Each file or folder on a Macintosh volume is identified by a unique catalog node ID (CNID) in the Catalog file. The CNID is assigned at file creation time.
The catalog node ID (CNID) is a unique 32-bit number between 10h and ffffffffh. Each CNID is unique per disk. If a file is deleted, its CNID may never be reused.
CNIDs are usually assigned sequentially. If you reach CNID ffffffffh, no matter that the disk be empty, it can no longer be used.
If this should ever happen (with a file server, whatever), you have to back up and restore the drive, whereupon each file gets a new CNID.
This is the weird code you'll see in the file name, of a file with a long name (over 32 chars) created on Mac Os X (or other system) when accessing it with Mac Os 9. (and this banned me from translating some Gigasampler libraries
)
This system allows to move a file where we want, even if an app is using it. On other systems the file is identified by his name and access path, so it can't be moved while it is used.
A lot of software authorization techniques uses this unique CNID of a file to detect that the file is the original or a copy.
... and this is the background of the ASR trick! LOLASR when restoring a image file, it restore the original CNID of each file also, so most of the software can't detect that we are running a copy!