**** BEGIN LOGGING AT Tue May 29 02:59:57 2007 May 29 03:29:42 ipkg.py..... May 29 03:30:01 * * OE Bug 2391 has been created by shane.oe@hathawaymix.org May 29 03:30:03 * * glibc fails locale generation due to bad dependency, May 29 03:30:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2391 May 29 03:39:29 * JPatrin-mac_ attempts a build from scratch to see if these ipkg.py errors go away (I doubt they will...) May 29 03:43:24 JPatrin-mac_: I started a build from scratch about 2hrs ago, no issues seen so far.. it has not yet reached the point where it creates the rootfs images though. May 29 03:43:42 yeah, that's the part that fails May 29 04:36:52 Does oe support package management on the target? May 29 04:49:24 xl0: Yes, using ipkg ("ipkg update", "ipkg install " etc) May 29 04:50:34 Is the ipkg format in some way related to the Debian packages? May 29 04:56:02 xl0: Yeah, based on the same file format (I think one of them you can rename to the other and they'll work fine). May 29 04:56:47 There has been some people working on actually using .deb packages in OE, not sure of the status of that. All of the deb tools are a lot larger then the ipkg versions. May 29 04:58:50 Cool. May 29 05:00:04 I see oe is still using monotone. Any plans of switching to git? May 29 05:02:34 xl0: Nope, no plans to switch away from monotone. Monotone is a *lot* better now than it was a few months ago so I doubt a move would even been considered anymore. May 29 05:04:55 How much smpler it things were just a few years ago, when there was only cvs... ;) May 29 05:08:13 xl0: I'd use almost anything to avoid having to use cvs! I use monotone, subversion, git and perforce pretty much equally now and they are all a lot better than cvs ;) May 29 05:09:14 No complains of course. ;) May 29 05:11:41 Does Monotone really beat git in some way, or the reason behind it is more historical? May 29 05:14:49 xl0: No idea, the whole selection thing happened before I was using oe May 29 05:15:38 Yes, but you are using Monotone... ;) May 29 05:17:54 Hmm, they are different... I'm not sure which I'd prefer if I had the choice... I'd probably have to think about it more than I have so far ;) May 29 05:42:26 good morning May 29 06:48:14 morning all May 29 06:49:36 good morning May 29 06:53:51 i have a strange file not found error when I try to compile an angstrom on an omap5912. (rafb.net/p/vwwFVy47.html) anybody know what could be the cause ? May 29 07:55:02 * * OE Bug 2392 has been created by frederic.devernay@m4x.org May 29 07:55:04 * * synergy for diet-x11 May 29 07:55:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2392 May 29 08:08:19 morning May 29 08:11:50 morning May 29 08:18:47 good one. May 29 08:26:09 morning all May 29 08:41:41 hey May 29 08:42:05 does anybody know what the difference between meta-toolchain & meta-sdk is supposed to be? May 29 08:42:46 tkp: One is just a toolchain, the other contains development libraries (at a guess) May 29 08:42:50 morning all May 29 08:42:59 hey RP May 29 08:43:34 RP: you might think... but actually they both build an identical set of packages May 29 08:43:54 neither of those packages build anyway... May 29 08:44:17 I have done some tweaking to get meta-toolchain to build May 29 08:44:47 I'll send to the bugtracker now so someone can vet it May 29 08:51:01 * * OE Bug 2393 has been created by stv@kirkconsulting.co.uk May 29 08:51:03 * * meta-toolchain/sdk does not build. May 29 08:51:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2393 May 29 08:58:24 hi koen May 29 08:58:38 good morning all May 29 08:58:43 hi koen May 29 08:58:53 hi psokolovsky May 29 08:59:37 i succeed to compile minimo for mipsel, but failed on my board's rootfs (due to old libc).. May 29 08:59:50 there. Patch attached.. I hope that's ok May 29 09:01:12 so i am build new rootfs(bitbake bootstrap-image), but it failed on task do_populate_staging May 29 09:01:26 tkp: the patch looks pretty ok May 29 09:02:09 It builds ok now and I have used the resulting toolchain to build my entire distro May 29 09:13:18 03polyonymous 07org.oe.dev * rc119c6a4... 10/ (1 classes/qpf.bbclass): classes/qpf.bbclass: removed set -e from postinst/prerm scripts so that it doesn't fail on sourcing in /etc/profile. Closes #2388 May 29 09:14:02 * * OE Bug 2388 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 09:14:04 * *  qpf-unismall's postinst script fails [...] May 29 09:14:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2388 May 29 09:15:02 * * OE Bug 2385 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 09:15:04 * *  nunome prevents opie from starting up [...] May 29 09:15:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2385 May 29 09:19:11 hey koen May 29 09:19:17 hey XorA May 29 09:40:20 can i build rootfs without kernel? May 29 09:41:52 kernel compiled but when task do_populate_staging, bitbake failed.. May 29 09:42:52 crink: more details please May 29 09:43:13 my build had ~600 do_populate_staging tasks May 29 09:46:33 hi hrw May 29 09:48:51 i replaced my real mipsel linux(linux au1200 r58) with linux-2.6.21, it compiled well, but when it has to populate_staging, it failed without any log at all May 29 09:49:32 * crink need the new rootfs to test new minimo ipk. May 29 09:51:11 crink: tmp/work/*/linux*/temp/log.do_stage* May 29 09:51:14 ~pastebin May 29 09:51:27 somebody said pastebin was a place to paste your stuff without flooding the channel - try http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/, or http://pastebin.com is usually painfully too slow and unresponsive to use, use one of the other pastebin sites, or dpaste.com is a very nice pastebin as well May 29 09:53:44 ok, I am being a bit dumb, I built a angstrom-minimal, and I can find the kernel image, but not sure whether it made a rootfs, I think it did May 29 09:53:55 hrw: aisb log.do_stage's size is 0, log.do_install/configure/compile, run.do_* has some size, May 29 09:54:46 aa_: checked in tmp/deploy/*/images/ dir? May 29 09:55:01 hrw: yes, there is just a 2mb image that looks like the kernel May 29 09:55:36 bzImage-2.6.20-ecsh70-20070528154205 (machine is ecsh70) May 29 09:55:42 is that the whole thing? May 29 09:56:17 aa_: thats the kernel May 29 09:56:26 should I be able to test the resulting image with qemu? May 29 09:56:33 right, I thought that was just the kernel May 29 09:57:01 lots of ipk's in the sister directory May 29 09:57:01 aa_: if you want to test in qemu then I suggest qemuarm or qemux86 machine build (or atleast their kernels) May 29 09:57:13 hrw: oh ok May 29 09:58:16 hrw: but still if I built angstrom-minimal, there should be a root fs lying around somehwere? May 29 09:59:41 in images dir May 29 10:00:31 hmm, only a kernel there May 29 10:06:06 03koen 07org.oe.dev * rbaf576cb... 10/ (3 files in 3 dirs): xserver-kdrive: add patch to speed up pixbuf compositing on some 16bit displays May 29 10:06:40 koen: thats the one which you show me few days ago? May 29 10:06:44 yes May 29 10:07:00 although it needs some debugging why it only works for the n800 lcd May 29 10:07:10 I started uclibc angstrom build May 29 10:07:33 I suspect angstrom needs some stuff from poky merged back May 29 10:07:36 I think uclibc angstrom should be called angst May 29 10:08:02 koen: thats what I will do too May 29 10:08:11 hrw: thanks May 29 10:08:12 uclibc stuff from poky -> oe May 29 10:08:36 without OE there would not be poky May 29 10:08:42 I was thinking about a piece of python in the angstrom configs that error out if DISTRO isn't angstrom May 29 10:08:51 * koen is getting sick of cl0n3rz May 29 10:09:06 koen: do not do it May 29 10:09:27 koen: as long as they are release own stuff but want to be compatible with angstrom it is ok rather May 29 10:09:48 some people just do not like to cooperate May 29 10:10:04 right May 29 10:15:40 wpa-supplicant guys moved to git May 29 10:20:03 hrw crink : 2.6.21 puts kernel images to a different dir than kernel.bbclass expects May 29 10:20:08 git damn-it May 29 10:20:17 hrw crink : stamble on it yesterday May 29 10:20:29 ~lart kernel devs May 29 10:20:30 * ibot shoves a crumpet down kernel devs's throat, happy now?! Huh? Want some JAM with that? May 29 10:21:07 hrw crink : now the images are in /arch/${ARCH}/boot/images/ May 29 10:21:48 hrw crink : check the sequoia 2.6..21 kernel i posted yesterday for a quick "fix" May 29 10:22:17 hrw crink : copy the kernel image where kernel.bbclass excpects it to be May 29 10:23:17 we have to fix kernel.bbclass though to handle 2.6.21 kernels correctly May 29 10:28:23 hm May 29 10:29:07 koen: what codec is on your efika? May 29 10:29:07 linux-rp 2.6.21 in Poky does not have arch/arm/boot/images/ dir May 29 10:31:17 03pfalcon 07org.oe.dev * reac29782... 10/ (3 files in 3 dirs): May 29 10:31:17 ipkg-utils-native: Revert previously applied patch regarding "control" vs May 29 10:31:17 "./control" matter to the state which works with tar-native. May 29 10:32:54 steliosk: my 2.6.22-rc host kernels also lack /arch/${ARCH}/boot/images/ dir May 29 10:33:39 hrw : let me check again May 29 10:37:07 hrw : -> http://rafb.net/p/pqAQZl82.html May 29 10:37:13 from the workdir May 29 10:37:30 steliosk: i did it yesterday and early today, and now, but it didn't help May 29 10:37:40 /s/help/helped/ May 29 10:38:08 This is a DENX git kernel, same as the one used on the Sequoia dev boards May 29 10:38:15 0 size log.do_stage.xxxx made me nut.. May 29 10:38:32 * crink using mipsel board. May 29 10:39:50 steliosk: I build 2.6.21+ for amd64/host, i386/host and OE/Poky: progear, fic-gta01, c7x0 and none of them have images/ dir May 29 10:40:40 hrw : i am not sure this is the first 2.6.21 kernel i build/tested to tell you the truth and stamble on the problem May 29 10:40:47 ok May 29 10:41:31 hrw : maybe is a DENX "thing" but it looks strange May 29 10:42:41 steliosk: you have vendor patches applied to kernel? May 29 10:44:40 if yes then check them May 29 10:48:56 hrw : hmmm it would be strange for the patches to move the image dir around... can't see what that would benefit someone, but i'll have a look and see if anything changes that May 29 10:49:19 steliosk: I saw more strange things in vendor patches May 29 10:49:25 :) May 29 10:49:42 machine which I have here has patch which do not install modules in DESTDIR but in other place. May 29 10:54:01 * * OE Bug 2394 has been created by bugs.openembedded.org@rolf.leggewie.biz May 29 10:54:03 * * RFE: hibernate function May 29 10:54:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2394 May 29 11:02:55 bye May 29 11:06:55 03pfalcon 07org.oe.dev * r7a8b58eb... 10/ (4 files in 3 dirs): May 29 11:06:55 ipkg-utils: Support reading both "control" and "./control" from tar archives. May 29 11:06:55 * Fixes #2381. May 29 11:07:21 XorA: "Sigmatel STAC 9766 AC97" May 29 11:13:35 koen: cheers May 29 11:55:58 someone here use src_distribute? May 29 12:03:20 hi there, I don't know why but the bitbake is not building the RDEPENDS of my task... it just build the toolchain and the task and finish... It was working fine building all depends but suddenly it stopped to work... any idea? May 29 12:12:07 hrw: I do May 29 12:13:17 Laibsch: I am working on extending it a bit May 29 12:28:44 Laibsch, tried podbox yet? May 29 12:29:10 pobox, I mean. May 29 12:32:56 polyonymous: Not yet. I'll let you know May 29 12:33:05 hrw: In what way? May 29 12:33:12 ok. Just was reading my mtn status :) May 29 12:33:13 * Laibsch likes extensions of good stuff May 29 12:34:50 Laibsch: recurrency, distributing all sources May 29 12:34:57 03polyonymous 07org.oe.dev * r70443562... 