**** BEGIN LOGGING AT Thu Mar 08 02:59:57 2007 Mar 08 03:06:17 hmm xterm won't build on x86. oe_runmake is using an include from my host system Mar 08 03:09:50 ahh i see there's an open bug Mar 08 03:09:54 and a fix in the bug. Mar 08 03:10:04 just not applied nor marked as fixed in bugzilla Mar 08 03:59:58 http://bugs.openembedded.org/show_bug.cgi?id=1949 Mar 08 04:00:18 any one have any idea about this? Mar 08 04:30:39 03mwester 07org.oe.dev * r746269e8... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: add tzdata Mar 08 04:30:43 03mwester 07org.oe.dev * r9b341ad2... 10/ (1 packages/tasks/task-slugos.bb): task-slugos: drop kernel-module-ntfs from rootfs; added kernel-module-nfs back in. Mar 08 05:02:53 these microtouch touchscreens are a pain to calibrate manually via minicom. Anyone know of any programs that will do the same thing without the hassle of explaining how to press Ctrl and A at the same time to a tech over the phone? :) Mar 08 05:39:49 other than dgen (it's default EXTRA_OECONF has --without-opengl) are there any opengl apps in org.openembedded.dev? Mar 08 05:40:04 I just need something other than glxgears/glxinfo Mar 08 06:00:14 no? no opengl applications at all? Mar 08 06:04:27 ahh frozen-bubble Mar 08 06:09:49 http://bugs.openembedded.org/show_bug.cgi?id=1949 Mar 08 06:09:58 how to fix it? Mar 08 06:13:21 Readon: dunno - it works here on a build from scratch on Debian Etch, Mar 08 06:23:01 need i clean the buildings before? Mar 08 06:23:35 rwhitby, need i clean the buildings before? i use Debian Sarge Mar 08 06:24:09 rwhitby, someone else also met the problem. Mar 08 06:24:45 Readon: yes, there are multiple reports of it, but I cannot reproduce it, and it's a generic OE problem, not one related to MokoMakefile, so I'm not able to fix it for you. Mar 08 06:25:37 rwhitby, it's oe channel here.:) Mar 08 06:26:07 rwhitby, do you notice that ./configure did not find arm-linux pkg-config Mar 08 06:26:09 Readon: yep, but you won't get much response until Europe comes on line. Mar 08 06:29:50 hmm Mar 08 06:30:05 r200_dri.so isn't being produced by xorg.. funny Mar 08 06:34:08 can i bootstrap-image without doing kernel-builld? Mar 08 07:04:35 good morning all Mar 08 07:19:11 hi Mar 08 07:36:11 03koen 07org.oe.dev * r2e1d1df3... 10/ (1 conf/machine/fic-gta01.conf): fic-gta01: disable SERIAL_CONSOLE so we can actually use the gsm modem Mar 08 07:57:43 hi Mar 08 07:57:58 hey hrw Mar 08 07:57:59 jselect: which microtouch touchscreen you have? Mar 08 07:58:40 man Mar 08 07:59:08 the second person in a week unable to grasp the 'db migrate' message Mar 08 07:59:47 (altough this one is a bug in mtn) Mar 08 07:59:59 http://maps.google.com/maps?f=q&hl=en&q=15.298683+19.429651&layer=&ie=UTF8&z=23&ll=15.298684,19.429651&spn=0.001291,0.002698&t=k&om=1&iwloc=addr Mar 08 08:01:16 morning all Mar 08 08:01:38 has anyone tried to make an ekiga.bb ? Mar 08 08:03:24 Ifaistos: iirc we have a gnome-meeting .bb Mar 08 08:03:40 Ifaistos: but the sofware it depends on (opal, ptlib) is *horrible* Mar 08 08:07:46 koen|lab: that's why i am asking :) Mar 08 08:08:54 morning Mar 08 08:19:35 mtn 0.33 fine here Mar 08 08:57:22 koen|lab: i just uploaded the efika packages from the weekend clean build, but the sort scipt seems to fail Mar 08 09:05:16 Ifaistos: where does it fail? Mar 08 09:05:22 could you pastebin the error? Mar 08 09:07:22 koen|lab: cleared the window :( but now seems to have moved the files Mar 08 09:09:25 koen|lab: btw how do you initially create the feed dirs in the web server ? Mar 08 09:09:56 koen|lab: if they are not allready created the sort/index scripts fail Mar 08 09:11:15 I create them by hand Mar 08 10:10:35 hi, how can i add packages to already builded bootstrap-image ? On my first build i have only DISTRO_EXTRA_RDEPENDS += " minimo" Mar 08 10:12:34 now i add additional packages in DISTRO_EXTRA_RDEPENDS += " matchbox-desktop minimo " and they builds but not present in $(OE)/tmp/rootfs Mar 08 10:19:47 lkm`: rebuild task-base Mar 08 10:23:31 koen|lab: 10x :), now it is ok Mar 08 11:02:10 03koen 07org.oe.dev * r3a81a922... 10/ (4 files in 3 dirs): glib: add 2.12.10 Mar 08 11:05:20 XorA|gone: I suspect glib 2.12.0 should fix the crashers in abiword 2.5 Mar 08 11:12:58 just added a new lirc-modules package which will compile with the 2.6.20 kernel: http://bugs.openembedded.org/show_bug.cgi?id=1952 Mar 08 11:22:54 is it possible to use overrides with export? like export_via-c7 DISABLE_OPENGL=1 Mar 08 11:23:14 tkp: no export is a keyword Mar 08 11:23:29 hmm. ok Mar 08 11:23:56 tkp: but you could do_configure_prepend() { if [ some shell codehere ]; export DISABLE_OPENGL=1 } Mar 08 11:26:23 morning all Mar 08 11:26:32 hey Richard Mar 08 11:27:47 zecke: yeah, thanks Mar 08 11:28:20 ~fish CIA-3 Mar 08 11:28:34 * ibot slaps CIA-3 around with a large trout Mar 08 11:33:03 good morning all Mar 08 11:33:18 hyh, how can i install xserver-kdrive-fbdev ? Mar 08 11:33:56 xserver-kdrive is build ok, and create package xserver-kdrive-fbdev.ipkg, but install nothing in rootfs Mar 08 11:45:14 hi Mar 08 11:54:27 hmmm some people still think only x86 cpu exist in the world Mar 08 11:54:36 Word16 extract_h(Word32 L_var1) Mar 08 11:54:48 var_out = (Word16) (L_var1 >> 16); Mar 08 11:57:07 ~lart mickeyl for bugs in python packaging Mar 08 11:57:07 * ibot beats mickeyl over the head with a microkernel for bugs in python packaging Mar 08 11:59:14 hi koen Mar 08 12:00:32 