**** BEGIN LOGGING AT Tue Apr 18 09:59:57 2006 Apr 18 10:05:13 http://cgi.ebay.com/Onset-Model-8-SBC-computer-processor-C-and-Basic_W0QQitemZ7610891176QQcategoryZ4660QQssPageNameZWDVWQQrdZ1QQcmdZViewItem#ebayphotohosting Apr 18 10:10:18 03koen 07org.oe.oz354x * rcea035c2... 10/packages/epdfview/epdfview_0.1.2.bb: epdfview: add 0.1.2 from .dev Apr 18 10:35:13 sws: offhand no idea Apr 18 10:37:05 do13_: thanks, i will see what I can learn, maybe I can assist you in supporting 2.6 on tosa Apr 18 10:50:12 NOTE: package libtool-native-1.5.10-r3: task do_configure: started Apr 18 10:50:12 ERROR: function do_configure failed Apr 18 10:50:24 | /usr/bin/autoreconf: unrecognized option `--exclude=libtoolize' Apr 18 10:50:49 I tried to build openzaurus kernel 2.4 for tosa Apr 18 10:51:07 sws: For some reason its using your system's autoreconf rather than the one bitbake builds Apr 18 10:51:25 sws: Try manually building autoconf-native Apr 18 10:51:42 bitbake -b ?? Apr 18 10:55:41 03xora 07org.oe.dev * ree780a36... 10/packages/abiword/abiword-plugins_2.4.4.bb: abiword-plugins_2.4.4.bb : prevent from picking up hosts libwmf Apr 18 11:00:06 what is the best way to mirror the OZ feed? do they have an rsync server? Apr 18 11:08:57 bitbake autoconf-native want to build libtool-native, which breaks Apr 18 11:10:16 doing bitbake bootstrap-image with ZKERNEL=2.6 in local.conf, the toolchain works? Apr 18 11:16:46 maybe the problem is quilt-native and m4-native aren't build? Apr 18 11:22:44 sws: quilt will be the problem - its DEPENDS are incorrectly calculated by some versions of bitbake. Which version of bitbake are you using? Apr 18 11:23:21 03xora 07org.oe.dev * rfbc67e6f... 10/packages/epdfview/epdfview_0.1.2.bb: epdfview_0.1.2.bb : fix .desktop file to point to right exe and icon Apr 18 11:24:06 koen: bump that to branch? Apr 18 11:25:04 RP: 450, did a svn up this morning, 4 hours ago Apr 18 11:27:13 sws: Can you pastebin the exact error you're seeing please? Apr 18 11:29:36 XorA: done Apr 18 11:30:24 bbl Apr 18 11:31:25 http://pastebin.com/666932 Apr 18 11:31:52 03koen 07org.oe.oz354x * r9b257597... 10/packages/epdfview/epdfview_0.1.2.bb: epdfview: sync with .dev Apr 18 11:31:56 03koen 07org.oe.oz354x * re405c750... 10/packages/abiword/abiword-plugins_2.4.4.bb: abiword-plugins: sync with .dev Apr 18 11:37:11 now if only I could get the plugins actually working Apr 18 12:29:48 03koen 07org.oe.dev * re617332c... 10/conf/distro/angstrom-2006.9.conf: use pcmciautils for some more machines Apr 18 12:36:27 RP: could you have a look at http://ewi546.ewi.utwente.nl/tmp/pcmciautils_013.bb ? Apr 18 12:37:04 koen: Any part of it specifically? Apr 18 12:37:12 RP: it throws a warning in do_package and I'm not sure it the makefile is at fault or the .bb Apr 18 12:37:29 (apart from hardcoding ln to /usr/bin/ln) Apr 18 12:37:43 What warning? Apr 18 12:38:02 'dangling symlink to /pccardctl' Apr 18 12:38:34 "Udev has been shipping with a scaled down version of libsysfs since late February. Udevstart is now 35% faster with the newer version." Apr 18 12:38:39 from http://linux-diag.sourceforge.net/Sysfsutils.html Apr 18 12:39:02 stupid linux changes Apr 18 12:39:05 I'd suspect the makefile and it sounds like a bug Apr 18 12:39:19 Speedups in udevstart are really needed... Apr 18 12:39:56 koen: That .bb is horrible. If we can work out how to make it more cross compilable, I could probably get the patches applied upstream. Apr 18 12:41:22 my update frenzy got triggered by http://handhelds.org/hypermail/h2200-port/current/3950.html Apr 18 12:41:50 pcmciautils 0.13 seems to have integrated sane defaults from ubuntu and opensuse Apr 18 12:42:28 pcmciautils is quite new and will welcome help to improve it Apr 18 12:43:16 I just pushed the new sysfsutils Apr 18 12:43:28 although udev 0.89 shouldn't need them anymore Apr 18 12:44:05 I'll have to try 0.89 as udev is one of the slowest parts of the system boot on spitz Apr 18 12:47:41 03koen 07org.oe.dev * r55fdf0ec... 10/packages/sysfsutils/sysfsutils_2.0.0.bb: sysfsutils: update to the latest stable version (2.0) Apr 18 12:47:43 I thought pcmciautils would detect and use udev automagcally by installing a rules file? Apr 18 13:06:27 quick everyone hide from mallum Apr 18 13:06:52 * koen hides Apr 18 13:07:14 ;P Apr 18 13:08:29 mallum: hal needs to be taught about pda hardware to make gnome-power-manager being remotely usefull Apr 18 13:08:40 drat, I suck at hiding Apr 18 13:08:51 koen: yeah I know Apr 18 13:09:38 koen: we could get a lot of cool stuff in Hal could be made to play a bit better with PDA's etc Apr 18 13:09:48 yeah Apr 18 13:11:57 might be nice to hack up something before guadec Apr 18 13:12:07 (which I won't be attending, *again*) Apr 18 13:12:25 koen: we are all pretty much too busy to do much about it at least atm. but could throw at least some money at the problem if someone wanted to hack it Apr 18 13:12:48 koen: :( how come no GUADEC ? Apr 18 13:13:03 examweek.... Apr 18 13:13:12 hey zecke Apr 18 13:14:54 koen: drop out ! Apr 18 13:15:10 koen: then you could go Apr 18 13:16:39 mallum: hey Apr 18 13:16:52 hey zecke zecke Apr 18 13:16:59 mallum: I wonder why you are mentioned as member of the pdaxrom team? Apr 18 13:16:59 how are you sir ? Apr 18 13:17:08 *bored* Apr 18 13:17:09 zecke: because i am elite Apr 18 13:17:12 on the simpad, /etc/resolv.conf is not set coreclty. it contains search fqdn but has no nameserver specified, if i edit this it works for a while but is then set back to how it was, i assume by the udhcpc that is still running. How can i fix this? i ran a search which turned up some comments saying it should infact be a symlink to /var/run/resolv.conf or some such, but as that doesnt exist i asuume it is irelevant in this case Apr 18 13:17:12 uni started Apr 18 13:17:18 mallum: true Apr 18 13:17:51 zecke: i think they just added me as I answered a question about how they should run mb about 3 years ago Apr 18 13:18:00 I dont think they even run MB nowadays Apr 18 13:18:12 mallum: only mb-panel Apr 18 13:20:07 mallum: pdaxrom are "having" to switch to 2.6 kernels as "users eat" sashz... Apr 18 13:20:57 RP: its the future Apr 18 13:21:04 oh the lecture continues Apr 18 13:21:08 RP: perhaps now you will be part of the pdaxrom tram also Apr 18 13:21:11 *bored* *kill* *destroy* Apr 18 13:21:14 mallum: More like present now ;-) Apr 18 13:21:23 RP: if they can find a picture of you Apr 18 13:21:31 hmm... Apr 18 13:21:43 RP: I discovered why the pdax people want the cpuscaling (and overclocking) Apr 18 13:21:58 koen: why? Apr 18 13:22:14 RP: libvorbis, mpg123 and other fp intensive stuff Apr 18 13:22:31 koen: OT: where did the tinderbox go? Apr 18 13:22:46 zecke: no idea Apr 18 13:22:47 wow programming Embedded Sensors with Visual Basic Apr 18 13:22:52 *idiots* Apr 18 13:22:53 zecke: afaik I didn't touch it Apr 18 13:23:12 koen: I forgot the location of it :} Apr 18 13:23:31 koen: We need to work out a driver and will get one eventually. Compared to all the other stuff we have working, getting cpufreq working is relatively straight forward Apr 18 13:25:39 cya Apr 18 13:27:02 RP: did they want to switch to 2.6 before or after they notices their users switched to OZ/gpe? ;) Apr 18 13:27:58 koen: I don't read the forums so I don't know what users are doing. I know what I'd like them to be doing though :) Apr 18 13:28:11 koen: I also keep suggesting to sash he tried OE ;-) Apr 18 13:28:40 RP: no, their buildsystem is superior Apr 18 13:28:45 (so they keep telling us) Apr 18 13:29:22 koen: He didn't actually say that to me. He just said it was better than OE when he started it and as he has it, he won't change Apr 18 13:31:56 * chouimat is back. Apr 18 13:32:08 hey chouimat Apr 18 13:32:23 morning Apr 18 13:44:31 bye Apr 18 13:44:47 ~kill matlab Apr 18 13:44:49 * ibot shoots a charged electron gun at matlab Apr 18 13:45:34 :-D Apr 18 13:45:47 ~cast Animate Dead on matlab Apr 18 13:49:15 hi Apr 18 13:56:17 hi all Apr 18 13:56:30 hey cedric Apr 18 13:59:14 hi cedric Apr 18 13:59:28 cedric: Did you have a chance to test the LEDs? Apr 18 13:59:43 just wondering I've got simple program with an autoconf structure, it compile fine under my local host. so I've copied the tarball to the my "sources" directory and create a bb file which inherits from "autotools pkgconfig". the problem is the tarball is not unpacked in the work directory Apr 18 14:00:04 RP: nope yet Apr 18 14:02:10 cedric: forgot to touch the foo.tar.gz.md5? Apr 18 14:04:23 koen: I've done that a same thing... Apr 18 14:04:36 s/a/and Apr 18 14:14:01 NOTE: Unpacking /home/cedric/oe/org.openembedded.dev/packages/foo/foo_1.11.bb to /home/cedric/oe/build/tmp/work/armv5te-linux/foo-1.11-r0/ Apr 18 14:14:31 it should say foo_1.11.tar.gz, no? Apr 18 14:25:16 ok found it, I need to provide a SRC_URI... I though is going to look by magic in the sources folder... Apr 18 14:55:52 hi Apr 18 14:56:20 RP: ping Apr 18 14:57:27 CoreDump|home: pong Apr 18 14:58:25 RP: I was playing w/ udev this morning and found out that USB events are not detected at all. Is this known? Apr 18 14:58:45 CoreDump|home: How recent was the udev? Apr 18 14:58:52 from .354x Apr 18 14:59:01 084 IIRC Apr 18 14:59:01 CoreDump|home: Does .354x have the fixes from .dev? Apr 18 14:59:08 * CoreDump|home checks Apr 18 14:59:23 I know there were some usb fixes that went into .dev Apr 18 14:59:50 hmmm, .oz has actually more patches than .dev =) Apr 18 15:00:06 CoreDump|home: Which ones does it have as extras? Apr 18 15:00:09 there's a 089 in .dev but it's disabled Apr 18 15:00:25 CoreDump|home: Its not seen much testing which is why Apr 18 15:00:44 Its supposed to be a bit faster at starting Apr 18 15:01:08 noasmlinkage.patch udevsynthesize.patch udevsynthesize.sh Apr 18 15:01:34 RP: It's creating a metric ton of devices,like 20 ttys and stuff =) Apr 18 15:02:18 aehem, I didn't check "files" yet Apr 18 15:02:59 gna Apr 18 15:03:30 RP: checked the .bb, .oz and .dev are identicaö Apr 18 15:04:15 CoreDump|home: What about the udev.rules and local.rules ? Apr 18 15:04:48 identical file size Apr 18 15:05:26 CoreDump|home: ok, I thought they were in sync. Looks like a potential problem them (although udev does load the right usb drivers for me) Apr 18 15:06:08 RP: udev doesn't load any drivers AFAICT, thats /etc/hotplug/usb$something's work Apr 18 15:06:26 CoreDump|home: No, hotplug is totally dead with udev - udev turns it off Apr 18 15:06:27 and even w/ b0rked rules, udevmonitor should see something Apr 18 15:06:46 udev calls modprobe - see the rules files Apr 18 15:06:50 the hotplug thingy is run by init on boot, loads USB drivers Apr 18 15:07:28 CoreDump|home: udev disables hotplug. I know that as fact. Apr 18 15:07:33 that hotplug thingy messed up NFS boot over usb0, that's why I know Apr 18 15:07:41 03koen 07org.oe.dev * rf3fdff92... 10/packages/sysfsutils/sysfsutils_2.0.0.bb: sysfsutils: default pref -1 till we get a newer pcmciautils Apr 18 15:07:51 RP: It's a simple script run by init, no deamons involved Apr 18 15:08:17 CoreDump|home: I'm talking about the hotplug binary that the kernel calls for hotplug events Apr 18 15:08:44 And I am about a hotplug initscript Apr 18 15:09:03 which loads ohci*, usbcore and friends on boot Apr 18 15:09:05 CoreDump|home: hotplug should probably get removed from images with udev Apr 18 15:09:35 righto, but currently it doesn't cause problems =) Apr 18 15:10:54 CoreDump|home: Its nearly totally inactive apart from this init script Apr 18 15:11:56 right, I say screw that script heh. Our problem is udev based anyway. I'll do some more test w/ USB host devices (usb sticks etc) Apr 18 15:12:54 udev should be taking care of ohci itself anyway (and if anything needed usbcore, it would get loaded as a depend) Apr 18 15:12:56 ok, I'm getting events for Host mode Apr 18 15:13:33 I'm getting no events for client mode (g_ether) Apr 18 15:15:57 CoreDump|home: Which event would you expect? Apr 18 15:16:32 dunno, hotplug handled this stuff in any case so I assumed udev would be the same Apr 18 15:17:01 hotplug handled what "stuff"? Apr 18 15:17:30 you could script hotplug to react to a usbd0 connection (doing ifup usbd0 etc) Apr 18 15:17:52 in addition you could hack host-mode handling like mounting of usb-storage cards Apr 18 15:17:57 CoreDump|home: network.sh doesn't see a usb network device creation? Apr 18 15:18:07 yes it does Apr 18 15:18:17 so udev must see the event Apr 18 15:18:22 the network device itself is added. Apr 18 15:19:02 RP: what I was trying to do this morning was auto-modprobe g_ether for client mode Apr 18 15:19:10 mounting of usb-stroage cards would work if the was an entry in fstab Apr 18 15:19:39 CoreDump|home: There is currently no kernel detection of use host/client mode Apr 18 15:19:47 ok :\ Apr 18 15:27:20 hey zecke Apr 18 15:33:15 hmm usb-storae devices are found but no SCSI devices are assigned to them Apr 18 15:33:50 ho