke|baking, I don't think there's something we actually argue about. We agree on most importants matters here. Apr 14 20:36:11 right Apr 14 20:36:20 I just _had_ to make that remark in bug :) Apr 14 20:36:31 polyonymous: KDE is a really bad example for autotools and libtool altogether Apr 14 20:36:47 there aren't many good examples. of anything :) Apr 14 20:38:39 polyonymous: I never felt the preassure to look into that pcre bug Apr 14 20:38:48 polyonymous: as a last resort your patch should be applied Apr 14 20:39:39 zecke|baking, well, it just prevented me from doing build opie-image, that's it :) Apr 14 20:39:54 gerwinin: hey, got enough feedback? Apr 14 20:39:59 yes Apr 14 20:40:05 if it works for other as well I don't see why not apply it. Apr 14 20:40:27 For now I have , zecke: would a sempron 64 help you in compiling ? Apr 14 20:41:30 gerwinin: not much, I have access to a couple of P4 racks for now Apr 14 20:42:11 Zecke :) Apr 14 20:42:20 hmm what is the monotone view program? Apr 14 20:42:34 Zecke: I am going to hang a couple of sempron machines in a rack Apr 14 20:42:34 Crofton: viewmtn? it is a 'webapp' Apr 14 20:42:50 hmmm Apr 14 20:42:54 gerwinin: do you know how to do surveys? Apr 14 20:42:57 no command line version? Apr 14 20:43:01 Crofton: http://ewi546.ewi.utwente.nl/tmp/viewmtn/ Apr 14 20:43:06 Zecke: yes Apr 14 20:43:21 Crofton: sure, don't know how though Apr 14 20:43:53 gerwinin: we have to use survey.openembedded.org ;) Apr 14 20:45:15 zecke :) Apr 14 20:45:32 zecke : I will have a look when the servers come in Apr 14 20:45:50 gerwinin: it runs phpsurveyor.org Apr 14 20:45:59 gerwinin: I just need help with the questions soon Apr 14 20:47:03 gerwinin: http://survey.openembedded.org/index.php?sid=1 Apr 14 20:47:15 gerwinin: do you think generating a token will be a issue for the people? Apr 14 20:47:28 gerwinin: the survey is anonymous, generating a token not :} Apr 14 20:47:45 heheheh Apr 14 20:48:22 Zecke I can help you Apr 14 20:50:08 Zecke: Shall I make a list of questions ? Apr 14 20:50:48 gerwinin: that would be awesome Apr 14 20:51:12 gerwinin: I think the focus should be to find out who uses OE, and how OE is used Apr 14 20:51:27 gerwinin: this should be the basis for a newly designed website etc... Apr 14 20:51:27 * chouimat is away: OUT Apr 14 20:51:34 okay Apr 14 20:51:46 gerwinin: or what would you want to ask? Apr 14 20:52:06 gerwinin: and do you think that generating a token will be an issue? Apr 14 20:52:10 Zecke: the same and I would like to know if people want to donate and what they want to donate Apr 14 20:52:54 Zecke: I don't think so but people should not be doing too much work to take the survey Apr 14 20:52:55 gerwinin: yeah like the linuxdevices.com survey Apr 14 20:53:05 gerwinin: "Do you plan to use OE in the next year(s)" Apr 14 20:53:14 gerwinin: Would you be willing to pay XXX to zecke Apr 14 20:53:36 but more important is to find out... who uses OE Apr 14 20:53:46 Zecke: lol , but I want to make a list of commercial oe people as well Apr 14 20:53:57 Zecke : oe consultants Apr 14 20:54:19 uh oh, consultants mentioned Apr 14 20:54:42 Cockroach: are you a consultant for oe ? Apr 14 20:55:38 Do I get cash'n'hoes & BMW 700 series? Apr 14 20:56:21 But I was very suprised that OpenEmbedded is on some "check this out" list at work. Apr 14 20:56:53 Zecke: I start making the questions now Apr 14 20:56:56 hehe Apr 14 20:56:57 awesome Apr 14 20:57:00 I try to fix bitbake Apr 14 20:57:02 I haven't mentioned it myself even when my thesis work uses it Apr 14 20:57:26 time for beer Apr 14 20:58:42 come on OE & python gurus, help me out here. :) ftp.gnupg.org seems to be down and I would hate to manually change each required package to use a working mirror Apr 14 21:00:02 Cockroach-: package/gnupg/gnupg.inc Apr 14 21:03:52 zecke|baking: well libgpg-error libgcrypt packages don't seem to use that .inc, so would it help ? Apr 14 21:04:39 zecke|baking: "would you object to making the OE devs filthy rich?" Apr 14 21:05:12 hey gerwinin Apr 14 21:05:47 is there a monotone command to return the branch of the current worktree? Apr 14 21:11:43 CoreDump|home: cat MT/options? Apr 14 21:12:14 RP: doh, thanks =) Apr 14 21:13:12 or monotone status | grep old_rev and fill in the rev with 'monotone ls certs ' Apr 14 21:14:05 hey Koen Apr 14 21:16:21 LOL: http://times.hankooki.com/lpage/tech/200604/kt2006041317503911780.htm Apr 14 21:16:26 koen: if I'm more filthy and fluffy I'm fine Apr 14 21:17:25 koen: lol Windows XP and Touchscreen outsmarted people? Apr 14 21:17:54 it seems that way Apr 14 21:19:13 mickey|dinner: yeah, I don't think TARGET_SYS is going to cut it Apr 14 21:19:30 Cockroach-: too bad, you could add your PRE MIRROR as well Apr 14 21:19:46 mickey|dinner: linux-gnu-arm != arm-linux Apr 14 21:19:49 Cockroach-: this will use a regexp and if you match with the original site, your PRE MIRROR gets used Apr 14 21:20:12 anyone know how I can get the sys in the format linux-gnu-arm from OE? Apr 14 21:20:46 TARGET_OS + TARGET_ARCH ? Apr 14 21:21:49 koen: is that it? /me tries Apr 14 21:22:06 TARGET_OS + TARGET_ARCH = TARGET_LOVE Apr 14 21:23:01 * JustinP kicks himself Apr 14 21:23:13 need to stop doing this in one cmd: rm -rf tmp && cp -R gcc.tmp tmp && bitbake -i Apr 14 21:24:08 koen: I can confirm, the author of methodpool was on crack Apr 14 21:24:23 no he wasn't :} Apr 14 21:24:24 bleh Apr 14 21:24:32 he still is Apr 14 21:24:41 * koen removes another arm assumption from angstrom.conf Apr 14 21:25:46 I liked the link. Shows how badly the industry sucks. Apr 14 21:26:19 03koen 07org.oe.dev * rccbffbb6... 