hey chris144 Mar 08 12:03:55 koen|lab: how can i add XFbdev bin in rootfs? Mar 08 12:04:29 * chris144 is just testing bb 1.8... Mar 08 12:06:37 anybody saw bb1.8 say "preferred version 2.5 of glibc not available" and the circular dependency at the end http://pastebin.ca/386131 ? Mar 08 12:07:52 ...when bb'ing bootstrap-image Mar 08 12:08:38 using generic distro Mar 08 13:10:18 re Mar 08 13:12:56 Is anybody able to successfully crossbuild gcc on an x86_64 host? Mar 08 13:13:25 gcc and not gcc-cross? Mar 08 13:13:33 gcc, not gcc-cross Mar 08 13:13:38 Trying to build the ipk Mar 08 13:13:46 !oebug 1951 Mar 08 13:13:48 * * Bug 1951, Status: NEW, Created: 2007-03-07 15:45 Mar 08 13:13:49 * * nail(AT)nslu2-linux.org: gcc fails to build on x86_64 Mar 08 13:13:50 * * http://bugs.openembedded.org/show_bug.cgi?id=1951 Mar 08 13:13:55 I confirmed that bug Mar 08 13:13:59 I kno Mar 08 13:14:00 w Mar 08 13:14:10 when will get some cpu free I will try for i686 Mar 08 13:14:55 From what I gathered when googling, it's specific to x86_64 Mar 08 13:23:49 heh Mar 08 13:23:56 I keep running into the odd problems Mar 08 13:23:58 /bin/sh: line 1: 16529 Floating point exception../src/util/makekeys ks_tables_h Mar 08 13:24:59 this is in bugtracker Mar 08 13:30:51 ah Mar 08 13:30:55 !oebug 1940 Mar 08 13:30:56 * * Bug 1940, Status: NEW, Created: 2007-03-04 14:59 Mar 08 13:30:58 * * angus(AT)handheldshell.com: libx11 and diet-libx11 build fails with floating point exception on amd64 Mar 08 13:30:59 * * http://bugs.openembedded.org/show_bug.cgi?id=1940 Mar 08 13:45:00 hrw: thanks. Fixes my build atleast. Mar 08 13:57:01 what is the difference between setting DISTRO_TYPE = "release" and "debug"? Mar 08 13:57:41 xuumbi: misc things Mar 08 13:57:57 xuumbi: dropbear allow root-no-pass for debug but not for release Mar 08 13:57:58 xuumbi: e.g. password les root login for dropbear and other misc things Mar 08 13:59:26 ah ok, interesting, I thought it might have something to do with the way things are compiled Mar 08 13:59:58 that's good to know thanks, I couldn't find any documentation on it Mar 08 14:02:24 hi :) Mar 08 14:02:52 right place :) Mar 08 14:04:49 I am looking for a dreambox developer Mar 08 14:06:24 tmbinc there you are :) Mar 08 14:07:20 cool, the work directory now has host and target dirs! Mar 08 14:08:18 need help with some keymapping - nordic characters - ist only one I cant get working Mar 08 14:08:48 are the characters also in a lib file ? Mar 08 14:28:54 is there any way to rebuild all files with dependancy dconf by example Mar 08 14:36:19 hrw,NAiL: Can't cross compile gcc on powerpc gcc 4.1.1/glibc2 for sometime now Mar 08 14:36:44 hrw,NAiL: /s/glibc2.5 Mar 08 14:37:05 ok Mar 08 14:37:13 hrw,NAiL: both on x86 and amd64 Mar 08 14:39:27 same error? Mar 08 15:08:39 cu Mar 08 15:19:01 where debian renaming handled for dependencies? I don't see that in debian.bbclass Mar 08 15:19:52 psokolovsky_: See package.bbclass Mar 08 15:21:53 RP: Thanks! Mar 08 15:24:17 psokolovsky_: There is a generic renaming mechanism debian.bbclass uses, namely PKG_ variables Mar 08 15:25:07 RP: I wonder why I have stale stuff in staging/pkgmaps/debian, investigating... Mar 08 15:29:05 RP: It seems that if package was debian-renamed, but later it was disabled, stale map file is left in pkgmaps, which causes packages to continue to be renamed. Mar 08 15:29:21 RP: Weird, right? but hardly can do anything... Mar 08 15:30:20 psokolovsky_: Yes, that is a bug and a common problem Mar 08 15:30:36 psokolovsky_: The problem is half the build packages probably had their depends renamed too :/ Mar 08 15:32:32 psokolovsky_: IF we could ensure the pkgmaps files were removed, that would half fix the problem, the dependent packages would then only need to be rebuilt Mar 08 15:33:02 psokolovsky_: The other half of the problem is then having bitbake rebuild all dependencies which is actually fairly easy to do with the 1.8.x+ code (but not 1.6.x) Mar 08 15:33:39 psokolovsky__: Did you see what I said? Mar 08 15:34:22 RP: I exactly has that - package is not renamed, but in other ipk there's renamed dep... Mar 08 15:35:32 We need to rethink pkgdata in staging at some point, try and combine it... Mar 08 16:05:46 * koen|lab stabs matlab IIR filter functions Mar 08 16:06:13 rofl Mar 08 16:11:04 after 8 hours of matlabbing functions named 'butt ord' aren't speeding things up Mar 08 16:27:10 our matlab guru is in las vegas this week Mar 08 16:40:56 Does someone want to bump madwifi to r2182-20070308? I'm having some trouble with the current version but the newest version seems to work. Mar 08 16:44:00 apaulsen, do you have a patch you can post on bugzilla? Mar 08 16:45:44 Crofton|home: I could do that, but you just have to copy madwifi-ng_r2156-20070225.bb to madwifi-ng_r2182-20070308.bb Mar 08 16:46:53 hmm Mar 08 16:46:53 Do you know who cares about madwifi? Mar 08 16:47:36 I have commit access, but am reluctant to do it without knowing anyhthing about madwifi Mar 08 16:47:42 good morning all Mar 08 16:48:12 Crofton|home: as far as I know madwifi is just for expansion cards Mar 08 16:50:06 Is there a bitbake command option to skip certain build steps? for example when preparing a cross-compile enviroment, it would be safe to stop after the staging step but to build a toolchain, I do someting like this: 'bb virtual/armeb-linux-gcc virtual/libc' but I don't really need to package anything, just stage it Mar 08 