10/ (3 files in 3 dirs): apmd: prevent build from interferring with host kernel headers. Closes #1257 May 29 12:35:01 * * OE Bug 1257 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 12:35:04 * *  local linux headers interfere with apmd build [...] May 29 12:35:05 * * http://bugs.openembedded.org/show_bug.cgi?id=1257 May 29 12:36:52 hrw: What do you mean recurrency? What re(oc)curs? Why do you want to distribute all sources? May 29 12:37:07 Laibsch: do you know fetchall? May 29 12:37:34 Maybe that was mistyped recursion? May 29 12:37:48 * polyonymous makes guesses having no knowledge of the subject whatever :) May 29 12:38:44 Laibsch, is splashscreen properly oriented for you? May 29 12:39:43 hrw: Yes, really cool. I extended your nice work on it to patchall, unpackall and compileall ;-) May 29 12:39:52 ;DD May 29 12:40:17 polyonymous: On the spitz it is, yes May 29 12:40:26 Have not flashed the collie in a few days May 29 12:40:28 Laibsch, haven't tested on collie? May 29 12:40:38 I am building an image for collie now May 29 12:40:42 a few days? The patch is there for months. May 29 12:40:46 mont. May 29 12:40:47 month. May 29 12:40:58 Just ran into a problem using a meta-toolchain toolchain to build libtool-cross... May 29 12:41:23 it fails because it can't find libgcc_s.so May 29 12:41:32 polyonymous: 1 month = 28-31 days May 29 12:41:39 libgcc_s.so is packaged up in the libgcc-dev package May 29 12:41:49 which is not part of the sdk-bare package May 29 12:41:53 Laibsch, well, I might have exaggerated it :) May 29 12:42:09 polyonymous: /me too May 29 12:42:12 Anyway, I just thought that you know the answer already, if you don't - I don't mean to hurry you. May 29 12:42:15 I fixed by adding a line to create the symlink in meta-toolchain.bb May 29 12:42:24 In any case, I am not sure about the screen orientation. May 29 12:42:53 but would it be better to add libgcc-dev to sdk-bare, or to libtool-cross itself? May 29 12:42:56 Just pay attention on your next build based on anything after May, 6th May 29 12:43:11 on your next boot that is. May 29 12:43:54 Is there a way to xmodmap certain mouse buttons to /dev/null? May 29 12:47:17 polyonymous: I removed all previous collie images so my next build will certainly be "> May 6th" May 29 12:47:38 Laibsch, would be strange otherwise :) May 29 12:48:15 huhu May 29 12:58:37 XorA: weren't you looking for an L3 driver? May 29 12:59:43 koen: notice Liam and I are in that email chain :-) May 29 13:00:24 not in the [rfc][patch] one :) May 29 13:03:29 koen: nice fellow saves me writing codec driver for competitiors chip May 29 13:03:51 heh May 29 13:09:23 Why is oe.org always so slow recently? Can't we do something about it? May 29 13:09:30 * Laibsch is talking about website, not monotone May 29 13:10:20 its same machine ;( May 29 13:11:53 yes, I was aware of that. And I assume the openmoko euphoria is making a huge dent. Couldn't we point monotone.openembedded.org to another machine? May 29 13:12:08 round-robin DNS or something May 29 13:12:40 or restrict access to monotone.oe.org and require people to use one of the mirrors? May 29 13:16:56 blooming scottish rail system May 29 13:18:31 koen: Just the Scottish one? :) May 29 13:18:49 koen: we release mediatomb 0.9.1, could you please update the package? May 29 13:19:08 nothing has changed in terms of configure settings, so a simply version renaming should be enough May 29 13:20:35 ~curse nokia for 770-bt-crap May 29 13:20:39 May you be reincarnated as a Windows XP administrator, nokia for 770-bt-crap ! May 29 13:21:41 koen: whats wrong with out railways? May 29 13:22:02 * * OE Bug 2233 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 13:22:03 * *  prismstumbler compiling failed on gpe-image [...] May 29 13:22:05 03polyonymous 07org.oe.dev * re4cb7961... 10/ (1 packages/prismstumbler/prismstumbler_0.7.3.bb): prismstumbler: added missing dependency on libxpm May 29 13:22:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2233 May 29 13:22:19 XorA: trying to plan trips to inferness and kyle of logash May 29 13:22:27 XorA: the pricing scheme is insane May 29 13:22:49 koen: yeah, it depends some stations are subsidised by government May 29 13:23:17 2 single trips are cheapers than one return tickets for example May 29 13:23:20 -s May 29 13:24:50 koen: it really doesnt help that its partially privatised May 29 13:25:01 koen: the government gets to choose the wierd pricings May 29 13:25:28 JEZ!!! May 29 13:25:34 nokia devels are MORONS May 29 13:25:46 took you that long to realize? May 29 13:25:46 5 minutes bt scan... why so long? May 29 13:26:01 bt scan does not check does it is UP.. May 29 13:26:12 koen: no. I just got another thing to list May 29 13:26:57 claudio and johan have some neat stuff in the queue, but as usual, that will take 6 months to get through the legal and build team May 29 13:26:58 hrw: you can meet them at guadec May 29 13:27:29 XorA: I know May 29 13:27:36 XorA: that does not change anything May 29 13:27:45 XorA: I have unsupported device May 29 13:28:09 * koen wonders when hildon-1 will have a tarball release May 29 13:28:25 hrw: beat them with it May 29 13:30:24 XorA: and lost mahjongg device? May 29 13:30:56 hrw: claim on insurance, unexpected encounter with nokia dev May 29 13:31:52 koen: BTW are you doing any trains on Sundays? May 29 13:32:02 ehm May 29 13:32:23 the planner just left the building :) May 29 13:32:57 is there a variable that holds all package names that get installed into the root file system? I'm looking through image.bbclass and rootf_*.bbclass and nothing looks obvious May 29 13:33:22 jsheldon: PACKAGE_INSTALL is the closest thing May 29 13:33:30 koen: have to warn you about UK sundays and trains, dont expect your train to a) actually run, and b) you may end up on buses May 29 13:33:32 03koen 07org.oe.dev * r710b9cc1... 10/ (1 packages/mediatomb/mediatomb_0.9.1.bb): mediatomb: update to 0.9.1 May 29 13:33:47 XorA: ah, I see May 29 13:34:00 * koen hates buses May 29 13:34:17 koen: Engineering Works they call it :-D www.jsr.com/fsr should have all engineering works for the next month May 29 13:34:25 koen: but its not always all inclusive May 29 13:34:37 www.jchech.com/fsr sorry May 29 13:34:43 www.jcheck.com/fsr May 29 13:34:48 yay that ones right May 29 13:37:02 koen: right, but PACKAGE_INSTALL doesn't include dependencies, does it? May 29 13:37:23 no, the package-manager handles that May 29 13:38:20 heh.. no luck. May 29 13:38:35 openmoko lack any pairing stuff.. May 29 13:39:36 koen: ok that's too bad, thanks for the help May 29 13:42:27 koen: thanks May 29 13:54:34 morning May 29 13:54:45 hi May 29 13:54:50 hi chouimat May 29 13:55:06 hm little flash and thunder and the internetconnection goes crazy May 29 13:56:38 hi mr_nice May 29 13:56:42 woglinde: your flux capacitor needs replacing May 29 13:56:51 woglinde: hi May 29 13:57:01 woglinde: what about the uclibc opie image ? May 29 13:57:06 mr_nice hm something goes wron with linux-2.6.21 didnt find touchscreen anymore May 29 13:57:32 but mabyee its my fault with playing defconfig May 29 13:57:42 woglinde: this bug ist fixed. mrdata has the source but didn't found time jet to upload it May 29 13:57:49 uh May 29 13:57:59 hm than I cannt test May 29 13:58:01 anyfurther May 29 13:58:13 woglinde: but i will get the code today May 29 13:58:18 he okay May 29 13:59:17 woglinde: do you know if usbnet should work with familiar 8.4? May 29 13:59:29 mr_nice???? May 29 13:59:46 usbent only needs ifcofing or iproute May 29 14:00:17 woglinde: I want to test if is the host or clinet which causes my problem with 2.6 and usbnet May 29 14:00:30 hm May 29 14:00:39 2.4 kernel works May 29 14:00:55 but the driver is older an other than gadget May 29 14:01:28 but even the 2.4 driver wasnt the best May 29 14:01:37 often stopped the connection May 29 14:01:51 woglinde: hm May 29 14:02:23 yeah the gp2x people worte there usbnet driver on that May 29 14:02:47 so shity software gets mutiply May 29 14:02:51 multiply May 29 14:04:21 hm, I will try 2.4 and than I will try 2.6 a second. May 29 14:06:18 woglinde: did you mean that it won't find the touchscreen on booting or on wake up? May 29 14:06:24 booting May 29 14:06:35 woglinde: ah , on booting it should work May 29 14:06:37 touchscreen and switches NULL May 29 14:06:46 hm May 29 14:06:47 hm May 29 14:07:00 okay May 29 14:07:03 I will compile May 29 14:07:08 another May 29 14:07:22 with defconfig from slackpad May 29 14:07:55 this should work May 29 14:11:09 mr_nice which patches do you have applied on your kernel? May 29 14:14:54 woglinde: all which are in 2.6.21/main/ May 29 14:15:38 okay May 29 14:16:22 woglinde: but the netshared patch in testing is quite useful. I have it not im my kernel because i have no wlan card May 29 14:16:39 hm the patch isnt right May 29 14:16:45 ah May 29 14:16:48 ok May 29 14:17:01 i wasn't able to test it May 29 14:17:02 because it has the complete driver May 