10/conf/distro/angstrom.conf: angstrom.conf: only use softfloat on arm Apr 14 21:29:48 http://linuxdevices.com/articles/AT8247255296.html Apr 14 21:30:33 directFB Apr 14 21:30:38 Interesting. Without the cache, bitbake takes 3 minutes to parse. When building the cache it takes 5 minutes... Apr 14 21:31:23 and http://code.google.com/soc/ Apr 14 21:32:16 1.5 minutes to parse with a valid cache Apr 14 21:33:36 gerwinin: what do you think about OE being a mentoring organization for google SoC 2006? Apr 14 21:34:34 I am having a look :) Apr 14 21:35:19 I think that would be cool Apr 14 21:35:35 maybe I have some time this summer to either mentor a student or be a student :) Apr 14 21:35:56 Well I think it would be awesome Apr 14 21:36:02 gerwinin: could you mail your thought on google soc to oe@ ? Apr 14 21:36:07 'night all Apr 14 21:36:09 I will Apr 14 21:36:12 night koen Apr 14 21:36:27 Zecke do you still have some good technical questions for the survey ? Apr 14 21:36:37 gerwinin: sadly not Apr 14 21:37:02 okay need to have three questions more than I am finished Apr 14 21:37:21 The questionaire is set up in a way that we can do some correlation as well :) Apr 14 21:39:12 good :) Apr 14 21:45:52 zecke the survey is on its way Apr 14 21:47:44 gerwinin: awesome ;) Apr 14 21:47:56 gerwinin: just the window manager question is 'wrong' Apr 14 21:48:34 okay you can change that one hehehe Apr 14 21:49:16 I have on sunday some time to work on websites and t-shirt etc ... Apr 14 21:49:25 gerwinin: awesome Apr 14 21:49:49 I am thinking of some slogans Apr 14 21:50:03 gerwinin: once again, do you think generating a token will be an issue for some people? Apr 14 21:50:37 I don't think so Apr 14 21:51:37 Would this be a good one: if bitbake results in a succesfull built ......... Apr 14 21:52:18 gerwinin: what is built, what is succesfull ;) Apr 14 21:52:27 Zecke : hehehe Apr 14 21:53:02 gerwinin, Better "Wanna be a .bb king?" Apr 14 21:53:10 ha ha Apr 14 21:53:28 polyonymous that is a really good one ! Apr 14 21:53:30 "Stop pulling your hairs, use OE" Apr 14 21:53:40 you're free to use it :) Apr 14 21:54:08 /home/jabberhosting/code/framework-ruby/lib/db/row.rb:150:in `method_missing': undefined method `class' for # (NoMethodError) Apr 14 21:54:12 to_s Apr 14 21:54:43 I noted them down Apr 14 21:56:01 OE what do you want to create today ? Apr 14 21:56:12 I can come up with a great line for OE in russian too... But I don't think it's translatable... Apr 14 21:56:38 Well poly I think we can do them in russian as well ! Apr 14 21:57:24 Poly we can make t-shirts in russian as well Apr 14 21:57:25 Then it's just a quote from Pushkin (the most well known russian poet, just in case you don't know). Looking for translation on the web now :) Apr 14 21:57:41 Poly I know pushkin :) Apr 14 21:57:54 who doesn't =D Apr 14 21:58:04 gerwinin, well, just in case. You know Eugene Onegin is abbreviated as 'OE' :) Apr 14 21:58:13 hahahaha Apr 14 21:58:25 oe the hermitage of linux :) Apr 14 21:58:45 hi Apr 14 21:58:56 hey bernardo Apr 14 21:59:06 Do you speak spanish ? Apr 14 21:59:12 gerwinin, The thing is that Tatiana was drawing this OE thing on the window at some point :) Apr 14 21:59:28 ok Apr 14 21:59:54 gerwinin: no, Portuguese Apr 14 22:00:04 Bernardo can you help with translation ? Apr 14 22:00:39 what translation? opie, oe, gpe? Apr 14 22:00:49 Bernardo no for the website of oe ? Apr 14 22:00:58 * Bernardo is typing on a akita Apr 14 22:01:13 ok, but not today Apr 14 22:01:36 Bernardo no but I am looking for persons that can help us with translations on the long term :) Apr 14 22:01:44 :) Apr 14 22:02:10 translations are in the few things I can do while at work Apr 14 22:02:32 03coredump 07org.oe.oz354x * ra7724b50... 10/packages/xserver-common/ (files/softkeys-slcxxxx-xmodmap.patch xserver-common_1.8.bb): xserver-common: Add keymapping F20-F30 for softkey handling with keylaunch Apr 14 22:02:36 03coredump 07org.oe.oz354x * rb946784b... 10/packages/xserver-common/ (files/load-xmodmap-k26.patch xserver-common_1.8.bb): xserver-common: Enable xmodmaps for kernel 2.6 on c7x0 and cxxxx Apr 14 22:02:40 03coredump 07org.oe.oz354x * rf649e7da... 10/packages/keylaunch/ (3 files in 3 dirs): keylaunch: Move the softkeys to F20-F30 to avoid Opie problems Apr 14 22:02:45 03coredump 07org.oe.oz354x * r53d50dc2... 10/packages/keymaps/ (3 files in 3 dirs): keymaps: Move softkeys to F20-F30 to avoid Opie problems Apr 14 22:03:50 well. I've now checked opie irc works, see you guys later Apr 14 22:03:53 i have a CFLAGS += line that only applies to non-eabi arm toolchains, how do i setup a conditional for that Apr 14 22:03:55 gerwinin, http://www.pushkins-poems.com/Yev313.htm XXXVII - the last two lines :) Apr 14 22:04:42 gerwinin, hrw has my email, and I show up almost every day Apr 14 22:04:50 :) Apr 14 22:04:57 Bernardo cool Apr 14 22:04:57 I'll help what I can Apr 14 22:05:03 The same here Apr 14 22:05:04 bye Apr 14 22:05:08 bye Apr 14 22:05:22 cya Apr 14 22:05:30 poly cool :) Apr 14 22:06:02 way to literary, though, I'm afraid.; Apr 14 22:06:08 hhehee damn some weired people on the radio :) Apr 14 22:06:28 would CFLAGS_append_arm work ? Apr 14 22:06:30 anyone looked into setting up binfmt_misc with qemu to do the autoconf tests in oe? Apr 14 22:06:31 (yet) Apr 14 22:06:43 Polyonymous I will note all suggestions down Apr 14 22:06:55 And than make some suggestions to the mailing list Apr 14 22:07:09 Sure, keep it in case russian ones come up.; Apr 14 22:07:46 polyonymous my idea is to appoint some people in every country who can do the distri and pr in their local language Apr 14 22:08:03 and that can visit events and stuff Apr 14 22:08:16 Yup, sounds like a good idea. Apr 14 22:09:29 polyonymous we need to put oe a bit on the map in the world Apr 14 22:11:15 since kergoth` doesn't lik me Apr 14 22:11:21 someone forward that url to him Apr 14 22:11:54 There was an old man in O.E, // Who used to say "non" and say "oui". // Let's pardon his french // For the code he could wrench // Was so fine that it went in O.E. Apr 14 22:22:20 Anyone know the