Apr 18 15:34:34 CoreDump|home: I think there is something in bugzilla about that - needs a fix committing to udev Apr 18 15:34:47 * CoreDump|home checks Apr 18 15:35:13 CoreDump|home: #811 Apr 18 15:36:48 hi Apr 18 15:36:54 hi mickeyl Apr 18 15:37:01 NOTE: multiple providers are available (linux-openzaurus, hostap-modules); Apr 18 15:37:01 NOTE: consider defining a PREFERRED_PROVIDER to match runtime hostap-modules Apr 18 15:37:06 aaaaargh Apr 18 15:37:16 looks good to me Apr 18 15:37:28 koen: btw., i fixed efl++, did you see? Apr 18 15:37:30 it is building linux-oz.... Apr 18 15:37:32 koen: Note its a runtime so it really already knows the answer Apr 18 15:37:47 and I need hostap-modules for hh-pxa-2.6 Apr 18 15:37:51 or it can usually get them right... Apr 18 15:38:14 RP: excellent, that fixes it. Should I add that to udev.rules or local.rules? Apr 18 15:38:15 linux-oz dies with 'no config found for h2200' Apr 18 15:38:19 btw., hey i'm really excited about the parsing speed... can't think how we could live without it Apr 18 15:38:41 mickeyl: I noticed, but I'm in libtool hell with ecore Apr 18 15:38:49 koen: It picked linuz-oz? :-/ Apr 18 15:39:43 RP: it did Apr 18 15:39:43 koen: Thinking about it. we evaluate RDEPENDS before DEPENDS. I had a very good reason for doing that but if we did the opposite, it would probably avoid that Apr 18 15:40:09 bye Apr 18 15:41:00 mickeyl: I'm quite pleased with the parsing speed - poky (with 367 .bb files) takes 0.57s for time bitbake foo :) Apr 18 15:41:29 mickeyl: I enabled psyco and it went 0.05s faster... Apr 18 15:41:41 hehe Apr 18 15:41:49 RP: you could try freeze as well Apr 18 15:41:59 RP: then one could distribute static bitbake binaries Apr 18 15:42:11 NOTE: Update cvs://anonymous@cvs.sourceforge.net/cvsroot/gaim;module=gaim Apr 18 15:42:11 cvs [update aborted]: end of file from server (consult above messages if any) Apr 18 15:42:11 NOTE: Task failed: Fetch failed: gaim Apr 18 15:42:13 yipee Apr 18 15:42:24 ~stab sf.net Apr 18 15:42:29 * ibot runs at sf.net with an origami Swiss Army knife, and inflicts a nasty paper cut. Apr 18 15:43:43 koen: GPE sucks, it has so many compile errors Apr 18 15:43:56 many = one Apr 18 15:44:08 RP: question is, do we want to auto-mount usb-storage? (IMO: yes) And if so, all partitions (could be mounted in /media/usb[0-*]) or one hardwired partition via fstab? Apr 18 15:44:31 RP: if we can get it even faster, then I will call reparse in front of every shell command and don't spit out a progress indicator. that would make tracking changed stuff much more easier from within the shell... Apr 18 15:44:36 CoreDump|home: we want! Apr 18 15:45:10 though so heh Apr 18 15:45:10 CoreDump|home: I looked into pmount a while back which helps a lot with this (doesn't need fstab entries) Apr 18 15:45:34 \gone Apr 18 15:45:51 We need to rewrite pmount a bit to do what we want though Apr 18 15:46:28 It won't mount devices that aren't removable for a really broken definition of removable for a start - we'd want a globable whitelist of devices Apr 18 15:46:35 pmount is in the bugtracker, btw Apr 18 15:46:53 it in .oz Apr 18 15:47:02 koen: pmount was in OE last time I looked - I also have a hacked version of my own :-/ Apr 18 15:47:19 RP: TheMasterMind1 added an updated version with a few patches Apr 18 15:47:25 hi everybody Apr 18 15:47:32 hi wolverine Apr 18 15:47:37 I'll check it out Apr 18 15:47:42 hi coredump Apr 18 15:47:55 koen: I'll have to have a look at that... There were a few things I wanted to fix before we could use it Apr 18 15:48:18 can anyone tell me why there are two packages glibc-intermediate and glibc ? Apr 18 15:48:24 RP: http://bugs.openembedded.org/show_bug.cgi?id=867 Apr 18 15:48:28 which one should I pick? Apr 18 15:48:31 or both? Apr 18 15:49:07 CoreDump|home: Feel free to check the udev fix into local.rules btw Apr 18 15:51:01 koen: Looks like its in hand. It doesn't address the problems I saw using it on a handheld though :-/ Apr 18 15:52:12 the bugtracker pmount build fine Apr 18 15:54:04 wolverine: glibc Apr 18 15:55:03 then what's glibc-intermediate for? Apr 18 15:55:30 seems glibc-intermediate has linuxthread addon enabled Apr 18 15:55:54 and glibc is for NPTL Apr 18 15:57:20 to install glibc with NPTL enabled, glibc-intermediate must be built, then gcc-cross Apr 18 15:57:40 ugh, pmount / pumount are ~48k binaries. methinks a small script could do the same Apr 18 16:00:21 CoreDump|home: They also make their own mount system calls and hence don't update /etc/fstab Apr 18 16:01:44 and udevd runs as root, who can do the same thing for auto-mounting. The big advantage of pmount is that a non-root user can mount / umount Apr 18 16:02:09 CoreDump|home: and the dynamic creation of mount points Apr 18 16:02:36 CoreDump|home: We should write our own version :) Apr 18 16:02:38 RP: any idea why OE insists on building hostap-modules 0.3.9? Apr 18 16:02:44 RP: =) Apr 18 16:03:07 koen: based on the information provided, no idea Apr 18 16:03:13 gah Apr 18 16:03:14 won't PREFERRED_PROVIDER_hostap-modules = ... do it? Apr 18 16:03:16 ~lart local.conf Apr 18 16:03:16 * ibot holds local.conf to the floor and spanks him with a cat-o-nine-tails Apr 18 16:04:00 mickeyl: I'd guess its a version issue now but koen isn't being very specific :) Apr 18 16:04:12 RP: my local.conf had some bogus entries Apr 18 16:06:23 RP: I'll hack up a basic auto-mounter for the next 3.5.4.* release. And if you or someone else finds the time to hack on pmount, we throw it away Apr 18 16:06:59 CoreDump|home: ok. Keep in mind this is what mount.sh in dev was meant to become Apr 18 16:07:01 aaargh Apr 18 16:07:08 Greetings ! Apr 18 16:07:08 ~stab atmel-wlandriver Apr 18 16:07:12 hey Ifaistos Apr 18 16:07:13 * ibot runs at atmel-wlandriver with an origami Swiss Army knife, and inflicts a nasty paper cut. Apr 18 16:07:17 Ifaistos: hey Apr 18 16:07:20 hi Ifaistos Apr 18 16:07:25 Ifaistos: you are a PPC guy? Apr 18 16:07:30 I was going to hack just that file =) Apr 18 16:08:07 zecke : working on it :) Apr 18 16:09:41 Ifaistos: do you have the board from the PPC hack zone (or whatever this site is called) Apr 18 16:10:31 RP: I want to branch bitbake now, what do you want to see fixed before the release? Apr 18 16:11:47 zecke: No, they have not send it yet. From what i read on the forum they are waiting to get a new version from production Apr 18 16:12:18 Ifaistos: lets hope for a version with video hardware Apr 18 16:12:22 * zecke wants another PPC as well :} Apr 18 16:12:31 zecke: Hope it's not one of the products that gets like 5 prototype runs but no product at the end Apr 18 16:12:52 03koen 07org.oe.dev * rc717ea22... 