Message-Id: <199705161309.JAA04721@delorie.com>
Date: Fri, 16 May 1997 08:09:59 -0500
From: "Jonathan E. Brickman" <brickman AT cjnetworks DOT com>
Reply-To: "Jonathan E. Brickman" <brickman AT cjnetworks DOT com>
To: opendos AT delorie DOT com
Subject: Re: 16 vs. 32-bit performance
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Precedence: bulk

> So although 32-bit systems and programs are usually somewhat faster
> than 16-bit systems and programs, real-life results vary quite a bit,
> depending on the application, the skill of the programmer, and the
> quality of the compiler.

It's actually even worse than that.  Real-life speed results from
the combination of quality of programming of OS, applications,
and compilers.  And a huge number of ports are not done very well
at all in terms of OS and CPU optimizations.  I saw the
full source code to "zip" (the original Unix ZIP file handler)
a while back.  It is horribly complicated.  Trying to rewrite
that source optimized for a very different processor would take
many hours.

Jonathan E. Brickman         River City Computing, Inc.            (913) 232-6663
http://www.cjnetworks.com/~rivercity                    brickman AT cjnetworks DOT com
It seems to me that men usually think more about carburetors, and women
think more about doors.  I think the world needs really good carburetors...and
really good doors.