16:50:34 seems to me, skipping the packaging step would save quite a lot of time in the toolchain build process Mar 08 16:51:22 apaulsen, I wish I knew who was bumping the version Mar 08 16:52:03 Crofton|home: who has been bumping it in the past? Mar 08 16:52:10 not sure Mar 08 16:52:29 Crofton|home: sorry, is that what you were asking? Mar 08 16:52:32 I am in a workshop, so it is hard for me to figure htat out :) Mar 08 16:52:44 yeah, who is commiting Mar 08 16:53:33 hillct: well don't INHERIT package_ipk, pakage_tar? Mar 08 16:53:45 hillct: then no packages will be created Mar 08 16:54:22 can that be done on the bitbake command line? I want to do this on a one off basis, not by modifying all the recipes Mar 08 16:55:07 something like --ignore package_ipk ? Mar 08 16:58:01 Ifaistos, you around? Mar 08 16:58:32 apaulsen, Ifaistos did the last version update Mar 08 17:02:27 Hi Mar 08 17:03:52 Crofton|home: ok, I'm going to lunch now, I'll check in a little later Mar 08 17:03:57 apaulsen, Mar 08 17:04:05 you must be EST :) Mar 08 17:04:49 CST, but our pregnant secretary wants to get lunch now. Can't argue :) Mar 08 17:04:52 Does anyone care if I create yet another version of madwifi to fix a problem apaulsen is having? Mar 08 17:04:56 ah Mar 08 17:05:30 basically, the version is getting bumped evey few weeks and we a re accumulating a lot of versions Mar 08 17:05:56 03florian 07org.oe.dev * r40bad367... 10/ (4 files in 3 dirs): gpe-scap: Add patch to fix device information on ARM devices. Mar 08 17:06:05 03coredump2 07org.oe.dev * rcf8c280a... 10/ (3 files in 3 dirs): openmoko-pim: Add temporary compile-fix to work-around a missing file Mar 08 17:06:11 03coredump2 07org.oe.dev * ra133abcb... 10/ (6 files in 4 dirs): alsa: Add alsa-state for machines w/o zaurus-updater.... Mar 08 17:06:18 03coredump2 07org.oe.dev * r20e69e89... 10/ (1 conf/machine/fic-gta01.conf): fic-gta01.conf: Add alsa-state to MACHINE_EXTRA_RRECOMMENDS Mar 08 17:06:24 03coredump2 07org.oe.dev * rf2c871ba... 10/ (2 files in 2 dirs): xserver-kdrive-common: Bump PR from previous commit Mar 08 17:06:30 03nail 07org.oe.dev * rb51f5304... 10/ (1 packages/meta/openturbostation-packages.bb): openturbostation: Remove ifupdown from feed Mar 08 17:06:39 03nail 07org.oe.dev * rd112c842... 10/ (3 files in 3 dirs): busybox: Add openturbostation config Mar 08 17:06:46 03nail 07org.oe.dev * r7051e417... 10/ (1 conf/distro/openturbostation.conf): openturbostation: Add libx11 preferred provider Mar 08 17:06:53 03nail 07org.oe.dev * rd3e19212... 10/ (1 packages/meta/openturbostation-packages.bb): openturbostation-packages: Remove non-building packages and clean up Mar 08 17:06:57 03nail 07org.oe.dev * r7fc5093b... 10/ (1 packages/images/openturbostation-image.bb): openturbostation: Add xfs-support in root, remove duplicated utils Mar 08 17:07:04 03nail 07org.oe.dev * rdfcdb523... 10/ (4 files in 4 dirs): netbase: Add openturbostation-stuff Mar 08 17:07:10 03nail 07org.oe.dev * r679a82d6... 10/ (1 packages/images/openturbostation-image.bb): openturbostation: Correct xfs-utils to xfsprogs Mar 08 17:07:34 oh, someone woke CIA-3 Mar 08 17:08:39 CoreDump|afk: coredump2? you lost your key as well? Mar 08 17:10:05 * koen makes those hippie zaurus owners a bit happier Mar 08 17:10:17 03koen 07org.oe.dev * r0038f020... 10/ (5 files in 4 dirs): tslib: disable grab-events thingy by default, turn it on for fic-gta01 Mar 08 17:11:53 ~lart mtn for being slow Mar 08 17:11:54 * ibot installs a bad bootloader on mtn and turns mtn into a brick for being slow Mar 08 17:13:54 apaulsen, I am goign to push the new rev, then get lunch Mar 08 17:14:03 I should be back in an hour or so Mar 08 17:16:01 03crofton 07org.oe.dev * r63be5372... 10/ (1 packages/madwifi/madwifi-ng_r2182-20070308.bb): madwifi-ng_r2182-20070308.bb : Bump version to solve a problem for apaulsen. Mar 08 17:24:33 apaulsen: which architecture do you use madwifi on? Mar 08 17:25:51 Is there touchscreen support for the ep93xx ? Mar 08 17:25:52 zecke:my machine is based on a compulab-x270 Mar 08 17:26:14 pxa-270 Mar 08 17:41:48 hi Mar 08 17:52:36 * chris144 just completed a bootstrap-image for x86 with bb 1.8 Mar 08 17:52:43 whoooo...cool Mar 08 17:53:26 can anyone tell me why I might be getting this when trying to do a git fetch: ? Mar 08 17:53:26 fatal: Needed a single revision Mar 08 17:53:27 Pulling into a black hole? Mar 08 18:05:15 Anyone think it's a bad idea to create a minix fs on an mtdblock? I need a tiny filesystem (768kb on 5 128kb flashblocks), and jffs2 doesn't cut it. Mar 08 18:07:29 It won't be used a lot, mainly to store config stuff and such Mar 08 18:12:10 ~lart work :) Mar 08 18:12:10 * ibot says "boot to the head" and knocks work :) over Mar 08 18:12:12 apaulsen, ping? Mar 08 18:12:28 Crofton|home: pong Mar 08 18:12:43 I commited a new madwifi bb file Mar 08 18:12:54 I saw that Mar 08 18:13:09 um Mar 08 18:13:12 does it work? Mar 08 18:14:30 looks like it Mar 08 18:15:28 * mwester goes to sync up and build... Mar 08 18:16:35 yep, it builds correctly for me Mar 08 18:18:55 ~convert 1 brl to eur Mar 08 18:19:15 ~change 1 brl to eur Mar 08 18:19:21 1.00 Real (BRL) makes 0.362100 Euro (EUR) (from http://www.xe.com/) Mar 08 18:19:44 NAiL: cramfs, or does it have to be rw? Mar 08 18:20:52 03nail 07org.oe.dev * rcc551d0b... 10/ (3 files in 3 dirs): turbostation: Update defconfig Mar 08 18:21:46 No luck with the madwifi: Mar 08 18:21:51 