Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!shaman!jiro
From: jiro@shaman.com (Jiro Nakamura)
Newsgroups: comp.sys.next
Subject: Re: NO MORE Squash... (was Re: Using OLD 1.0 Disk as datadisk)
Message-ID: <1991Jun15.170228.2479@shaman.com>
Date: 15 Jun 91 17:02:28 GMT
References: <78577@brunix.UUCP>
Sender: jiro@shaman.com (Jiro Nakamura)
Organization: Shaman Consulting
Lines: 65

In article <78577@brunix.UUCP> rca@cs.brown.edu (Ronald C.F. Antony) writes:
> Squash is the most stupid software ever seen on a NeXT. Instead of
> writing a program that is compatible with hundreds and thousands of
> existing UNIX machines, i.e. writing a graphical user interface to
> (gnu)tar, compress, ar, et al. the author of this program thinks that
> by using a incompatible scheme and trusting the stupidity of naive
> users in combination with distributing for free (haha) a UnSquash only
> program, he can create a new standard by means of which he can cream
> off the money in the NeXT market. 

    I agree that the incompatibility of Squash with standard *NIX software
is a major drawback. However, remember that most of the Squashed software
will be NeXT-specific anyway (I think) or user-specific (archives, backups
etc).
    Whatsmore, the author did an excellent job with the compression
algorithm. Many many *NIX compression methods can handle NeXT sound
files? Squash does a pretty good job on these. It also beats compress
in most cases, I've found.
    The author is also not unprecendented in his distribution method.
Sit for the Mac, I believe, was distributed in the same way (albeit
the original versions were shareware).

> Sorry, I only hope that someone will find the time and write a similar
> program that interfaces to at, tar, ar, compress and uncompress. For
> such a program I would pay, but not for someone who tries to destroy
> standards for personal profit. (I think however that many people
> proficient enough to produce such a program would make it publicly
> available anyway...) 

    What I hope for is that the author writes command line versions
of Squash and also ports it over to other machines.
    If you want commercial quality compression (ie., better than
compress), then you'd better pay for it. Let's hope that some really
smart people (ie., not me) at the FSF make the next generation
file compression package.
    One problem though, is that most of the really good compression
algorithms are not publically available. Compress is based on the 
LZW algorithm, which is patented (as everyone knows ad nauseum) but
apparently isn't being rigorously enforced in the *NIX kingdom as a
file compression algorithm. Most of the good algorithms are spin-offs
of LZW. Huffman just doesn't kick it, although some variants are quite
good. 
   It takes more than a good programmer to write a compression package.
It takes a excellent mathematician and a superb lawyer (sigh...). I am
waiting for a publicly available good archiving-compressing program,
but until then, I'll use Squash.

> Another neat thing would be a program that creates file packages for
> the NeXT. 

    Agreed. Note, I'm not disagreeing with you totally. I'm just
taking a more realistic (?pessimistic?) view of what is possible today.


   - Jiro Nakamura
     jiro@shaman.com

// Disclaimer: No connections with Agog (makers of Squash), although
// I was a very active beta tester (I think) and have communicaed with
// the author on many occasions. My general feeling is that he deserves
// to be able to have bread on his table. 
-- 
Jiro Nakamura				jiro@shaman.com
Shaman Consulting			(607) 256-5125 VOICE
"Bring your dead, dying shamans here!"	(607) 277-1440 FAX/Data