On 21 Mar, Peter Garay wrote:

> maybe I should keep my mouth shut if I am not sure what I am talking about but
> my memories are that you had to reformat every drive when you moved them
> around with those "old" controllers, where the controller was generating the
> signal that actually went to the head. The clocking of the controller changed,
> the signal changed slightly, so you low level formatted every time you moved
> around (also most times worked without that but was not reliable).
This is correct for MFM, RLL, ESDI, SMD, ???.

> on SCSI the drive gets a command, and pretty much formats the drive the same
> way, so once the drive is formatted, now the best thing the format does is
> rebuilding the bad block table. If the disk is from an other type of
> system, the bad block list might have been used for data storage and it should
> be rebuilt.
The bad block list of SCSI disks can not be used for data storage in
any way. The disk hides this storage and the spare blocks from the
outside. You need special SCSI commands to read the defect list. 

> New drives transparently map errors, still some systems like to see the empty
> bad block list, some systems verifies all blocks and  reserve bad areas
> during the initialization and do not rely on bad block info from low level
> format.
You have to distinguish between the disk-internal bad block management
/ list / spare blocks and the bad block management of the OS / file
system. The disk-internal bad block management (e.g. of SCSI disks) is
OS and file system independent, because the OS does not see any bad
blocks in normal operation. Old MFM, RLL, ... drives could not do
automatic, transparent bad block remaping, so the OS had to do this in
a sepatate table. This table is placed on the normal storage blocks of
a disk and completly independent from the internal bad block list of a
SCSI-Drive. 
-- 



tschüß,
         Jochen

Homepage: http://www.unixag-kl.fh-kl.de/~jkunz/