**** BEGIN LOGGING AT Thu Apr 13 09:59:57 2006 Apr 13 10:01:21 where are the pcmcia conf files now we have pcmciautils? Apr 13 10:01:40 anyone know where FILE is declared? Apr 13 10:05:05 TheMasterMind1: make sure they use the same distro, otherwise, paths to binaries might change Apr 13 10:05:35 hrm, probably won't work then Apr 13 10:05:36 darn Apr 13 10:06:13 I just know that going from slackware to ubuntu doesn't work Apr 13 10:09:09 :4: RuntimeWarning: tmpnam is a potential security risk to your program Apr 13 10:09:09 ERROR: Package 'kernel-module-hostap' exists but has no files Apr 13 10:09:09 ERROR: Error in executing: /home/sander/openembedded/org.openembedded.dev/packages/linux/linux-openzaurus_2.6.16.bb Apr 13 10:09:40 sws: I've broken the zaurus kernel .bb. I can push a fix for that... Apr 13 10:11:10 Bernardo: now card mapping is in kernel driver Apr 13 10:13:45 what header file includes the definition for FILE? Apr 13 10:14:42 TheMasterMind1: I went from Redhat to Gentoo Apr 13 10:14:49 mithro: stdio? Apr 13 10:14:50 mithro: stdio.h? Apr 13 10:14:58 TheMasterMind1: the main problem is libstdc++ changing ABI with every minor gcc release Apr 13 10:15:17 mithro: I thought it was a bitbake variable you were asking about the first time :) Apr 13 10:15:27 ahh okay :P Apr 13 10:15:51 hrw|work: so how do I tell a if I want to use the flash or the wifi part of a sandisk connect plus card? Apr 13 10:16:17 Bernardo: XorA disabled ide part Apr 13 10:17:52 RP: uncached with r412: http://oe.pastebin.com/657334 Apr 13 10:17:56 I know, it will only work with one or the other, so how do I enable the ide part and disable the wifi? with pcmcia it only needed editing a conf file Apr 13 10:18:38 Bernardo: ide is built into kernel, so you need to remove the -HACK.patch from kernel.bb and rebuuild Apr 13 10:18:47 koen: When doing that, did it still write out the cache files? Apr 13 10:18:58 yes Apr 13 10:19:13 Bernardo: you will also need to remove its ID from hostap modules in kernel if you are using pcmcia-utils, to udev doesnt load the modules Apr 13 10:19:17 RP: cached from /dev/shm/OE with r412: http://oe.pastebin.com/657338 Apr 13 10:19:19 koen: We should stop it doing that - save 400MB of disk IO Apr 13 10:19:33 XorA: so it isn't easy... :) Apr 13 10:19:48 Bernardo: as easy as deleting a line and adding a line Apr 13 10:19:52 koen@bitbake:~/OE/build$ df -h | grep shm Apr 13 10:19:52 tmpfs 1014M 431M 584M 43% /dev/shm Apr 13 10:20:09 koen: df -h /dev/shm works too ;) Apr 13 10:20:13 whats addtask in bitbake file do? Apr 13 10:20:15 XorA: I mean at runtime Apr 13 10:20:37 mithro: it adds a task to execute, e.g. do_configure Apr 13 10:20:49 XorA: one more reason for me to get colinux to work here... Apr 13 10:21:11 koen: Was the run in memory creating the cache or just running from it? Apr 13 10:21:19 RP: running from the cache Apr 13 10:21:31 Bernardo: There should be a way to use the bind entries in sysfs to force pcmcia to use a particular driver for a card Apr 13 10:21:34 with python doing 99% - 100% cpu Apr 13 10:21:58 RP: you could cut that time in half by using both cpus Apr 13 10:22:56 koen: I'm still favouring the stop writing 400MB of files approach :) Apr 13 10:23:10 Bernardo: yes, cant be dont runtime due to IDE being built in Apr 13 10:23:27 RP: we now know that the 1:05 time is cpubound Apr 13 10:23:30 RP: the best way is to get the damn card to work with both interfaces up at the same time Apr 13 10:23:59 RP: which means I do need to get colinux working... At home I only have ~30 minutes/day in front of a pc... Apr 13 10:24:18 koen: writing 400MB of data to memory still takes a while ;-) Apr 13 10:24:19 Bernardo: someone on the Sandisk Connect+ driver forum once said that both will work if kernel uses IO ports to communicate with card, at moment I beleive IDE uses MMIO Apr 13 10:24:43 RP: would be nice :) Apr 13 10:25:12 XorA: http://www.dsitri.de/phpBB2/viewforum.php?f=21 ? Apr 13 10:25:35 Bernardo: yeah Apr 13 10:26:32 RP: about 0.98 seconds on ewi :) Apr 13 10:26:41 419430400 bytes transferred in 0.979497 seconds (428209995 bytes/sec) Apr 13 10:27:18 XorA: I recall some stuff about true-ide versus regular ide from that forum Apr 13 10:27:30 RP: where will you push the fix for kernel.bb? vanille or ewi.utwente.nl Apr 13 10:27:31 03rpurdie 07org.oe.dev * r1d81433f... 10/packages/linux/linux-openzaurus.inc: linux-openzaurus: It appears we can't set runtime variables for dynamic packages (or at least not include their names in PACKAGES). Change to work around this for now Apr 13 10:27:43 RP: now trying r423 Apr 13 10:28:07 sws: Its gone to vanille Apr 13 10:28:16 thanks Apr 13 10:28:59 RP: fwiw, the disks on ewi seem to be fast enough to keep up with the cache monstrosity Apr 13 10:29:20 koen: mine aren't ;-) Apr 13 10:29:31 on my home machine it crawls Apr 13 10:29:52 * koen tests r423 Apr 13 10:30:05 It must admit it does fly on my Xeon box (with all its ram) Apr 13 10:30:39 * Philippe is back (gone 12:12:49) Apr 13 10:32:02 RP: with r423: http://oe.pastebin.com/657352 Apr 13 10:32:14 first one is uncached, second one reads cache from /dev/shm Apr 13 10:33:04 koen: So zecke has speeded up the cpu bound part by 33% Apr 13 10:33:12 koen: not bad :) Apr 13 10:33:15 it seems that way Apr 13 10:33:25 i'll bblr Apr 13 10:33:27 koen: We just need to sort the disk bound part Apr 13 10:33:47 and if mithro can get some more cpu time back, that will be ideal :) Apr 13 10:33:53 mithro: bfn Apr 13 10:36:17 bfn? Apr 13 10:36:48 mithro: bye for now Apr 13 10:38:50 can FILES_DIR have multiple overrides? Apr 13 10:39:03 like files/linuxeabi/c7x0 ? Apr 13 10:39:13 or files/c7x0/linuxeabi Apr 13 10:39:53 FILEPATH is for this Apr 13 10:40:20 How to i fix this? How do i get it to compile? NOTE: package xanadux-ba-2.6-2.6.12-r0: task do_compile: failed Apr 13 10:40:27 RP: thanks for the fix, kernel2.6 compiles again :) Apr 13 10:41:21 is there a way to reset a card with pcmciautils after suspending ? Apr 13 10:42:26 tell zecke to ping me if you see him Apr 13 10:42:49 Bernardo: eject/insert Apr 13 10:42:58 ibot: botmail for zecke: 12:50 < mithro> tell zecke to ping me if you see him Apr 13 10:43:08 Bernardo: you can do the commands, dont actually need to remove the card Apr 13 10:44:27 XorA: ok, time to see where will I put them as the sandisk needs the firmware reloaded after a suspend Apr 13 10:44:51 Bernardo: RP was talking about that yesterday Apr 13 10:45:00 ah, didn't see it Apr 13 10:45:09 Bernardo: Currently, what is handling the firmware upload for that card? Apr 13 10:45:19 Bernardo: mickey|bbl had problems with that if I recall, dont know what outcome was Apr 13 10:45:35 The kernel now has a nice mechanism for it but I don't know if that card uses the kernel mechanism Apr 13 10:46:12 XorA, Bernardo: The problem mickeyl was having looked like a card crashing upon suspend so it wasn't firmware related that I could see Apr 13 10:46:14 RP: I think it's hostap_fw_load, and not the kernel mechanism Apr 13 10:46:46 Bernardo: hostap doesn't appear to use the standard kernel mechanism, no :-/ Apr 13 10:47:09 ~lart hostap non-standard fw loading Apr 13 10:47:09 * ibot frags hostap non-standard fw loading with his BFG9000 Apr 13 10:47:14 lets hope that that ieee80211 kernel stuff catches on Apr 13 10:47:45 I should really write/submit a patch to convert it... Apr 13 10:47:52 there are a couple alignment traps on card insertion Apr 13 10:48:35 and a few more after resuming Apr 13 10:49:00 Bernardo: I saw them in mickeyl's log. Are there any other interesting messages from hostap? Apr 13 10:49:12 hostap_cs shutdown fails at suspend, so the card is probably left in a unstable state Apr 13 10:49:24 Bernardo: This is the same error mickey; had Apr 13 10:49:57 Bernardo: Its not that the firmware isn't loaded - its that the shutdown fails Apr 13 10:50:15 yes, looks like it Apr 13 10:50:37 a eject followed by a insert resets the card and hostap_cs, so it works again Apr 13 10:51:05 the strange part is that the screen flickers after the "pccardctl insert" command Apr 13 10:51:16 Its a mainline kernel bug so perhaps we should report it... Apr 13 10:51:35 