wordsize for the Intel XScale PXA270 processor (I think it is an arm based CPU) Apr 14 22:22:24 or know where I can find this out Apr 14 22:22:53 cyrus__: intel's sheets? arm site file Apr 14 22:23:10 cyrus__: but I think it is governed by the ARM ABI Apr 14 22:27:53 gerwinin: I speak Spanish and I'd love to collaborate in translations Apr 14 22:28:07 PROVIDES is only used by O.E or is it put into ipks too? Apr 14 22:30:38 TheMasterMind1: do you think swt could be run with jamvm in an arm machine? Apr 14 22:31:23 katossi: yea it should Apr 14 22:32:08 RPROVIDES is wahts used for ipks, tmm. PROVIDES is the build end of things Apr 14 22:32:22 TheMasterMind1: but then some parts should be compiled, right? Apr 14 22:36:55 ah, ok that makes sense, the R prefix is consistent Apr 14 22:41:59 i need to read up on hotplug, its confusing me. gotta figure out how to get rid of it completely and use udev Apr 14 22:44:25 it depends on what do you do on hotplugging... Guess the best reading is in /etc/udev/rules.d/ :) Apr 14 22:54:19 * JustinP hacks Apr 14 22:54:30 TheMasterMind1: Recent udevs will disable hotplug Apr 14 22:55:16 RP: do you know how I can get linux-gnu-arm from OE? Apr 14 22:55:28 RP: I'm using ${TARGET_OS}-gnu-${TARGET_ARCH} but that seems wrong... Apr 14 22:56:53 JustinP: I'm not sure you can get that easily. To search our existing variables, you could grep the output of bitbake -e for gnu... Apr 14 22:59:50 kdepimipi build breaks with gcc4... Apr 14 23:00:17 zecke|baking: ping Apr 14 23:00:33 pong Apr 14 23:01:21 zecke|baking: Does pkgdata contain the data for each parsed file? Apr 14 23:01:54 RP: it used to Apr 14 23:02:04 zecke|baking: but doesn't now? Apr 14 23:02:06 RP: it contains enough information to load the right bb file for a provider Apr 14 23:02:39 This was where our memory problem was? Apr 14 23:02:45 RP: ah, yeah it is a magic dict Apr 14 23:02:51 RP: right, now it is a proxy Apr 14 23:03:02 "magic dict"? Apr 14 23:03:14 RP: see bb/data.py Apr 14 23:03:26 zecke|baking: It does the COW magic? Apr 14 23:03:27 RP: DataDictFull and DataDictCache Apr 14 23:03:34 "magic dict" sounds like a bunch of magic words. Apr 14 23:03:52 RP: no this magic is happening in the data_smart.py (which should be named data_magic.py) Apr 14 23:05:00 zecke|baking: Can we pickle pkgdata? Apr 14 23:05:16 RP: it won't bring us much Apr 14 23:05:32 RP: it is just Provider -> Filename Apr 14 23:05:49 RP: we would still have to lot the 'other' data Apr 14 23:06:10 RP: We can kill DataDictFull now... Apr 14 23:06:52 I think I'm totally failing to realise what you've changed the data store to do :- Apr 14 23:06:55 :-/ Apr 14 23:07:12 RP: BBFILES -> list of filenames Apr 14 23:07:26 RP: for each file name: see if it is cached (by comparing mtime) Apr 14 23:07:43 RP: if cache is current -> load cached version Apr 14 23:07:50 RP: else parse it Apr 14 23:08:06 RP: in both cases extract the information we need into the BBParsingStatus Apr 14 23:08:26 RP: throw the data away (DataDictCache) after having written the cache Apr 14 23:09:14 zecke|baking: So when collect_bbfiles says self.pkgdata[f] = bb_data, does it get a copy of all the data, or just some reference which lets the backend look up data if that reference is passed to it? Apr 14 23:09:59 DataDictCache::__setitem__ gets called Apr 14 23:10:57 and we just save 'f' in our case Apr 14 23:11:08 Hi, I built an GPE image and can't seem to get it read touchpanel input from /dev/input/event1. Apr 14 23:11:14 anybody has an idea? Apr 14 23:11:31 jsun: try ts utilities first Apr 14 23:11:41 root@mainstone:/bin# cat /etc/profile.d/tslib.sh Apr 14 23:11:41 export TSLIB_TSDEVICE=/dev/input/event1 Apr 14 23:11:58 zecke|baking, the gpe-image does not seem to have any ts utilities. Apr 14 23:12:07 you mean something like ts_test? Apr 14 23:12:14 jsun: well... Apr 14 23:12:18 jsun: Install the tslib-tests and tslib-calibrate packages Apr 14 23:12:26 jsun: what are you building your image for? Zaurus, iPAQ? Apr 14 23:12:33 mainstone. Apr 14 23:12:38 Intel's dev board. Apr 14 23:12:50 jsun: you can also try TSLIB_TSDEVICE=/dev/input/event1 Xfbdev :0 Apr 14 23:13:00 jsun: well, how is your ts.conf looking? Apr 14 23:13:17 jsun: what kernel are you using? how does it report touchscreen events? Apr 14 23:13:20 default; Apr 14 23:13:40 zecke|baking, kenerl reports through /dev/input/event1. Apr 14 23:13:55 jsun: I don't have a mainstaone, so I don't know what defaults means Apr 14 23:14:00 I saw some craps show up when I touch scrreen and do "cat /dev/input/event1". Apr 14 23:14:09 jsun: then try the ts utilities to work with that device Apr 14 23:14:25 zecke|baking, sounds reasonable .. Apr 14 23:14:34 mreimer, I will try yours first. :) Apr 14 23:15:04 RP: we could save the BBParasingStatus Apr 14 23:15:24 RP: but the issue is with changing files Apr 14 23:15:31 RP: we could force something like this Apr 14 23:15:40 RP: bitbake --only-cache foo Apr 14 23:15:47 RP: bitbake --update-cache Apr 14 23:15:51 zecke|baking: Changing files is easy to detect Apr 14 23:15:58 zecke|baking: See __depends Apr 14 23:16:12 RP: only by iterating over each file Apr 14 23:16:31 mreimer, your suggestion did not work. The screen does not even show the calibration screen. Apr 14 23:16:33 RP: and we would need a way to remove the old data, and set the new one Apr 14 23:16:40 Its just a stat call = fast Apr 14 23:16:52 RP: this makes caching BBParsingStatus quite hard to use Apr 14 23:17:12 jsun: it won't show a calibration screen, just a cursor and the stippled background Apr 14 23:17:14 zecke|baking: Yes, I'm not going to directly cache BBParsingStatus Apr 14 23:17:43 RP: we could cache BBParsingStatus to have some ultra fast build Apr 14 23:17:49 zecke|baking: I just worry the caching in the data backend is going to conflict in a nasty way with what I'm trying to do... Apr 