From: Ray Cromwell <rjc@clark.net>
Date: Thu, 20 Jul 95 13:45:51 PDT
To: tcmay@sensemedia.net (Timothy C. May)
Subject: Re: Netscape the Big Win
In-Reply-To: <ac3402ec0a0210046af5@[205.199.118.202]>
Message-ID: <199507202044.QAA14974@clark.net>
MIME-Version: 1.0
Content-Type: text/plain


Tim May wrote:
> I am sorry that some folks heavily committed to the Linux route, or to
> Emacs, or to GNU/FSF, or to other approaches feel that their work is
> technically superior and deserves to be as popular as Netscape and simiar
> approaches, but reality is reality.
> 
> (And I could be wrong on the way things will unfold. All I'm saying is that
> technology is a moving target, that plans have to change, and that ease of
> use will likely win out over technical sophistication. Folks who think the
> stronger technology will inevitably win should pick up a copy of a
> 15-year-old book called "The Soul of a New Machine," by Tracy Kidder.)

   I agree wholeheartedly with this. When General Magic first released
the Telescript white paper, I was really hot for the technology. I tried
to become a developer, I sent mail to every General Magic employee on the
net I saw posting (one guy even CC'ed me accidentally to his manager
saying they should hire me). I did searches in the media for any
mention of it. Harry Hawk even had dinner with the VP of Product Development
at General Magic. Alas, they would not give out alphas/betas of the 
development environment, which is all the same, because they don't know
how to market Telescript and make it a defacto standard. Instead of charging
for the interpreter/server, they should have given away the servers and
development stuff for free, or near free, and made their money by selling
services and clients (personal digital assistants using Magic Cap and
Telescript). The result is that no one uses Telescript except AT&T.
If I had gotten my hands on Telescript, I would have wasted lots of time
and effort on a failed product (failed in my eyes, because of its potential)
[lesson: proprietary programming languages fail unless they come embedded
within a killer consumer application]

    Then I got into Safe-Tcl, which is a little more promising, but still
a failure because there was no "killer app" which used it and which would
encourage its incorporation into other servers and clients. HTML would
have failed were it not for Mosaic. I was on the Web when it only had a line
mode browser and it was about as exciting as Gopher.

   I think Sun has taken the right approach with Java. Giving out Alphas
and Betas for free with source code. Encouraging heavy porting, and
incorporating it into a "killer app" (HotJava). They will make money
by licensing and selling tools and environments for Java, but their biggest
success will be that it will become the defacto "enabled content" language.

   Java still lacks what Telescript has (the ability to checkpoint execution
state and migrate execution across servers seamlessly), but what Telescript
has that Java doesn't isn't enough to make people wait for it, or pay
lots of money to be developers for.

   I could be wrong about how successful Java will be, but my confidence
factor is high.

-Ray