Bernardo: The power drain from the card probably upsets the backlight power supply Apr 13 10:52:40 * Philippe is away: visual contact - melancholic dreams Apr 13 10:52:40 well, time to go now, see you guys later Apr 13 10:52:53 thanks RP Apr 13 10:53:29 Bernardo: np. bfn :) Apr 13 10:59:45 | Applying patch patches/usb-owner.patch Apr 13 10:59:45 | patching file src/prism2/driver/prism2_usb.c Apr 13 10:59:45 | Hunk #1 FAILED at 313. Apr 13 10:59:46 | 1 out of 1 hunk FAILED -- rejects in file src/prism2/driver/prism2_usb.c Apr 13 10:59:46 | Patch patches/usb-owner.patch does not apply (enforce with -f) Apr 13 10:59:57 is this patch already needed? Apr 13 11:01:28 sws: that patch is for umbaumba 17.3? Apr 13 11:03:43 hrw|work: no 5.6 Apr 13 11:04:02 NOTE: package wlan-ng-modules-0.2.3+svn20060413-r0: task do_patch: failed Apr 13 11:08:04 sws: which MACHINE? Apr 13 11:09:22 What was the name of the variable which contains the path to the linux headers files? Apr 13 11:09:27 STAGING_DIR? Apr 13 11:10:03 STAGING_INCDIR Apr 13 11:10:48 thanks Apr 13 11:10:48 hrw|work: tosa Apr 13 11:12:15 sws: 2.4 kernel use wlan-ng 0.2.1pre Apr 13 11:12:24 sws: 0.2.3-svn are for 2.6 kernel Apr 13 11:13:33 with STAGING_INCDIR I get: Apr 13 11:13:34 configure: error: No Linux kernel tree in /var/tmp/bitbake/staging/i686-oe-linux/include Apr 13 11:13:45 is there another variable I could try? Apr 13 11:14:45 Cobelius: kernel headers you want Apr 13 11:14:51 ? Apr 13 11:14:59 Cobelius: STAGING_KERNEL_DIR iirc Apr 13 11:15:07 okay I try that one Apr 13 11:15:37 inherit module-base iirc Apr 13 11:16:01 i try to build a bootstrap-image with ZKERNEL=2.6 for tosa, I looked at the patch and the to be patched .c file, the bug in prism2_usb.c is already corrected in the svn trunk Apr 13 11:16:31 sws: then remove that patch locally Apr 13 11:16:39 Do I need 'inherit module-base' or is this variable also visible if I don't inherit my .bb file from this class? Apr 13 11:17:00 | configure: error: Unable to retrieve compilation flags for kernel modules out of /var/tmp/bitbake/staging/i686-oe-linux/kernel/Makefile Apr 13 11:17:02 hmn Apr 13 11:17:59 Cobelius: iirc all kernel stuff needs to inherit module-base Apr 13 11:18:03 (but I could be wrong) Apr 13 11:18:30 I try it Apr 13 11:18:58 the variable is visible an has a value but he can't process the makefile of the kernel Apr 13 11:19:36 are you sure that the staging is working correctly? I had similar problems by trying to insert driver modules for the virtual hardware of vmware Apr 13 11:20:37 hmn Apr 13 11:20:40 same error again Apr 13 11:20:47 | configure: error: Unable to retrieve compilation flags for kernel modules out of /var/tmp/bitbake/staging/i686-oe-linux/kernel/Makefile Apr 13 11:21:59 hmn Apr 13 11:22:08 this line is above the error: Apr 13 11:22:08 | checking for module compilation flags... make: *** No rule to make target `i686-oe-linux-gcc'. Stop. Apr 13 11:22:09 hrw|work: i commented out the patch, now its build Apr 13 11:24:20 does somebody have some experiences with rtai an cross compilation? Apr 13 11:24:34 when I compile it outside of bitbake with the compiler on the machine is runs Apr 13 11:26:38 jnc: ping Apr 13 11:26:44 ~seen jnc Apr 13 11:26:48 jnc is currently on #oe (4h 47m 27s), last said: 'darmou: please do some testing to discover the cause of your C3100's trouble and let me know'. Apr 13 11:29:52 FYI, will talk to people working in porting Linux to embedded systems next week : http://www-inf.int-evry.fr/~olberger/weblog/2006/03/27/accepted-papers-for-international-track-of-wfs-2006-in-porto-alegre/ Apr 13 11:30:08 if some of you are in Porto Alegre, I'll be glad to say Hi Apr 13 11:30:50 obergix[work]: its a bit far to swim Apr 13 11:31:17 XorA: sure... and not directly by the sea ;) Apr 13 11:37:12 does no on have an idea how can i fix this error?:| checking for module compilation flags... make: *** No rule to make target `i686-oe-linux-gcc'. Stop. Apr 13 11:41:26 RP: ping Apr 13 11:41:36 Cobelius: looks like the makefile is broken Apr 13 11:42:00 Cobelius: a 'target' should be something like 'all' or 'rtai', not the compiler version Apr 13 11:42:07 (at least on most packages) Apr 13 11:43:30 hrw|work: pong Apr 13 11:44:26 RP: how goes poodle thing? Apr 13 11:44:51 hrw|work: The one John Lenz had is on its way to me Apr 13 11:46:23 RP: that guy which was on #gpe does not respond so we forget it - right? Apr 13 11:46:30 forget/ignore Apr 13 11:46:52 hrw|work: correct - never got a response back Apr 13 11:46:57 ok Apr 13 11:47:29 RP: I'm asking because users from oesf got enough cash to give us poodle Apr 13 11:48:46 hrw|work: ok. I should be getting this device sometime next week. I guess we need to decide who has time to work on the user support side of poodle Apr 13 11:49:35 RP: jnc offerred help Apr 13 11:50:15 hmm... Apr 13 12:02:55 RP: one off the OH hackers wouldn't happen to have a .bb for gnome-power-manager? Apr 13 12:03:50 koen: There was plans to make one but nobody ever did as far as I know Apr 13 12:04:07 03koen 07org.oe.dev * r4c9177ab... 10/packages/gnome/orbit2_2.14.0.bb: orbit2: disable parallel make for the non-native version as well Apr 13 12:04:09 RP: ok Apr 13 12:04:55 * koen bitbake gnome-power-manager Apr 13 12:09:27 prelink makes things like X and GPE load a lot faster :) Apr 13 12:12:03 nice Apr 13 12:15:22 what's the downside? Apr 13 12:18:16 I did build a bootstrap-image, what do I need to connect to my zaurus using the usb net interface? Apr 13 12:19:17 sws: kernel 2.6? modprobe g_ether Apr 13 12:24:20 RP: tried tskeys on c7x0? Apr 13 12:24:44 hrw|work: thanks, interface usb0 on tosa is up and running, but my host does not recognize it after i put it in the cradle Apr 13 12:24:56 sws: modprobe usbnet on desktop Apr 13 12:26:31 usbnet is already loaded Apr 13 12:26:59 another usb device connected to the same usbport is recognized Apr 13 12:27:06 sws: lsusb? Apr 13 12:27:12 sws: Bus 002 Device 032: ID 0525:a4a2 Netchip Technology, Inc. Linux-USB Ethernet/RNDIS Gadget Apr 13 12:28:24 lsusb does not show the Z Apr 13 12:29:05 sws: then no ideas - my c7x0 is visible Apr 13 12:29:27 thanks Apr 13 12:29:48 sws: you're using a 2.6.x kernel on a tosa? Apr 13 12:30:52 lardman: yes Apr 13 12:31:03 sws: just checking, cool Apr 13 12:32:58 do I need another module, that my Z is recognized by the cradle? Apr 13 12:34:15 I've not tried connecting my Z to a Linux box since I've been running a 2.6 kernel so I can't help I'm afraid Apr 13 12:35:52 03hrw 07org.oe.oz354x * re213842b... 10/packages/gpe-windowlist/gpe-windowlist_0.1.bb: added gpe-windowlist 0.1 from .dev Apr 13 12:35:55 03hrw 07org.oe.oz354x * r77e69361... 10/packages/keymaps/ (files/c7x0/keymap-2.6.map keymaps_1.0.bb): keymaps: added LCD Softkeys for c7x0 Apr 13 12:41:04 hrw|work: I haven't tried tskeys for a while on c7x0 Apr 13 12:41:29 RP: updated keymaps get them working Apr 13 12:41:47 hrw|work: Good, that's not too bad then :) Apr 13 12:42:01 koen: The disadvantage of prelink is it takes a short while to run its postinst and the binaries get slightly bigger Apr 13 12:42:26 koen: It can also break binaries but you can undo prelinking so its easy to test/indentify Apr 13 12:44:49 RP: good thing angstrom binaries are slightly smaller due to -Os :) Apr 13 12:46:00 koen: Its not a big increase Apr 13 12:46:14 RP: does that mean people can do 'ipkg install prelink' and enjoy a faster system? Apr 13 12:47:14 RP: if I recall, prelinking need to re-run every ipkg upgrade as well if libs were involved Apr 13 12:47:48 koen: correct Apr 13 12:48:09 XorA is also correct - you need to rerun it if you change libs Apr 13 12:48:24 * XorA remebers his KDE pre-linked system Apr 13 12:48:30 It will detect if libs have changed and binaries will still run just without the speed increase Apr 13 12:48:59 what about running it on boot as well? Apr 13 12:49:55 It would be better just have ipkg run it after installing packages or hook it into ldconfig somehow Apr 13 12:50:31 bye Apr 13 12:51:14 RP: does ipkg support suchs hooks (or rpm and other packages OE supports)? Apr 13 13:02:17 koen: I'm not sure. Maybe not. We do know about running ldconfig