14 23:18:48 zecke|baking, I think the tpanel driver is likely working fine as the kernel worked with qpe before. Apr 14 23:19:04 zecke|baking, I suspect it is more of interoperability issue between kernel and GPE. Apr 14 23:19:24 jsun: is that Monta Vistas qpe? Apr 14 23:19:31 no. Apr 14 23:19:39 RP: don't worry, fail and fix it :) Apr 14 23:19:39 it is qpe 2.1.1 that I made by myself. Apr 14 23:19:56 and it also worked with an older one (1.1.2?) Apr 14 23:20:11 jsun: why didn't you use Qtopia4.1? Apr 14 23:20:22 anyway Apr 14 23:20:35 Ok, I'm off back to Denver and then on to vacation Apr 14 23:20:36 is that freely available? Apr 14 23:20:38 see you all on Tuesday Apr 14 23:20:38 jsun: that is a stock 2.6er kernel? Apr 14 23:20:45 jsun: not that I know Apr 14 23:20:49 2.6.9 with intel hacks. Apr 14 23:21:30 jsun: Find out that eventX provides that what you expect it to provide Apr 14 23:21:40 jsun: and see it matches with the info you have in ts.conf Apr 14 23:22:41 hm, I only have these in ts.conf: Apr 14 23:22:44 RP: thanks, no cigar, though. I'll just leave it -gnu- for now....no time to try to "fix" this right right now... Apr 14 23:22:48 module pthres pmin=1 Apr 14 23:22:48 module variance delta=30 Apr 14 23:22:48 module dejitter delta=100 Apr 14 23:22:48 module linear Apr 14 23:23:08 jsun: no input module? Apr 14 23:23:20 zecke|baking, there are a bounch of input modules all commented out. Apr 14 23:23:34 I suppose I should pick one to uncomment? Apr 14 23:23:49 jsun: it should match your input ;) Apr 14 23:24:02 module_raw input sounds right Apr 14 23:24:21 * jsun is trying that now Apr 14 23:24:34 do I need restart the whole system or just gpe? Apr 14 23:24:44 jsun: just XfbDev Apr 14 23:24:59 root@mainstone:/bin# XfbDev Apr 14 23:25:00 -sh: XfbDev: not found Apr 14 23:25:10 jsun: whatever your X server is named ;) Apr 14 23:25:32 Ahh, Xfbdev. :) Apr 14 23:26:14 hmm, that did not work. :( Apr 14 23:26:22 once again Apr 14 23:26:31 try the ts utilities first... Apr 14 23:26:58 the driver is ucb1x00. Apr 14 23:26:59 check that raw input is working, calibrate, check that it is calibrated, and do some drawing.. Apr 14 23:27:03 maybe I should uncomment that one? Apr 14 23:27:16 zecke|baking, good idea. Apr 14 23:27:19 send me your mainstone, and I will find out for you Apr 14 23:27:26 * jsun actually check the old ts.conf file that worked. Apr 14 23:27:39 jsun: that is a good idea as well Apr 14 23:28:01 jsun: and you need to differentiate Apr 14 23:28:18 jsun: You might have a ucb1x00 device, but if you use the Linux Input System Apr 14 23:28:36 jsun: you have a generic 'gateway' to the User Space regardles of the hardware you use Apr 14 23:29:12 Darn. The old one does not even talk about input module. Apr 14 23:29:20 alright I am going to do the ts test thingy ... Apr 14 23:29:23 jsun: module_raw Apr 14 23:31:01 hmm, ts_test does not work. Apr 14 23:31:39 jsun: what is wrong testing the raw part? Apr 14 23:31:44 I mean the program runs but the test failed. :) Apr 14 23:31:54 no response to my touches. Apr 14 23:31:56 :) Apr 14 23:33:01 jsun: what could cause that? Apr 14 23:33:02 I swear the kernel and drivers are good. Apr 14 23:35:24 jsun: tslib is a chain (filter and pipe architecture) Apr 14 23:35:43 jsun: it starts with a module_raw just above a 'device' and then adds filters (other modules) Apr 14 23:36:02 jsun: by testing with ts_print_raw (or whatever that tool is named) Apr 14 23:36:15 jsun: you check at the lowest level of the chain Apr 14 23:36:29 Just test that: no output when I touch screen. Apr 14 23:36:42 jsun: this means, your module_raw is wrong Apr 14 23:36:50 jsun: the TSLIB_TSDEVICE is wrong Apr 14 23:36:59 I suspect the program is using a difference device. Apr 14 23:37:08 jsun: or... Apr 14 23:37:28 cat $TSLIB_TSDEVICE Apr 14 23:37:34 actually responds to my touch. Apr 14 23:37:45 jsun: which protocol is it using? Apr 14 23:37:56 dunno. Apr 14 23:39:02 jsun: try running evtest /dev/input/event1 Apr 14 23:39:13 that will tell you if it's using the evdev driver Apr 14 23:39:46 which package has evtest Apr 14 23:39:48 ? Apr 14 23:40:11 btw, using module_raw ucb1x00 seems to return some meaningful data. Apr 14 23:40:26 * jsun is testing a little further.... Apr 14 23:40:46 opening /dev/input/event1 as ucb1x00 sounds like montavista to me... Apr 14 23:41:09 zecke|baking, well maybe the intel hacks are made by mvista. :) Apr 14 23:41:15 actually very possible. Apr 14 23:41:32 jsun: http://sermons.desiringGod.org/~mreimer/evtest Apr 14 23:43:50 mreimer, that is binary file! Apr 14 23:44:11 jsun: yes, run it Apr 14 23:44:16 ./evtest /dev/input/event1 Apr 14 23:45:15 cool - Apr 14 23:45:18 root@mainstone:/home# ./evtest /dev/input/event1 Apr 14 23:45:18 Input driver version is 1.0.0 Apr 14 23:45:18 Input device ID: bus 0x0 vendor 0x0 product 0x0 version 0x0 Apr 14 23:45:18 Input device name: "Touchscreen panel" Apr 14 23:45:18 Supported events: Apr 14 23:45:19 Event type 0 (Sync) Apr 14 23:45:21 Event type 3 (Absolute) Apr 14 23:45:23 Event code 0 (X) Apr 14 23:45:25 Value 625 Apr 14 23:45:27 Min 0 Apr 14 23:45:30 Max 0 Apr 14 23:45:31 Event code 1 (Y) Apr 14 23:45:33 Value 729 Apr 14 23:45:35 Min 0 Apr 14 23:45:37 Max 0 Apr 14 23:45:42 Event code 24 (Pressure) Apr 14 23:45:45 Value 0 Apr 14 23:45:47 Min 0 Apr 14 23:45:49 Max 0 Apr 14 23:45:51 Testing ... (interrupt to exit) Apr 14 23:45:53 it seems hanging there now ... Apr 14 23:46:50 mreimer, anything we can decipher here ? Apr 14 23:47:02 it definitely hangs now... Apr 14 23:47:20 * jsun ctrl-C'ed evtest. Apr 14 23:48:26 jsun: use pastebin for that Apr 14 23:48:29 ~pastebin Apr 14 23:48:31 well, pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.com/, or http://pastebin.ca, or http://channels.debian.net/paste Apr 14 23:48:42 not everyone has a mainstone and bandwidth enough Apr 14 23:49:00 