10/conf/machine/ (h2200.conf ipaq-pxa270.conf): Apr 18 16:12:52 h2200, ipaq-pxa270: remove ipaq-common include: Apr 18 16:12:52 * several drivers in that include don't compile against 2.6.16 Apr 18 16:12:52 * the apps in that include should be handled by a distro, not a machine include Apr 18 16:13:06 zecke : I have OE running on 405GP and got several of these boards around if interested Apr 18 16:13:37 zecke : http://www.elinux.org/wiki/DHT-Walnut Apr 18 16:14:22 btw anyone had tried to get linux on a Philips Nino 300 ? Apr 18 16:14:52 I was cleaning my office and found one (we had them a few years ago, but totaly forgot about it) Apr 18 16:15:28 Ifaistos: can one have video running on this Walnut? Apr 18 16:16:49 03koen 07org.oe.dev * r74449257... 10/packages/gpe-what/gpe-what_0.43.bb: gpe-what: add 0.43 Apr 18 16:17:02 zecke : I had an old matrox card and at least i could get console on it. Never tried another pci vga though Apr 18 16:17:51 zecke123@svn.berlios.de's password: Deleting oe Apr 18 16:17:55 what a feeling Apr 18 16:20:18 zecke: that's a strange password you have Apr 18 16:20:36 pb__: no, just a bad terminal emulation Apr 18 16:21:15 ah right Apr 18 16:21:39 the bsd terminal must be worse than I thought if it prints "Deleting oe" at random times. Apr 18 16:21:53 must be terribly confusing, it would be easy to think that you were actually deleting an oe directory. Apr 18 16:21:55 hehe, just omits the line endings Apr 18 16:22:24 * zecke hugs is mac Apr 18 16:22:29 his even Apr 18 16:22:59 zecke: Is there much else we need to fix before a release? Apr 18 16:23:21 zecke: Your annoying method "error" s one thing :) Apr 18 16:23:22 RP: else? Apr 18 16:23:33 RP: I will copy trunk to branch now Apr 18 16:23:42 great Apr 18 16:24:02 zecke: ok Apr 18 16:24:02 ipkg-native .159 segfaults like crazy on my buildmachine Apr 18 16:25:32 RP: I plan to do it by '#' Apr 18 16:26:18 this one is new though: file_md5sum_alloc: Failed to open file /data/build/koen/OE/build/tmp/angstrom/rootfs/etc/pcmcia/config.opts: No such file or directory Apr 18 16:27:44 zecke: Funnily enough, I have that "fixed" like that locall :) Apr 18 16:28:35 RP: I have fixed it better Apr 18 16:28:44 RP: # XXX, hack hack hac k \nreturn Apr 18 16:36:00 03koen 07org.oe.dev * r40d5e9c9... 10/conf/distro/angstrom-2006.9.conf: angstrom 2006.9: some tweaks to get things to build properly Apr 18 16:37:07 okay we have a bitbake-1.4 branch Apr 18 16:37:15 please test it, please test installing it Apr 18 16:37:21 please test the manual... Apr 18 16:42:21 Hello, How can I communicate with an external I2C bus? Can I use serial or what? Thanks Apr 18 16:42:48 look in the kernel Apr 18 16:43:12 theyre are quite a few I2C devices and a generic I2C bus Apr 18 16:43:19 emte: yes, I know there are some i2c drivers, but aren't these drivers for the motherboard i2c bus? Apr 18 16:43:20 there* Apr 18 16:43:34 I2C is I2C Apr 18 16:43:47 Erm. Is it just me, or is sf anoncvs cooked? Apr 18 16:44:06 cooked? Apr 18 16:44:19 Necronom: yes, it is sf.net Apr 18 16:44:27 sf anoncvs has a load limiter, you probably need to retry lots of times Apr 18 16:45:01 Meh. Apr 18 16:45:06 they should use monotone insteaf of anoncvs... Apr 18 16:45:14 sf load limiter = they turn off the servers :) Apr 18 16:45:30 Well, monotone would make mirroring a lot easier. Apr 18 16:45:48 Necronom: even svn would be a lot better Apr 18 16:45:59 They claim to have svn access, now.. Apr 18 16:46:13 but neither moinotone or svn are default in most development tools on any platform Apr 18 16:46:20 emte: what do you mean by I2C is I2C? Apr 18 16:46:32 .. speaking of monotone, I've updated monoton-5 and created a monotone-6 package. What's the proper process to submit them? Apr 18 16:46:41 Send patches to the maintainer? Apr 18 16:46:44 fromvega, it doesnt change where you use it Apr 18 16:46:45 Necronom: via bugtracker please Apr 18 16:47:04 don't bug the maintainer directly unless there's something important between you and him ;) Apr 18 16:47:06 hi andersee Apr 18 16:47:11 mickeyl; okay.. Apr 18 16:47:47 emte: yes, but I it change how can I reach it! And what I would like to know is how can I connect my software to an external I2C bus. Apr 18 16:48:47 mickeyl: morning Apr 18 16:49:23 * andersee need to get a working oe build so he can (finally) get nethack into the oe build tree Apr 18 16:49:43 heh Apr 18 16:49:49 the holy grail of packages Apr 18 16:49:52 my 6 year old boy was complaining that nethack was missing Apr 18 16:49:59 you know you've arrived when nethack is in yourrepository Apr 18 16:50:03 andersee: right, that'd be cool Apr 18 16:50:14 later all Apr 18 16:50:21 c'ya koen Apr 18 16:50:36 CosmicPenguin: oddly enough, I bunch of years ago when I did the gnome nethack port, I (and Jeff Garzik) converted nethack to use automake/autoconf Apr 18 16:51:03 CosmicPenguin: I bet if I resurructed that code, building nethack in oe would be pretty easy Apr 18 16:51:12 I'll bet it is Apr 18 16:51:13 much as I hate auto* Apr 18 16:51:25 I looked at nethack once, but the config wasn't very OE friendly at all Apr 18 16:54:07 CosmicPenguin: that's an understatement Apr 18 16:55:04 CosmicPenguin: koen claims we could have a build box hosted ;) Apr 18 16:55:06 * zecke runs Apr 18 16:55:20 fromvega, it would depend on what your doing Apr 18 16:56:36 fromvega, currently i am accessing an external i2c bus using an FPGA designed PCI controller Apr 18 16:58:21 OpenEmbedded? Apr 18 16:58:31 but you could use serial, usb, wireless, etc Apr 18 16:59:33 emte: let me explain, I have some devices that have an i2c interface. And I want to make a software to communicate with those devices. I would like to use some PC port like parport or serialport. A PCI controller is beyond my time and knowledgement for now, I think. But in your project, are you converting from I2C to PCI? Apr 18 17:00:21 fromvega: linux already has support for i2c on parallel port Apr 18 17:00:23 yo ggilbert Apr 18 17:00:37 hey treke Apr 18 17:00:54 ggilbert: can I return some of the money you once gave me to get a disk for oesources.org? Apr 18 17:01:54 zecke: We don't so much need a disk as I need to clean the crap already there Apr 18 17:02:22 The CVS builds piled the hell up, so now