29 14:17:06 not the changes May 29 14:18:21 thx for hinting I will change that May 29 14:18:31 ;) May 29 14:18:34 no thank you May 29 14:19:08 but I wonder why nobody fix this for pcnet_cs.c May 29 14:19:15 I am the only user or what? May 29 14:20:29 hm, don't know May 29 14:20:45 I think it is because of 16 bit cards? May 29 14:21:47 yes ;) May 29 14:22:34 mr_nice you could try to get them upstream May 29 14:23:57 I can try it. the changes are from mrdata. May 29 14:24:17 hm oh okay May 29 14:24:30 dont forget the sign at tinks May 29 14:24:37 things May 29 14:24:39 args May 29 14:24:41 my english May 29 14:25:13 ~ping Laibsch May 29 14:25:25 pong Laibsch May 29 14:29:31 Crazy thing: http://geektechnique.org/projectlab/781/slurpr-the-mother-of-all-wardrive-boxes May 29 14:30:05 woglinde: hopefully you will be able to get a new key or find out your old one May 29 14:30:10 :) May 29 14:30:26 hm dont know if koen likes me that much May 29 14:31:24 cm MiniPCI slot Six MiniPCI Type IIIA/IIIB slots May 29 14:31:27 that rocks May 29 14:31:53 I'm sure you could do lots of cool stuff with it :) May 29 14:32:04 he box will cost 999 euro, May 29 14:32:05 steliosk: I was just thinking about you May 29 14:32:06 hm May 29 14:32:10 thats May 29 14:32:15 a showstopper May 29 14:32:23 Definitely.. :/ May 29 14:35:41 mr_nice uh hm opie started May 29 14:35:48 but touch screen isnt usable May 29 14:37:48 woglinde: can you type echo 8 > /proc/sys/kernel/printk at a console and press on the touchscreen? May 29 14:38:11 hm works May 29 14:38:23 maybee something with tslib? May 29 14:38:41 no udev isnt runnig right May 29 14:38:49 maybee thats an problem too May 29 14:38:56 I will start new kernel now May 29 14:39:04 you can try ts_calibrate May 29 14:39:14 ah ok May 29 14:45:11 mr_nice hm May 29 14:45:57 th tmpfs error irrtates me May 29 14:46:25 /proc/filesystems says May 29 14:46:27 nodev tmpfs May 29 14:46:37 but mount claimed it isnt there May 29 14:47:31 woglinde: mom May 29 14:47:33 ah May 29 14:47:35 no May 29 14:47:40 strace May 29 14:47:41 helped May 29 14:47:49 stat64("/sbin/mount.tmpfs", 0xbe9fd8f8) = -1 ENOENT (No such file or directory) May 29 14:48:42 hm May 29 14:48:45 hm May 29 14:48:46 no May 29 14:49:02 this isnt the error May 29 14:49:43 maybe it was mounted and the /proc/mounts file was owerwritten May 29 14:50:16 I have the same behaviour wich you described here on my pad May 29 14:50:19 hm May 29 14:50:48 how to find out if it is really is mounted or not? May 29 14:50:53 hm can you test with a kernel 2.6.20? May 29 14:50:59 when it works May 29 14:51:13 is kernel stuff when not maybe util-linux-mount is wrong May 29 14:52:02 2.6.20 vanilla? May 29 14:52:07 no May 29 14:52:15 2.6.20 from oe May 29 14:52:40 hm I have no working oe here May 29 14:53:43 the machine which I normaly use for oe seams to have a crupted ram May 29 14:54:13 and on this machine I have not enought space May 29 14:54:18 :( May 29 14:54:36 I can try a vanilla kernel if you want May 29 14:55:20 or if you can tell me which oe 2.6.20 kernel I can extract the patches from oe and build it May 29 14:56:47 XorA: I have found a volunteer to work out the netbook flash format :) May 29 14:57:23 koen: that so linux can advance :-) May 29 14:57:31 hi koen May 29 14:57:44 hm May 29 14:57:46 SIGILL (Illegal instruction) @ 0 (0) --- May 29 14:57:47 Can someone take a look at these an make sure they work for me please? 2361,2368,2370 May 29 15:00:07 woglinde: net-shared-irq.patch should be now fixed May 29 15:05:08 hm May 29 15:05:09 mount("tmpfs", "/dev", "tmpfs", MS_MGC_VAL, "size=10M,mode=0755") = -1 EINVAL (Invalid arg May 29 15:07:07 ????? May 29 15:07:09 hm May 29 15:07:14 no clue May 29 15:07:26 why the hell /proc/fs claims to have tmpfs May 29 15:07:36 and in defconfig its disabled May 29 15:09:10 * mr_nice shame May 29 15:09:17 heh May 29 15:09:23 mr_nice not your fault May 29 15:09:33 /proc/fs shouldnt report it May 29 15:09:39 if it isnt there May 29 15:11:26 he zecke May 29 15:11:36 hey May 29 15:11:41 what a rain... May 29 15:11:46 yes May 29 15:11:51 now sun May 29 15:11:55 you guys getting out scottish rain? May 29 15:12:00 s/out/our/ May 29 15:12:08 * zecke gets free teaching of how to be an admin May 29 15:12:25 zecke he I can bill you May 29 15:12:26 * zecke just moved one kvm and installed a shiny new one May 29 15:12:49 woglinde: and had to put the tape drive at another place... May 29 15:12:55 xora hm dont know the scottish one May 29 15:13:42 woglinde: where it rains so hard it also bounces off ground and goes up again, roads flood to depths of curbs (4inches), and it wont bloomin stop May 29 15:13:46 I have seen cats and dogs falling from the sky? May 29 15:13:59 xora hm yes and flashes and thunder May 29 15:14:28 woglinde: we never get the cool shit like thunder in Edin May 29 15:14:42 ah we have it since 4 days May 29 15:14:52 woglinde: thx now it works. I will change the config May 29 15:14:54 2 times a day May 29 15:15:09 woglinde: cool May 29 15:15:18 hm dont cool May 29 15:15:29 if your inetconnection gets whacky becaus of it May 29 15:16:00 woglinde: heh heh, when I lived near london we would get spectacular storms that lasted 7 days at a time. Damaged quite a few of my computers May 29 15:16:23 hi mrdata May 29 15:16:32 hi mrdata May 29 15:16:34 woglinde: hi May 29 15:16:39 mr_nice: hi May 29 15:17:01 lart kernel for reporting filesystems it dont have support for May 29 15:17:01 * * OE Bug 2395 has been created by bugs.openembedded.org@rolf.leggewie.biz May 29 15:17:04 * * RFE: package manager GUI May 29 15:17:05 woglinde: bang, burning smell, exclamation of "shit I forgot to unplug that" May 29 15:17:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2395 May 29 15:17:21 xora oh hehe never see such a think happen May 29 15:17:35 maybay we german has more thoughts an safety May 29 15:17:52 woglinde: I think it depends if your area has lots of overhead power and such stuff May 29 15:18:11 hm blitzableiter May 29 15:18:14 *g* May 29 15:18:27 woglinde: what is wrong with paule? May 29 15:19:23 Can someone take a look at these an make sure they work for me please? 2361,2368,2370; I would really like this stuff commited... May 29 15:19:44 zecke: do you have a WIP webkit recipe? I don't mind checking out 300MB of svn May 29 15:21:34 koen: none, DEPENDS = flex-native, qmake4-native, uic, libxml2, sqlite3 (no kidding), cairo, gdk, gtk May 29 15:22:05 koen: building is calling qmake-qt4 CONFIG+=gdk-port CONFIG-=qt CONFIG-=release CONFIG+=debug and then typing make May 29 15:22:49 and watch it fail? May 29 15:22:51 what tools does bitbake/oe have in to see when a package's version/release number has been raised, and then (re)build that package? May 29 15:23:34 koen: no, check build.webkit.org it builds most of the time May 29 15:23:46 koen: two patches makes your live a bit better May 29 15:24:16 hi. I'm working on asterisk 1.4 bb files (I'm using openSLUG/NSLU2). Right now there is an asterisk package into the repo and it is pointing to 1.2. Both 1.2 and 1.4 are maintained by the asterisk project. Should I create a new asterisk14 package for asterisk 1.4 or update the existing one? May 29 15:24:17 Zero_Chaos: I will commit them May 29 15:24:35 zecke: look at the comments in the bugs first May 29 15:24:46 zecke: did you make sure they build properly? I believe they do, but I like others to check May 29 15:24:51 yes May 29 15:24:58 opie runs on uclibc May 29 15:25:01 koen: everything you (and others) requested should be fixed. May 29 15:25:05 koen: sir May 29 15:25:06 I'm hoping that I don't have to write a script to parse generated package version/release #'s, then read through .bb files and see if the version numbers have increased May 29 15:25:07 * Zero_Chaos bows May 29 15:25:12 Zero_Chaos: no, it isn't May 29 15:25:28 koen: I will push the ones were it is fixed May 29 15:25:38 koen: I beg forgiveness, I will check again. May 29 15:25:51 Zero_Chaos: #2361 seems fine now May 29 15:26:32 zecke: 2361 and 2368 are set. May 29 15:26:39 Zero_Chaos: #2368, what was wrong with your mtn diff? May 29 15:26:56 zecke: I messed up my repo for a little while so it wasn't working right May 29 15:27:03 zecke: sorry May 29 15:27:24 koen: can you be more specific than "bogus"? May 29 15:27:34 "really bogus" May 29 15:28:03 koen: come on, there is progress May 29 15:28:05 nylon stuff was added by nylon guys and Ive never known if it worked/was needed May 29 15:28:22 XorA: exactly what I was about to say May 29 15:28:26 CONFFILES are always needed May 29 15:28:30 sounds stuff is old crud I probably should have fixed/removed ages ago May 29 15:28:38 using distro overrides for those is bogus May 29 15:28:58 what is it that makes a spitz image have bluetooth related stuff installed by default (with deivce name ipaq BTW). I believe bug 1353 has been fixed so where does this come from? I cannot spot it by looking at the spitz machine conf file. May 29 15:29:03 XorA: doesn't the sounds stuff make an ipk just for the sounds to not waste space? May 29 15:29:18 Zero_Chaos: maybe, I never checked did it work May 29 15:29:32 XorA: yeah, the sounds stuff is cool. May 29 15:29:34 zecke: http://rafb.net/p/cjRYbi43.html May 29 15:29:51 Do we have a way to specify whether a device has built-in Wifi or bluetooth for example? I think we should. May 29 15:30:08 Laibsch: machine features I believe. May 29 15:30:18 Laibsch: see the oe-devel mailinglist from the past few days May 29 15:30:19 Laibsch: something like that anyway. May 29 15:31:04 koen: if you really want me to remove the nylon override I will, I don't care. But what is wrong with making the sound files a seperate package? it's a waste of space for most people. May 29 15:31:31 koen: almost, there is some way to select qmake4, I don't remember it May 29 15:31:50 Zero_Chaos: read the bugreport comment again May 29 15:32:05 Zero_Chaos: and for once start thinking instead of balking May 29 15:32:20 Zero_Chaos: no ${D} May 29 15:33:13 aha, would have been nice to put THAT in the comment May 29 15:33:43 so ${datadir} is /usr/share/ ? May 29 15:33:50 Zero_Chaos: yeah May 29 15:34:01 or something else May 29 15:34:08 whatever you set it to in bitbake.conf May 29 15:34:20 CONFFILES += "${sysconfdir}/kismet.conf" May 