ok - thanks. Apr 14 23:49:08 you are welcome Apr 14 23:49:54 jsun: looks okay, Pressure of Min/Max looks okay Apr 14 23:49:56 * Philippe is away: visual contact - melancholic dreams Apr 14 23:50:20 jsun: not having a min/max for x/y sounds a bit wrong though - but acceptable Apr 14 23:51:57 more output from evtest after I touched the screen Apr 14 23:52:00 http://channels.debian.net/paste/2359 Apr 14 23:52:35 looks good right? Apr 14 23:53:01 start ts_print_raw using strace and see what file gets opened Apr 14 23:53:10 looks good to me - : Apr 14 23:54:32 jsun: try http://sermons.desiringGod.org/~mreimer/Xfbdev Apr 14 23:54:46 hmm, I need to bitbake strace ... starting ... Apr 14 23:54:57 mreimer, anything special in that? Apr 14 23:55:20 not really, just an Xfbdev built by OE that I know has evdev support Apr 14 23:55:26 maybe yours doesn't Apr 14 23:57:08 mine is built by OE tho. Apr 14 23:57:43 should be good then Apr 14 23:59:00 zecke|baking, when I uncomment ucb1x00, I got output from ts_print. Apr 14 23:59:21 when I uncomment module_raw input, I don't. Apr 14 23:59:28 jsun: that doesn't sound right Apr 14 23:59:38 jsun: and evtest works with the device as well Apr 14 23:59:40 I am puzzled. Apr 14 23:59:58 jsun: ucb1x00 just reads 32 bit IIRC and prints the values Apr 15 00:00:16 jsun: module_raw input tries to open it as a special device and does a sanity check Apr 15 00:04:02 * jsun got strace going ... Apr 15 00:04:49 zecke|baking, if I have .bb that has include and modify inc and then reparse it, why being smart and say file has not been updated instead of reparsing? :) Apr 15 00:05:27 polyonymous: it doesn't know about internal file dependencies Apr 15 00:05:36 polyonymous: can emerge do that? Apr 15 00:05:43 yes, but if I say reparse why not just reparse? Apr 15 00:05:55 polyonymous: ah the shell, ask mickeyl ;) Apr 15 00:06:02 I see :) Apr 15 00:06:03 'night all Apr 15 00:06:15 Well, that's no prob to touch, it's just that it's a bit weird. Apr 15 00:06:16 RP: cya Apr 15 00:07:01 I think emerge doesn't need it, it parses needed ebuilds at each startup... Apr 15 00:08:01 zecke|baking, another question... I'm trying to build pwmpi and I have gcrypt stuff in staging. pwmpi has its own copy, but it gets in the include list after the staged version. Can I prepend -I ? Apr 15 00:08:41 polyonymous: actually this is what should always happen... Apr 15 00:08:57 polyonymous: we slightly abuse C(XX)FLAGS and place -I${STAGING_INCDIR} there Apr 15 00:09:11 polyonymous: packages should always put their includes in front of that :} Apr 15 00:09:14 Well, in the resulting build it has staging before local includes... Apr 15 00:09:29 I'm worse with qmake than I am with autotools... Apr 15 00:09:48 Not sure how it builds those makefiles... Apr 15 00:09:58 polyonymous: ah that is qmake Apr 15 00:10:02 but it definitely goes after staging stuff. Apr 15 00:10:05 yes, that is it. Apr 15 00:10:17 polyonymous: and it is from lutz (the pi guy)... Apr 15 00:10:23 should be something like prepending to OE_QMAKE_CFLAGS ? Apr 15 00:10:35 what is from the pi guy? Apr 15 00:10:47 the .bb? Apr 15 00:10:49 polyonymous: what is wrong with the not forked version of libgcrypt? Apr 15 00:11:03 it even has different prototypes Apr 15 00:11:05 zecke|baking, with "module_raw input" uncommented, I see ts_print open /dev/input/event1 (which is correct), and also read data when I touch screen, however, it does not output anything. Apr 15 00:11:20 make any sense? Apr 15 00:11:36 I'm not sure that porting it to the staged gcrypt is the way to go. Apr 15 00:11:36 polyonymous: ouch, pester lutz to use the official one Apr 15 00:11:52 could it be that he's had a reason for using local copy? Apr 15 00:12:05 polyonymous: well why would you want to keep two copies of a security critical library? Apr 15 00:12:21 polyonymous: which version do you trust more? Lutz one? or the upstream source code? Apr 15 00:12:28 (from a distribution point of view) Apr 15 00:12:44 zecke|baking, you're perfectly right, but I don't know his reasons. Apr 15 00:12:56 polyonymous: he is not a team player at all Apr 15 00:13:16 well, neither I am ;-) Apr 15 00:13:22 polyonymous: he has forked most of Opie just to implement Button Order for tab focus (well that is not completely true) Apr 15 00:13:43 :))) Apr 15 00:13:59 polyonymous: he has forked ~6 Opie applications for no good reason Apr 15 00:14:00 well, then it doesn't make sense to fight over crypt issue, easier to just make it work. Apr 15 00:15:06 or do you suggest that we fork kdepim to make it work with stock gcrypt? :) Apr 15 00:15:08 polyonymous: try QMAKE_VARSUBST_PRE += "INCLUDEPATH+=private-gcrypt/include" Apr 15 00:15:17 Aha, thank yuo Apr 15 00:16:21 and thank you too :) Apr 15 00:16:31 polyonymous: if that doesn't work, we need to play tricks to QMAKE_CC or QMAKE_CXX Apr 15 00:16:37 didn't work though... Apr 15 00:16:39 using the VARSUBST_PRE thing Apr 15 00:16:40 okay Apr 15 00:17:33 OE_QMAKE_CC_append = "-I...." Apr 15 00:17:37 OE_QMAKE_CXX_append = "-I...." Apr 15 00:17:44 and see what is hapening Apr 15 00:17:52 skip the first two lines of my reply Apr 15 00:18:04 maybe OE_QMAKE_CFLAGS_append? is there such thing? or is it too late? Apr 15 00:18:10 just in your .bb file (like any other variable) doe OE_QMAKE_CC_append = "-I" Apr 15 00:18:32 I'm currently looking at packages/qmake/files/linux-oe-qmake.conf Apr 15 00:18:49 in fact, I'm already trying my version, I enetered it before you suggested yours Apr 15 00:18:52 entered Apr 15 00:19:03 ah good, keep that then ;) Apr 15 00:19:22 and well, I got different error, must be another fork :) Apr 15 00:19:40 this seems to have worked: OE_QMAKE_CFLAGS_prepend="-I${S}/pwmanager/libcrypt/mpi/crypt" Apr 15 00:19:50 jsun: http://cvs.arm.linux.org.uk/cgi/viewcvs.cgi/tslib/plugins/input-raw.c?rev=HEAD&content-type=text/vnd.viewcvs-markup