we have every cvs checkout of mozilla for a year and a half Apr 18 17:02:29 along with X, and everything else Apr 18 17:02:45 yo treke Apr 18 17:02:48 good to see you around Apr 18 17:02:59 mickeyl, heya Apr 18 17:03:20 pb__: hum... Apr 18 17:03:46 fromvega, i wouldnt say converting .... pushing the i2c data into the PCI controller memory banks and read it from there on the PCI bus Apr 18 17:03:55 * pb__ hums, as instructed Apr 18 17:04:10 its pretty hard to convert a 2 wire bus to 32 wires Apr 18 17:04:16 ggilbert: ah smells, like a need for a policy then :) Apr 18 17:04:34 ggilbert: only two weeks of cvs snapshots, distrtibutions should mirror these packages :) Apr 18 17:05:03 emte: is PCI difficult to implement? Apr 18 17:06:02 well, you need to trustthat you built the circuit right before you jam it into your computer ... Apr 18 17:06:08 zecke: Yeah. The only real tough part is sorting through 48,942 legacy files :) Apr 18 17:07:08 ggilbert, cant you just modify the old scripts taht you used for the LICENSE sort? Apr 18 17:07:34 emte: I'm trying to dig up my old script since I had something half written Apr 18 17:07:40 ah Apr 18 17:07:45 emte: so, what in fact do we have done for I2C? A linux kernel driver for the motherboard I2C, I2C on the parallel port and anything else? Apr 18 17:08:01 emte: Step one is to learn how bitbake works now Apr 18 17:08:12 emte: Since I'm going on year and a half old knowledge Apr 18 17:08:25 fromvega, you should really look at all the i2c drivers in the kernel ... tehre are quite a lot Apr 18 17:08:38 france_: hey :) Apr 18 17:08:55 and if you want something slightly different you can always write a new one Apr 18 17:09:03 pb_, is the hh.o cvs server entirely intact? Apr 18 17:09:51 fromvega, what are you trying to talk to out of curiosity? Apr 18 17:10:31 emte: some sensors, some A/D and digital potentiometers Apr 18 17:11:44 use a microcontroller and push the data to serial, parallel, usb, or wireless Apr 18 17:12:41 ggilbert: as far as I know Apr 18 17:12:55 k Apr 18 17:13:13 Stronly tempted to just wipe any cvs files from hh.o and let bb rebuild them Apr 18 17:13:16 emte: I think it's the best aproach... Apr 18 17:13:25 ggilbert: sounds reasonable to me Apr 18 17:14:24 there are a shed load of opie files and I don't have a clue which were actually shipped :p Apr 18 17:14:29 03coredump 07org.oe.oz354x * r36c7245d... 10/packages/udev/ (files/mount.sh udev-084/udev.rules udev_084.bb): Apr 18 17:14:29 udev: Add udev rule to auto-load the sd_mod module on demand, closing OE bug #811 Apr 18 17:14:29 Add basic auto-mounting support with automatic creation and deletion of mount-points Apr 18 17:14:32 emte: I was wondering, is there a way to connect anything to the motherboard I2C bus? I mean, any connector or something? Apr 18 17:14:34 03coredump 07org.oe.dev * ra2ab0d3f... 10/packages/udev/ (files/mount.sh udev-084/udev.rules udev_084.bb): Apr 18 17:14:34 udev: Add udev rule to auto-load the sd_mod module on demand, closing OE bug #811 Apr 18 17:14:34 Add basic auto-mounting support with automatic creation and deletion of mount-points Apr 18 17:14:53 ggilbert: Mail oe@handhelds.org what you are going to wipe Apr 18 17:15:05 ggilbert: then people can bb -cfetch from oesources Apr 18 17:15:22 fromvega, it would depend on your mobo .. i have a hunch that the "extra" temp/fan sensor blocks are probably i2c Apr 18 17:15:24 ggilbert: e.g. familiar is already mirroring sources at familiar.handhelds.org Apr 18 17:18:45 emte: tks! Apr 18 17:19:41 up to 14gig free, just clearing some of the temp files generated by bitbake Apr 18 17:20:11 fromvega if you are interested a PCI->i2c card does exist http://www.uxd.com/i2c.html#2 Apr 18 17:20:17 oh ... he is gone Apr 18 17:20:38 I think my logs for oesources.org go back a month or so, so I'll make sure to move aside any files that are actively being requested Apr 18 17:24:55 ggilbert, you running anything like mrtg? Apr 18 17:25:06 emte: nope Apr 18 17:25:15 that might graph/profile file Apr 18 17:25:24 :( Apr 18 17:34:51 ~lart me Apr 18 17:34:52 * ibot hurls dozens of incontinent, insomniac, hungry kittens with tiny little razor-sharp claws and a wide variety of contagious intestinal parasites at zecke Apr 18 17:58:57 the handhelds.org SSL certificate is pretty wonky :) Apr 18 18:29:53 wheee rm Apr 18 18:30:30 ? :P lol Apr 18 18:30:42 ... Apr 18 18:30:44 not you Apr 18 18:31:04 i know ;) Apr 18 18:44:17 zecke: hi Apr 18 18:44:23 hey Apr 18 18:51:48 hehe, made it 6% through the file list :p Apr 18 18:57:16 30% of the list is opie :p Apr 18 18:57:42 ggilbert: I think most of it can be removed Apr 18 18:57:46 ggilbert: we have all the data in cvs as well Apr 18 18:57:46 yeah Apr 18 18:58:00 zecke: I'm going to delete all the ones that aren't a specific version tag Apr 18 19:00:12 not even worrying about old fixed version stuff at this point :p Apr 18 19:06:05 * Philippe is back (gone 20:42:31) Apr 18 19:23:21 cya Apr 18 19:31:38 so I finished my list of files to delete, and it's 1.5mb of file paths :p Apr 18 19:37:33 crumbs Apr 18 19:37:45 actually that was the wrong file. 5.4mb Apr 18 19:37:48 sounds like some serious deleting Apr 18 19:55:32 that get's it to a very managable 4229 files and 4.6gig of space Apr 18 20:03:45 Hi what do you think from the following openobex_12.bb patch for the dev branch? Without it the configuration will faill for me to the complain that libusb.pc can not be checked in crosscompiler environment. http://pastebin.ca/49762 Apr 18 20:08:52 lamikr: are you working on the h6300 port? Apr 18 20:09:20 mreimer: Hi mreimer yes. Apr 18 20:10:11 lamikr: what do you think about shipping familiar 0.8.4 with udev 0.84? Apr 18 20:10:19 I have mostly spend my time with the kernel. Apr 18 20:10:20 the numbers match, so it must be right :-) Apr 18 20:10:59 do you have any objections? h2200 and hx4700 need 0.84, but I heard there was some problem with h6300. Apr 18 20:11:16 mreimer: I tested it about week ago (see bug 1525) and I had bad problems with the nfs mount over usb cable. Apr 18 20:12:00 The existing connection simply stopped to work after there came some hotplug agent messages in the console. Apr 18 20:12:11 lamikr: all nfs, or just nfs mount rootfs? Apr 18 20:13:29 I mounted the rootfs over nfs. The mount itself worked in the beginning but I think got broken after udev was started from the init.d. Apr 18 20:13:54 When I jumped back to older udev, everything worked again