[1275] in java-interest
re: The Future of Java (and Sun)
daemon@ATHENA.MIT.EDU (Robert Stephen Rodgers)
Mon Aug 28 16:27:55 1995
Date: Mon, 28 Aug 1995 12:34:38 -0400 (EDT)
From: Robert Stephen Rodgers <rsrodger@wam.umd.edu>
To: java-interest@java.sun.com
cc: java-interest-digest@java.sun.com
In-Reply-To: <199508280232.TAA28457@webrunner.neato.org>
> From: MikeDacon@aol.com
> Subject: Re: The Future of Java
>
> Glen stated that he did not see how Sun
> was going to make money off of Java.
Sun does not have to make money from Java for Java to have been a good
investment of resources. Rather, if _Java_ prevents Sun from _losing_
money (in the form of further marketshare and competitiveness in the form
of apps available on Solaris but not, e.g., AIX) it will have been a good
investment.
> I think they will do well making money in the most
> obvious way - a really cool IDE for java.
> (And a less-technical series of Web authoring
> tools based on Java).
Expanding on what I wrote above, I think that whatever Java's actual
history, the above paragraph entirely misses how Java fits into the
world today and how, if development proceeds and doesn't stagnate, it's
going to affect the world tomorrow.
A more obvious explanation for the way Sun has begun handling Java's current
development is as a hedge. If Solaris loses the operating system war, which
is going to happen if the SPARC continues to lose (and badly) the CPU war,
Sun will be in big, big trouble. Unix has long been heralded as a way to
get away from petty issues like CPU architecture, but in reality, binary
compatibility is as important in the workstation market as in the PC market --
there is no other way to explain Sun's continued dominance when they are
clearly outclassed and have been for at least two years in terms of
performance.
Sun is attempting a number of strategies to combat slipping WS marketshare,
the most obvious having been releasing non-SPARC versions of Solaris 2.x
(Intel and soon, PowerPC). Unfortunately, they suffer from the same problem
as Windows NT, which is equally portable, and available on much faster
hardware, and has at its disposal the development going on in what promises
to be the largest single platform since DOS and Windows 3.1 (Win32 in the
form of Windows 95). It's also less expensive, has more features (in a
single-user workstation environment, including OpenGL, out of the box SMP,
higher-scaling hardware availability, better connectivity, more end user
apps and threading that isn't F-ed). Windows NT isn't exactly setting the
world on fire, but it's carving out new niches (in particular, 3d), something
Solaris isn't.
That's the world Sun is looking at. Sun is one of my favorite companies,
and it's sad to see them steadily crawling more tightly between that rock
and hard place.
Java can help.
In the long term, from Sun's point of view, if developers transition from
C++ to Java, Sun will be able to abandon SPARC if the need arises and move
to one or more CPUs produced by third parties (PowerPC, most likely). If
they had to do it today, they'd be hurt *badly* in the transition -- Sun's
strength isn't their performance (though their balanced performance is what
has made them a holdout against CPU-heavy IO-light Intel systems), it's
in their software base. Portability -- making Sun's system design
choices more flexible -- is one solution to this problem, and with the
hottest chips on the market and arriving soon outclassing even the
upcoming SPARCs by a good margin (MIPS, PA and AXP), it's got to be
a solution on Sun's mind. In two years they're going to have cheap, very
fast PowerPC systems pounding them on the low end, and unlike the WinDOS
Intel systems of today, they're not going to be running a toy operating
system (except the Mac OS..).
It explains their interest in OpenStep, in Windows NT (with Intergraph) and
now Java. A popular OS (Solaris, ported to multiple platforms, or Windows
NT) with a common interface and compile-time portability is the first half
of the solution -- to avoid becoming marginalized (e.g., apps available
for Intel Solaris vs. SPARC, apps available for MIPS NT vs. Intel NT),
Java is the second half. Once the developer-side effort to port is gone,
systems _running the same platform_ _running Java_ wont be marginalized ever
again.
That may be the opposite of Sun's market moves over the past decade (which,
if anything, encouraged the marginalization of non-Sun workstations -- the
smart marketing which made Sun's compatibility/software base as important
and useful as it is today), but it's the only obvious way for them to come
out of this decade a strong and growing company.
> And of course, since SUN gave us such a great
> language as a gift. We will be loyal to SUN and
> buy their IDE when it comes out. Sunsoft has
> announced that they will be putting an IDE out in
> the first quarter of 96.
Dunno. I'm pretty mercenary -- I'll buy the best Java IDE, not just
the one from Sun. Price and features count more when your paycheck is
on the line.
RSR
-
Note to Sun employees: this is an EXTERNAL mailing list!
Info: send 'help' to java-interest-request@java.sun.com
