Update: YES, it can happen... If
the operation below had been a Move,
I would have lost all of the files...

I recently did an alt-Copy from a
large directory with a large number
of small files with a 2GB Jaz disk
as the target... ZTree dumped a
LARGE number of blocks into the
windows cache, and the cache software
proceeded to overload the Jaz firmware,
which started throwing errors back
(at least, I think it was the Jaz
firmware...-- It could have been
windows...), including READ errors
(I presume, against the FAT or a
directory). The upshot is that the
Jaz disk is now unusable, it won't
even low-level format...!!
(fortunately I know some people and
will be able to get a new one in
trade...).

I WOULD LIKE TO STRONGLY REQUEST
IMPLEMENTATION OF SOME KIND OF
PACING PARAMETER, e.g., "don't dump
data faster than x MB/sec
(calculated using the actual cluster
size, NOT the file size -- my guess
is that part of the problem I
encountered was the very small
files into a 32K cluster...)