Apr 15 00:20:09 polyonymous: maybe add a ' ' Apr 15 00:20:13 to the end Apr 15 00:20:19 makes sense. Apr 15 00:21:06 Although I think there was a space, otherwise build would be shorter, not longer :)) Apr 15 00:21:14 hehe Apr 15 00:21:59 weird.... after adding space it's back to the error #1... Apr 15 00:22:04 but it's clearly impossible Apr 15 00:22:13 and the -I is there. Apr 15 00:23:07 nevermind, I put the wrong path :) Apr 15 00:30:37 built it. Apr 15 00:38:56 well, goodnight everybody Apr 15 00:49:43 gosh I hate svk Apr 15 01:54:30 anyone alive still? Apr 15 01:54:45 Yep Apr 15 01:57:43 are the assumptions in the following patch correct http://sourceware.org/ml/newlib/2004/msg00046.html Apr 15 01:58:48 No idea.. Apr 15 01:59:10 Ask RP? Apr 15 02:00:17 when's he around i guess Apr 15 02:43:11 * TheMasterMind1 bangs his head against a wall Apr 15 02:56:36 bugger, debian sid (unstable) is currently with an broken Xorg transition period Apr 15 02:56:44 heck of a time for me to try to upgrade :P Apr 15 02:56:55 jnc: Yeah, I've been putting off my apt-get upgrades for a few days.. Apr 15 02:57:12 good idea. neat stuff though, it's not executing or doing useful work yet Apr 15 02:57:36 the split has happened. Xorg modularization is now in effect, and /usr/X11R6 is disappearing ;) Apr 15 02:58:43 most of the parts are arranged as of tonight, except video and input modules Apr 15 02:58:58 kind of makes it tough to run X11 without video or input modules, y'know =) Apr 15 02:59:07 True =) Apr 15 02:59:21 i see my nick highlighted while i was away Apr 15 02:59:26 happen to know what i missed? Apr 15 02:59:53 Nah, can't see it from flicking back Apr 15 03:00:08 nothing in my history here Apr 15 03:00:18 thanks Apr 15 03:00:25 how y'all been? Apr 15 03:00:48 Not too bad; and yourself? Apr 15 03:01:04 depriving myself of sleep sometimes. okay though! Apr 15 03:01:27 watching bitbake build its cache, as usual Apr 15 03:01:37 heh Apr 15 03:01:41 that annoys me Apr 15 03:01:58 i wait all this time for bitbake to build a cache, and then it seems like the cache never speeds anything up Apr 15 03:03:02 TheMasterMind1: which embedded hardware are you using? Apr 15 03:03:59 gumstix Apr 15 03:04:06 i also have a 5500 Apr 15 03:04:38 if i have a-type1.bb and a-type2.bb and both have provides=a in them, how can i make one preferred Apr 15 03:08:13 ooh gumstix. don't know much about it Apr 15 03:08:29 PREFERRED_PROVIDER_a = "a-type2" should do Apr 15 03:08:34 could you explain an example use of gumstix hardware in real world application? Apr 15 03:11:04 jnc: there's many projects out there, just look at gumstix.com or search google Apr 15 03:11:14 oh Apr 15 03:11:25 i meant, more specifically what /you/ use it for Apr 15 03:11:59 robotics Apr 15 03:12:47 with the robostix attachment we can write linux progs in c/java/python/whatever to control inputs and motors and use bluetooth to ssh in wirelessly or communicate amongst robots Apr 15 03:13:24 holy laz0r beam. Apr 15 03:15:01 they're being developed primarily for graduate research on swarm robotics Apr 15 03:15:53 i.e. how swarms of fish and flock of birds move in groups without hitting each other, and how ants work collectively to gather food Apr 15 03:16:16 i've often wondered that, watching the gulls gather outside a datacenter parking lot Apr 15 03:16:32 why is it they poop on everyone's car except mine? Apr 15 03:16:41 i feel so left out Apr 15 03:17:59 (***) /oe/org.openembedded.dev/packages/classpath/classpath-minimal_0.90.bb Apr 15 03:18:01 finally Apr 15 03:39:47 I am trying to compile glibc for an ARM processor. I am getting an error since an include file (link.h) has one line in it that says it needs to be defined for the target platform. My question is, how do I find out what definitions are suppose to be in link.h Apr 15 03:40:28 * JustinP merges Apr 15 03:40:48 jnc: the cache makes it *much* faster Apr 15 03:41:03 jnc: the first run is very slow, then subsequent ones are much faster Apr 15 03:41:12 jnc: may not be if you didn't install psyco, though Apr 15 03:41:23 JustinP: psyco is no-go on my platform Apr 15 03:41:28 :/ Apr 15 03:41:45 could cache be disabled if psyco is not there? Apr 15 03:42:44 jnc: just try blowing away the cache and see if it slows down... Apr 15 03:43:10 it's about the same, i've tried that out of curiosity Apr 15 03:43:25 anyone? Apr 15 03:43:40 my CPU is not the limiting factor, disk IO seems to be the limiter Apr 15 03:43:40 03justinp 07org.oe.dev * rb387bb17... 10/packages/e17/ (entrance/disable-autodetect.patch entrance_0.9.0.006.bb): entrance: remove useless autodetection script and fix packaging with e.bbclass Apr 15 03:43:44 03justinp 07org.oe.dev * re879a0da... 10/packages/efl/ (6 files): evas: stage modules to arch-based directories Apr 15 03:44:07 caching uses tons of disk IO w/o psyco Apr 15 03:44:11 i'm not sure about that Apr 15 03:44:17 it appears to be the case though Apr 15 03:47:58 jnc: I copied cache to a tmpfs and got no performance boost on an Athlon 1900+ with 768MB RAM, but that machine is just for building Apr 15 03:48:09 5400 RPM disk, BTW Apr 15 03:48:22 hmm Apr 15 03:48:32 that's an 1900+ though Apr 15 03:49:03 you'd have to try it on a rediculously fast box to do a fair test Apr 15 03:49:31 never goes above 70% CPU usage Apr 15 03:49:48 er...spiked to 80% for a second Apr 15 03:49:48 jnc: it's not the "cache" that's slowing things down. The cache speeds it up. It's the *parsing* that slows it down. Apr 15 03:50:00 jnc: with psyco the cache speeds things up considerably Apr 15 03:50:04 oh Apr 15 03:50:14 jnc: and the newer versions of bitbake are supposed to be even faster Apr 15 03:50:20 jnc: although they have other issues Apr 15 03:50:23 the requirement is on psyco? Apr 15 03:50:27 I believe so Apr 15 03:50:42 bugger ;/ Apr 15 03:50:51 if your first parse and subsequent parses are the same speed it's because you have *no* cache Apr 15 03:51:03 I suggest using bitbake -i more often ;-) Apr 15 03:55:50 meh. reiserfs failing already Apr 15 05:53:02 03justinp 07org.oe.dev * r379d3632... 