Path: utzoo!mnetor!tmsoft!torsqnt!news-server.csri.toronto.edu!bonnie.concordia.ca!uunet!munnari.oz.au!metro!macuni!sunb!ifarqhar
From: ifarqhar@sunb.mqcc.mq.oz.au (Ian Farquhar)
Newsgroups: comp.sys.amiga.graphics
Subject: Re: New IFF compression technique
Message-ID: <1225@macuni.mqcc.mq.oz>
Date: 18 Feb 91 00:36:19 GMT
References: <1991Feb17.015223.25784@uokmax.ecn.uoknor.edu>
Sender: news@macuni.mqcc.mq.oz
Organization: Macquarie University, Sydney, Australia.
Lines: 38

In article <1991Feb17.015223.25784@uokmax.ecn.uoknor.edu> drtiller@uokmax.ecn.uoknor.edu (Donald Richard Tillery Jr) writes:
>I've also heard that all the other values up to 9 are used as well.  I
>even heard that 0x02 was some type of Huffman encoding (like LZH).

No offense, but if you're going ti play about with compression, you'd
better study a it bit more.  Huffman and LZW are entirely different
systems.  Huffman maps a fixed-length character set into a variable
length character set on the provision that the more commonly used
characters in the source are the smaller ones in the destination, and
LZW builds a dictionary of recently used character streams which can be
referenced to reduce redundancy.  BTW, LZW is under a bit of a legal
cloud, owing to a patent reportly obtained upon it.  I have not heard
the outcome of the case.

[ Description of the system deleted ]

>That's it.  Any comments or suggestions are welcome.  And anyone who knows
>how I can register this please let me know.  

One of the areas where IFF will get into real trouble, and this is
starting to happen now, is if standardised formats end up with an
enormous number of variants that require continual modification of the
reader program.  One such danger is if ILBM starts to have a large
number of different compression systems assigned to it.

Now I would be the first to agree that the current compression system in
IFF is inefficient and poorly thought out, but I would strongly resist
any attempt to improve it without a careful consideration of the options
available (ie. huffman, lzw, jpeg etc.) to choose a *SMALL* subset of
efficient and suitable compression systems.  I just don't think that
the system described above has enough advantages to warrant its official
registration.

--
Ian Farquhar                      Phone : + 61 2 805-9400
Office of Computing Services      Fax   : + 61 2 805-7433
Macquarie University  NSW  2109   Also  : + 61 2 805-7420
Australia                         EMail : ifarqhar@suna.mqcc.mq.oz.au