Please note: the author has performed only minimal testing, and that under QEMU rather than KVM itself, so would welcome feedback from anyone who makes serious use of KVM.
KVM is a kernel-based virtualisation system for Linux, which you can invoke from the command line. If you prefer a graphical front-end, you could try out AQEMU — there are source and binary packages.
KVM requires hardware virtualisation extensions to function efficiently.
If your Linux distribution does not include KVM, you can obtain it from the KVM Downloads page. You will need the kvm-kmod package.
You will require the OVF Tool program, available from VMware's Open Virtualization Format Tool page.
Then:
Use ovftool to convert GNUstep-VM-0_9.ovf from OVF to VMX format, e.g.:
ovftool http://www.rstonehouse.co.uk/extras/GNUstep-VM-0.9/GNUstep-VM-0_9.ovf \
~/GNUstep-VM-0_9.vmx
(change the destination for the VMX image to suit your needs — note that it will be about 1.5GB in size).
Start KVM, using the '.vmdk' part of the VMX as the 'disk image' e.g.:
qemu-kvm -m 256 ~/GNUstep-VM-0_9.vmdk
KVM offers a large number of command-line options (see man kvm) and it is likely that you will want several of these. In particular, it is recommended that you increase the memory allocated to the VM to at least 256M (as shown in the above example) — with the default setting of 128M, the system thrashes badly. If you have a lot of memory, allocating more than 256M to the virtual machine, up to a maximum of about half the machine size, will improve performance even further.