29 15:34:28 osas: Here is what I would do. "mtn mv -e packages/asterisk/asterisk_1.2.18.bb packages/asterisk_1.4.4.bb;mtn mv -e packages/asterisk/asterisk_1.2.18/ packages/asterisk_1.4.4". Then try to compile and use it. Fix up the bb file until everything is to your satisfaction. Then do an "mtn diff packages/asterisk/" and attach the output to the bug tracker. May 29 15:35:26 XorA: add that and remove the nylon specific one? when I build I do not need this CONFFILES line May 29 15:35:28 so in other words, trash the 1.2 out May 29 15:35:44 I already have the bb files (as asterisk14) May 29 15:35:56 and tested ok on NSLU May 29 15:36:34 osas: check who feels repspsonsible for asterisk and ask him (hint check the MAINTAINER file) May 29 15:37:15 last time I worked with NAiL in getting asterisk 1.2 in May 29 15:37:23 osas: You can attach that to the BTS as well. The other way would be easier to apply. but either is fine. May 29 15:37:32 The important part is getting it into OE May 29 15:38:04 If possible, make it easy for committers and they will apply your stuff faster and with increased likelihood ;-) May 29 15:38:11 ok. once I polish them, I will attach the files May 29 15:38:23 :-) May 29 15:38:23 Cool, thanks for your contribution May 29 15:38:30 np May 29 15:38:36 Zero_Chaos: yeah May 29 15:39:11 zecke: I think that is up to the person commiting more so than the person submitting updated recipes. May 29 15:39:13 Laibsch: don't advice people to do 'mtn mv' and then submit a patch May 29 15:39:14 Zero_Chaos: that just makes ipkg ask before overwriting the file May 29 15:39:37 XorA: is that really needed? as I said, it doesn't seem to be needed, works for me without May 29 15:39:37 XorA: and show a diff May 29 15:39:50 XorA: ahhh May 29 15:39:51 koen: Why not? If that is what we do, I think it is better than having a patch you cannot "just apply" May 29 15:39:56 XorA: that's a good thing then. May 29 15:40:00 XorA: thanks ;-) May 29 15:40:19 Zero_Chaos: I would have answered long ago if my home machine worked May 29 15:40:50 XorA: no problem, I have been learning a lot. to be honest, it's been fun... May 29 15:40:51 Laibsch: doing 'mtn mv' before 'mtn diff' generates a patch you can't apply May 29 15:40:53 Laibsch: I'm a bit tired. You say removal of recipes is up to the guy comitting and not the one who Maintains it? May 29 15:42:25 IMO non maintainer patches shouldnt remove old versions May 29 15:42:54 * Zero_Chaos seconds XorA's statement May 29 15:43:21 koen: ? May 29 15:43:29 koen: I think I do that all the time. May 29 15:43:47 koen: But probably you are right because I can commit the stuff. May 29 15:43:57 I will heed the advice. May 29 15:45:07 mr_nice, mr_data opie is working now May 29 15:45:18 but the battery information isnt tight May 29 15:45:20 right May 29 15:45:30 for now, I will stick with asterisk14 alternative (so ppl can choose between 1.2 and 1.4) May 29 15:45:31 zecke: No, I say that submitters of updated patches should make sure that a) it compiles and b it runs. Committer should make sure of at least a) and preferably b), but also c) think about consequences for OE as a whole and that includes asking MAINTAINER and other interested parties May 29 15:45:52 hehe May 29 15:45:57 zecke: c) in my opinion is not so important for submitters to BTS without RW access May 29 15:46:22 Laibsch: mtn is sucky since it can't apply all patches mtn diff generates May 29 15:46:22 If they go to the trouble of doing c) as well, I won't chide them for it, though ;-) May 29 15:46:34 koen: It is rather funny that the apple engineers know Gtk much better than I do May 29 15:46:53 woglinde: awsome!! May 29 15:46:54 koen: Is there a bug report against mtn? May 29 15:47:06 woglinde: can you please send me the image? May 29 15:47:06 Laibsch: the developers are aware of it May 29 15:47:15 Laibsch: see org.oe.dev/contrib/mtnpatch.py May 29 15:47:32 the OE devs or the mtn devs? May 29 15:48:25 woglinde: the battery stuff is in working progress. May 29 15:48:33 hm oh May 29 15:48:40 touchscreen isnt working after resume May 29 15:48:44 sorry May 29 15:48:57 Laibsch: both May 29 15:49:13 OK May 29 15:49:41 woglinde: this is a known bug and fixed. but i don't have the code May 29 15:49:46 polyonymous: Your patch fixed the crash from poboxserver, thank you! May 29 15:49:51 mr_nice okay May 29 15:49:53 woglinde: touchscreen works !!! two extra lines of code are neccessary ;-) May 29 15:50:00 polyonymous: It still does not work but that seems to be a font issue. May 29 15:50:02 mrdata provide it *g* May 29 15:50:10 hehe May 29 15:50:14 polyonymous: I'll commit it in a minute May 29 15:50:24 woglinde: damm, "erwischt" May 29 15:50:46 mrdata: do you have time today? May 29 15:51:05 mr_nice: for what? May 29 15:51:19 mrdata: I would also be pleasured if you will be able to find time to send me the kernel changes :) May 29 15:52:02 zecke: where did you hide qmake4-native? May 29 15:52:05 polyonymous: Well, of course it could also be that something is wrong with UTF-8 on opie? May 29 15:52:24 mr_nice: you know, i need install k3diff and must upgrade over 900 packages befor i can do that? May 29 15:53:15 mr_nice: but okay, i can try it, but without garanty May 29 15:54:03 mrdata: hm, I would also take a "diff -uNr linux-2.6.21.vanilla/ your_kernel > changes.patch May 29 15:54:06 :) May 29 15:54:08 thx! May 29 15:57:12 koen: uicmoc/ ? May 29 15:58:01 zecke: uicmoc4-native builds qmake4-native? May 29 15:59:05 koen: no May 29 15:59:09 koen: qmake/ then? May 29 15:59:23 nope May 29 15:59:30 only qmake1 and qmake2 May 29 15:59:44 DEPENDS qmake2-native ;) May 29 16:00:19 anyone knows something like xmlspy for GNU/LInux? May 29 16:02:17 zecke isn't xmlspy java? there are a few good java based xml editor. are you editing docbook? May 29 16:02:20 xee or xxe i can't remember May 29 16:02:34 keesj: OMA Bcast May 29 16:03:05 zecke: I give up on webkit for a while May 29 16:03:23 * koen gets some food and goes to watch the a-team May 29 16:03:26 I got a xsd and want to create a valid minimal document by hand. I can validate using xmllint, but want some easy way to create May 29 16:03:39 koen: catch me at 20h May 29 16:03:45 koen: I will stop working then May 29 16:05:54 * keesj wispers eclipse perhaps, it uses xsd a lot. for the module loading May 29 16:06:18 mr_nice pcmcia needs some love too, to work after suspend May 29 16:06:36 I am also interested in what editor people use to create XML docs. I'd be looking for the mom-and-dad-would-know-how-to-use type. May 29 16:14:55 woglinde: pcmcia works also ;-) May 29 16:15:15 woglinde: one line of extra code May 29 16:17:41 mrdata please enable querys May 29 16:18:14 woglinde: Private messages from unregistered users are currently blocked due to spam problems May 29 16:18:25 woglinde: sorry for that May 29 16:19:09 woglinde: but #opensimpad.org is good a place to talk May 29 16:19:34 woglinde: yes ;) May 29 16:20:27 woglinde: i also want the changes ;) May 29 16:35:59 Jin^eLD: http://www.angstrom-distribution.org/repo/?action=details&pnm=mediatomb May 29 16:39:35 koen: cool, thanks! May 29 16:41:24 where can I find documentation on ipkg-cl? May 29 16:44:11 zecke: http://groups.google.de/group/alt.os.linux/browse_thread/thread/1568bfb792491cef/7ef2926626c49dd9?lnk=st&q=xml+editor+ubuntu&rnum=3&hl=en#7ef2926626c49dd9 May 29 16:44:33 quite a few editors mentioned May 29 16:45:40 Laibsch: vim+xmllint do for now May 29 16:57:00 interesting May 29 16:57:02 * * OE Bug 1272 has been RESOLVED (WONTFIX) by bugs.openembedded.org@rolf.leggewie.biz May 29 16:57:02 oops wrong window :) May 29 16:57:04 * *  Never resume after suspend [...] May 29 16:57:06 * * http://bugs.openembedded.org/show_bug.cgi?id=1272 May 29 17:08:46 hmm, guys, when working with NFS exported rootfs, is there anything that I could use to automatically populate my /etc/resolv.conf? right now the kernel is handling dhcp before mounting the rootfs, but of course that does not take care of /etc/resolv.conf May 29 17:08:51 any hints? May 29 17:15:37 zecke: I'm sucking down webkit now May 29 17:16:01 * * OE Bug 2396 has been created by bugs.openembedded.org@rolf.leggewie.biz May 29 17:16:03 * * flash utility should support repartioning of internal HD May 29 17:16:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2396 May 29 17:19:53 03koen 07org.oe.dev * r14e319e7... 10/ (1 packages/icu/icu_3.6.bb): icu: fix build May 29 17:31:49 * koen is waiting for a gazillion MB of LayoutTests to get downloaded from webkit svn May 29 17:42:17 ah, it finished fetching May 29 17:43:01 206M /home/koen/OE/downloads/trunk_svn.webkit.org_.repository.webkit__20070529.tar.gz May 29 17:43:35 allow me to say 'OMG' May 29 17:43:57 38M LayoutTestResults May 29 17:43:58 342M LayoutTests May 29 17:44:04 that explains a lot May 29 17:45:39 lol May 29 17:46:06 o wait, it's still extracting May 29 17:46:09 455M LayoutTests May 29 17:46:42 hm the application is a faktor 10 smaller than the tests? May 29 17:46:45 lol May 29 17:46:52 chouimat|sick: GWS May 29 17:47:01 GWS? May 29 17:47:07 get well soon May 29 17:47:38 koen I'm ok ... it was about a discussion (joke) in another channel ... May 29 17:49:22 did the resolveconf package work for anyone? May 29 17:51:56 koen: This is why Webkit is so fscking awesome May 29 17:52:21 WebKit can order pizza and bring your children to class May 29 17:52:36 wau May 29 17:55:36 zecke: http://rafb.net/p/K9vAyl95.html May 29 18:01:25 koen: remove Bakefiles May 29 18:01:34 WebKitSize May 29 18:01:36 Site May 29 18:01:57 koen: WebKitLibraries doesn't exist? May 29 18:01:58 done May 29 18:02:11 WebKitQt? heck we build the Gtk port May 29 18:02:11 http://trac.webkit.org/projects/webkit/browser/trunk/WebKitLibraries May 29 18:02:33 hehe, oops May 29 18:02:44 so webkit-qt can go? May 29 18:03:37 Hi. I am looking to hire an experienced OpenEmbedded developer on a project basis for something I have coming up. I apologize if this is not the correct place to make such solicitations. /msg me if interested May 29 18:04:11 yes May 29 18:05:19 http://www.vanille-media.de/site/index.php/freelancer/ http://haerwu.biz/ http://bec-systems.com/web/ and http://o-hand.com/ are a few options May 29 18:05:20 n2aag: I complained at