ok. Apr 18 20:15:16 mreimer: Are you familiar with udev or should I try to google if I can find similar kind of errors & solutions for others in the net? Apr 18 20:16:10 lamikr: I've spent a little time with udev, but I'm no expert. Google would probably be more helpful. All I could tell you about that problem would be to run udevmonitor and try to find out what it's doing Apr 18 20:16:30 lamikr: it would be good to get that fixed, but it doesn't sound like a serious enough problem to prevent shipping it in fam 0.8.4. If you agree, would you mind if I give reenoo the go-ahead to include udev 0.84 in the fam 0.8.4 release? Apr 18 20:16:54 ggilbert: heh, how much disk space was it using before? Apr 18 20:16:58 What does this mean? no buildable providers for bootstrap-image How can i solve this problem? Apr 18 20:17:08 pb, about 100gig Apr 18 20:17:53 goxboxlive: it means that either there are no .bb files providing "bootstrap-image", or they are all unbuildable Apr 18 20:18:39 pb: Thanks, i edited the machine.conf file and it's solved. Apr 18 20:18:51 mreimer: At least for me it is very serious proplem as it prevents the development by using desktop. fam083 is now otherwise pretty stable. How much I would have time to find the fix for this? Apr 18 20:19:56 lamikr: all you would have to do is "ipkg install udev_0.83.ipk Apr 18 20:20:08 lamikr: err, 0.70 Apr 18 20:20:12 or whatever works for you Apr 18 20:20:25 endusers aren't going to need to nfs mount root Apr 18 20:21:36 mreimer: Yes that's quite true. Would it be possible to install 070 by default for the h6300 image? Apr 18 20:22:20 yes, by adding PREFERRED_VERSION_udev = "070" to h6300.conf Apr 18 20:23:31 mreimer: Ok, then it is fine for me to upgrade to 084. I will try to test more how the 084 works with the SD/MMC cards and whether I could find a solution for the nfs problem. Apr 18 20:23:58 * chouimat is away: Sailing Away .... Apr 18 20:26:45 The mailing list appears to hate me, so I'll just post the list here Apr 18 20:26:55 http://treke.net/~ggilbert/delete.txt Apr 18 20:28:07 mreimer, lamikr: Your udev problem sounds like it could be the network.sh script being over zealous Apr 18 20:28:19 I'd be interested to know if removing that script helps Apr 18 20:28:52 RP: Let me check... Apr 18 20:31:18 ggilbert: That's some list! :) Apr 18 20:31:24 :) Apr 18 20:34:20 What is ipkg-linkudev, the task-bootstrap complains about it can not satidfy it's depedency Apr 18 20:36:52 The packace isent a *bb file either., i have googled for it, but no hits. Apr 18 20:37:47 goxboxlive: I suspect you have missed a space in an _append somewhere Apr 18 20:40:54 RP:Thx i'll look for it ( do u mean in a machine.conf file? ) Apr 18 20:41:12 afternoon Apr 18 20:41:38 goxboxlive: Probably. I suspect you have an _append = "udev ..." somewhere which should be an _append = " udev Apr 18 20:41:56 hi tmm1 Apr 18 20:42:36 RP: Thx i found it, it was in a module definition with an space before it. thx a lot. Apr 18 20:48:38 lamikr: ok Apr 18 20:58:19 ok, I am now booting with udev084 without /etc/udev/scripts/network.sh Apr 18 20:59:52 mreimer & pb: Without network.sh the boots seems to work with udev084 from the nfs rootfs. Apr 18 21:00:31 Hups, not pn but RP :-) Apr 18 21:00:53 lamikr: very good Apr 18 21:01:49 So it seems that the rootfs connection is lost because network.sh calls propably ifconfig down... Apr 18 21:03:35 mreimer: One problem I still have. For some reason /etc/modules is still not generated automatically from the files in /etc/modutils during the first boot. Apr 18 21:04:26 lamikr: do you have the right lines in h6300.conf, like: module_autoload_snd-pcm-oss = "snd-pcm-oss" Apr 18 21:06:36 I havent used bitbake in awhile now, and when i try it. I get a Traceback (most recent call last): Apr 18 21:06:36 File "/usr/bin/bitbake", line 26, in ? Apr 18 21:06:36 import bb Apr 18 21:06:36 ImportError: No module named bb Apr 18 21:06:43 no idea, what it might depend on Apr 18 21:07:03 try a newer version of bitbake, from svn Apr 18 21:07:14 sounds like its not installed, bb should be a part of bitbake Apr 18 21:07:42 oki, might have something to do with me upgrading the system last week Apr 18 21:09:57 anyone in here with upload access to oesources? We need a mirror of gpgme Apr 18 21:10:00 mreimer, lamikr: That network.sh script should really look at the status of a network device before fiddling with it... Apr 18 21:11:10 It would also be nice if bb/oe checked the sources dir for these files before fetching Apr 18 21:11:58 JustinP: It does? Apr 18 21:12:27 mreimer: I think so... h6300.conf has things like: module_autoload_omapts = "omapts" Apr 18 21:12:46 module_autoload_i2c-omap = "i2c-omap" Apr 18 21:13:41 RP: I haven't been able to fetch it for days.... Apr 18 21:14:34 JustinP: Check oesources.org before checking upstream? Apr 18 21:15:14 or do you mean checking the local file system for checking upstream? Apr 18 21:18:14 ggilbert: I mean checking the local DL_DIR before trying either the source or oesources Apr 18 21:18:27 lamikr: does update-modules do the right thing if you run it manually? Apr 18 21:18:28 hmm, it should do that already Apr 18 21:18:35 pb_: Yes Apr 18 21:18:42 ggilbert: yeah, it's always done that Apr 18 21:18:43 ggilbert: I found a mirror and downloaded it to my sources dir but it still went for the normal download... Apr 18 21:18:53 JustinP: It already does this. You probably didn't create the md5 file Apr 18 21:18:54 lamikr: in that case I guess you need to inspect the postinst script Apr 18 21:19:03 Probably wants the md5 file Apr 18 21:19:08 pb_: thanks for fixing apmd Apr 18 21:19:08 RP: ah, I didn't think of the md5, I see Apr 18 21:19:13 nm, then Apr 18 21:19:28 pb_: Is it working for others? Apr 18 21:19:57 lamikr: it works for me on h3900. I don't know of anybody else who's testing o h6300. Apr 18 21:20:02 That's the last thing tht's created for the file, so I believe do_fetch assumes that the file is incomplete if the md5 doesnt exist Apr 18 21:25:21 pb_: Do you know which script is supposed to call "update-modules"? Apr 18 21:27:22 lamikr: not offhand, but I can look it up Apr 18 21:27:26 postinst on all the kernel modules Apr 18 21:28:01 pb_: That would be nice Apr 18 21:28:29 tmm1: is that true at first boot? I had a vague feeling it was done differently there. Apr 18 21:29:11 oh, wait, no, I'm probably mixing it up with depmod Apr 18 21:29:28 yeah, I think it is indeed in the kernel