First of all, you need git to install git (always strange to install such software on a server, but who cares )

sudo aptitude install git-core

When you already have a Dokuwiki installation, you have to convert it to a git repository. Because git clone doesn’t allow using a non-empty directory, we have to work around this limitation. Warning: Using this method you will loose any changes made to Dokuwiki specific files. Configuration and data should be in their own files so you won’t loose such data.

git clone -b stable -n http://github.com/splitbrain/dokuwiki.git
mv dokuwiki/.git/ .
rmdir dokuwiki
git checkout -f

When you are going to install a new installation, use this command

git clone -b stable http://github.com/splitbrain/dokuwiki.git wiki

For security reasons its recommend to disallow the .git directory for HTTP requests by adding these lines to the .git/.htaccess file. You should also remove install.php.

rm install.php
vi .git/.htaccess

order allow,deny
deny from all

To update your installation, just use the command git pull, which will pull the latest commits from the original Github repository.

git pull

arm-elf-gcc: /lib/libc.so.6: version `GLIBC_2.11' not found (required by arm-elf-gcc)

Oh yeah, binary compatibility is sometimes hard in open source world. On the other side, why binary compatibility, we have the source! But of course, its easier to use binaries then compile the same on every machine…
I found out that the function mkstemps is used from the new glibc version:

$ readelf -a arm-elf-gcc | grep GLIBC_2.11
    11: 0000000000000000     0 FUNC    GLOBAL DEFAULT  UND mkstemps@GLIBC_2.11 (4)
   427: 0000000000000000     0 FUNC    GLOBAL DEFAULT  UND mkstemps@@GLIBC_2.11
  008:   3 (GLIBC_2.3.4)   0 (*local*)       2 (GLIBC_2.2.5)   4 (GLIBC_2.11)
  0x0030:   Name: GLIBC_2.11  Flags: none  Version: 4

The functions, namly mkstemps and mkstemps64, were added with glibc 2.11. Grepping showed that the function mkstemps is used inside the library “libiberty”. The configure script checks if the function is available on the host, if yes, it will use the hosts function, if not, it will build its own. As soon as the compiler descide to use the hosts mkstemps function, the resulting binary need glibc 2.11, which I would like to avoid.
The solution is pretty simple: Add the object file “mkstemps.o” to the variable REQUIRED_OFILES in Makefile.in. This forces the Makefile to compile its own mkstemps function, which will be prefered when gcc is linked.

$ vi libiberty/Makefile.in

REQUIRED_OFILES = ./mkstemps.o                                          \
        ./regex.o ./cplus-dem.o ./cp-demangle.o ./md5.o ./sha1.o        \
...

My gcc’s glibc requirement is now 2.4, which is much older then gcc 2.11! As a result my gcc runs even on very old linux systems

To convert the VMware images I used this How-To. This worked quite well for most of my Virtual Machines. But several had snapshots or were split into 2GB files. This cant be handled by qemu-img. I converted the files to single, flat disk file first using vmware-vdiskmanager. In the end, I wanted to have pre-allocated (single file) anyway, because it provides best performance:

$ vmware-vdiskmanager -r Ubuntu.vmdk -t 0 Ubuntu-copy.vmdk

This takes a lot of time, and doesn’t work for snapshoted files (I get always an error saying “parent of this virtual disk could not be opened”). Therefore I searched for another solution which I found using “vmware-mount”, an utility provided by VMware. Because I already set up a new Ubuntu Server Installation, I tried to avoid installing VMware Server 2 again. I had to copy some libraries around to start that utility direct out of the unpacked VMware directory.

$ tar xvzf VMware-server-2.0.2-203138.x86_64.tar.gz
...
$ cp -R vmware-server-distrib/lib/lib/ vmware-server-distrib/bin/libdir/

After that, I mounted the source VMware disk (“-0000002″ points to the current running snapshot).

$ sudo /home/user/vmware-server-distrib/bin/vmware-mount VirtMachine-0000002.vmdk /mnt

Using that command generates a flat view to the VMware disk using FUSE. This disk is then mounted using “losetup”. By using the command “losetup”, we can figure out where that flat FUSE file lies on the filesystem. We see that the loopback device created by vmware-mount uses a offset, which is the effective beginning of the partition which was mounted. But we want the whole disk, with the partition table in it. So we can ignore this offset.

$ sudo losetup -a
/dev/loop0: [0015]:2 (/var/run/vmware/fuse/6512247014271621407/flat), offset 32256

The file “/var/run/vmware/fuse/6512247014271621407/flat” contains the flat disk. This can be copied by just invoking the copy command “cp”. Its a good idea to unmount the disk before, to avoid writing new data to the disk during the copy process.

$ sudo umount /mnt
$ sudo cp /var/run/vmware/fuse/6512247014271621407/flat /path/to/image/VirtMachine.raw

This raw file can be used by KVM. If you want a qcow2 file, you can use qemu-img on the source file, and choose qcow2 as output…
If you move Windows Machines to KVM, also consider this Blog post to avoid BSOD’s. I had the 7b BSOD, and wiped VMware away already. I had to set the disk to IDE and use a Windows Server 2003 Disc to repair the installation (Using “Enter” to install, and Repair the detected installation with “R” then)

Definitiv erfreuliche Resultate. Die Performance Verbesserung konnte gemessen werden, insbesondere mit Windows 7. Der Bootprozess bei Windows XP ist bereits so schnell, das da nicht viel raus schaut, zumal das BIOS/Bootloader selber etwa 5 Sekunden in Anspruch nehmen… Die Messungen sind keineswegs exakt, jedoch zeigen sie dass sich mit VirtualBox 3.1.0 wirklich etwas getan hat!

$ sudo aptitude install xserver-xorg-video-vesa
$ sudo vi /etc/X11/xorg.conf

Section "Device"
    Identifier    "Bochs VESA VBE 2.0"
    Driver        "vesa"
EndSection

Eventuell müssen die Modes noch angepasst werden. Bei mir funktionieren damit Auflösungen bis 1920×1200!

with (GetSelectedFolderResource()) prompt("URI: " + URI, URI)

Danach habe ich ein neues Tastenkürzel definiert, welche dann die selektierten E-Mails verschieben soll:

MsgMoveMessage("imap://falstaff@mail.deheime.ch/user/spam/Spam")