Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!harpo!decvax!cca!ima!haddock!johna
From: johna@haddock.UUCP
Newsgroups: net.micro.pc
Subject: Re: Cluster size and Hard drives - (nf)
Message-ID: <167@haddock.UUCP>
Date: Wed, 16-May-84 23:43:15 EDT
Article-I.D.: haddock.167
Posted: Wed May 16 23:43:15 1984
Date-Received: Fri, 18-May-84 02:28:35 EDT
Lines: 21

#R:sri-arpa:-60000:haddock:13600007:000:866
haddock!johna    May 16 16:09:00 1984

> So much for the "yes" part.  The reason there's a trade-off between volume
> size and cluster size is because you have finite-sized FAT's.  Yes, you
> can increase the FAT size (i.e. the number of FAT's), but you have to be
> *damn* sure that DOS is clever enough to use such big FATS's.

This information is not correct!  The reason for the trade-off is because
each FAT entry is only 12 bits.  Expanding the number of FATs will do nothing.
Only one FAT is normally used, the other is there for recovery procedures.

FAT = File Allocation Table

Clusters (allocation units) are linked together using the FAT.  Reducing the
sectors per cluster will just hide part of your drive from DOS.

DOS for the PC contains a table on the first real sector of the fixed drive
which is used to partition a fixed drive.  It is documented in DOS 2.0 .

			    ima!haddock!johna