madwifi-ng-r2182-20070308/net80211/ieee80211_linux.c:165: warning: implicit declaration of function 'vlan_group_set_device' Mar 08 18:22:28 Whereupon it croaks. Mar 08 18:22:56 thanks apaulsen Mar 08 18:23:16 hmm Mar 08 18:23:27 works for apaulsen , but not mwester ? Mar 08 18:23:58 kernel versions? Mar 08 18:24:08 2.6.16 for me Mar 08 18:26:33 mwester, I assume going back a version does not solve your problem? Mar 08 18:26:53 btw, I need to stop payying attention, need to do my presentatoin in a minute Mar 08 18:27:14 good luck with it Mar 08 18:27:24 Crofton|home: thanks and good luck Mar 08 18:28:06 hi Mar 08 18:32:24 Crofton: the previous version of madwifi-ng worked very well. Mar 08 18:32:33 kernel version I'm using is 2.6.20 Mar 08 18:34:54 mwester, what version? Mar 08 18:37:09 Good evening Mar 08 18:38:35 I'm a bit unsure, what libc this crosscompiler use... "arm-angstrom-linux-uclibcgnueabi" Mar 08 18:40:00 Crofton: all previous versions of madwifi-ng,but most recently SlugOS has been building with madwifi-ng-r2156-20070225-r3 Mar 08 18:40:18 ok Mar 08 18:40:34 I'll be back in a bit Mar 08 19:05:48 anyone know the difference between samba and samba-essential? Mar 08 19:27:50 mwester, it looks like it fixes a problem for apaulsen Mar 08 19:28:07 it is annoying we are pulling in new madwifi-ng's every few weeks :) Mar 08 19:28:53 I can maintain my own version for now if you want to revert to 2156 Mar 08 19:30:54 I would suggest both projects use preferred version to pull in the suitable version Mar 08 19:36:03 mwester, Can you use preferred version to select the madwifi version? Mar 08 19:36:46 I think that would be a good idea since the madwifi code seems to get changed frequently Mar 08 19:37:47 Crofton: for the moment it seems to be a reasonable solution, until the madwifi upstream stabilizes. Mar 08 19:38:04 yeah Mar 08 19:38:31 hi Mar 08 19:38:40 I'll see how one selects a preferred version for the SlugOS feeds and see if we can resolve this in that manner. Mar 08 19:38:43 sorry commited something broke you, although it did fix another guys problem Mar 08 19:40:30 :) np -- the real issues with madwifi-ng all lie upstream; our difficulties merely reflect those problems. Mar 08 19:41:15 yeah Mar 08 19:44:01 later Mar 08 19:55:13 03nail 07org.oe.dev * r436213c8... 10/ (11 files in 3 dirs): uboot-utils: Make it usable on other machines than the storcenter Mar 08 20:00:08 03nail 07org.oe.dev * r71987d43... 10/ (1 packages/uboot-utils/uboot-utils_1.1.2.bb): uboot-utils: Bump PR Mar 08 20:31:53 nite folks Mar 08 20:40:23 Is there a standard method for dealing with a package of alternative config files that overwrite a default set? In my case, I'll have two packages. openpbx-1.2_rc3.ipk and openpbx-samples-1.2_rc3 .ipk, the latter of which will overwrite all the config files in /etc/openpbx.org with a sample set. Is there a means to back up the originals automatically as there is for RPM packages? Mar 08 20:50:52 hillct: CONFFILES += "/path/to/files" should alert the package manager and give you a prompt with what to do Mar 08 20:52:03 koen: cool. Is there a way to get it to take a default action (like backup any existing files) rather than prompting? Mar 08 20:54:24 d'oh Mar 08 20:59:30 * rwhitby guesses no-one has built DISTRO=openmoko with rm_work enabled yet ... Mar 08 21:00:45 * hvontres|poodle guesses its borked Mar 08 21:01:16 * rwhitby is testing a fix Mar 08 21:09:50 03rwhitby 07org.oe.dev * r6990fd1b... 10/ (1 packages/uboot/u-boot-mkimage-gta01-native_svn.bb): u-boot-mkimage-gta01-native: Fixed ordering of do_deploy in relation to rm_work. Mar 08 21:14:11 good evening all Mar 08 21:15:05 when try to build "freetype" it fails because of a 'i686-pc-linux-gnu' not being found Mar 08 21:15:30 what kind of compiler is that? Mar 08 21:15:50 Henkie: http://bugs.openembedded.org/show_bug.cgi?id=1906 Mar 08 21:16:12 there a distro config for an n2100 ? Mar 08 21:16:50 timtimred: thecus? no, atleast not yet. Mar 08 21:17:11 yep yep, and aww. Mar 08 21:17:27 cyrilRomain, tnx Mar 08 21:17:32 I was considering making one Mar 08 21:17:49 actually, not making a distro, just make a machine config and use angstrom Mar 08 21:18:09 shouldn't be too hard Mar 08 21:18:13 yeah. Mar 08 21:18:35 is there any crc checking of anything in the version of redboot installed on it? Mar 08 21:19:00 not that I can remember, no Mar 08 21:19:10 there is a debian-installer image for the n2100 though Mar 08 21:19:23 Although it's not working right now Mar 08 21:19:26 i saw that. it seems to not be on his site atm pending a new release Mar 08 21:19:27 i have another question, not sure it this is the right channel. But why is qemu being built? Mar 08 21:19:30 should be fixed in a week or two. Mar 08 21:19:38 seems to have trouble with a gcc 4+ compiler Mar 08 21:20:22 cool thanks. Mar 08 21:20:40 im more interested in having OE on it though. Mar 08 21:21:02 and i need to submit the linkstation/kuro stuff before i get started on that Mar 08 21:21:15 which i will do this weekend hopefully. Mar 08 21:21:26 what linkstation/kuro stuff? Mar 08 21:21:36 support for them in OE. Mar 08 21:22:09 kernel & uboot (non denx atm for dtb support - is this in mainstream uboot yet?) & native dtc tool Mar 08 21:22:39 ah Mar 08 21:22:46 i have a half assed distro config im not gonna submit until its done. Mar 08 21:22:47 I'm working on the qnap turbostation Mar 08 21:22:53 which is quite similar Mar 08 