though so perhaps we can handle prelink in a similar way? Apr 13 13:02:30 yeah Apr 13 13:03:06 add DISTRO_USE_PRELINK=1 and then add running prelink to lib* postinst/postrm like it is done with ldconfig? Apr 13 13:03:44 technically it would be needed for anything that installs a binary Apr 13 13:04:16 and running it on boot can't hurt Apr 13 13:04:17 The other issue I'm aware of is elfutils needs 0.89 for gcc3 and 0.108 for gcc4 Apr 13 13:04:36 koen: It does rescan a lot of things so it will hurt boot time slightly Apr 13 13:04:36 0.108 will not work with gcc3? Apr 13 13:05:32 hrw|work: I haven't pinned down exactly what's wrong with it 0.108 might only work with EABI Apr 13 13:05:39 if I say 'ifconfig usb0 192.168.6.127 netmask 255.255.255.0 up' , dmesg shows ADDRCONF(NETDEV_UP): usb0: link is not ready Apr 13 13:06:36 sws: Not all of tosa 2.6 is functional yet - i don't know which bits are/arent though Apr 13 13:06:39 hrw|work: http://ewi546.ewi.utwente.nl/angstrom/feed-manager.txt Apr 13 13:06:59 hrw|work: any more requirements? Apr 13 13:07:41 koen: "allows users to download sources via a webfrontend" maybe? Apr 13 13:08:30 that's a good one Apr 13 13:08:42 btw - how to get A from Angstrom from keyboard? Apr 13 13:09:19 Better still, why don't we have OE generate source tarball packages (including our patches) Apr 13 13:09:29 option-shift-a on my mac Apr 13 13:09:34 ehm Apr 13 13:09:40 command-shift-a Apr 13 13:09:51 and get the -dev and -doc packages working for that matter :-/ Apr 13 13:09:55 RP: that's the easiest way Apr 13 13:10:08 RP: but sourcepkg.bbclass needs some improvements Apr 13 13:15:10 A Apr 13 13:15:44 RP: angstrom having it's own source mirror like oesources would be a good start Apr 13 13:16:48 koen: Yes. I think oesources might even be preapared to help. What treke needs is a list of all the sources that were used in old OZ releases (and familiar?) so its known which verions need to be kept Apr 13 13:17:17 Particularly CVS versions of things that were released Apr 13 13:18:06 RP: we have classes for it Apr 13 13:18:24 hrw|work: We have now, but did we have then? Apr 13 13:19:09 ? Apr 13 13:19:36 hrw|work: We really need the info about what was in lder OZ releases as oesources is the source mirror for them Apr 13 13:19:45 s/lder/older/ Apr 13 13:19:48 I know Apr 13 13:20:17 wouldnt it be easier just to remove old releases and start now? Apr 13 13:20:21 RP: familiar has all the sources for it's older releases (except 0.8.1, which is not being distributed anymore) Apr 13 13:20:53 koen: ok, that's something less to worry about. That leaves the old OZ releases Apr 13 13:21:46 OT: I like the bugzilla look gnome has: http://bugzilla.gnome.org/show_bug.cgi?id=338341 Apr 13 13:33:18 morning Apr 13 13:35:05 hey chouimat Apr 13 13:35:26 koen: posted reply Apr 13 15:04:56 hi Apr 13 15:07:34 hi mickeyl|writing Apr 13 15:07:57 mickeyl|writing: we need to decide what to do with poodle Apr 13 15:08:12 which poodle? Apr 13 15:08:13 http://www.oesf.org/forums/index.php?showtopic=17632&st=40 - users collected money to get one Apr 13 15:08:23 users poodle Apr 13 15:08:40 mickeyl|writing: RP will get the one from John Lenz Apr 13 15:08:50 right. since RP got the other one, lets send it first to coredump, you, or me. Apr 13 15:09:34 I cannot work on it. maybe will get spitz soon so 4 Z at home will make Ania nothappy Apr 13 15:10:54 mickeyl|writing: so it looks like you and coredump left Apr 13 15:11:28 k Apr 13 15:11:32 * mickeyl|writing points to CoreDump|afk Apr 13 15:13:48 hi mickeyl|writing Apr 13 15:15:08 mickeyl|writing: no messages for you at freenode network? (not identified) Apr 13 15:15:19 hey mickeyl|writing Apr 13 15:16:30 hey phil and koen Apr 13 15:16:41 hrw|work: /msg should work Apr 13 15:18:01 mickeyl|writing: I msged you few minutes ago Apr 13 15:18:07 hrw|work: and i replied :D Apr 13 15:18:12 hrw|work: you're getting nothing? Apr 13 15:18:42 now? Apr 13 15:18:58 got Apr 13 15:22:18 wow Apr 13 15:22:26 * koen is impressed Apr 13 15:22:44 drw: nice concept Apr 13 15:22:53 ? Apr 13 15:23:22 mickeyl: http://draknor.net/projects/parcel/repository.php + http://handhelds.org/hypermail/angstrom-dev/current/0049.html Apr 13 15:24:08 koen: thanks, now if I only would get everything I visualize actually written... Apr 13 15:25:01 CoreDump|home: hi Apr 13 15:25:04 hi Apr 13 15:26:05 * chouimat is away: lunch Apr 13 15:26:26 ~lart users Apr 13 15:26:26 * ibot decapitates users conan the destroyer style Apr 13 15:26:43 jesus, how stupid can one single person be grrrrrr Apr 13 15:26:51 CoreDump|home: asked mickeyl - now your turn.. we need to decide what to do with poodle Apr 13 15:27:06 yep I read it Apr 13 15:28:05 CoreDump|home: I will reply in thread with my opinion (as users ask me) and then let you both finish. Apr 13 15:28:21 fire away Apr 13 15:28:34 more people to join my 'death to pxa250' club Apr 13 15:29:30 What did the poor pxa250 do? Apr 13 15:30:00 RP: the better question is "what doesn't it do" Apr 13 15:30:09 well, with HRW covering the kernel stuff, we are about 95% set ;) Apr 13 15:30:18 hehe Apr 13 15:30:50 what kernel stuff? Apr 13 15:30:56 which kernel stuff? Apr 13 15:31:11 Is "--eraseblock option of mkfs.jffs2" an important option for nor flash ? Apr 13 15:31:13 all the kernel stuff of course :D Apr 13 15:31:24 * RP is also confused Apr 13 15:31:25 well, all the resume probs people where mentioning for starters Apr 13 15:31:33 Genesis: yes Apr 13 15:31:36 yeah. and the wlan stuff Apr 13 15:31:41 righto Apr 13 15:31:43 yep Apr 13 15:31:46 hehe Apr 13 15:31:50 I even got WE19 on my collie Apr 13 15:32:02 there is no WE19 for 2.4 kernels AT ALL Apr 13 15:32:10 btw. anyone looked into the new WE system yet? I'm pleased that there is an attempt to renovate that thing Apr 13 15:32:25 you mean WE21 experimental? Apr 13 15:32:40 is that it? Apr 13 15:32:49 dont know Apr 13 15:32:53 I do not trace it Apr 13 15:33:02 * mickeyl out of the loop as well *sigh* Apr 13 15:33:26 mickeyl: Did you see my comments on your hostap problem? Apr 13 15:33:59 RP: yep. could you report that bug ? Apr 13 15:34:25 I will look into firmware loading - IIRC hostap can use the kernel hotplug mechanism, but I didn't use it since we had so many 2.4 machines Apr 13 15:35:05 drw: parcel need to be online to get info about installed packages or not? do I need to be online to check what packages are available or can I fetch list to device? Apr 13 15:35:17 mickeyl: ok, I'll see what I can do. I'm not very familiar with the wireless world though and should it go to the hostap or mainline kernel people? Apr 13 15:35:18 die 2.4! die! Apr 13 15:35:28 ~lart users Apr 13 15:35:28 * ibot offers users some herring Apr 13 15:35:37 RP: atleast CC hostap/netdev Apr 13 15:35:39 ~users Apr 13 15:35:41 so this is what we get for sharing our unpaid volunteer work with you? complaints and accuses? well done, this clearly supports our motivation to continue working on open source projects. Apr 13 15:35:49 RP: that should go via hostap. thanks Apr 13 15:35:51 mickeyl: With orinoco, I have a 2.6 firmware uploading package Apr 13 15:36:05 OT: Openzaurus.org looks better on vga devices now - css changed a bit Apr 13 15:36:11 mickeyl: (spectrum-fw iirc) Apr 13 15:37:31 someone know when I will find time to resolder 9-9 nullmodem into 9-25? I have zaurus<>25 cable ready for weeks... Apr 13 15:37:54 RP: cool. i'll take a look Apr 13 15:38:21 bbl, going to a comedy play Apr 13 15:42:07 drw: can parcel handle 'install package(s) from cf without net connection' ? Apr 13 15:46:08 koen: the idea would be to handle simple install/removal of a package if its requirements are met Apr 13 15:47:09 drw: or a static parceld + sqlite db on the cf :) Apr 13 15:47:44 koen: right Apr 13 15:48:05 hrw|work, why not buy an adapter ? Apr 13 15:48:20 my idea is to do as much of the dependency stuff on the server as possible Apr 13 15:48:28 emte: have to go to city etc.. Apr 13 15:48:35 ah Apr 13 15:49:23 i'd mail you ... but your in poland Apr 13 15:49:40 emte: maybe one day I will buy it Apr 13 15:49:53 the shipping would far exceed the value and it would take weeks Apr 13 15:50:03 yep and