Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!iuvax!bobmon
From: bobmon@iuvax.cs.indiana.edu (RAMontante)
Newsgroups: comp.sys.ibm.pc
Subject: Re: 12-bit or 16-bit FAT entry
Message-ID: <10207@iuvax.cs.indiana.edu>
Date: 3 Jul 88 22:50:48 GMT
References: <5930009@hpcupt1.HP.COM> <11454@steinmetz.ge.com>
Reply-To: bobmon@iuvax.UUCP (RAMontante)
Organization: Computer Science Dept., Indiana University
Lines: 14

I think you should infer the size of the FAT entry.  I have a 30Meg RLL drive,
currently split into four partitions.  MSDOSv3.21 chose a 16-bit FAT for the
25Meg (i.e., larger than the 20Meg "standard" limit) partition, and 12-bit
FATs for the other three partitions.

Norton's Programmer's Guide elaborates:  "If a disk format has more clusters
than [4080], then we need the 16-bit FAT....Our programs can learn the format
of a disk by reading and inspecting the FAT ID byte [the first bite of the FAT].
However, the official way of finding out the format is to use DOS function 27
(hex 1B)."  His description of this function indicates that it will report the
number of clusters, but not the FAT size itself.
-- 
bob,mon						(bobmon@iuvax.cs.indiana.edu)
"In this position, the skier is flying in a complete stall..."