21:23:00 i also did hfs, hfs+ and netatalk meta files on the way Mar 08 21:23:57 Henkie: qemu does not build with gcc 4.x, you have to use an older version of gcc Mar 08 21:24:35 koen: Fancy a laugh? Mar 08 21:24:47 cyrilRomain, sounds doable, but why is it build? Mar 08 21:24:51 Henkie: or install it in your host system and use ASSUME_PROVIDED in your conf/local.conf Mar 08 21:26:05 http://pxa27x-linux.svn.sourceforge.net/viewvc/pxa27x-linux Mar 08 21:26:07 Henkie: with ASSUME_PROVIDED += "qemu-native", bitbake will not compile it Mar 08 21:26:36 and http://orzlab.blogspot.com Mar 08 21:26:51 timtimred: I think u-boot 1.2.0 supports dtb Mar 08 21:26:58 atleast the readme mentions them Mar 08 21:27:16 where did you get the other u-boot? Mar 08 21:27:54 g'night all Mar 08 21:29:07 RP: heh Mar 08 21:29:53 RP: Hello. What do you think about changing MEM_INT_BASE_VALUE to 0x200000 in the w100fb driver? Mar 08 21:29:53 koen: Note the linux-rp-2.6.20 patch series Mar 08 21:29:54 RP: I guess they liked your work so much they decided to help you back it up offsite :) Mar 08 21:30:26 sirfred: Is that depended on by any legacy library? Mar 08 21:30:48 RP: their rationale for using OABI is funny Mar 08 21:31:01 hvontres|poodle: They added a zlib 1% speedup and their own logo (not even bothering to change some Openedhand references Mar 08 21:31:05 RP: I don't know. But well, the value can be obtained just reading the appropiate register. Mar 08 21:31:13 koen: Yes, I was amused by that too Mar 08 21:31:30 sirfred: I have nothing against it particuarly, I was just wondering Mar 08 21:31:57 RP: What software is using w100 accel, that you're aware of? Mar 08 21:32:21 sirfred: I was mainly thinking of the old aticore Mar 08 21:32:27 NAiL: git://www.jdl.com/software/u-boot-86xx.git u-boot-86xx Mar 08 21:32:50 RP: I'm not sure if the old aticore is assuming memory to be in a fixed location. Mar 08 21:33:31 sirfred: Anyhow, I have no objection Mar 08 21:33:43 RP: I'm going to send you a patch. Mar 08 21:33:48 cyrilRomain, tnx Mar 08 21:34:08 RP: Ximageon is supporting randr with resolution change using this patch, and using all the memory Mar 08 21:34:29 sirfred: cool Mar 08 21:34:32 sirfred: This includes the external memory switch? Mar 08 21:34:32 RP: Well, I mean under this patched kernel. Mar 08 21:34:36 RP: Yes. Mar 08 21:34:40 sirfred: w3220 still needs kernel work? Mar 08 21:34:49 sirfred: ok, we should get that into linux-rp :) Mar 08 21:35:24 koen: I've cleaned out all the randr stuff dealing directly with registers. Mar 08 21:35:49 koen: So, as far as w100fb is able to change the mode in the w3220, Ximageon also should work. Mar 08 21:36:06 koen: But, I'm not sure, of course. Mar 08 21:37:25 My husky is now running the game lemmings, on snes9x, on libSDL, on Ximageon at 320x240. It looks great Mar 08 21:37:44 timtimred: As soon as I get another hdd to put into the n2100, I'll be flashing it with *something*. I might take the openturbostation distro and turn it into a bit more generic distro for all my devices ;) Mar 08 21:38:29 cool. i'd be interested to hear how that gos :) Mar 08 21:39:15 sirfred: I can test it on w3220 in ~2 weeks Mar 08 21:39:21 sirfred: I used to enjoy lemmings :) Mar 08 21:39:25 * NAiL has a few similar devices (in function, not cpu) Mar 08 21:39:25 i know it has serial, but jtag? Mar 08 21:39:36 RP: A glorious game. Mar 08 21:39:47 timtimred: the n2100? I think it has jtag, but I don't think anyone has used it. Mar 08 21:39:53 RP: It's a pitty it does not support the mouse Mar 08 21:40:04 Well, touchscreen in this case Mar 08 21:40:21 sirfred: We're there patches around for SDL to make it work with tslib? Mar 08 21:40:26 thats reassuring :) Mar 08 21:40:36 heh Mar 08 21:40:55 I've replaced the redboot with a version that actually sets up the network interfaces correctly Mar 08 21:41:02 RP: Humm, what do you mean? SDL on X11 uses tslib, as X11 supports it. Mar 08 21:41:03 kernel status? i can see it has 2.6.9 in the sources and i can see vavgue evidence my google there might be better support? Mar 08 21:41:21 yes, afaik most of the stuff works on recent kernels Mar 08 21:41:42 there has been work on committing stuff upstream Mar 08 21:42:02 What I want to test is to build under OE libSDL >= 1.2.10 as I've read it support changing the mode automatically using randr. Mar 08 21:42:03 leds/buzzer doesn't work, usbcopy button doesn't work. That's about it, I think. Mar 08 21:42:38 With 1.2.9, I have to change previously the resolution manually using xrandr. Mar 08 21:42:45 great :) Mar 08 21:43:02 ah, sensor-stuff needs to be reviewed before it goes upstream Mar 08 21:43:19 timtimred: I would suggest you join #iop for any specific questions ;) Mar 08 21:43:54 hi hrw Mar 08 21:44:00 oh gone... Mar 08 21:44:10 NAiL: You know a tame LED maintainer ;-) Mar 08 21:44:19 RP: Indeed I do ;) Mar 08 21:44:58 koen: Great, I expect this time we will get something more than a black screen. Mar 08 21:45:19 sirfred: Could you document the valid values you discovered in the header file against the registers please? Mar 08 21:45:43 RP: For what registers? Mar 08 21:46:03 RP: You mean MC_FB_LOCATION and MC_EXT_MEM_LOCATION ? Mar 08 21:46:07 sirfred: yes Mar 08 21:46:13 RP: Of course. Mar 08 21:46:40 RP: By the register #define ? Mar 08 21:47:10 sirfred: or within the appropriate struct Mar 08 21:47:17 RP: OK Mar 08 21:49:48 RP: Like this ? Mar 08 21:49:58 u32 