days Apr 13 15:50:10 drw: make it use XMPP so I can IM new packages ;) Apr 13 15:50:13 or weeks from NZ... Apr 13 15:50:26 koen: so if you say 'install a', the server determines that 'a' depends on 'b', and would send package 'b' to the client first to install, then package 'b' - greatly reduces the work the client does Apr 13 15:50:49 koen: heh, that would be interesting... Apr 13 15:51:22 drw: but with server-client how to check offline what is available? Apr 13 15:52:13 hrw|work: yeah, right now it kind of depends on an internet connection, but maybe could cache metadata on the device if wanted Apr 13 15:52:52 drw: with parceld in place it wouldn't be very hard to write a 'replicator' to make mirrors of the feeds Apr 13 15:53:29 hrw|work, use some type of hash to reduce char and compress the "index" cache :) Apr 13 15:53:40 koen: right, it just stores package names in the DB, but could have those pulled from different locations Apr 13 15:53:51 I think an ipkg rewrite which could talk more effeicently to this online database might be the more desirable option... Apr 13 15:54:14 emte: we already compress index by removing Source: info Apr 13 15:55:02 hrw|work: to go back to what you were saying, devices are becoming more connected, so my assumption is that it is becoming less likely they will not have some kind of internet connection Apr 13 15:57:10 I rarely install from storage, but I'm not a typical user Apr 13 15:58:11 On the other hand I always do Apr 13 15:58:16 what would be a really cool tool, would be an active sync style tool where installs are driven from a desktop Apr 13 15:58:18 koen: same here Apr 13 15:58:25 hey mallum Apr 13 15:58:55 drw: I'm now adding new package to few feeds - its just cp and run one shell script to recreate Packages* - parcel-upload-tool would be nice Apr 13 15:59:57 This idea of a server side database is ideal for the distributor side of things but I can't see how its going to scale when you 100 users trying to download/upgrade to a new release Apr 13 16:00:17 hey koen Apr 13 16:00:30 200mhz pda vs 2GHz server Apr 13 16:00:55 hrw|work: yes, could be done with a web fornt-end Apr 13 16:01:11 RP: hence my note about a desktop tool Apr 13 16:01:28 XorA: abiword 2.4.4 is out ;) Apr 13 16:01:38 koen: Im on it :-) Apr 13 16:01:40 RP: Could be an issue, but it is easier to throw hardware at a server Apr 13 16:01:53 is burst flash = NOR flash ? Apr 13 16:02:10 ( i think i found the good datasheet ) Apr 13 16:02:21 XorA: perl doesn't build with eabi (yet), so I gave up on adding it Apr 13 16:02:21 but they don't speak of nor flash Apr 13 16:02:31 The problem is ipkg and its use of the Packages list. CHange that to a local database and I suspect a lot of the problems go away Apr 13 16:02:45 koen: where? its not on their site BTW Apr 13 16:03:08 koen: ah, its just not annouced :-) Apr 13 16:03:15 koen: That's a factor of 10 speed improvement. 100 users would break it :-/ Apr 13 16:03:38 XorA: http://uwog.net/blog/?p=32 Apr 13 16:03:47 koen: dont have time tonight, Ill try to add it tomorrow if you havent Apr 13 16:03:55 XorA: thanks Apr 13 16:04:16 RP: the problem is, the current Packages list for OZ/Familiar is a couple megs in size, so even if it is in a DB, you still have to download and store all of this info on the device Apr 13 16:04:35 RP: looking at the apache logs OZ seems to get ~10 concurrent users at most Apr 13 16:04:45 03hrw 07org.oe.oz354x * rcc80d868... 10/packages/matchbox-wm/ (matchbox-wm/c7x0/kbdconfig matchbox-wm_1.0.bb): matchbox-wm: added c7x0 kbdconfig Apr 13 16:04:54 anyway, it's just an idea I had and wanted to play around with to see if it is practical Apr 13 16:04:55 cu Apr 13 16:04:59 hrw|work: cu Apr 13 16:05:02 hrw|work: bye Apr 13 16:05:02 cu hrw|work Apr 13 16:05:14 CoreDump|home: send post in poodle thread Apr 13 16:05:19 drw: So we find a new way to distribute the packages data - the Packages file has always struck me as wrong... Apr 13 16:05:23 RP: but I can see the problem we we get thousands of users :) Apr 13 16:05:36 drw: I'm certain it can be stored in a more space friendly way Apr 13 16:05:37 s/we we/when we/ Apr 13 16:05:55 hrw|gone: I just did Apr 13 16:06:01 CoreDump|home: just read(ed) Apr 13 16:06:02 cu Apr 13 16:06:07 ;) Apr 13 16:09:53 has anyone tried the slack embedded system? Apr 13 16:10:34 found it when i was looking at scratchbox as an option for dropin web compiling for PIC stuff Apr 13 16:11:11 scratchbox, not embedded slack Apr 13 16:13:48 * emte takes that as a no Apr 13 16:15:03 bye Apr 13 16:47:23 aehem Apr 13 16:47:59 did anyone mess w/ collie jffs2 generation in .oz354x? Apr 13 16:48:30 CoreDump|home: 'monotone log' ? Apr 13 16:48:41 monotone log conf/machine/collie.conf Apr 13 16:49:28 trying that, thanks Apr 13 16:50:43 last commit was from the oz354fam083 branch which produced working images Apr 13 16:50:48 emte: which slack embedded system you referring to? Apr 13 16:50:57 curent immages can not be flashed Apr 13 17:24:14 Hello Apr 13 17:30:23 hey Ecco_ Apr 13 17:30:34 Hello Koen :-) Apr 13 17:50:36 hi Apr 13 17:53:53 hey hrw|tv Apr 13 17:55:08 wb hrw|tv Apr 13 17:56:51 we need to update wpa-supplicant to 0.4.8 Apr 13 17:57:35 or 0.5.1/gnutls Apr 13 17:58:26 heh Apr 13 17:58:28 good luck Apr 13 17:59:01 isn't wpa-supplicant on the 'touch it and see how a few days worth of work gets reverted' list? Apr 13 17:59:22 will look Apr 13 18:01:54 03coredump 07org.oe.dev * rbb4cd5cb... 10/packages/udev/ (8 files in 6 dirs): udev: Add changes from .oz354x Apr 13 18:01:59 03coredump 07org.oe.dev * rc0ab6801... 10/packages/matchbox-wm/ (matchbox-wm/c7x0/kbdconfig matchbox-wm_1.0.bb): matchbox-wm: Add c7x0 kbdconfig from .oz354x Apr 13 18:02:03 03coredump 07org.oe.dev * r78129b5b... 10/packages/xserver-common/ (4 files in 2 dirs): xserver-common: Add 200DPI fix and zaurusd/calibrate fix from .oz354x Apr 13 18:05:06 koen: the problem with that stuff is that it left gnutls in a horrible mess Apr 13 18:05:28 koen: as expected it still is in that sorry state Apr 13 18:06:14 koen: once broken stuff enters the repository it is likely to stay Apr 13 18:07:16 how unexpected Apr 13 18:07:22 "revert, not fix" Apr 13 18:07:29 and behold, it stays broken Apr 13 18:07:56 while killing the last bits of motivation people had Apr 13 18:08:23 afk Apr 13 18:11:12 it's hardly realistic for people to expect to be able to just dump broken patches into the repository and have someone else fix the problems. Apr 13 18:11:48 I didn't say that Apr 13 18:12:01 pb_: right. I asked jnc to fix things up and then push it again. Apr 13 18:12:22 koen: well, you said "revert, not fix", which kind of implied that you felt reenoo ought to have been fixing it. Apr 13 18:12:37 It's realistic to persuade the person to fix it themselves Apr 13 18:13:01 instead of scaring them away by reverting stuff blindly Apr 13 18:13:24 I'm not sure what you mean by "blindly". I think Rene was quite aware of what changes he was reverting. Apr 13 18:14:05 looking at oe@ it seems he wasn't Apr 13 18:14:10 as for scaring people away, yes, the presentation could have been better. But I thought at the time, and still think, that the basic principle of reverting broken changesets is and was a good one. Apr 13 18:14:50 reverting isn't bad per se Apr 13 18:15:18 the complete lack of communication and the collateral damage made it bad Apr 13 18:15:38 * Philippe is back (gone 07:23:01) Apr 13 18:15:50 hey Philippe Apr 13 18:16:05 hey koen Apr 13 18:22:39 basically gnutls_1.3.5 can be renamed to gnutls13_1.3.5 (other soname) like we did in past with libnet iirc Apr 13 18:25:49 hrw|tv: sounds like a plan, yeah Apr 13 18:27:17 wpa_supplicant 0.4.8 link to openssl Apr 13 18:28:44 GPL is incompatible with OpenSSL - how about BSD? Apr 13 18:30:16 03rw 07org.oe.oz354fam083 * rbc012a58... 10/packages/base-files/base-files_3.0.14.bb: base-files: fix PACKAGE_ARCH Apr 13 18:32:06 wpa-supplicant is GPL or BSD - dual license Apr 13 18:32:21 so it can be linked to openssl for distributing? Apr 13 18:34:27 03coredump 07org.oe.oz354x * r535c1294... 