somehow who send a (unrelated) Job Offer to our mailinglist. The general consensus though was that this is okay?! May 29 18:07:08 ws# May 29 18:07:22 wwwy4~~22222222222222q^12222222222222222222222222222222222222222222222222~~f May 29 18:07:36 tobi is typing May 29 18:07:54 zecke: yeah.. I didn't see an oe-jobs list, so I didn't know what to do. I apologize. I will try those sites, but I also thought I would present the opportunity directly to the awesome people who are already working on this stuff. May 29 18:10:20 (just in case anyone is interested, this involves contactless smartcard-based authentication, ipv6, physical access control, and possibly some video streaming. All work will be GPL'd when mature.) May 29 18:12:31 zecke: http://ewi546.ewi.utwente.nl/tmp/webkit-oe.tar.bz2 May 29 18:13:11 koen: I will run home now, will be back later May 29 18:22:05 zecke: http://rafb.net/p/8aVTEU70.html May 29 18:22:14 zecke: summary: /bin/sh: xcodebuild: command not found May 29 18:26:54 it there an oe initrd/ramdisk target? May 29 18:27:16 the angstorm minimal-image contains a kernel May 29 18:39:46 <[cc]smart> Do you think it should be possible to get OE to build it's packages for Infrant NV storage boxes ? May 29 18:40:04 <[cc]smart> Currently, i'd think so but am unsure... May 29 18:41:11 [cc]smart : What architecture is the cpu ? May 29 18:41:18 <[cc]smart> sparc based May 29 18:41:19 in software world nothing is inpossible :P May 29 18:41:39 <[cc]smart> sure. but we are technical so we verify with reality first :) May 29 18:42:19 oxo helped fix the sparc build a few weeks ago May 29 18:42:21 [cc]smart : someone was trying a build with sparc a few weeks ago. not sure how far they wend May 29 18:42:21 Good you have contact with the human beings! May 29 18:42:22 <[cc]smart> the other's are called sales. May 29 18:42:39 lol May 29 18:43:04 * oxo denies everyhing! May 29 18:45:27 cpu specs for those interested -> http://www.analogzone.com/iop_0418.htm May 29 18:47:42 what? sales person are like the marketing one ... not human but another advanced breed of monkey?? I knew it !!!! May 29 18:55:54 Hello May 29 18:56:31 Is there any way I can make Xfbdev load the bitmap fonts in /usr/lib/X11/fonts/75dpi automatically at startup? May 29 18:57:16 Cause some window managers cannot start without those, and have no support for startup commands themselves May 29 18:57:31 is there a variable to disable packages being moved into the "morgue"? May 29 18:58:05 If you install newer versions of the packages, it's natural that the older ones will go into morgue May 29 18:59:03 how do the new ipkg package performs when installing on a target board in terms of speed ? May 29 18:59:13 B_Lizzard: well a "debian" co-worker of mine saw that and thought it was very odd May 29 18:59:18 the older ones are really slow May 29 18:59:39 B_Lizzard: what if I want to build a root file system from a point in the past with older packages? May 29 18:59:59 You specify the packages to be used in the "disto" file May 29 19:00:06 And it uses those May 29 19:00:25 good evening May 29 19:00:36 Good evening May 29 19:00:47 hey mickeyl May 29 19:00:47 mickeyl hi ! May 29 19:01:01 * mickeyl solved telephony & dsl problem May 29 19:01:04 finally phew May 29 19:01:23 mickeyl: I'm afraid to ask..... May 29 19:02:06 B_Lizzard: so if I specified PREFERRED_VERSION_something = "1.0", and something_1.0-r1_all.ipk was in the "morgue", it would get picked up when I build my root image? May 29 19:02:22 Yeah, probably May 29 19:02:37 no, it wouldn't May 29 19:02:52 Would it have to rebuild it, then? May 29 19:02:55 koen: you would be surprised what a man with a screwdriver and access to the cellar can do May 29 19:02:56 :D May 29 19:02:59 Cause the stamps are there May 29 19:03:32 mickeyl: a huge mess? May 29 19:03:35 mickeyl: I usually run away screaming when a CS student presents a 'hardware fix' ;) May 29 19:04:20 koen: so the package would need to rebuilt in my example above? May 29 19:04:25 hehe May 29 19:04:27 well May 29 19:04:32 switching two cables was a 5min job May 29 19:04:48 jsheldon: no, since a newer version is present, so ipkg-make-index will move it to the morgue again May 29 19:04:49 overcoming my hesitation was much more of a problem May 29 19:04:55 heh May 29 19:05:11 "i shouldn't be doing that... well, who cares" May 29 19:05:32 mickeyl: didn't get caught by the 'germany uses the the other wires in rj11' problem? May 29 19:05:48 koen: interesting, I know we'll definitely be at a point soon where we'll want to build root file systems from old packages, I guess I'll have to find a way around that May 29 19:05:52 s/the the/the two/ May 29 19:06:14 Laibsch, here? May 29 19:06:21 koen: hehe, luckily not. once i indentified my cables in the main hub, it was a piece of cake May 29 19:11:02 * * OE Bug 2397 has been created by koen@dominion.kabel.utwente.nl May 29 19:11:04 * * libmimedir package depends on -dev package May 29 19:11:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2397 May 29 19:20:02 * * OE Bug 2227 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 19:20:03 * *  insane rpath check has troubles with funny [...] May 29 19:20:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2227 May 29 19:21:10 03polyonymous 07org.oe.dev * r181d0167... 10/ (1 classes/insane.bbclass): classes/insane.bbclass: quote filenames to avoid problems with funny characters. Closes bug #2227 May 29 19:23:23 polyonymous: yes? May 29 19:23:39 Laibsch, so, what didn't work for you? For me it works. May 29 19:23:55 I mean, I can type the first column from dictionary and see the second one. May 29 19:24:10 Have you rebuilt both packages? May 29 19:24:13 * koen prods atmel for some at91 and avr32 boards May 29 19:24:42 koen: the AT91 and AVR32 stuff looks pretty nice! May 29 19:25:01 cbrake: I was prompted by your blog entry :) May 29 19:25:09 koen: :-) May 29 19:25:37 polyonymous: yes, I have rebuilt both May 29 19:25:41 koen: I just realized last week the AVR32 included a MMU and was capable of running standard Linux. May 29 19:25:41 url? May 29 19:25:58 polyonymous: Do you see actual chars instead of just square boxes? May 29 19:26:06 chouimat|nap: http://bec-systems.com/web/index.php?option=com_content&task=view&id=70&Itemid=9 May 29 19:26:13 Laibsch, yes. May 29 19:26:26 polyonymous: OK, I will retry, then May 29 19:26:28 Laibsch, have you restarted pbserver? May 29 19:26:42 polyonymous: I have restarted opie May 29 19:26:56 Laibsch, pbserver is restarted separately. May 29 19:27:10 Either with /etc/init.d/pbserver stop/start or reboot. May 29 19:27:16 polyonymous: What font did you use? Was it readable? May 29 19:27:24 polyonymous: I'll just reboot May 29 19:27:31 cbrake: ever played with a blackfin? May 29 19:27:40 koen cbrake nice May 29 19:27:44 koen: no, I never have May 29 19:28:02 Laibsch, I've tried quite a few strings which I don't remember, of course. May 29 19:28:14 "Blackfin refers to a family of 16/32-bit microprocessors and incorporate Digital Signal Processor (DSP) capabilities and features more commonly found in micro-controllers in a power-efficient architecture. The result is a low-power, unified processor architecture that can run operating systems while simultaneously handling complex numeric processing tasks like real time H.264 video encoding, for example." May 29 19:28:40 koen: hey, how far are you? May 29 19:28:55 ~fishslap gtk for gdk_window_process_updates (layout->bin_window, TRUE); May 29 19:28:57 zecke: bombs out on missing xcodebuild May 29 19:28:58 * ibot slaps gtk up side the head with a wet fish for gdk_window_process_updates (layout->bin_window, TRUE); May 29 19:29:11 some guy in my lab has a blackfin May 29 19:29:21 koen: install it? May 29 19:29:35 mkdir -p WebKitBuilds/Debug May 29 19:29:39 cd WebKitBuilds/Debug May 29 19:29:43 * chouimat|nap does the happy guy dance May 29 19:30:50 polyonymous: Still nothing but squares after a reboot. May 29 19:30:57 polyonymous: I think this is a font issue. May 29 19:31:03 What font did you use? May 29 19:31:22 qmake -r OUTPUT_DIR=$PWD/ CONFIG-=qt CONFIG+=gdk-port $PWD/../../WebCore.pro May 29 19:31:29 koen: and then make May 29 19:33:48 Laibsch, unismall May 29 19:34:03 (set via appearance) May 29 19:34:20 zecke: sweet May 29 19:34:41 Laibsch, could it be because I have globally (via profile) enabled utf8 locale? May 29 19:34:57 zecke: http://ewi546.ewi.utwente.nl/tmp/webkit-oe.tar.bz2 updated May 29 19:35:15 <[cc]smart> somebody here who knows hx2000 arch well and can do uclibc.machine file for it ? May 29 19:36:11 [cc]smart: copy the c7x0 one May 29 19:36:19 ehm May 29 19:36:23 the hx4700 one May 29 19:36:26 koen qmake2? May 29 19:36:28 <[cc]smart> k'll do May 29 19:36:31 <[cc]smart> thx May 29 19:36:47 <[cc]smart> will try testrun first... May 29 19:36:51 zecke: ../../../JavaScriptCore/kjs/regexp.cpp:139: error: cannot convert 'const uint16_t*' to 'const char*' for argument '3' to 'int pcre_exec(const pcre*, const pcre_extra*, const char*, int, int, int, int*, int)' May 29 19:36:52 chouimat|nap: qmake-qt4 --version May 29 19:37:06 zecke: k May 29 19:37:20 koen: wrong include order... May 29 19:37:25 [15:36] new KDE4 on Qtopia-Core video: http://www.youtube.com/watch?v=T9ozRGmQPBs May 29 19:37:38 hehe, saw that yesterday by accident May 29 19:40:08 zecke how do you accidently search for Qtopia? May 29 19:40:56 cbrake: just sent you an email :) May 29 19:41:07 Kristoffer: he was looking for porn :) May 29 19:41:07 Kristoffer: I saw a video on beryl vs. vista May 29 19:41:20 Kristoffer: and related was bIp0lar May 29 19:41:29 Kristoffer: or I saw the plasma videos May 29 19:42:24 but actually my sith force made me know this even before bipolar decided to compile KDE4 for QtopiaCore May 29 19:42:42 what's the use? May 29 19:42:57 if your device can run kde, it can run it on regular qt May 29 19:42:58 polyonymous: That could be. How did you do that? /etc/profile? May 29 19:43:09 Laibsch, hang on. May 29 19:43:23 03koen 07org.oe.dev * rd35e5bf1... 