mc_ext_mem_start : 16; /* Mask 0xe000 */ Mar 08 21:50:28 RP: Only the *_mem_start are limited, so, I'm not saying nothing about *_mem_top, OK' Mar 08 21:51:28 sirfred: Maybe /* Valid bitbask: 0xe000 */ ? Mar 08 21:51:38 RP: Of course. :) Mar 08 21:51:46 RP: even bitmask Mar 08 21:51:51 sirfred: bitmask, right :) Mar 08 21:52:19 * chouimat|away is away: Going to The Pub ... (nice motorhead's song too) Mar 08 21:53:18 RP: Sent Mar 08 21:54:20 sirfred: looks good Mar 08 21:58:07 sirfred: I'll run a build, then add to linux-rp Mar 08 21:58:27 RP: Thanks. Mar 08 22:16:46 sirfred: pushed :) Mar 08 22:17:04 RP: Great! What version 2.6.20 I suppose? Mar 08 22:17:37 03rpurdie 07org.oe.dev * r685f49ac... 10/ (1 packages/linux/linux-rp_2.6.20.bb): linux-rp-2.6.20: Add w100 external memory patch (from sirfred) Mar 08 22:17:48 sirfred: I think CIA just answered that :) Mar 08 22:17:49 RP: ok Mar 08 22:18:10 RP: Yes, couldn't be more opportune Mar 08 22:19:15 RP: A curiosity about 2.6.20 Mar 08 22:19:43 RP: Looking at the kernel message queue, I found a log about some DSP coprocessor detected or something so, running 2.6.20 on my husky Mar 08 22:20:20 RP: I don't remember the exact log, and my current dmesg was wrapped, but I don't remember to have seen that log in 2.6.19 Mar 08 22:22:24 sirfred: What DSP coprocessor? I'm not aware of one (dmesg should show you the log) Mar 08 22:22:57 RP: I remember seeing that message on a pxa250 poodle as wall Mar 08 22:23:05 s/wall/well Mar 08 22:23:08 RP: The message fifo was wrapped. Mar 08 22:23:26 RP: I'm going to restart it to tell you the exact message. Mar 08 22:25:42 RP: <6>XScale DSP coprocessor detected. Mar 08 22:26:31 re Mar 08 22:27:03 * florian triggered a nice compilation problem Mar 08 22:27:09 fs/squashfs/inode.c:586: fatal error: internal consistency failure Mar 08 22:27:29 sirfred: Ah, the mmx extensions. They've a lot of nasty alignment constraints Mar 08 22:27:41 florian: There's something in the OE bugzilla about that Mar 08 22:28:11 RP: Oh, so the xscale has some kind of mmx ? Mar 08 22:28:13 RP: oh ok, so i'm not the only one Mar 08 22:28:38 florian: no, its been seen before Mar 08 22:28:57 I wonder if they could be used in idct mplayer implementation Mar 08 22:29:15 sirfred: possibly yes... Mar 08 22:29:33 Perhaps we could gain some cycles Mar 08 22:32:46 * mmp would also love if mplayer could be fine-tuned for pxa25x :) Mar 08 22:33:27 hm learning arm assembler? Mar 08 22:33:32 he laibsch Mar 08 22:33:44 hi woglinde Mar 08 22:35:17 RP: i can't find it... do you remember in which area the number was? Mar 08 22:36:47 florian: 1084 Mar 08 22:37:01 RP: many thanks Mar 08 22:40:44 I have a problem compiling gpe-ownerinfo because its Makefile use 'ar', and it use my host system one, not the oe generated :( Mar 08 22:40:56 Should 'ar' be generated into ${STAGING_BINDIR_CROSS} ? or should the bitbake.conf's PATH_prepend include something like ${CROSS_DIR}/${TARGET_SYS}/bin ? or it is something else ? .. Mar 08 22:46:04 RP: oh well, i guess we have to fix that in all newer kernel with squashfs Mar 08 22:51:33 florian: Ideally, we should fix the compiler Mar 08 22:54:23 cyrilRomain: It should be using arm-linux-ar Mar 08 22:54:55 RP: true... i always wanted to take a look inside gcc :-) Mar 08 22:55:01 but not today Mar 08 22:55:12 RP: yes, but I don't know what is the right way to fix Mar 08 22:55:34 cyrilRomain: The right way is to make it use arm-linux-ar ;-) Mar 08 22:56:16 hi Mar 08 22:56:27 hi CoreDump|home Mar 08 22:56:53 * CoreDump|home wonders if it is worth upgrading u-boot yet Mar 08 22:57:26 RP: :) I can do it with a silly Makefile.patch, or changing bitbake.conf file, but maybe should ar (or a symlink) be created in STAGING_BINDIR_CROSS , ... Mar 08 22:57:39 cyrilRomain: that's a bug in the makefile, it should use $(AR) instead if "ar" Mar 08 22:57:45 CoreDump|home: hello Mar 08 22:58:10 florian: thanks, simple :) Mar 08 22:59:55 cyrilRomain: silly! Mar 08 22:59:57 ;) Mar 08 23:00:31 Good night. Mar 08 23:00:36 florian: ;) Mar 08 23:01:26 florian: worked like a charm. I'll file the bug and patch. Thanks for your help Mar 08 23:01:42 cyrilRomain: yw Mar 08 23:01:50 * florian fixes upstream Mar 08 23:02:21 is it possible to send ir-signals to a TV in linux/OE? (with an ipaq h2200) Mar 08 23:03:13 florian: http://rafb.net/p/sr2u1P96.html Mar 08 23:03:35 Henkie: i'm not sure... yes if the h2200 supports consumer IR. Some iPaq models do and some don't Mar 08 23:04:00 florian, well it supports it, because in winCE the program nevo worked like a charm Mar 08 23:04:35 only question is, must i use lirc or something else? Mar 08 23:04:38 Henkie: i remember there is an application for linux as well Mar 08 23:04:42 LaF0rge: are you available? Mar 08 23:05:03 hey zecke Mar 08 23:05:33 zecke: yes, I kinda "lost" the key as well =) Mar 08 23:05:34 ,Hello. Im trying to compile angstrom for my ipaq h5550, but I get this error: /openembedded/tmp/work/armv5te-angstrom-linux-gnueabi/glibc-2.5-r4/locale-tree/bin/localedef: error while loading shared libraries: libc.so.6: cannot open shared object file: No such file or directory Mar 08 23:05:37 any ideas? Mar 08 23:06:00 CoreDump|home: I proposed this to liam. I store your private key just for 99$ a month Mar 08 23:06:12 =) Mar 08 23:06:15 :-) Mar 08 23:06:41 cyrilRomain: ok, many thanks - i guess i can check that in :-) Mar 08 23:07:29 florian: yep, thanks Mar 08 23:08:15 http://www.rit.edu/~tfs1812/cgi-bin/document.pl?document=opie-remote Mar 08 23:10:51 our robots can drive http://robotics.mi.fu-berlin.de/videos/autonom-08-03-2007.wmv Mar 08 23:14:31 * hvontres|poodle envisions robot cars "drafting" down the autobahn...just like NASCAR without all the advertising on the cars Mar 08 23:17:27 zecke: Have you any objection to changing python function execution to redirect to a file like shell functions do? Mar 08 23:22:31 03florian 07org.oe.dev * rfb386b14... 