10/packages/gpe-bootsplash/ (4 files in 2 dirs): gpe-bootsplash: Do not write on RO FS, remount RW instead + unite all patches to the init script Apr 13 18:34:32 03coredump 07org.oe.dev * rdbd31bea... 10/packages/gpe-bootsplash/ (4 files in 2 dirs): gpe-bootsplash: Add changes from .oz354x Apr 13 18:35:21 CoreDump|home: if you can then avoid logs like 'add changes from .oz354x' because they are useless for those who track only .dev Apr 13 18:37:16 true, will kep it in mind Apr 13 18:38:41 CoreDump|home: I usually use midnightcommander with .oz354x in one panel and .dev in second. commit in left, tab, commit same from history (--message option to mt commit) Apr 13 18:39:37 ~praise mc Apr 13 18:39:38 All hail mc! Apr 13 18:39:46 ;) Apr 13 18:40:03 CoreDump|home: opie-image/collie has too many stuff probably? Apr 13 18:40:46 hrw|tv: nope, size is correct. The initrd doesn't get flashed at all Apr 13 18:41:12 CoreDump|home: iirc collie flasher ignore too big images and flash eveyrthing other Apr 13 18:41:30 reenoo: are you building 0.8.4 from .oz354fam083, or is there another branch? Apr 13 18:41:53 I'll check with an older image tomorrow, thanks for the hint ;) Apr 13 18:41:53 mreimer: no, there is no other branch Apr 13 18:42:18 reenoo: images from feeds? Apr 13 18:42:37 reenoo: so if I build gpe-image from .oz354fam083, I'd get the (rough) equivalent of your 0.8.4 images? Apr 13 18:43:55 mreimer: a fairly precise equivalent Apr 13 18:43:59 ok, thanks Apr 13 18:45:29 hrw|tv: the current feeds were built from scratch. prior to the release there were no matching feeds. Apr 13 18:45:43 aha Apr 13 18:49:25 03hrw 07org.oe.oz354x * rca0098fe... 10/packages/wpa-supplicant/ (6 files in 2 dirs): wpa-supplicant: added 0.4.8 (latest stable) Apr 13 18:50:12 CoreDump|home: gpe-bootsplash added to upgrades Apr 13 18:57:55 CoreDump|home: http://ewi546.ewi.utwente.nl/tmp/hrw/t/ contain collie 3.5.4.1-alpha3 built from feeds Apr 13 18:59:00 CoreDump|home: opie flavour Apr 13 18:59:16 thanks, but I really need to track down this issue here :\ Apr 13 18:59:50 ok - but can test does those work ;) Apr 13 19:21:51 wtf Apr 13 19:22:04 that was the strangest thing I have seen Apr 13 19:22:42 a SMS subscription scam for bible texts Apr 13 19:26:12 who would pay $4 per message for phrases out of the bible you could legally download for free? Apr 13 19:30:50 heh Apr 13 20:06:48 * chouimat is back. Apr 13 20:10:29 <[g2]> what the status of building OE with large values of make -j N where N is large ? Apr 13 20:10:59 hey [g2] Apr 13 20:10:59 <[g2]> large being like 8-12 Apr 13 20:11:03 it should be ok Apr 13 20:11:15 I was using -j8 for while Apr 13 20:11:16 <[g2]> hey koen|tv ! Apr 13 20:11:34 <[g2]> koen I hope the paper went well Apr 13 20:11:34 * koen has a nice long weekend to work on fun stuff :) Apr 13 20:11:39 How you guys may have 10 CPUs in your box ?? Apr 13 20:11:40 [g2]: it did Apr 13 20:11:46 <[g2]> excellent Apr 13 20:12:07 [g2]: the problem isn't OE, its just that Make sometimes breaks down with so many threads Apr 13 20:12:08 <[g2]> Ecco_ it'd be on a network Apr 13 20:13:12 <[g2]> CosmicPenguin I remember reading something about when N get larger than 5 or 6 to use a different make tool ( iirc the samba guys wrote up something about that) Apr 13 20:13:34 yeah Apr 13 20:13:49 Its all about constructing your makefiles right, I thik Apr 13 20:14:02 a multithreaded bitbake would probably yield a better result Apr 13 20:14:13 * koen dreams of an OE buildfarm Apr 13 20:14:22 I've had times when I've started out at like -j 6 and started dropping processors until the app completes Apr 13 20:14:37 koen : Get yourself a Quad G5 : OE builds soooo fast on mine :-D Apr 13 20:14:46 :) Apr 13 20:15:03 I can't justify buying even more hw right now Apr 13 20:16:11 although quad g5's aren't too expensive nowadays Apr 13 20:17:09 <[g2]> koen what's the status of the prebuilt toolchains ? Apr 13 20:17:15 Given their raw power, they aren't that expensive Apr 13 20:17:23 plus they're magnetic-fluid cooled Apr 13 20:17:24 :-D Apr 13 20:17:36 [g2]: still some weird issues with packaging them Apr 13 20:17:47 <[g2]> Ecco_ I'd put an AMD X2 up against them Apr 13 20:18:01 [g2]: my plan for the weekend was to build a native toolchain + distcc Apr 13 20:18:14 * koen has enough arm hw to make that effective Apr 13 20:18:18 Your AMD thing will neve get the style of a Pmac :-D Apr 13 20:18:33 * CosmicPenguin hugs his dual Opteron 880 box Apr 13 20:18:35 <[g2]> Ecco_ I don't need style I need function :) Apr 13 20:18:46 * koen pets his powerbook g4 Apr 13 20:18:49 <[g2]> i'ts going in the garage any way Apr 13 20:18:58 Duh Apr 13 20:19:17 Then ok Apr 13 20:19:32 Stick with this overheating noise-prone ugly AMD :-D Apr 13 20:19:33 lol Apr 13 20:20:08 the opteron 242 I use for builds is nicely tucked away in a lab Apr 13 20:20:12 <[g2]> 2006 is the of mostly fanless computing in the how Apr 13 20:20:15 <[g2]> house Apr 13 20:20:22 with it's 3 hot swappable PSUs... Apr 13 20:20:34 its* Apr 13 20:20:41 That's a great gear Apr 13 20:20:48 However you have to work remotely on this Apr 13 20:20:54 which may be a bit annoyning sometime Apr 13 20:21:05 only when I need to use meld to merge stuff Apr 13 20:21:35 CosmicPenguin: that i43 is the only alchemy handheld out there? Apr 13 20:21:36 I do all my builds on remote machines Apr 13 20:21:44 google isn't being helpfull on that Apr 13 20:21:48 my desktop hasn't ever felt the pain of a OE build Apr 13 20:21:50 koen: I believe so Apr 13 20:22:07 koen: previous versions of the Alchemy were not targeted at handheld markets Apr 13 20:22:22 And the AU1200 is relatively new Apr 13 20:23:23 By the way guys, I'm a bit new to Zaurus hardware : I got mine yesterday Apr 13 20:24:37 Which kernel sources does OE use ? Apr 13 20:24:44 vanilla ones ? Apr 13 20:25:41 Ecco_: proper ones Apr 13 20:25:48 Ecco_: ls packages/linux/ -la Apr 13 20:25:58 ok, thanks Apr 13 20:26:06 Are those sources available on their own ? Apr 13 20:26:21 To tell you the truth, I'm pretty happy with the current state of my OpenZaurus distro Apr 13 20:26:27 <[g2]> CosmicPenguin koen how much memory does your build servers have ? Apr 13 20:26:30 but I quite like to build my own kernel Apr 13 20:26:45 [g2]: 2GB + 2GB swap Apr 13 20:26:48 Ecco_: feel free to do it Apr 13 20:27:05 Great : where can I find the sources alone, then ? Apr 13 20:27:11 [g2]: I have 2G on one, and 4G on the other Apr 13 20:27:11 * koen loves budget surplusses Apr 13 20:27:32 but the 4G box runs a simulator too, so thats not a reasonable comparision Apr 13 20:27:38 8Gb in a Quad, _that_ rules :-D Apr 13 20:27:48 1G in one (no swap), 768M in second (512M swap) Apr 13 20:27:49 <[g2]> koen what's happening with the 4.1 for the ep93xx ? I was getting Seg faults from the build a week or two ago Apr 13 20:28:23 * [g2] wonders what Ecco_ is building :) Apr 13 20:28:33 [g2]: I think that should be fixed now with the latest syscall patch Apr 13 20:28:56 RP: did you get the csl glibc to compile in OE? Apr 13 20:28:59 [g2]: his hipersuper kernel for zaurus probably Apr 13 20:29:26 <[g2]> hrw|tv yeah, just because he can :) Apr 13 20:29:34 <[g2]> in 2 femtoseconds Apr 13 20:29:38 Indeed I have "only" 3Gb on my Quad. I use it for rendering raytraced movies - that probably the most power-hungry application I'm aware of :-D Apr 13 20:29:51 gr : lol :-D Apr 13 20:30:29 No, indeed the QuadG5, when used without an external RAID SATA card has a pretty annonying bottleneck on the harddrive Apr 13 20:30:41 Not a problem when rendering Apr 13 20:30:50 but it could be somehow limitating when compiling stuff Apr 13 20:31:02 <[g2]> Ecco_ I pull 60MBs from a single sata drive Apr 13 20:31:19 Yes, but that's when reading one sequential file Apr 13 20:31:26 try reading 20 files in // Apr 13 20:31:27 I did Apr 13 20:31:36 for example (has I have the 30" screen) Apr 13 20:31:40 <[g2]> NCQ goes a long way to helping that Apr 13 20:31:47 i tried opening 30 DivXes at the same time Apr 13 20:32:06 the CPUs were vaguely tickled Apr 13 20:32:16 while the disk was spinning and scratching like crazy Apr 13 20:32:48 which eventually caused the movies to be all laggy Apr 13 20:32:55 but well, this test is insane anyway :-D Apr 13 20:33:56 <[g2]> I'd think a simple process queueing the file access woud help a lot Apr 13 20:34:10 <[g2]> but that's application optimizing Apr 13 20:34:17 yep :-) Apr 13 20:34:41 Indeed that was just "Ok, let's try this brand new beast with the most insane test ever" Apr 13 20:34:43 isn't that what ncq is suppossed to do? Apr 13 20:34:55 (and tcq as well) Apr 13 20:34:59 cu Apr 13 20:35:02 hrw|tv: cu Apr 13 20:35:06 <[g2]> yeah ncq works well on the AMD :0 Apr 13 20:35:28 or even better: a raid 5 array of >4 disks Apr 13 20:35:51 yeah, the RAID array blows everything Apr 13 20:36:00 including your wallet :-D Apr 13 20:36:20 * koen wants a thecus Apr 13 20:36:37 <[g2]> I think that's a bad idea Apr 13 20:36:39 http://www.thecus.com/products_over.php?cid=1&pid=2&PHPSESSID=a935d4e95a786e0e2da254a55ffc86d6 Apr 13 20:36:54 yeah, blatant gpl violations, etc Apr 13 20:37:05 <[g2]> I think they released source Apr 13 20:37:29 wow Apr 13 20:37:33 I want one of those too Apr 13 20:37:44 <[g2]> bad idea :)) Apr 13 20:38:00 Ok - I want something similar that isn't making baby Linus cry Apr 13 20:38:11 <[g2]> the performance numbers didn't seem that great Apr 13 20:38:40 CosmicPenguin: http://linuxdevices.com/articles/AT3184179979.html Apr 13 20:39:00 <[g2]> my old AMD64 nforce4 ultra mobo servers up 400+ Mbs from a single disk Apr 13 20:39:24 <[g2]> it's probably < $300 for that box now Apr 13 20:39:28 * koen wants something small, quiet with gbit and hotswappable disks Apr 13 20:39:50 <[g2]> they've got hot swappable disk holders Apr 13 20:40:14 <[g2]> the point is that the extra $500 isn't worth it for suckey performance Apr 13 20:40:20 <[g2]> IMHO Apr 13 20:40:21 true Apr 13 20:40:32 I also want it to be <$200 Apr 13 20:40:41 otherwise I can't afford to put disks in it Apr 13 20:40:50 which would kinda defeat it's purpose Apr 13 20:42:53 <[g2]> I've got an external SATA2 enclosure for $35 that I'm playing with Apr 13 20:54:43 Just being curious here Apr 13 20:54:56 but does anyone in this channel live in the SF south bay areay ? Apr 13 20:56:28 I'm currently *in* Santa Clara, but I don't live here Apr 13 20:56:38 oh, great ! Apr 13 20:56:46 Where do you live ? Apr 13 20:56:51 Colorado Apr 13 20:56:52 In fact I'm a french student Apr 13 20:57:03 Wow, that's far away :-) Apr 13 20:57:18 not as far away as say - boston Apr 13 20:57:32 And not as far away as I currently am - France :-D Apr 13 20:57:40 Yet I'm coming to Cupertino in a few days Apr 13 20:57:48 And I still have no place to live in Apr 13 20:58:01 So I was wondering : maybe you have some tips to find some housing Apr 13 20:58:09 Not me Apr 13 20:58:21 I've been mailing people on CL like crazy, but I couldn't find anything Apr 13 20:58:55 By the way, is Santa Clara a nice place ? I'm going to stay there for 4 months, so I'm a bit anxious :) Apr 13 21:32:56 XorA|gone: ping me when you're bakc Apr 13 21:33:31 mickeyl: vaguely here now Apr 13 21:33:42 excellent. quick check Apr 13 21:33:54 open qt4.bb Apr 13 21:34:12 to the ./configure line add the following Apr 13 21:34:25 -platform ${TARGET_OS}-g++ Apr 13 21:34:43 then try to rebuild qt4 Apr 13 21:36:20 mickeyl: trying Apr 13 21:38:01 mickeyl: well it hasnt bailed instantly this time :-) Apr 13 21:38:56 excellent Apr 13 21:40:15 * XorA|gone waits patiently for C++ code compiling Apr 13 21:43:01 mickeyl: good evening Apr 13 21:44:08 good evening pb_ Apr 13 21:44:19 * mickeyl still impressed by the play he just visited Apr 13 21:44:22 Blithe Spirit Apr 13 21:44:33 theatre is so much more intense than cinema Apr 13 21:44:44 http://english-theatre.org/et_new/html/main.php?nav=170&url=15Blithe_spirit.html Apr 13 21:45:40 besides me admiring one of the female actors Apr 13 21:45:47 but ssshhh... don't tell anyone :D Apr 13 21:47:32 hehe Apr 13 21:47:41 mm, looks good Apr 13 21:47:52 :)) Apr 13 21:49:36 ok Apr 13 21:49:36 I haven't been to the theatre for ages, though I think I'm going to see a play next weekend. Apr 13 21:49:47 mickeyl: can you help with this staging problem again? Apr 13 21:49:57 JustinP: sure, what's wrong? Apr 13 21:50:08 mickeyl: thanks for the evas fix, BTW, but the modules are supposed to be staged into arch specific dirs Apr 13 21:50:22 mickeyl: I "fixed" it by basically hard-coding them but that won't realy work Apr 13 21:50:28 mickeyl: (hack not checked in) Apr 13 21:51:00 what's the pattern? Apr 13 21:51:04 mickeyl: still building :-D Apr 13 21:51:08 mickeyl: as in "linux-gnu-i686" Apr 13 21:52:08 right. so you need to add a ${TARGET_SYS} somewhere Apr 13 21:56:52 bbiab Apr 13 21:57:26 mickey|bbl: ${TARGET_SYS} si the correct var? I'll try it then Apr 13 21:57:52 I had thought that the TARGET_SYS wasn't in the right format.... Apr 13 22:07:06 damn bitbake....it cleaned glibc on me... Apr 13 22:13:17 mickey|bbl: NOTE: package qt4-x11-free-4.1.1: completed Apr 13 22:20:44 koen: I didn't attempt csl glibc Apr 13 22:21:17 koen: The glibc tarball would be included with that the gcc build downloads though - try feeding that through the glibc files? Apr 13 22:46:41 Is there an OE mailing list Apr 13 22:53:38 Hrm. libetpan-0.43 compile fails. Apr 13 23:13:44 this is probably a very stupid question. Apr 13 23:13:52 I am trying to compile 2.6 kernel for ARM, Apr 13 23:14:09 but build environment keeps thinking i386 as the target. Apr 13 23:14:32 I was basically doing the old drill like "make oldcnfig" etc. Apr 13 23:14:36 anything changed in 2.6? Apr 13 23:17:51 ahem Apr 13 23:17:51 do you have a cross-compiled installed ? Apr 13 23:17:51 did you patch linux/Makefile to use the cross-compiler instead of your native one ? Apr 13 23:18:23 ok, this fixes it: Apr 13 23:18:26 export ARCH=arm Apr 13 23:18:34 export CROSS_COMPILE=arm-linux- Apr 13 23:18:40 ;-) Apr 13 23:18:45 not sure if this is the recommanded practice now. Apr 13 23:19:05 2.4 used to be able to figure this out by looking at ARCH= line inside .config file. Apr 13 23:19:35 you can now make menuconfig as usual :-) Apr 13 23:20:41 it appears you still need to set up those env. Apr 14 00:30:56 is CVS_TARBALL_STASH used only for cvs sources? Apr 14 00:31:41 Well, I think some of the other fetchers support it Apr 14 00:31:47 like SVN Apr 14 00:31:58 And is there a fallback for wget ? Apr 14 00:32:06 ? Apr 14 00:32:07 or whatever is the tarball fetcher? Apr 14 00:32:22 I mean, if I have a local mirror with distfiles? Apr 14 00:32:47 I think there it tries to fetch from someplace if it can not access source uri, no? Apr 14 00:32:48 The http and ftp fetchers support a different method Apr 14 00:32:52 but basically yes Apr 14 00:33:20 is there a variable I can set to my local mirror and what is the variable if it is? Apr 14 00:33:29 There is, see the config files Apr 14 00:33:36 that's what I'm doing... Apr 14 00:33:58 it should be in there somewhere Apr 14 00:34:05 probably near the CVS_TARBALL_STASH Apr 14 00:34:27 no, I don't think anywhere near that one... But I'll keep looking, thought you might now. Apr 14 00:34:29 or maybe it was in a bbclass Apr 14 00:34:40 that could have been it Apr 14 00:34:54 I Think there is a list in base.bbclass Apr 14 00:35:03 That sounds right Apr 14 00:35:27 seems to be hardcoded. oh wel... Apr 14 00:35:29 Haven't touched any of this stuff in a year and a half now Apr 14 00:35:34 you can override it Apr 14 00:35:46 Yes, thank you. Apr 14 00:36:34 base.bbclass sets the initial value, but any .conf file further up the path can change the value Apr 14 00:36:52 yes, I understand. Apr 14 00:37:13 if you have the stuff archived locally, you probably want PREMIRRORS so that it checks there before going out over the network Apr 14 00:37:24 I take it mirrors are tried _after_ the source fetch fails? Apr 14 00:37:28 right Apr 14 00:37:34 yes, that's what I was asking for :) Apr 14 00:37:35 PREMIRRORS are before Apr 14 00:37:36 Thank you Apr 14 00:37:49 at least as of a year and a half ago :) Apr 14 00:37:54 heh Apr 14 01:23:08 anyone using any of the hotplug implementations other than linux-hotplug Apr 14 01:34:26 okay, really annoying question... I have a collie, and I need to pass options to hostap-modules when they load... how can I do that? Apr 14 01:35:10 