10/packages/e17/e-wm/fix-configure.patch: e-wm: remove unneeded fixes from patch Apr 15 06:07:27 woohoo! e-image-core builds again Apr 15 06:22:50 * jnc applauds Apr 15 06:24:09 hey mickey|dinner you about? Apr 15 06:24:12 * mithro pokes koen|sleep Apr 15 06:46:16 morning all Apr 15 06:48:03 * jnc waves Apr 15 06:58:23 hey RP Apr 15 06:58:35 did you see that there is another Summer of Code? Apr 15 06:59:11 will be interesting to see what happens regarding it Apr 15 06:59:24 hi mithro Apr 15 06:59:40 mithro: I did notice it being mentioned, yes Apr 15 06:59:55 I offered to mentor last year... Apr 15 07:06:19 the kernel module ipks have too much overhead. each control file is 2K because it contains all the patches used in the sources line, and they have the exact same postinst scripts.. couldn't we set it up so they all shared a common post inst atleast Apr 15 07:07:43 file it as a bug and offer to fix it :) Apr 15 07:08:05 TheMasterMind1: In theory that's a nice idea. In practice, its rather tricky to do without rewriting ipkg Apr 15 07:08:27 it shouldnt contain all the patches tho Apr 15 07:09:13 the names of all the patches that is, the entire SRC_URI Apr 15 07:09:25 ah Apr 15 07:09:36 i can see why thats a good idea, but i only have a 3.8mb jffs2 root =\ Apr 15 07:10:14 TheMasterMind1: Add an option to disable the Sources: line in the packaging Apr 15 07:11:12 yea planning on doing that. trying to think of a decent hack to get the kernel modules sharing scripts Apr 15 07:11:27 i keep thinking symlinks but that just can't be pretty Apr 15 07:13:23 RP: can you confirm the patch at http://sourceware.org/ml/newlib/2004/msg00046.html is incorrect Apr 15 07:15:02 TheMasterMind1: I'm not sure. Paul Brook usually knows what he's doing... Apr 15 07:16:35 well its an old patch, but "__VFP_FP__ will be defined, even if soft-float" is certainly not true for me with 3.4.3 Apr 15 07:17:23 03justinp 07org.oe.dev * rfb37f14b... 10/packages/meta/e-image-core.bb: e-image-core: fix DEPENDS Apr 15 07:18:20 TheMasterMind1: That suggests it isn't quite right then... Apr 15 07:19:10 well the ieeefp.h in gnu classpath has the same code causing floats not to work with any jvm using classpath Apr 15 07:19:42 i filed a report on the classpath bugzilla but i'm not really sure where upstream it needs to be reported Apr 15 07:21:13 It should be reported to the people providing that header I guess Apr 15 07:21:24 yea, how do i find out who that is Apr 15 07:21:53 Where did the header come from? Apr 15 07:22:23 its in the classpath source tree under the fdlibm directory. Apr 15 07:24:16 TheMasterMind1: So classpath is one place it needs fixing I guess. If they get it from anyone, hopefully they'll pass the fixes on... Apr 15 07:24:48 alright, sounds good. atleast its reported somewhere, hopefully google with crawl it Apr 15 07:30:01 RP: how does that sound, a flag to go through kernel-module*.postinst and create symlinks to one file, and another flag to perhaps clean out all the sources lines Apr 15 07:31:06 TheMasterMind1: What happens when you remove the package they all symlink to? Apr 15 07:32:43 well i was going to take the first one in the list and copy it to kernel-modules-common.postinst and symlink to that Apr 15 07:33:27 So how does kernel-modules-common.postinst get removed? Apr 15 07:33:41 it doesn't Apr 15 07:33:54 You just broke package management then :-/ Apr 15 07:34:27 yea, its a hack and a tradeoff. Apr 15 07:34:38 the sources thing will save me more space so i'll do that first Apr 15 07:34:59 I'm not sure we want hacks in OE itself... Apr 15 07:36:52 well the other way would be to have them all depend on an empty package that contained only the scripts, but then you need to have broken symlinks into the ipks which is nastier Apr 15 07:37:32 The only correct solution is a more intelligent ipkg :-( Apr 15 07:38:07 ok, fair enough Apr 15 07:38:12 how would ipkg ideally handle it Apr 15 07:38:43 All its metadata was in a database and it would recognise the postinsts were identical and only store one copy Apr 15 07:39:27 I have given this thought in the past... Apr 15 07:39:42 It would also work for your Sources: problem Apr 15 07:40:04 ok so something nicer than the files we have now Apr 15 07:40:34 yes Apr 15 07:40:49 sqlite comes to mind Apr 15 07:41:04 It might be seen as too much of a change from traditional debian behaviour so you might have to fork ipkg :-/ Apr 15 07:41:17 sqlite is what I was thinking of using Apr 15 07:42:37 well thats definitely not something that can be hacked on.. not easily anyway Apr 15 07:43:23 and the idea of forking ipkg makes me shudder Apr 15 07:43:58 You can see why it might be needed though? Apr 15 07:44:13 oh yea definitely Apr 15 07:44:31 I agree, its not ideal. Its also very heavy changes to ipkg Apr 15 07:45:00 you could do something similar and rudimentary, perhaps compare md5sums to find identical files, but thats not nearly as beneficial as doing it on a field by field basis Apr 15 07:46:07 well ideally i think its something that could be achieved in the ipkg backend without any changes to the ipk format Apr 15 07:46:17 data comes in the same way, its just stored more efficiently Apr 15 07:46:42 Yes, the ipk format wouldn't change Apr 15 07:49:04 well if neither the users nor the developers/packagers have to change anything they're doing it should be a lot easier to get accepted. reminds me of when ethernet bridges came out, replace your hub, do nothing else and all of a sudden you have dedicated bandwidth link to every computer on the network Apr 15 07:51:15 TheMasterMind1: The only issue might be that people