10/ (3 files in 3 dirs): gpe-ownerinfo: Add patch to fix compile issue contributed by Cyril Romain Mar 08 23:22:37 he zecke Mar 08 23:28:29 RP: where will bb.note end up? Mar 08 23:29:31 zecke: Being sent as an event for the ui to deal with Mar 08 23:29:55 zecke: It idea would be to catch any stdout/stderr output from the function Mar 08 23:30:06 ,Hello. Im trying to compile angstrom for my ipaq h5550, but I get this error: /openembedded/tmp/work/armv5te-angstrom-linux-gnueabi/glibc-2.5-r4/locale-tree/bin/localedef: error while loading shared libraries: libc.so.6: cannot open shared object file: No such file or directory Mar 08 23:30:07 any ideas? Mar 08 23:30:18 This is mainly the "created package foobar" messages atm as far as I can see Mar 08 23:31:06 rds: There might have been something about locale generation on the mailing list recently Mar 08 23:31:54 RP: hmm, for trunk this would be okay Mar 08 23:32:06 zecke: Its not for 1.8 Mar 08 23:32:57 zecke: Basically, on a bitbake server, where is stdout going to go otherwise? I'm not sure we actually have much choice... Mar 08 23:33:13 RP: right Mar 08 23:33:25 RP: BTW: I just saw http://people.openmoko.org/henryk/bitbake-1.6.6-om2.patch Mar 08 23:34:30 zecke: Interesting Mar 08 23:35:41 RP: http://lists.openmoko.org/pipermail/commitlog/2007-March/001024.html Mar 08 23:37:04 RP: looks clean and sane Mar 08 23:37:10 florian: we applied the patch for gpe-ownerinfo_0.28, but we should also apply the patch for previous version (if they also use 'ar' instead of $(AR)) Mar 08 23:37:45 zecke: I'm not sure I agree. I'm not sure why OE itself should have to have to worry about things... Mar 08 23:38:18 zecke: and it should be PV that changes, not PR Mar 08 23:38:23 RP: I agree with getting the revision info from the fetcher Mar 08 23:38:37 zecke: Right, I agree with that, that is fine Mar 08 23:39:36 RP: the question is how you force a rebuild. And technically this is either by removing a stamp, refetching to invalidate the dependant stamps, or change PV|PR Mar 08 23:39:41 cyrilRomain: i don't think anyone is using the old ones anymore Mar 08 23:39:58 florian: by the way, same error ('ar' instread of $(AR) in Makefile) for iptable. I checked the last version of iptable on the web they still use 'ar'. Mar 08 23:40:11 zecke: The SCM revision should be in PV somehow Mar 08 23:40:18 great :/ Mar 08 23:40:19 I'll do a patch for oe and suggesting the patch upstream then Mar 08 23:40:34 that would be good Mar 08 23:40:48 RP: I would have to read our version policy Mar 08 23:42:16 03rwhitby 07org.oe.dev * rb4827e09... 10/ (3 files in 2 dirs): sjf2410-linux-native: Fixed ordering of do_deploy to work with rm_work. Mar 08 23:42:35 zecke: I think we'd have to propose moving from cvs${SRCDATE} and svn${SRCDATE} to x.y.z+cvs-r${REVISION} Mar 08 23:42:49 zecke: and use revision even when a date was specified Mar 08 23:43:01 RP: looking at our policy I agree that it must be PV Mar 08 23:43:29 RP: "+cvs-r" but we should avoid hypens in PV Mar 08 23:43:52 zecke: right ;-) Mar 08 23:43:54 debian use ~ now don't they? Mar 08 23:44:15 rwhitby: In certain cases, yes. It would be nice to support that Mar 08 23:44:28 but that is a different problem for a different day Mar 08 23:44:33 * rwhitby goes into a meeting ... Mar 08 23:44:38 RP: date based and revision based PV will bite each other in some way, I don't think I bother enough to create an example Mar 08 23:45:28 RP: oh http://people.openmoko.org/henryk/bitbake-1.6.6-om3.patch Mar 08 23:45:37 zecke: I think we have to force a switch from one to the other... Mar 08 23:46:56 florian: iptables patch (that applies on 2/3 of iptable version) http://rafb.net/p/hMlQK183.html , with PR++ this time ;) Mar 08 23:48:45 zecke: Are cvs revisions always unique? Mar 08 23:49:23 cyrilRomain: oh cool, i'll apply it as soon as my local database is up to date Mar 08 23:50:46 RP: the question is which file is monitored, let me look at the code more closely Mar 08 23:51:45 RP: obviously there is no patch for cvs.py. Back with cvs every file has a separate revisions Mar 08 23:51:54 RP: there was nothing like cset :) Mar 08 23:52:40 zecke: So there is no such thing as a unique revision? Mar 08 23:53:09 RP: for cvs, no. You could build a hash over all revisions of all files Mar 08 23:53:22 florian: cool, thx. I'll file the patch upstream as soon as they get my account activated. Mar 08 23:53:31 zecke: So a date and time is as good as it gets Mar 08 23:53:32 RP: this is not practiable though Mar 08 23:53:55 RP: for cvs yes, there are tags though Mar 08 23:54:30 RP: e.g. the cvs2svn script tries to find revs with the same commit message and forms a cset from them if the time is not too far away (shows the lack of cset concept in cvs) Mar 08 23:54:49 zecke: I guess bitbake should have a cache somewhere where is caches the last built time - now sees if any changes were checked in since the last date/time (for cvs) Mar 