Zero_Chaos: Edit the hotplug config? Apr 14 01:35:42 NAbyss: collie uses hotplug? (I have no clue) if so, I can figure it out.... Apr 14 01:36:01 Zero_Chaos: No idea TBH... have a look around, see if it does Apr 14 01:36:24 NAbyss: I suppose I can do that. Apr 14 01:37:29 /etc/modutils i think you can setup options that are sent to modules Apr 14 01:37:41 modules.conf Apr 14 01:37:52 TheMasterMind1: I'll check that too, thanks Apr 14 02:31:29 Can someone take a look at this http://pastebin.com/658946 and tell me what is going wrong here Apr 14 02:40:28 glibc build problems? May kergoth have mercy on your soul Apr 14 02:41:19 haha..I know..can't figure out what it is referring to since the last file stdio_lim.hT does not exist Apr 14 03:13:38 cyrus__, what are you building for? Apr 14 03:14:22 and stdio_lim.hT does not exist because itas trying to assemble the file Apr 14 03:14:27 was* Apr 14 03:16:14 odd... it looks liek you have a very weird setup .. Apr 14 03:17:13 which distro uses /usr/.home/cyrus/development ? Apr 14 03:19:13 emte - sorry, stepped away for a second. I am trying to build package virtual/kernel Apr 14 03:19:21 i have distro set to familiar-0.8.3 Apr 14 03:19:27 i mean host Apr 14 03:19:36 machine is intel-pxa270 Apr 14 03:19:40 i cannot recall any that use /usr/.home/ Apr 14 03:20:11 so unless that is normal for your distro ... you may have config issues Apr 14 03:20:14 oh...that is just my computer setup. I have the org.openembedded.dev and build directory in /home/cyrus/development Apr 14 03:20:30 home is link to /usr/.home since I screwed up my partitioning when I first installed linux Apr 14 03:20:40 yes that makes sense but why is it referencing /usr/.home Apr 14 03:20:49 aha Apr 14 03:21:51 where did you get intel-pxa270 from ? Apr 14 03:22:06 maybe its in a branch i dont have ... Apr 14 03:22:17 otherwise its an invalid machine ... Apr 14 03:22:41 org.openembedded.oz354fam083/conf/machine/ Apr 14 03:22:48 gives a list Apr 14 03:22:52 or Apr 14 03:22:55 sorry, ipaq-pxa270 Apr 14 03:23:01 not intel Apr 14 03:23:25 did you recently update bitbake? Apr 14 03:23:30 yes Apr 14 03:24:04 (not that i am sure it would cause it, but there are issues with the recent svn) Apr 14 03:25:36 hmm Apr 14 03:25:52 maybe i am mistaken ... i thought i read taht somewhere on the mailing list ... Apr 14 03:27:07 aha Apr 14 03:27:21 BitBake Status and Future Development Apr 14 03:29:22 http://article.gmane.org/gmane.comp.handhelds.openembedded/10606/match=bitbake+status+future+development Apr 14 04:03:59 Charming. The signing key for gdb-6.4 isn't on any public keyservers. Apr 14 04:04:40 http://rtgo.hopto.org/th99/ Apr 14 04:05:21 http://www.talkaboutcomputing.com/group/comp.sys.ibm.pc.classic/messages/5589.html Apr 14 06:27:52 Out of curiousity, why is CONFIG_CPU_FREQ not set for spitz on the 2.6.16 config? Apr 14 06:49:19 salut Apr 14 06:53:42 hi all Apr 14 06:59:48 hi han Apr 14 07:43:13 morning Apr 14 07:45:05 hey XorA Apr 14 07:47:06 koen: abiword 2.4.4 is almost ready for me to submit, unfortuneately they have made --enable-embedded and --enble-hildon inter dependant :-( I shall have to file a bug Apr 14 07:49:11 to reenable the Japanese keys on a borzoi, do I need to recompile the kernel? Apr 14 07:49:38 Opie sees the Ctrl key and the ZenHan key as being the same Apr 14 07:49:51 and keymap seems to have nothing to do with it Apr 14 07:49:55 ashikase: keymaps should be a userspace thing Apr 14 07:50:09 although with opie you never know Apr 14 07:50:18 edit /etc/keymap* Apr 14 07:50:57 XorA: I tried removing the definitions for the keys from the keymap, thinking they would revert to defaults Apr 14 07:51:05 but it doesn't seem to have made a difference Apr 14 07:51:15 ashikase: try assigning them to a different code instead then Apr 14 07:51:31 hrm.. I wonder what codes Sharp uses by default Apr 14 07:51:34 ashikase: "defaults" for OZ might not be the same as "defaults" for sharp Apr 14 07:51:45 right Apr 14 07:56:40 koen: what was the problem with perl in angstrom? Im pretty sure I had it working Apr 14 07:58:06 ERROR: log data follows (/data/build/koen/OE/build/tmp/angstrom/work/armv5te-linux-gnueabi/perl-5.8.7-r14/temp/log.do_configure.30766) Apr 14 07:58:09 | sed: can't read config.sh-arm-linux-gnueabi: No such file or directory Apr 14 07:59:32 koen: bizarre Apr 14 07:59:46 I could have sworn I compiled abiword under angstrom, maybe I forgot Apr 14 08:00:06 I ran out of diskspace for doing test builds of angstrom though so I had to deleet the results Apr 14 08:04:59 morning Apr 14 08:05:05 hey hrw|work Apr 14 08:07:17 angtrom is already compilable ?? do you devs heppen to take some rest ? ^^ Apr 14 08:07:28 happen* Apr 14 08:07:59 XorA: I think I 'solved' the perl problem Apr 14 08:08:09 http://pinouts.ru/ Apr 14 08:08:12 koen: what was it? Apr 14 08:08:28 XorA: cp config.sh.arm-linux config.sh.arm-linux-gnueabi Apr 14 08:08:43 koen: heh heh Apr 14 08:09:45 ok, keymap fixed... F25 was the proper value Apr 14 08:15:03 03koen 07org.oe.dev * r114d71d1... 10/packages/perl/perl_5.8.7.bb: perl: make 5.8.7 compile for arm EABI Apr 14 08:15:54 koen: if your in the mood for fixing perl I have a couple of fixes in bugs.treke.net Apr 14 08:16:39 this was just a oneliner Apr 14 08:16:52 you're welcome to apply the perl patches yourself, though Apr 14 08:17:14 koen: I really need someone to apply then and build on 32bit to confirm I got them right, they are 64bit fixes Apr 14 08:17:24 koen: http://bugs.openembedded.org/show_bug.cgi?id=639 Apr 14 08:18:57 WTF ipkg randomly decided to replace dropbear with openssh-sshd Apr 14 08:19:08 those patche look OK Apr 14 08:19:23 XorA: check conflict/replaces/provides fields there Apr 14 08:20:10 hrw|work: Replaces: dropbear Apr 14 08:20:19 hrw|work: that'll be wrong then Apr 14 08:21:56 yeah, that is indeed wrong Apr 14 08:22:20 openssh-sshd might be entitled to Conflict with dropbear, but it shouldn't Provide it and it definitely shouldn't Replace it. Apr 14 08:23:28 even conflicting is a bit debatable, since there's no particular reason you couldn't have both installed at once. Apr 14 08:23:53 u-a is supposed to do the heavy lifting on that, right? Apr 14 08:24:48 not really in the case of daemons, no. u-a would be appropriate for managing the /usr/bin/ssh client binary though. Apr 14 08:25:06 pb_: it manages the client already Apr 14 08:25:15 XorA: okay, very good Apr 14 08:25:28 this REPLACES is new in ssh 4.0 bb Apr 14 08:25:43 and RCONFLITS as well Apr 14 08:26:32 righto Apr 14 08:26:34 who added them? Apr 14 08:26:54 coredump? Apr 14 08:27:51 Revision: 9fabec84d1319c3b4c559119d1a5b2024ad83adb Apr 14 08:27:56 Author: coredump@openembedded.org Apr 14 08:27:56 ah Apr 14 08:27:56 Date: 2006-03-27T16:31:02 Apr 14 08:28:00 openssh: Make openssh-sshd and the meta-package openssh RREPLACE and RCONFLICT dropbear. Apr 14 08:29:31 to make things worse, the post install scripts dont even restart sshd so you end up with connection lost and no way to re-connect Apr 14 08:30:05 03xora 07org.oe.dev * r4951b3dd... 10/packages/abiword/abiword_2.4.4.bb: Apr 14 08:30:05 abiword_2.4.4.bb : new version, now support cdump in perl so I have Apr 14 08:30:05 removed the need for cdump-native. Also autotools stuff works now without Apr 14 08:30:05 hacks so I have removed those. Apr 14 08:30:28 koen: there you go Apr 14 08:30:40 that does sound bad Apr 14 08:30:46 is openssh not using update-rc.d.bbclass? Apr 14 08:31:06 XorA: thanks Apr 14 08:31:09 anything that uses the class ought to get automatically (re)started after installation Apr 14 08:31:37 pb_: I dont think so with a quick scan of .bb Apr 14 08:31:46 XorA: will abiword 2.4.4 get spellchecker functionality in OE distros? Apr 14 08:32:51 ah, right, yeah, it runs update-rc.d by hand from its postinst Apr 14 08:32:53 that's no good Apr 14 08:33:03 hi RP Apr 14 08:33:13 hrw|work: Ill have to look into that Apr 14 08:33:34 * koen will try to build the ODT plugins Apr 14 08:33:41 also, rather bizarrely, it seems to unconditionally do an "addgroup sshd" in the postinst, which seems like it would probably blow up after the first time Apr 14 08:33:45 koen: I assume I can