I'm having fun with Btrfs, it 's a nice FS and it's got fancy features.
I won't explain it to you as some many sites have really good documentation about btrfs (i.e.
http://www.funtoo.org/BTRFS_Fun). I will just introduce it. First, Btrfs is not a standard FS, it's a container. It can contains files, directories (like other standard fs as extX) but it can also contain sub-containers. These sub-containers are named subvolumes and they can be mounted individually.
What the need for subvolumes?
I wanted to install 2 distributions with their homes in a 100 GiB HD with an Windows already installed. Don't ask for Windows. I didn't want to waste more HD space with extra slices so I created a big Btrfs partition, plus a swap partition, and create 4 subvolumes (2 for the roots and 2 for the homes). And all subvolumes share the same disk space. No 5 GiB wasted here, 7 GiB wasted here, every free byte on the Btrfs partition is available for every subvolumes.
But it's not all. if I want to change something in a installed distribution and I'm not sure of the result, I can create a snapshot of the subvolume containing the distribution and restore it if my modification has gone wrong or boot on the snapshot.
A Btrfs parition can be composed of two or more HD (or another mass memory). Each drive can be dynamically mounted or unmounted.
I have not tested every possibility of Btrfs yet but i already like it.