module postinsts. Apr 18 21:32:31 pb_: I have for example a following post-inst script for the omapts. Apr 18 21:33:32 http://pastebin.ca/49780 for kernel-module-omapts-2.6.postings Apr 18 21:34:32 that looks fine Apr 18 21:34:36 does it actually get run at startup? Apr 18 21:35:05 oh, wait, no, it doesn't look fine Apr 18 21:35:30 Is the "2.6" thing in the end of the name incorrect? Apr 18 21:35:35 that script will run offline without doing the update-modules thing, so there will be nothing to cause the postinst to run again at boot time Apr 18 21:36:38 I think actually it's update-modules's own postinst script that does the business on first boot. Apr 18 21:36:54 Maybe that isn't happening for you for some reason. Do you see update-modules getting configured? Apr 18 21:37:31 the -2.6 thing is fine, by the way Apr 18 21:38:30 ok... update-modules.postints looks following: http://pastebin.ca/49782 Apr 18 21:38:59 I do not understand what if [ "x$D" != "x" ]; then checks. Apr 18 21:39:15 that's to make sure it does't run in offline mode during image construction Apr 18 21:39:30 $D will be set to a non-empty string at that point Apr 18 21:40:22 hmm i tried to bitbake mplayer-atty and it started to build quilt Apr 18 21:40:37 So is this postinst script supposed to be called during the image creation time or during the first boot? Apr 18 21:40:44 lamikr: both Apr 18 21:41:03 it should be called during image creation, return an error, then be called again during first boot and succeed at that point Apr 18 21:41:50 ok... Are h3900 images executed from the SD/MMC card or from the flash? Apr 18 21:42:15 flash Apr 18 21:42:54 In h6300 there is no support for the flash boot. Could that make the difference? Apr 18 21:43:05 I doubt it. The scripts should be just the same. Apr 18 21:43:28 If you watch the console output during initial boot, does it say "Configuring update-modules"? Apr 18 21:44:34 pb_: I will try again from fresh image. Apr 18 21:44:40 okay, good plan Apr 18 21:45:24 you could also inspect /usr/lib/ipkg/status in the image before booting. It should show update-modules as "Status: install ok unpacked". Apr 18 21:45:42 If it says "install ok installed", that means that ipkg thinks the postinst has already run. Apr 18 21:47:25 pb_; That may be a one step closer to solution. For omapts module I have for example "install ok installed" Apr 18 21:47:42 lamikr: that's ok for the modules Apr 18 21:47:48 they don't need reconfiguring at boot time Apr 18 21:48:48 Update modules, status: install ok unpacked Apr 18 21:49:32 that looks good Apr 18 21:49:54 I guess you need to watch during bootup now to see what happens Apr 18 21:50:19 Yes, I will disable the bootsplash. Apr 18 21:50:43 I have empty /etc/modules could that prevent it to be overwritten? Apr 18 21:52:02 trying to build gpe-image I get: ERROR: Cannot satisfy the following dependencies for task-bootstrap: kernel-image-2.6.16-hh0 Apr 18 21:52:23 but I see the kernel has been built: tmp/deploy/images/zImage-2.6.16-hh0+cvs20060418-h2200-20060418005901 Apr 18 21:52:46 does gpe-image not know how to deal with kernels built from cvs? Apr 18 21:53:03 lamikr: mm, not sure. the h3900 images don't seem to have that file, but I wouldn't expect it to really be a problem. Apr 18 21:53:51 oh, actually, yeah, it would be a problem. update-modules won't automatically overwrite a file that it didn't create. Apr 18 21:54:03 I guess you need to figure out where that file is coming from, and... stop it. Apr 18 21:54:09 i thought somewhere an ipkg flag was run Apr 18 21:54:16 so that postinst ran again on first boot Apr 18 21:54:23 yes, it is Apr 18 21:55:00 mreimer: that is rather odd. Apr 18 21:55:11 I've tried rebuilding task-bootstrap but that didn't help Apr 18 21:55:36 where's task-bootstrap getting "kernel-image-2.6.16-hh0" from? Apr 18 21:55:48 I guess it must be in one of your .conf files someplace Apr 18 21:55:56 good question; I'll look Apr 18 21:56:20 * pb_ go to bed now Apr 18 21:56:31 PREFERRED_PROVIDER_virtual/kernel = "handhelds-pxa-2.6" Apr 18 21:56:43 probably the _cvs.bb gets chosen Apr 18 21:56:51 yeah, that's fine Apr 18 21:56:54 how do i clean a package ? i want to rebuild it Apr 18 21:57:01 something went wrong Apr 18 21:57:09 you need to figure out where the entry in task-bootstrap's RDEPENDS is originating Apr 18 21:57:14 bitbake -i, then 'parse', then 'rebuild package' Apr 18 21:57:26 thanks pb_. sleep well Apr 18 21:57:27 bedtime now, really Apr 18 21:57:29 night night Apr 18 21:57:38 pb_; Good night Apr 18 21:58:27 mreimer: do you know if wpa for the h5550 is still unsolved ? Apr 18 21:58:35 CSMan: I don't know Apr 18 22:07:34 i'm trying to build mplayer and it starts to build quilt , is it trying to update everthing? because i already have that package and others built Apr 18 22:08:01 it's probably updating everything that's changed Apr 18 22:08:21 right, but after a while it fails Apr 18 22:09:20 koen|away: same problems with rdepends in http://logs.nslu2-linux.org/buildlogs/autobuild-nudi-last.txt Apr 18 22:09:44 ERROR: Nothing provides runtime dependency glibc-gconv Apr 18 22:09:44 NOTE: Apr 18 22:09:44 NOTE: no buildable providers for bogofilter Apr 18 22:10:01 ERROR: Nothing provides runtime dependency python Apr 18 22:10:01 NOTE: Apr 18 22:10:01 NOTE: no buildable providers for ipkg-utils Apr 18 22:10:18 Are these metadata bugs, or bitbake bugs? Apr 18 22:32:27 Hey guys Apr 18 22:32:45 do you know if it's worth anything to buy an high-speed SD card on a Zaurus (C-1000 or c)760) Apr 18 22:57:39 03xora 07org.oe.dev * rf52978a7... 