clean the old versions Apr 14 08:34:16 XorA: yes, I kept them to avoid the crash bug, but that was related to ClearLooks Apr 14 08:37:17 how do you override the autoconf site/* for one package only? Apr 14 08:38:13 03xora 07org.oe.dev * r10a1dd8d... 10/packages/abiword/ (6 files in 3 dirs): packages/abitword/* : remove old versions and uneeded patches. Apr 14 08:40:42 morning all Apr 14 08:40:50 abitword? ;) Apr 14 08:41:05 TheMasterMind1: you don't. the contents of the site file, by definition, are meant to be common to all packages. Apr 14 08:41:13 hrw|work: my own keyboard is 2/3 size, I get confused at work :-) Apr 14 08:41:26 TheMasterMind1: if something in the site file disagrees with one of your packages then either the site file is wrong or your configure script is faulty. Apr 14 08:42:10 the configure script is faulty, it uses a big_endian check but only applies it to float code Apr 14 08:42:17 so patch the package, i got it Apr 14 08:43:08 right Apr 14 08:44:31 sweet Apr 14 08:44:37 i just fixed jikes to work with floats as well Apr 14 08:46:36 rock Apr 14 08:46:54 * koen bitbake -b abiword-plugins_2.4.4.bb Apr 14 08:48:05 hi all, may i savely delete all in work directory or do I rely on paths to work? Apr 14 08:48:16 koen: http://bugzilla.abisource.com/show_bug.cgi?id=10188 Apr 14 08:48:57 * koen reads Apr 14 08:50:57 sws: can - probably some gtk related stuff will break after this but in general tmp/work/ is not needed after build Apr 14 08:52:52 XorA: http://oe.pastebin.com/659176 Apr 14 08:58:42 koen: looks like some TLC needed there Apr 14 09:00:33 after building goffice and poppler: http://oe.pastebin.com/659192 Apr 14 09:07:07 ls Apr 14 09:08:55 XorA: 424K /data/build/koen/OE/build/tmp/angstrom/deploy/ipk/abiword-plugins_2.4.4-r0_armv5te.ipk Apr 14 09:10:06 koen: should probably split that into seperate packages Apr 14 09:10:17 yeah Apr 14 09:10:46 koen: but those be nice to get in OE before hrw|work doesn 3.5.4.1 Apr 14 09:11:08 drat, you guessed my secret plan Apr 14 09:11:42 hrm i wish hotplug got a 'i'm starting to pull out the mmc card' signal instead of just one when the card is gone completely Apr 14 09:12:03 http://www.alltronics.biz/cgi-bin/category.cgi?item=04C002&type=store Apr 14 09:13:06 wow 5 bucks Apr 14 09:13:46 I dunno what cpu that is if it's 32 bit or not Apr 14 09:13:57 but mabye if they didn't rape the boards it's possible to use them for something Apr 14 09:14:26 oh shit it's a arm7 cpu Apr 14 09:14:58 anyways goodnight Apr 14 09:15:27 n8 fishhead Apr 14 09:18:27 maybe if i do something crazy, like tie an interrupt to the ro switch detector Apr 14 09:21:50 how come there's no fs similiar to jffs2 for nand flash Apr 14 09:22:23 TheMasterMind1: jffs2 works on nand as well Apr 14 09:22:37 TheMasterMind1: google for yaffs Apr 14 09:22:47 wish I could speed up ipkg-make-index Apr 14 09:23:03 XorA: and ipkg too... Apr 14 09:23:14 i spoke too soon Apr 14 09:23:31 koen: so i could theoretically use jffs2 on a mmc card? Apr 14 09:23:57 TheMasterMind1: very theoretically Apr 14 09:24:01 TheMasterMind1: apart from the fact mmc is a block device? Apr 14 09:24:33 that makes a difference huh? Apr 14 09:25:11 mmc cards should do their own wearleveling Apr 14 09:26:54 hrw|work: could you put libwpd, abiword and abiword-plugins in .oz354x after lunch? Apr 14 09:26:59 ah, gotcha. so raw nand flash + wear leveling, bad blocks, etc = block device Apr 14 09:27:15 sort of Apr 14 09:28:03 a block device is what most filesystems expect (e.g a hard drive, floppy disk, etc) Apr 14 09:28:27 with flash chips you can poke at some lower levels -> mtd device Apr 14 09:28:28 koen: can do Apr 14 09:28:34 hrw|work: thanks Apr 14 09:28:41 hrw|work: I'll clean it up and push in a minute Apr 14 09:29:09 ok Apr 14 09:29:33 koen: if you can then push this into .oz354x too Apr 14 09:31:43 hrw|work: abiword has spell checking Apr 14 09:32:01 koen: Ive just done an updated enchant.bb Apr 14 09:32:02 XorA: yes, but it's greyed out Apr 14 09:32:03 anyone remember a kernel option a few years ago under fat in filesystems- i seem to remember very distinctly an option for a variant of vfat that supported permissions and symlinks Apr 14 09:32:27 XorA: there certainly are some things you could do to speed up ipkg-make-index. Apr 14 09:32:47 for example, stop it from recalculating the md5sums for every package as it scans them: it only really needs to inspect the control file. Apr 14 09:33:14 that said, given that you don't run it all that often, the speed has never really bothered me too much. Apr 14 09:33:16 koen: thats because inline spelling is enabled in prefs Apr 14 09:33:21 koen: I just worked that out Apr 14 09:33:24 03koen 07org.oe.dev * r834fd290... 10/packages/libwpd/libwpd_0.8.4.bb: libwpd: add 0.8.4 Apr 14 09:33:42 pb_: I run it pretty much everytime I upload a new package to the Z for testing Apr 14 09:33:55 XorA: right, but presumably 99% of the index is already prebuilt in that case. Apr 14 09:34:03 koen: disbale spelling as you type in prefs and the menu becomes enabled Apr 14 09:34:04 you don't re-index your feeds from scratch every time, right? Apr 14 09:34:34 pb_: I do as I have found it doesnt re-index stuff it has already seen but you changed the ipk underneath it Apr 14 09:34:57 pb_: I might produce upto 10 .ipk's for testing while fiddling Apr 14 09:35:21 XorA: ah, right. yeah, it doesn't expect you to re-issue the same package with the same version number and different contents. Apr 14 09:35:40 though, if you are bent on doing that, you could easily enough patch it to check the datestamp on the .ipk file and compare it to the datestamp on the old index Apr 14 09:35:44 pb_: I know the speed issue is me missusing it, so its upto me to look into it one day Apr 14 09:36:28 heh, right Apr 14 09:36:48 koen: of course the spell checking says that all words are spelt right, so there is something odd there Apr 14 09:37:15 XorA: I just pushed abiword-plugins Apr 14 09:38:26 koen: found the abiword spelling problem, now going to fix Apr 14 09:39:00 XorA: ok, I'll wait with pushing abi* in to the branch Apr 14 09:41:47 03koen 07org.oe.oz354x * r4c981ad0... 10/packages/libwpd/libwpd_0.8.4.bb: libwpd: add 0.8.4 from .dev Apr 14 09:41:51 03koen 07org.oe.dev * r85ba8b2c... 10/packages/abiword/abiword-plugins_2.4.4.bb: Apr 14 09:41:51 abiword-plugins: add plugins package for abiword Apr 14 09:41:51 * needs to split plugins in subpackages Apr 14 09:41:51 * needs some more deps added to OE to build all plugins Apr 14 09:41:51 * apart from the above it works and people now have OpenDocument support Apr 14 09:48:53 ~lart gdome2 Apr 14 09:48:54 * ibot stabs gdome2 Apr 14 09:50:32 Does anyonw know which "standard" packages replaces tinylogin ? Apr 14 09:50:48 (i.e. which one contains "passwd" for example) Apr 14 09:51:05 (or su) Apr 14 09:51:06 Ecco_: look in any desktop distro? Apr 14 09:51:22 Yeah, but I thought someone out there would already know Apr 14 09:51:45 su in login, passwd in passwd (Debian) Apr 14 09:52:18 ok Apr 14 09:52:22 thanks a lot Apr 14 09:52:31 (I'm not a debian user myself, so...) Apr 14 09:52:37 what about "which" for example ? Apr 14 09:54:31 Ecco_: is it hard to check? Apr 14 09:54:46 Does anyone have 2.6.15 on a zaurus handy? Apr 14 09:54:54 RP: Yep.. Apr 14 09:54:56 dpkg -S `which which` or equivalent on rpm or other packagesystem? Apr 14 09:55:42 NAbyss: Could you see what happens if you run "dd if=/dev/zero of=/dev/fb0" for me please? It should blank the screen and then return to the command prompt Apr 14 09:55:53 RP: From opie, or normal tty? Apr 14 09:56:17 NAbyss: A normal tty is probably best Apr 14 09:57:11 dd: /dev/fb0: No space left on device Apr 14 09:57:15 RP, : do you know if loadvm is possible with qemu arm ? Apr 14 09:58:37 Genesis: I don't see why it wouldn't work Apr 14 09:58:56 NAbyss: THanks, that's exactly what I needed to know :) Apr 14 09:59:34 RP: Okies :) Apr 14 09:59:39 RP, : because -loadvm is only supported by qemu ( not qemu-system-arm ) and qemu only support PC machine Apr 14 09:59:52 hrw|work, I don't have any package-based Linux PC under my hand **** ENDING LOGGING AT Fri Apr 14 09:59:57 2006