10/ (8 files in 4 dirs): WebKit: add WIP recipe to build the webkit-gdk port May 29 19:43:36 zecke: there you go May 29 19:43:44 wish some kernel hacker could fix my usbnet not working after suspend :) May 29 19:43:49 koen propably to make sure you can use some of the kde4 stuff with qtopia 4 ... May 29 19:43:51 * koen hopes people will get annoyed and fix that recipe May 29 19:45:06 koen: but maybe they want a cut down size of qt as well as minimal kde. and perhaps they dont need no stinkin' X :) May 29 19:45:14 koen bipolar told me that the next step is to integrate the qtopia ui :) May 29 19:45:58 Laibsch, but basically I have something like /etc/profile.d/locale with export LC_CTYPE=en_US.utf8 May 29 19:47:10 hi all May 29 19:47:30 Laibsch, just checked - exactly what I've said. May 29 19:48:12 Laibsch, I'll try to run pbserver without locale now. May 29 19:48:47 someone was asking about my kde4 on qtopia-core work? May 29 19:48:59 polyonymous: Any locale packages to install for that? May 29 19:49:37 Laibsch, I think locale-base-en-us, but lemme see its postinst.. May 29 19:49:47 zecke: what were those 2 webkit patches you were talking about? May 29 19:50:22 Laibsch, yes, I think that's it, but it may by different if you have binary locale generation on. May 29 19:50:36 yes, I have it on May 29 19:51:18 koen: http://bugs.webkit.org/attachment.cgi?id=14750 May 29 19:51:25 Laibsch, then try putting the locale file into your profile.d. I'll try to do the opposite :) May 29 19:51:52 koen: I might need to rediff this patch (or git can) May 29 19:52:20 http://bugs.webkit.org/attachment.cgi?id=14268&action=view May 29 19:53:20 Laibsch, with utf-less pbserver it works fine. May 29 19:53:43 polyonymous: for me it is completely broken May 29 19:53:53 And I believe this is more than a font issue. May 29 19:54:03 What happens if you type ni? May 29 19:54:07 n+I May 29 19:54:17 one character or two? May 29 19:54:55 heckofalot of characters if I get your right May 29 19:55:23 yes, of course. May 29 19:55:24 Laibsch, lemme enable screenshot applet :) May 29 19:55:31 type nihongo, please May 29 19:55:38 and put up a screenshot May 29 19:55:45 ok May 29 19:56:09 Laibsch, some are just square boxes, hang on. May 29 19:57:53 Laibsch, I have problems with ssh on my server, would you mind if I dcc it? May 29 19:58:51 dcc? May 29 19:58:57 Send it straight to me? May 29 19:59:03 like this. May 29 19:59:13 that would of course be OK May 29 19:59:17 can you accept it or not? May 29 19:59:30 great, you can. May 29 20:00:44 was someone in here asking about my work of porting kde4 to qtopia core? I'm here to answer questions if anyone has any.... May 29 20:01:42 polyonymous: The first entry is OK. Not sure why the others show as boxes. May 29 20:01:52 But it seems to be working to a certain extent May 29 20:02:01 Laibsch, I'm totally clueless about it... May 29 20:02:10 Lemme see what it looks like in dictionary... May 29 20:02:51 Laibsch, well, the dictionary only has first three characters. May 29 20:02:55 http://pastebin.ca/519924 May 29 20:03:03 I don't knoe where does the rest come from. May 29 20:03:08 anyone have any thoughts on this u-boot linking issue? May 29 20:03:29 This is the line from staticdic: nihongo 日本語 May 29 20:03:45 Pardon my Japanese :) May 29 20:03:50 Crofton: iirc a broken linkerscript May 29 20:04:05 hmmm May 29 20:04:17 the symbols aren't defined in the u-boot libs May 29 20:04:28 they define some, but not all May 29 20:04:39 it is very curious May 29 20:04:59 polyonymous: Where did you look that up? As I said, those glyphs are OK. The square boxes are another story. May 29 20:05:02 Crofton: iirc I had the same problem with the neo1973 uboot, but I can't recall how I solved it May 29 20:05:14 Laibsch, /opt/QtPalmtop/pobox/staticdic May 29 20:05:16 zecke: feel free to add those patches to the recipe :) May 29 20:05:24 zecke: any hint on the compile error? May 29 20:05:28 interesting May 29 20:05:34 koen: yes, remove pcre ;) May 29 20:05:38 the gta ones? May 29 20:06:04 koen: or make .pc files have -isystem in the includes May 29 20:10:30 svn.openmoko.org, connection refused? May 29 20:12:11 polyonymous: I think I might be reexperiencing May 29 20:12:15 !oebug 1673 May 29 20:12:16 * * Bug 1673, Status: NEW, Created: 2006-12-10 17:15 May 29 20:12:17 * * bugs.openembedded.org(AT)rolf.leggewie.biz: busybox sh not utf-8 safe May 29 20:12:18 * * http://bugs.openembedded.org/show_bug.cgi?id=1673 May 29 20:12:35 although this might after all not be a busybox issue May 29 20:13:02 Laibsch, have you enabled LC_CTYPE? May 29 20:13:04 Can you "grep nihongo /opt/QtPalmtop/pobox/staticdic" and receive readable glyphs? I don't. May 29 20:13:09 yes May 29 20:13:24 it is there when doing "set|grep LC" May 29 20:13:34 Laibsch, I can. But I have fullblown grep, I think May 29 20:14:11 Laibsch, yes, that should do. I'd just do a locale instead, though :) May 29 20:14:17 hi all, please, have a look into my problem and try to help me to figure it out: I have a task-test that has PACKAGES = "\ task-test-base \" and a list of depends of this task-test-base: RDEPENDS_task-test-base = "......" /*the same structure as task-openmoko*/ But the bitbake is not building the packages inside this RDEPENDS_task-test-base... Any idea? May 29 20:14:59 polyonymous: locale is not installed. Maybe that is the problem? May 29 20:15:02 the opnemoko svn has changed from svn to http May 29 20:15:10 Laibsch, I don't think so. May 29 20:15:22 Laibsch, it's in glibc-utils May 29 20:15:57 But please, do install it and type locale -a May 29 20:16:47 It seems I have just C and POSIX May 29 20:17:21 bipolar: apparently not... :) May 29 20:17:58 Laibsch, That may be a problem :) May 29 20:18:30 Although, like I've said, I've tried disabling poserver locale... I'll try it with opie now. May 29 20:18:54 jpwhiting: hehe... yeah. everyone hid :P May 29 20:19:52 Laibsch, with localeless opie it also works... May 29 20:23:21 Laibsch, where do you grep your staticdic? May 29 20:24:33 Laibsch, I get squares in opie-console. In ssh from urxvt it's okay, though. May 29 20:25:13 Laibsch, but, of course, I have non-jp font in opie-console May 29 20:26:07 in the standard terminal app May 29 20:26:25 Laibsch, on handheld? May 29 20:26:36 do you see utf in terminal at all. May 29 20:26:49 which of the thousand locale packages do I need to install (en_US ones, of course) May 29 20:27:01 polyonymous: Of course on the Z. May 29 20:27:22 polyonymous: I am not sure I do see utf8. that is what I am trying to find out. May 29 20:27:47 Laibsch, is the application opie-console? May 29 20:28:07 (That is 'console' icon in applications) May 29 20:31:41 yes May 29 20:32:18 Laibsch, Connection/Configure profiles/Edit/Terminal I think is where you set the font. May 29 20:33:58 interesting. May 29 20:34:07 've never seen and used that May 29 20:34:54 Laibsch, you can also just ssh from utf-enabled terminal on your desktop :) May 29 20:36:40 no, unfortunatley, I can't May 29 20:36:47 no network connection to the Z May 29 20:37:05 Laibsch, no usb cable? May 29 20:37:17 a Terminal with unismall is acting very strange May 29 20:37:25 no, no usb cable. May 29 20:37:33 I think unismall is proportional font? May 29 20:37:40 Is that part of the standard package? May 29 20:38:23 well, let's not bother with unismall being strange (it looks like as if it was introducing random spaces here and there in the console) May 29 20:38:38 usb cable? It was packaged with me by xtops.de, but not by sharp. May 29 20:43:13 Laibsch, this misbehaviour is common with non-monospace fonts, I think May 29 20:43:48 wow, I really need to create a bigger unismall font May 29 20:43:53 This is impossible to read. May 29 20:45:12 polyonymous: Are you sure en_US.utf8 is correct and not en_US.UTF-8? May 29 20:45:21 Laibsch, yes. May 29 20:45:32 Laibsch, what does locale -a say? May 29 20:46:08 Laibsch, btw, can you try to md5sum the staticdic? May 29 20:46:11 polyonymous: Told you, C and POSIX May 29 20:46:11 afc8337d5ecf4a679da1858091758810 /opt/QtPalmtop/pobox/staticdic May 29 20:46:32 Laibsch, you're missing something May 29 20:47:18 Laibsch: any news on the cabels? May 29 20:47:26 Laibsch, do you have binary locale generation on? May 29 20:47:33 My staticdic returns a different md5sum May 29 20:47:43 It starts with 6507 May 29 20:47:56 hvontres|home: nope, you'll be the first to know. May 29 20:48:03 polyonymous: yes May 29 20:48:48 Laibsch, I don't know what are the names for binary locales packages, but you should install one... May 29 20:49:25 And this difference may mean you have non-utf staticdic somehow. Are you sure you have patched pbserver installed? May 29 20:50:24 I think I have PR bumped but I'm not sure how did you go about building/installing with no network and whether you overlooked it somehow. May 29 21:03:38 Laibsch, 6507dd282c4757cce90eca2374553821 staticdic ? May 29 21:03:50 It's euc-jp. You don't have utf-8 staticdic. May 29 21:04:30 upgrade your pbserver and you're all set, I think. May 29 21:26:07 polyonymous: I had to make the changes from your patch manually since I had moved around a few files. Made a mistake doing that and thus OE took the unicode.patch to be a file instead of applying it as a patch. May 29 21:26:12 * Laibsch is rebuilding now May 29 21:26:23 Sorry for waisting your time May 29 21:26:27 wasting May 29 21:30:31 Laibsch, aha, I see. have you moved them into the same dir? May 29 21:30:49 Thought that might make sense. May 29 21:51:16 polyonymous: Yes, that is exactly what I did May 29 21:51:54 Laibsch, and how is it going now? May 29 21:52:03 Have to check later May 29 21:52:07 ah ok May 29 22:05:02 * * OE Bug 2323 has been RESOLVED (FIXED) by shane.oe@hathawaymix.org May 29 22:05:04 * *  angstrom collie: no suspend support [...] May 29 22:05:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2323 May 29 22:05:35 morning May 29 22:06:09 polyonymous: NICE May 29 22:06:27 Working now. But there are still some issues that I did not witness previously May 29 22:06:55 I'll commit your current work and we can take it from there. It think it is basically usable. May 29 22:07:30 heh May 29 22:07:39 Laibsch: be a pal and commit two of mine as well? May 29 22:07:54 Laibsch: 2361 and 2368 May 29 22:08:26 Well, it's hard for me to debug since I'm no potential user of it... May 29 22:11:18 Zero_Chaos: it is possible to be too pushy, yes I believe that. May 29 22:11:59 Laibsch: sorry, I was merely reminding you