10/packages/gnome/gnome-vfs-dbus_svn.bb: gnome-vfs-dbus_svn.bb : Add meaningful PV and make default Apr 19 00:09:47 That super-highspeed cache stuff is so damn cool Apr 19 00:12:55 Indeed. Apr 19 01:15:56 XorA|gone: how do i select mplayer to be built only with x11 and not with qt? Apr 19 02:41:36 what should I do to get some apps built with x11 instead of qt ? Apr 19 02:55:39 CSMan: what apps Apr 19 02:57:11 TheMasterMind1: mplayer-atty Apr 19 02:57:15 for example Apr 19 02:58:41 hrm Apr 19 02:58:51 well it depends on sdl so i think if you compile sdl for x11 itl Apr 19 02:58:53 l wrok Apr 19 02:59:11 libsdl-x11 instead of libsdl-qte Apr 19 02:59:50 how can i specify that when building mplayer-atty ? Apr 19 03:00:10 you don't Apr 19 03:00:23 but it's building with qte as default Apr 19 03:00:29 you have to build libsdl beforehand. set PREFERRED_PROVIDERS for virtual/sdl to libsdl-x11 Apr 19 03:00:50 right Apr 19 03:01:01 PREFERRED_PROVIDERS in local.conf, right ? Apr 19 03:03:30 ok, changed. i'll build libsdl Apr 19 03:04:04 virtual/libsdl that should be Apr 19 03:38:54 jnc: I don't suppos eyou have a copy of JavaScriptCore_cvs.sourceforge.net__20060212.tar.gz Apr 19 03:41:51 JustinP: err. nope Apr 19 03:42:12 jnc: didn't I check in the bbfile for that for you? Apr 19 03:42:18 yeah. :) Apr 19 03:43:00 :-P Apr 19 03:43:19 sf.net anoncvs isn't working for me for that package... Apr 19 03:43:33 ah, shame Apr 19 03:49:46 03mwester 07org.oe.dev * rdc7c705f... 10/packages/nslu2-binary-only/ (3 files in 2 dirs): unslung-rootfs: passwd patch, updated NOTES Apr 19 03:52:01 anyone else have that in their sources dir? Apr 19 03:59:20 javascriptcore? what uses that Apr 19 04:12:01 TheMasterMind1: minimo, i think Apr 19 05:42:28 moin Apr 19 06:07:24 evening Apr 19 07:05:35 HI there. I am trying to build a bootstrap-image and it wouldnt pass trhue. This is the only message i get ( ERROR: Nothing provides dependency boostrap-image) and i also had som ecomplains about ipkg-linkudev. Can someone helpme out here? Apr 19 07:09:02 oï ! Apr 19 07:12:00 goxboxlive: looks like a missing apce ipkg-link|udev Apr 19 07:12:23 goxboxlive: did you edit something or, is this default OE? Apr 19 07:14:15 It's not a default one, i am trying to build for the HTC-universal. It worked two days ago, but i had to move the workdirectory because of missing space. I have just adopted the conf file for blueangel. Apr 19 07:15:03 goxboxlive: smells like a wrong (BOOTSTRAP_)EXTRA_DEPENDS Apr 19 07:15:11 (missing space ' ') Apr 19 07:15:36 should it be spaces in every " bla bla bla " ? Apr 19 07:15:54 goxboxlive, : i've some pb with file with mod 000 for move them ( GCC visibilitypatch for example ) , how did you do ? Apr 19 07:15:54 goxboxlive: not every, but more spaces the better Apr 19 07:16:18 (only with OVERRIDES avoid spaces) Apr 19 07:17:03 Genesis: I dont follow, what do you mean? Apr 19 07:17:12 let it be ... Apr 19 07:17:32 hrmf, what installs stropts.h anyone have any ideas? Apr 19 07:17:35 goxboxlive: += adds a space Apr 19 07:18:26 ok thx Apr 19 07:23:41 Here is my htcuniversal.conf, can you see something wrong in it: Apr 19 07:23:44 #@TYPE: Machine Apr 19 07:23:45 #@NAME: htcuniversal Apr 19 07:23:45 #@DESCRIPTION: Machine configuration for the HTC Universal device Apr 19 07:23:45 KERNEL ?= "kernel26" Apr 19 07:23:45 OVERRIDES =. "${KERNEL}:" Apr 19 07:23:45 TARGET_ARCH = "arm" Apr 19 07:23:47 IPKG_ARCHS += "all arm armv4 armv4t armv5e armv5te ipaqpxa htcuniversal" Apr 19 07:23:49 PREFERRED_PROVIDER_xserver = "xserver-kdrive" Apr 19 07:23:51 PREFERRED_PROVIDER_virtual/kernel_kernel26 = "xanadux-un-2.6" Apr 19 07:23:52 goxboxlive: pastebin Apr 19 07:23:53 PREFERRED_PROVIDER_virtual/kernel = "xanadux-un-2.6" Apr 19 07:23:55 PREFERRED_PROVIDER_kernel = "xanadux-un-2.6" Apr 19 07:23:57 MODUTILS = "26" Apr 19 07:23:59 BOOTSTRAP_EXTRA_RDEPENDS += "kernel ipaq-boot-params ${@linux_module_packages('${UN_MODULES}', d)}" Apr 19 07:24:01 BOOTSTRAP_EXTRA_RDEPENDS_append_kernel26 += "udev module-init-tools" Apr 19 07:24:05 UN_MODULES_kernel26 += "htcuniversal_bt htcuniversal_leds snd-pxa-i2sound-htcuniversal htcuniversal_power htcuniversal_phone htcuniversal_wlan" Apr 19 07:24:08 SERIAL_CONSOLE = "115200 tts/2 vt100" Apr 19 07:24:10 USE_VT = "0" Apr 19 07:24:12 include conf/machine/include/handheld-common.conf Apr 19 07:24:14 GUI_MACHINE_CLASS = "smallscreen" Apr 19 07:24:16 ROOT_FLASH_SIZE = "32" Apr 19 07:24:18 #OPIE_EXTRA_INSTALL += "opie-embeddedkonsole opie-ftp opie-i18n-no libgaim0 libgaim-protocol-msn libgaim-protocol-oscar libgaim-protocol-irc" Apr 19 07:24:21 #module_autoload_htcuniversal_leds = "htcuniversal_leds" Apr 19 07:24:23 #module_autoload_htcuniversal_bt = "htcuniversal_bt" Apr 19 07:24:25 #module_autoload_htcuniversal_wlan = "htcuniversal_wlan" Apr 19 07:24:27 #module_autoload_snd-pxa-i2sound-htcuniversal = "snd-pxa-i2sound-htcuniversal" Apr 19 07:24:29 #module_autoload_snd-pcm-oss = "snd-pcm-oss" Apr 19 07:24:31 #module_autoload_snd-mixer-oss = "snd-mixer-oss" Apr 19 07:24:35 #module_autoload_bnep = "bnep" Apr 19 07:24:37 # optimize images, but build feeds without using tune-xscale so as to retain backwards compatibility Apr 19 07:24:39 include conf/machine/include/tune-xscale.conf Apr 19 07:25:04 lol Apr 19 07:25:05 ~pastebin Apr 19 07:25:07 somebody said pastebin was 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 19 07:25:12 goxboxlive: do not flood me Apr 19 07:25:35 goxboxlive: I'm glad to help, but I want to decide what information I consume ;) Apr 19 07:25:36 not only you ... Apr 19 07:25:37 :) Apr 19 07:25:57 lucky me, I'm not too selfish then Apr 19 07:26:37 use pastebin. Apr 19 07:27:05 ok i am so sorry, i'll paste it in the pastebin.com. Here is the link: http://pastebin.com/668797 Apr 19 07:27:42 BOOTSTRAP_EXTRA_RDEPENDS_append_kernel26 += "udev module-init-tools" Apr 19 07:27:49 try '" udev" Apr 19 07:28:01 touch conf/local.conf to reparse Apr 19 07:29:06 hmm it should add a space though :} Apr 19 07:30:50 ok i'll try, hold on. Apr 19 07:31:46 RP: bitbake gaim-plugin-foo would rock as well Apr 19 07:31:58 RP: building packages to get a RDEPENDS ;) Apr 19 07:32:33 cya later ;) Apr 19 07:34:52 Do i have to run 'bitbake -c clean bottstrap-image' before i try to build every time bootstrap-image failes? Apr 19 07:41:59 probably Apr 19 07:50:05 wow Apr 19 07:50:12 lfs bootscripts do *not* like busybox Apr 19 08:06:06 morning Apr 19 08:09:25 morning Apr 19 08:09:36 well one could argue this was the stupidest idea ever Apr 19 08:10:43 in case anyone was curious, lfs' bootscripts are completely incompatible with busybox Apr 19 08:23:06 whoa. kernel.inc creates .conf files using the aliases in the modules Apr 19 08:23:09 that's impressive Apr 19 09:30:44 I dont get it. I still have problem building the bootstrap-image. Now it says, maybe run 'ipkg update'. I dont have ipkg at my computer, i am using debian. Apr 19 09:31:03 goxboxlive: hmmm? Apr 19 09:40:21 morning al Apr 19 09:40:23 all Apr 19 09:52:55 This is was happening: http://pastebin.com/668909 Apr 19 09:56:31 and this is what 'cat /usr/src/universal//work/bootstrap-image-1.0-r0/temp/log.do_rootfs.11490' says. http://pastebin.com/668911 Apr 19 09:57:24 vim /usr/src/universal//deploy/ipk/Package Apr 19 09:57:53 grep for bootstrap **** ENDING LOGGING AT Wed Apr 19 09:59:56 2006