Congratulations, now Linux has (basic) rollback of the system too! :)
As the years pass, BTRFS will get more mature. It is said it takes at least five years before a filesystem matures to the point it is trustworthy - after it has been released officially as v1.0. Look at Enterprise file system ZFS with all Sun's excellent technology and engineers, it still has bugs - seven(?) years after official release! It takes many years to iron out bugs after a filesystem goes live.
As of now, BTRFS snapshots is quite messy to deal with. In OpenSolaris and ZFS, it is much easier to handle. But this is a cool feature. Sometimes when I upgraded to a alfa version, I had some problems in OpenSolaris. I just rebooted in GRUB and rollbacked. Nifty.
Originally Posted by movieman
Planned features include:
* Object-level (RAID5-like and RAID6-like) parity-based striping
I take that to mean the BTRFS volume will have, or be able to have, parity information spread around appropriately, so that BTRFS is like a RAID container, without the RAID container, i.e. file-system-level data redundancy.
Please correct me though if I'm wrong. I take that sentence also to mean that of course all file writes and reads will be striped, but that's sort of duh since it simply means making full use of all the drives which are part of the BTRFS volume.