that two of mine are ready for commit. May 29 22:12:10 polyonymous: Two things. One is that there should not really be those square boxes. May 29 22:12:48 Laibsch, I think I'd have to read the whole source to see what there _should_ be. I mean, I haven't used it before. May 29 22:12:56 second is that it should not act prematurely. Japanese is a double-byte language and it seems that pobox sometimes now acts on single bytes. May 29 22:13:07 hmm... May 29 22:13:21 It should not offer a glyph when you hit just one key May 29 22:13:33 Are you sure it was different before? May 29 22:13:53 I think previously it just put that key into the "waiting area" where you can see your input. May 29 22:14:12 polyonymous: I am pretty sure. But I need to verify with my OZ installation May 29 22:14:17 I don't think the lookup part is latin, so it is NOT double-byte. May 29 22:15:04 Zero_Chaos: I think we have about at least two hundred ready for commit. One I committed the first half either this morning or yesterday which had been in the tracker for almost two years!!! May 29 22:15:20 Zero_Chaos: So a little patience helps us all get stuff done. May 29 22:16:24 Laibsch: I wasn't trying to ride you, just trying to be commited with your next batch. May 29 22:16:45 OK, understood May 29 22:17:09 trying to make it easier, not be a pain in the ass ;-) May 29 22:17:16 I think it is good you are pushy so as not to let your stuff rot in there for two years until it eventually becomes worthless eventually. May 29 22:17:17 ah, pbserver has a debug mode, nice. May 29 22:17:34 But coming here more than once a day does not help and will not get a faster commit. May 29 22:18:11 Laibsch: I actually am in the channel all day everyday, and I have been for about 2 years now.... May 29 22:18:16 I think you can sit back now for at least a week. If there is still stuff left (I am sure there will be), some back May 29 22:18:37 being here and asking for "be my commit pal" is different May 29 22:18:45 * Laibsch thinks nuff said about this May 29 22:18:47 back to work May 29 22:19:19 I reiterate, I appreciate you being pushy, but don't overstretch it May 29 22:19:23 Just my 2 cents May 29 22:19:29 understood May 29 22:20:06 Laibsch, I'm pretty sure this two-character things isn't new. May 29 22:21:12 polyonymous: Could be. Just like you discovered before that something which was working was working by accident really. May 29 22:21:31 I#ll try and reflash my collie May 29 22:21:40 I mean, not all of the latin words have even number of characters May 29 22:21:54 ??? May 29 22:21:58 What do you mean? May 29 22:22:11 Like, 'sow', for instance. May 29 22:22:28 hi May 29 22:22:28 polyonymous: You are going in the wrong direction May 29 22:22:35 hi, CoreDump|home May 29 22:22:36 hi, wb! May 29 22:22:45 CoreDump|home: re May 29 22:23:03 Laibsch, well, I think you just described the problem in a strange way :) May 29 22:23:18 polyonymous: Japanese is double-byte. Meaning you need two bytes to encode it since with one byte you can only do 256 different chars May 29 22:23:31 polyonymous: might be ;-) May 29 22:23:32 yes, but what you enter is not a japanese charset. May 29 22:23:43 Of course not May 29 22:23:50 so this doesn't really apply. May 29 22:23:58 And maybe this is not about charset May 29 22:24:17 But just h does not exist in Japanese for example May 29 22:24:23 morevoer, you also have stuff like dot, numbers, which are all one-byte (in latin representation) May 29 22:24:31 yes, that's possible May 29 22:24:47 but still, it should do a lookup after a single byte May 29 22:25:05 You have ha は hi ひ hu ふ he へ ho ほ May 29 22:25:32 Laibsch, what about these: May 29 22:25:38 畔 May 29 22:25:38 a 畦 May 29 22:25:47 a i u e o also exist, as well as n. But other than that it is always transcribed with at least two characters. May 29 22:25:47 both are 'a', just miscopied May 29 22:26:10 polyonymous: that is utter rubbish, produced by the incorrectly working pobox May 29 22:26:12 Yes, but that essentially means that lookups should be fired when you only have one byte. May 29 22:26:25 no, I'm copying that from staticdic :) May 29 22:26:29 but there should mostly be no returns May 29 22:26:57 the thing is that pobox itself is dictionary-agnostic. May 29 22:27:11 and pbserver does not always do an exact lookup May 29 22:27:17 polyonymous: That is a very obscure glyph. May 29 22:27:32 Let me check in an online dict if it really is being read as a May 29 22:27:47 Laibsch, Well, it's all Japanese to me :) May 29 22:28:05 as I said a can stand by itself May 29 22:28:25 and that's enough to justify an immediate lookup May 29 22:28:28 polyonymous: You're lucky it's not greek;-) May 29 22:28:35 I couldn't help you with that May 29 22:28:43 yes, for a May 29 22:28:56 But not for h k p and a lot of others May 29 22:28:58 I once received a fanmail about my software saying "computers are like russian cyrillic to me" :) May 29 22:29:05 Had to answer, "same here" :) May 29 22:29:26 Laibsch, again, pobox applet is dictionary-agnostic. May 29 22:30:15 the strange glyph would read aze, not just a. So maybe the conversion was not OK. May 29 22:30:50 I am not a unicode expert, but I think unicode is also double-byte May 29 22:31:11 Laibsch, it wasn't exactly a conversion, I just changed output format. But yet, maybe the problem lies somewhere in dictionary generation. May 29 22:31:19 unicode is variable-length. May 29 22:31:27 And I vaguely remember that some portions (at least in UTF-16) were single-byte, others double-byte and some even triple-byte May 29 22:32:10 well, unicode is just an enumeration and utf8 is a variable-length transmission. May 29 22:32:32 That is it only has 127 single-byte sequences. May 29 22:32:34 you are more of an expert than me May 29 22:32:51 I don't know where in unicode exactly japanese script is, but it doesn't matter. The lookup key is latin. May 29 22:33:47 Laibsch, in tmp/work/poboxserver/... there is a fugodic.txt file with the "original" dictionary. May 29 22:33:55 But I'm not sure what encoding is it in yet :) May 29 22:34:43 There may be a problem in some ruby dic filter used in build as well... (you did say it's rubbish, didn't you?:)) May 29 22:35:15 Laibsch, commit your work, anyway, it would be easier to play with it then. May 29 22:35:28 guess you can't break it if it ain't fixed. May 29 22:37:02 * * OE Bug 2323 has been REOPENED by thommy@tabao.de May 29 22:37:04 * *  angstrom collie: no suspend support [...] May 29 22:37:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2323 May 29 22:37:40 Laibsch: When you're at reflashing you collie, you could test the suspend-patch at 2323 ;) May 29 22:40:57 thesing: I was reflashing to OZ May 29 22:41:17 but I have another collie and I am building an image for it now. May 29 22:41:39 Do I need to apply the patch or is it in the main tree? May 29 22:42:52 you need to apply it. May 29 22:44:19 OK, I will do that May 29 22:46:39 Thanks. May 29 22:48:02 03polyonymous 07org.oe.dev * r7cf4ac5d... 10/ (1 packages/m4/m4_1.4.8.bb): m4: removed -1 preference from 1.4.8, closes bug #837 May 29 22:48:02 * * OE Bug 837 has been RESOLVED (FIXED) by polyonymous@klever.net May 29 22:48:03 * *  m4 upgrade would let automake on handheld [...] May 29 22:48:05 * * http://bugs.openembedded.org/show_bug.cgi?id=837 May 29 22:48:32 you might want to play with kexec to avoid reflashing for kernel testing. May 29 22:52:12 Laibsch, can you read the jis 7 bit encoded files? May 29 22:53:03 thesing: I don't mind reflashing. But if you provide me with a pointer to some easy to understand instructions, I'd take a look May 29 22:53:12 * Laibsch knows close to nothing about kexec May 29 22:53:20 polyonymous: which ones? May 29 22:53:33 ipkg install kexec-tools :) May 29 22:53:44 The fugodic.txt in pbserver's workdir May 29 22:53:51 I guess, I'd need to bake them first May 29 22:54:08 If you could check whether this 'a' issue is in the dictionary that might be of help. May 29 22:54:14 and I guess I still would not know what the kexec-tools package does May 29 22:54:16 And yes, you'd need to bake them first :) May 29 22:54:21 Laibsch: kexec -l /media/cf/zImage; kexec -e will do May 29 22:54:32 polyonymous: nono, was talking about kexex May 29 22:55:11 I was talking about both :) May 29 22:55:29 but I'd rather let thesing talk about kexec to avoid confusion :) May 29 22:55:54 thesing: OK, I'll try to remember that. How does that play along with altboot? And how do I interrupt the boot process to issue that command? Or do you want me to build the image with that command hard-wired in? May 29 22:56:34 just build a console image. it doen't contain altboot afaik. May 29 22:56:53 flash it and install kexec-tools May 29 22:57:34 no, it doesn't but it is usually one of the first installs I made/make May 29 22:57:43 when you want do try a new kernel later you can put it on a CF and use "kexec -l /media/cf/zImage" to load it and "kexec -e" to start it. May 29 22:57:53 I like altboot very much May 29 22:58:49 I don't know if it supports kexec on collie. May 29 22:59:05 I don't think it supports kexec anywhere yet May 29 22:59:14 hvontres|home: is working on it AFAIK May 29 23:01:02 btw you may want to restart the kernel build as I just found a bug with the suspend key. (be sure to refetch locomokeyb-suspendkey.patch ) May 29 23:02:59 OK May 29 23:03:01 * * OE Bug 2398 has been created by bugs.openembedded.org@rolf.leggewie.biz May 29 23:03:03 * * RFE: altboot should not require shift key for selection May 29 23:03:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2398 May 29 23:21:40 guys, what is "RFE" please= May 29 23:21:41 ? May 29 23:22:01 radio free europe? :) May 29 23:22:10 * zecke schaut peinlich beruehrt May 29 23:22:34 * polyonymous thinks it's request for enhancement, though. May 29 23:22:54 unless you're into russian far east :) May 29 23:23:11 Is there any way I can make Xfbdev to load the fonts in /usr/lib/X11/fonts/75dpi? May 29 23:23:19 autoload May 29 23:23:37 I can do it with xset, but it's not the most optimal solution May 29 23:24:12 And as we don't have a xorg.conf file, I was wondering if there's another way of doing it May 29 23:24:37 (Even if it is just a script called by Xfbdev that calls xset) May 29 23:25:00 