From owner-freebsd-hackers  Tue May  7 00:34:38 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id AAA21276
          for hackers-outgoing; Tue, 7 May 1996 00:34:38 -0700 (PDT)
Received: from eac.iafrica.com (h196-7-192-157.iafrica.com [196.7.192.157])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id AAA21270
          for <hackers@freebsd.org>; Tue, 7 May 1996 00:34:30 -0700 (PDT)
Received: (from rnordier@localhost) by eac.iafrica.com (8.6.12/8.6.12) id JAA02147; Tue, 7 May 1996 09:30:34 +0200
From: Robert Nordier <rnordier@iafrica.com>
Message-Id: <199605070730.JAA02147@eac.iafrica.com>
Subject: Re: dosfsck anyone?
To: msmith@atrad.adelaide.edu.au (Michael Smith)
Date: Tue, 7 May 1996 09:30:33 +0200 (SAT)
Cc: hackers@freebsd.org
In-Reply-To: <199605070604.PAA20167@genesis.atrad.adelaide.edu.au> from "Michael Smith" at May 7, 96 03:34:15 pm
X-Mailer: ELM [version 2.4 PL24 ME8a]
Content-Type: text
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

On Tue, 7 May 1996, Michael Smith wrote:

> It's fairly easy to determine which sectors in the FAT have had the Big A;
> if you look at a "typically" corrupted filesystem, there'll be problems
> related to entries in one or more contiguous sectors.  
> 
> If I recall correctly, DOS' normal allocation strategy tends to use the 
> clusters towards the end of the filesystem last. 
> 
> A combination of these two would tend to imply that you could generally 
> expect to pick up free clusters from the end of the filesystem, and that
> it should be possible to determine whether the FAT in that area had been
> damaged.

OK.  I guess that should do it.  Which at least opens up the possibility
of conjuring up '/lost.fnd', if necessary.  And I'll give the other stuff
a go, and see how much additional complexity is entailed, if any.

> 
> > hierarchy, and this just doesn't apply in the DOS world ... where
> > designing "a better disk defrag utility" seems like a neat learn C++
> > project. 8(
> 
> Yecch.  I tried (several times) to write a 'dosfsck'-style program for
> the Atari ST, which uses essentially the same filesystem layout.  It
> was one of the most frustrating and scarring parts of my early
> programming days 8(  (None of the commercial tools was particularly good,
> and the filesystems were sufficiently different that you couldn't just
> move it to a DOS machine and hit it with 'chkdsk'.)

Wish I'd know this before: maybe some subtle hints, along the lines
of "unfinished business", could have gotten a BSD 'dosfsck' from you,
and saved me the trouble. 8)

-- 
Robert Nordier