08 23:56:09 RP: do we still have some cvs packages, I would like to deprecate cvs (oh opie, ipkg are still cvs... hmm) Mar 08 23:56:22 zecke: We have cvs packages... Mar 08 23:56:23 * mwester notes that the whole argument of csets remains open; it is still the case that csets have implied dependencies based on time, and essentially degenerate into the same situation as cvs in complex worst cases. Mar 08 23:56:39 zecke: what about git? Here, no revision numbers that sort. Cache the last tag, rebuild if changed? Mar 08 23:57:03 RP: yes for CVS we would need to save date+time and check the log against HEAD of the branch Mar 08 23:57:33 mwester: maybe, but we care about atomic operations. If we build atomic10 and we have atomicSomething we know the source changed Mar 08 23:58:00 RP: om3 patch has git support Mar 08 23:58:28 RP: and it looks like it is doing ls-remote against its master Mar 08 23:59:40 zecke: Getting the HEAD of a git tree is easy, just cat refs/heads/master Mar 09 00:00:23 RP: ls-remote might just do it, and refs/head/master only changes after a pull Mar 09 00:00:55 RP: looking at the man page it is doing a cat on the server side :) Mar 09 00:01:11 zecke: I mean that file on the remote repository and compare with the local one Mar 09 00:01:24 zecke: right :) Mar 09 00:01:51 RP: the funny part with the patch is, we don't do any compares in BitBake directly Mar 09 00:03:02 RP: just due a changed PV we update, now we need to make sure that PV is evaluated each time (I wonder about a performance penalty of having too many SRCDATE=now packages) Mar 09 00:03:06 zecke: If PV/PR are set correctly, we never have to do a compare... Mar 09 00:03:33 Lots of SRCDATE=now packages will be a pain but one that people want/need... Mar 09 00:06:22 cu Mar 09 00:07:36 nite Mar 09 00:07:41 woglinde: hey? Mar 09 00:09:15 yes? Mar 09 00:16:13 03florian 07org.oe.dev * r1bea34ea... 10/ (5 files in 3 dirs): iptables: Apply patch to use cross ar instead of host ar. (by Cyril Romain) Mar 09 00:20:26 03mwester 07org.oe.dev * rf8e82883... 10/ (3 files in 3 dirs): Mar 09 00:20:26 slugos.inc, slugos-native.bb: nail madwifi-ng to a specific version Mar 09 00:20:26 and remove monotone-6 in hopes of getting the feeds to build once again. Mar 09 00:20:37 RP: hehe, the om3 patch is not workin on debian stable. svn info does not support remote URLs there Mar 09 00:21:53 zecke: oops :) Mar 09 00:22:14 RP: eek and svn log has no --limit option Mar 09 00:22:20 zecke: btw, I think we can have - in PV, its PR its illegal in? Mar 09 00:22:39 03coredump2 07org.oe.dev * r75b7f018... 10/ (5 files in 3 dirs): alsa: alsa-state: Load asound.state on boot... Mar 09 00:22:44 * RP thinks of 2.6.20-mm3 Mar 09 00:23:18 RP: in PR it is not legal, in PV it is decouraged according to your policy Mar 09 00:24:30 zecke: I was told it was discouranged, I'm not sure I ever agreed with that bit :) Mar 09 00:24:41 hehe Mar 09 00:25:31 zecke: ANy ideas on a good format for svn: x.y.z+svn+r123 ? Mar 09 00:26:16 RP: sorry I've shown up in the middle of this. Are you discussing svn checkouts and package names for them? Mar 09 00:26:17 now '+'? Mar 09 00:27:15 jselect: yes Mar 09 00:27:29 We've just been through this for some products at work. You might want to note if it's from a branch or tag .. the revision number doesn't tell you where in the checkout it was from. I suppose that's not as much of a problem if the svn path is hardcoded in the recipe :) Mar 09 00:27:48 just my AUD$0.02 Mar 09 00:28:15 zecke: You mean include now or you don't like the +? Mar 09 00:28:42 RP: I don't know if I like the '+', I just noted a difference Mar 09 00:29:11 jselect: Its a good point and we need to keep that in mind for say DL_DIR (which is already covered). For the package itself, its a package name issue really Mar 09 00:29:32 zecke: I prefer - myself... Mar 09 00:31:01 RP: to match with SRCDATE, I would leave the '+' out Mar 09 00:31:20 RP: I don't know the impact on readability Mar 09 00:32:22 I don't like the impact on readability :/ Mar 09 00:32:54 RP: change it to +svn+${SRCDATE} and +svn+r${REV} Mar 09 00:33:19 RP: put 'svn' is just a protocol, the real information is within the last number Mar 09 00:35:10 zecke: Yes, its just simply to tell us it was an SCM built package rather than a proper version Mar 09 00:36:14 * chouimat is back. Mar 09 00:37:08 zecke: I've replied Mar 09 00:37:11 Quick question.. is anyone on at the moment using mesa-dri and mesa-dri-drivers-*? Mar 09 00:37:47 I've got a problem where mesa-dri-drivers-* builds (specifically mesa-dri-drivers-r200) but it depends on mesa-dri and of course that is an empty package.. Mar 09 00:38:19 * mwester ponders "SCM-built package *rather than* a proper version" and scratches his head... Mar 09 00:38:28 RP: replied where? Mar 09 00:39:03 bah.. i'll submit a patch Mar 09 00:39:04 ALLOW_EMPTY_mesa-dri = "1" Mar 09 00:43:46 stupid spam filter... Mar 09 00:44:56 RP: cvs you can put a timezone (the current one) into the date Mar 09 00:45:34 RP: ack on SVN behaviour change Mar 09 00:46:11 RP: good thing with using the SRCDATE is we can use the tarball from the mirror without hitting the upstream server Mar 09 00:47:13 anyway, I need some sleep now Mar 09 00:49:02 zecke: I'm assuming its all GMT timestamps as cvs uses Mar 09 00:49:14 anyhow, I need sleep too :) Mar 09 00:49:19 'night all Mar 09 02:25:14 Hello all. **** ENDING LOGGING AT Fri Mar 09 02:59:57 2007