I tried .xsession and .xinitrc, but as expected, I got nothing May 29 23:27:31 polyonymous: that makes sense, thanks May 29 23:28:09 CoreDump|home: heh...I think I have a solution for that...:) May 29 23:28:24 hvontres|home: ideed =) May 29 23:29:20 CoreDump|home: Do you remember how to set a local dir as SRC_URI ? May 29 23:29:38 hvontres|home: file://blah May 29 23:31:14 CoreDump|home: and that will grab the whole dir? May 29 23:31:41 exactly May 29 23:31:47 CoreDump|home: cool...thanks May 29 23:32:06 it would grap packages/${PN}/blah May 29 23:32:54 Hi, I built angstrom-x11-image for the h5000 with the 2.6.21-hh5 kernel ant it panics. Bootlog in http://rafb.net/p/m7BGPY20.html . Help is apreciated. May 29 23:36:37 should I file a bug? May 29 23:44:38 * thesing -> sleep May 29 23:53:58 CoreDump|home: yes, RFE is the intended (and I thought well-known) meaning here May 29 23:54:13 rfe = request for enhancement May 29 23:54:53 Laibsch: that one was new to me =) May 29 23:55:08 do you know what pcmcia stands for? May 29 23:55:43 (pcmcia = people can't memorize computer industry acronyms) May 29 23:57:51 heh May 30 00:01:08 is there a way to setup an OE env for 2 different machines? May 30 00:01:42 for two different targets? yeah May 30 00:01:52 summatusmentis: What do you mean? May 30 00:01:54 multimachine.inc or something like that May 30 00:02:00 One OE tree, two build machines? May 30 00:02:02 * * OE Bug 1729 has been RESOLVED (FIXED) by polyonymous@klever.net May 30 00:02:04 Laibsch: collie and spitz May 30 00:02:04 * *  ssh-agent dies when opie daemonizes [...] May 30 00:02:05 * * http://bugs.openembedded.org/show_bug.cgi?id=1729 May 30 00:02:16 03polyonymous 07org.oe.dev * r99133b6a... 10/ (3 files in 3 dirs): opie-init: spawn ssh-agent separately. Closes #1729 May 30 00:02:19 Well, time to sleep now. May 30 00:02:20 summasmentis: Easy May 30 00:02:26 Laibsch: I wanna play w/ angstrom on the collie too May 30 00:02:28 how? May 30 00:02:38 MACHINE ?= "spitz" in local.conf May 30 00:03:00 and env MACHINE=spitz bitbake opie-kdepim-image :) May 30 00:03:05 MACHINE="collie" bitbake $foo when building for collie May 30 00:03:12 not sure if this is exactly the image you'd want to bake, though :) May 30 00:03:19 goodnight everyone. May 30 00:03:34 But collie is oabi while spitz is eabi May 30 00:04:09 will that make a difference? I've got limited space on my drive, so... May 30 00:04:18 So you need to also set distro differently. An attempt to do this like DISTRO_collie=blah failed May 30 00:04:33 DISTRO May 30 00:06:00 Laibsch: DISTRO="angstrom-2007.1" MACHINE="collie" bitbake angstrom-console-image ? May 30 00:06:32 You need to have both DISTRO and MACHINE with ?= not = in local.conf, then May 30 00:06:42 That should work almost May 30 00:06:53 DISTRO="angstrom-2007.1-oabi" MACHINE="collie" bitbake angstrom-console-image May 30 00:07:17 and May 30 00:07:17 03lenehan 07org.oe.dev * rbbc47aa3... 10/ (3 files in 3 dirs): May 30 00:07:17 perl 5.8.8: Fix the CFLAGS being used for perl modules. Instead of replacing May 30 00:07:17 the CFLAGS from the perl configuration with the OE CFLAGS we prepend the OE May 30 00:07:17 CFLAGS. This is needed to ensure that compiler defines (such as _GNU_SOURCE) May 30 00:07:17 DISTRO="angstrom-2007.1" MACHINE="spitz" bitbake angstrom-console-image May 30 00:07:18 are correctly definied when building perl modules. This primarily effects May 30 00:07:18 modules that link against the core perl library, such as libxml-parser-perl. May 30 00:07:22 03lenehan 07org.oe.dev * rf33c14e3... 10/ (8 files in 2 dirs): May 30 00:07:24 perl: Bump PR on modules that include shared libraries since these may have May 30 00:07:26 been effected by the revent CFLAGS changes to perl 5.8.8. May 30 00:07:28 03lenehan 07org.oe.dev * rcfdef5c7... 10/ (6 files in 2 dirs): May 30 00:07:30 perl: Add some more modules which I've been using locally incase someone May 30 00:07:32 else happens to want to use them. May 30 00:07:34 03lenehan 07org.oe.dev * r8837110e... 10/ (3 files in 3 dirs): May 30 00:07:38 uclibc 0.9.29: Move the angstrom configuration to the global configuration. May 30 00:07:40 This allows uclibc to build for other distros and it'll make life a lot May 30 00:07:42 easier for everyone if we can have a single configuration for all distros. May 30 00:07:44 Long term we may need to re-add distro specific configurations, but for now May 30 00:07:46 lets see if we can all use the same one. May 30 00:08:03 Laibsch: do the DISTRO and MACHINE vars need to be set for spitz? if that's the default in local.conf? May 30 00:08:23 no May 30 00:08:33 that is how I do it May 30 00:08:37 ok, thanks May 30 00:22:22 mplayer_svn.bb:TARGET_CC_ARCH_spitz = "-march=iwmmxt -mtune=iwmmxt" May 30 00:22:22 mplayer_svn.bb:PACKAGE_ARCH_spitz = "iwmmxt" May 30 00:22:22 m May 30 00:22:25 ?! May 30 00:24:34 I don't even know ( or care ) what iwmmxt is supposed to be, but PACKAGE_ARCH_spitz should stay "spitz" May 30 00:26:01 CoreDump|home: iwmmxt is for acceleration... Let's mplayer run faster on PXA27x XScales. Which "spitz" has. May 30 00:26:48 well, making it MACHINE specific is cool, but inventing a whole new arch? May 30 00:27:42 It is a whole new arch. You can't run it on a standard arm system, since there are iwmmxt instructions inside the app. May 30 00:27:53 "MPlayer was compiled without TV interface support." meh May 30 00:28:28 Futaris: I understand that. In the past we set PACKAGE_ARCH to ${MACHINE} in these cases May 30 00:30:22 But you can run it on another iwmmxt system, not necessarily spitz. There is a machine/include/tune-iwmmxt.conf for a reason. aximx50, akita, asus730, htcuniversal, magician, hx2000, hx4700 all accept iwmmxt. May 30 00:31:57 if everyone were to create a new arch in such a case, we would end up with tons of them very soon ;) May 30 00:32:26 Look in spitz.conf: PACKAGE_EXTRA_ARCHS += "iwmmxt" May 30 00:35:05 Zero_Chaos: being overly pushy also includes marking your own bugs as blocking 2194! May 30 00:35:19 Please stop doing that. May 30 00:41:47 Futaris: I saw that, but my point is another one ;) May 30 00:43:23 CoreDump|home: I know there is a problem with no "iwmmxt" arch. I don't think "iwmmxt" arch gets set if you are building for akita, etc. I don't think there is an easy way to deal with it from mplayer bb unfortunately. May 30 00:43:58 PACKAGE_EXTRA_ARCHS += "iwmmxt" should take care of that May 30 00:44:36 CoreDump|home: I thought that only takes care of the dependency side of things. Not the compilation side... May 30 00:44:43 true May 30 00:45:08 w/o it you can't install this package arch onto the machine May 30 00:45:23 Yep. But that's the way its supposed to be. May 30 00:45:44 You don't want mplayer w/ iwmmxt on a standard arm machine. May 30 00:46:24 I don't want new arches to group a single package, but I guess that's just me May 30 00:47:34 I think long term, they just want to build mplayer once for several systems, and stick it in the iwmmxt directory. May 30 00:54:42 crap... I think ctrl+z kills building something, it's throwing errors now May 30 00:55:38 you do not want to do that May 30 00:56:31 CoreDump|home: heh.... will restarting it be ok? that is, running bitbake again? May 30 00:56:41 I guess so May 30 00:56:52 or have I completely borked it? I'm hoping it'll be alright May 30 00:57:18 summatusmentis: ctrl + z is fine. May 30 00:57:23 I do that all the time May 30 00:58:20 Laibsch: what about 'bg 1' afterwards? May 30 00:59:29 the blasted mplayer would need heavy patching for the v4l TV driver. May 30 00:59:34 maybe the building just failed, and it happened to time itself to my backgrounding it May 30 00:59:40 I'm done trying to get the cam to work for today.... May 30 01:00:11 I guess you mean 'bg %1' May 30 01:00:27 erm... no, I ran 'bg 1' May 30 01:00:52 in any case I only ever stop and foreground it May 30 01:01:42 mm, I like watching code compile most of the time, so I usually let it go May 30 01:01:54 I was trying to append a command at the end, so it'd text my phone when it was done May 30 01:02:18 but it died, hopefully running 'bitbake php' again fixed it May 30 01:02:59 duh, someone changed bugzilla to add loads of headers... so my mail->news gateway has thrown away the last 700 or so emails... I thought things looked quite recently ;) May 30 01:03:03 I usually do bitbake -c clean $lastbbthatfailed May 30 01:03:28 v8jlene: bugzilla is up to version 3 now May 30 01:03:52 summatusmentis: Did you start another instance of bitbake when it failed? May 30 01:04:08 Laibsch: yeah, I had 'bitbake php' running May 30 01:04:13 Laibsch: I noticed it looked different ;) May 30 01:04:42 now I've got 'bitbake php && mail @email.uscc.net < /home/jake/its_done' running May 30 01:04:47 summatusmentis: another, like "a second one" May 30 01:05:04 I don't know what you mean... May 30 01:05:23 what is so difficult? May 30 01:05:31 bitbake foo on console 1 May 30 01:05:40 bitbake bar on console 2 May 30 01:05:59 Or bitbake foo, then bg; bitbake bar May 30 01:06:11 that was my question May 30 01:06:48 l8er May 30 01:07:02 * * OE Bug 2399 has been created by oe@hentges.net May 30 01:07:04 * * linux-rp-2.6.20 is lacking the V4L device module and+ May 30 01:07:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2399 May 30 01:15:02 * * OE Bug 2400 has been created by oe@hentges.net May 30 01:15:04 * * mplayer is lacking the tv / v4l driver May 30 01:15:06 * * http://bugs.openembedded.org/show_bug.cgi?id=2400 May 30 01:19:10 ljp: I enjoy reading your responses to france on opie-devel May 30 01:43:27 :) May 30 02:02:40 can someone lend me a hand with an odd error? May 30 02:02:41 ERROR: /home/zero/oe/build/zc/kismet/kismet_2007-01-R1b.bb:18: unparsed line: May 30 02:02:42 'fakeroot do_install() { oe_runmake "DSTROOT=${D}" install }' while parsing /home/zero/oe/build/zc/kismet/kismet_2007-01-R1b.bb May 30 02:03:05 looks correct to me, but I get this everytime I have that line May 30 02:03:16 (so obviously it's not correct) May 30 02:18:10 can someone point me in the right direction? http://pastebin.ca/520964 May 30 02:19:37 summatusmentis: bitbake version? May 30 02:19:49 summatusmentis: building against .dev or .oz354x? **** ENDING LOGGING AT Wed May 30 02:59:57 2007