If you store much data at all, for most any purpose, having ZFS versus not having it is like developing software with source control, compared with life before

If you store much data at all, for most any purpose, having ZFS versus not having it is like developing software with source control, compared with life before source control. The main benefits without going nuts with some features you probably don't want on a workstation with anything short of crazy-high specs (mainly de-duplication, it eats lots of memory) are that it: 1) tells you when data has been corrupted even if the disk isn't erroring (and tells you exactly what was damaged) 2) has lightweight snapshots (it's copy-on-write so snapshots basically only cost the difference between the snapshot and either the next snapshot or the current FS) with a very handy access model (you can read from snapshots at will, as if they were a live FS) so you can snapshot before risky file operations and have an easy way to revert without having to go to backups; and 3) Makes various types of disk-pooling and RAID available purely in software. It also provides a common UI to all this across multiple operating systems, so there's less per-platform trivia to remember if you work cross-OS a lot. However, the UI is obtuse enough that it's firmly power-user-only, and a fairly advance power user at that.