are attached to the debian like file structure on the device Apr 15 07:51:39 yes, it certainly has its advantages Apr 15 07:53:08 It's easy to sh script Apr 15 07:53:50 what is? Apr 15 07:54:23 i need a variable name for toggling the source: line.. IPK_DISABLE_SOURCE? IPK_CONTROL_NO_SOURCE? Apr 15 07:54:51 the file-based structure is easy to hack on based on its flat file structure Apr 15 07:55:00 I agree, however, that sqlite would be a better choice Apr 15 07:55:05 yep Apr 15 07:55:32 i'm trying to think if it'd be possible to have a separate utility that did it.. if only you could create a symlink to multiple files that were to be appended Apr 15 07:55:53 morning Apr 15 07:56:26 ~logs Apr 15 07:56:27 apt/ibot/jbot/purl all log to http://ibot.rikers.org// where channelname is html encoded ie: %23debian | lines that start with a space are not shown | some channels have stats at http://ibot.rikers.org/stats/.html.gz, or updated "nightly" Apr 15 07:59:02 TheMasterMind1: yeah, sharing the postinsts between kernel packages would be a neat idea. It probably ought to be as easy as putting the postinst in a dedicated package that all the others depend on, and inventing some new control field to set the path of that script. Apr 15 07:59:58 heck a new control field with a path could just point to update-modules itself since thats all the script does Apr 15 08:00:16 TheMasterMind1: as for the verbose Source: lines, yeah, they're a bit of a nuisance. Including all the patch URIs has only dubious utility anyway; in the medium term it would be better to move to some kind of source package system for that stuff (even if the "source package" is just a tarball) Apr 15 08:00:16 *** Error: CONTROL/control is missing field Source Apr 15 08:00:17 last xstroke Apr 15 08:00:21 oops :) Apr 15 08:01:50 TheMasterMind1: oh, heh. if that's really all it does then yeah, that would work. But, equally, if that's all it does then it's only going to occupy a few bytes of storage for each copy of the postinst, so maybe it isn't worth worrying about. Apr 15 08:03:33 yea agreed, its not that big of a space hog. it has a conditional in there currently to see if its running on the OE build host, but still Apr 15 08:03:42 the Source: lines are the ones taking 2K per module Apr 15 08:03:52 yeah Apr 15 08:04:09 well i'm just going to include blank Source lines for now then Apr 15 08:04:19 we already strip the Source lines out of the Packages file for familiar for space reasons. Apr 15 08:04:34 ah, its not just me then Apr 15 08:04:46 well i just added a flag not to put them there in the first place Apr 15 08:04:48 (though we do leave them in the individual .ipks) Apr 15 08:05:14 oh packages file, misread that Apr 15 08:05:41 I think the best thing might be to patch package_ipk.bbclass to only put the URI of the upstream tarball in Source:, not all the patch stuff. Apr 15 08:06:08 in a sense that's a bit misleading, but all those file:// URIs are next to useless anyway for someone who just has a copy of the .ipk file. Apr 15 08:06:13 yea Apr 15 08:06:36 the name of the .bb file might make sense Apr 15 08:06:50 mithro: pong Apr 15 08:07:01 hi pb_ , koen Apr 15 08:07:18 longer term, package.bbclass should be putting all the sources in a tarball that gets distributed along with the .ipk, and the Source field in the latter should refer to the name of that tarball Apr 15 08:07:20 hi gremlin[it] Apr 15 08:08:41 TheMasterMind1: or, yet another idea would be to invent some flag that says "this package has the same source as ", so that all the kernel-modules could refer to kernel-image. Apr 15 08:08:56 and then kernel-image would have a real Source: field with all the URIs in Apr 15 08:09:27 hrm, yea but at that point it almost makes sense to go with RPs idea Apr 15 08:10:29 changing ipkg? yeah, I guess, but that isn't terribly appealing. Apr 15 08:11:27 well i meant more doing that for all fields, but you're right that would require changing ipkg Apr 15 08:12:26 mm. there aren't very many other fields that are routinely shared, at least not big ones. Apr 15 08:12:41 obviously stuff like Architecture: is almost always the same, but that one isn't big enough to worry about Apr 15 08:13:12 well even stuff like 'Maintainer: OpenEmbedded Team ' can add up Apr 15 08:14:09 but again i agree, sources is the big one and another control field sounds like a good solution for the moment Apr 15 08:16:48 hey koen Apr 15 08:20:00 mithro: you pinged me earlier? Apr 15 08:21:34 ahh about the SoC stuff Apr 15 08:24:26 can ibot give time in germany? Apr 15 08:24:58 meh. everytime i go to use ipkg i end up writing a patch for it Apr 15 08:31:16 mithro: 10:33 Apr 15 08:31:37 10:40 that is Apr 15 08:31:37 * mithro ponders what time zecke will get up Apr 15 08:42:20 mithro: I think OE can benefit from the GSoC Apr 15 08:45:46 good morning Apr 15 08:45:56 hey Bernardo Apr 15 08:46:11 hi koen Apr 15 08:47:04 I've been fighting to get my sandisk plus connect to work on my laptop in wifi mode, but no luck yet... Apr 15 08:49:06 and the specs for it have disapeared... I was thinking on starting hacking it on the laptop then if I had any luck getting it to work in wifi+ide porting the patch to the z Apr 15 08:57:54 smart little ipkg, detects if i leave Source: out or even if i have it blank! Apr 15 09:16:18 heh Apr 15 09:34:09 well that seems to have works, i don't know how much it really helped. guess tomorrow is for unionfs and udev. gotta get rid of sysvinit and hotplug Apr 15 09:35:27 okay, very good Apr 15 09:36:07 night Apr 15 09:36:13 'night TheMasterMind1 Apr 15 09:36:18 night tmm1 Apr 15 09:40:58 ~oe Apr 15 09:40:59 i heard oe is OpenEmbedded (see http://www.openembedded.org ), or "Opportunistic Encryption" (see http://www.wavesec.org ), or an email client which is not to be spoken of. **** ENDING LOGGING AT Sat Apr 15 09:59:57 2006