The following was written by bob_furick@MSN.COM on the dotnet mailing
list. Note, this is provided as an interesting bit of
information/commentary. It may or may not represent my views and I refuse
to be called to account for its contents. So there. Have fun. :-)


Some facts on Microsoft and SUN to help in the JUMP religious wars. Might
as well start from the facts.

(1) Server Sales
Microsoft does not compete with SUN. Microsoft is a threat to SUN.
There is a big difference between the two concepts.
In the third quarter of 2000 ( July - September 2000 ) SUN sold
96,000 units of servers and the entire UNIX market ( HP, IBM, SUN, etc. )
sold 195,000 units. Microsoft sold 750,000 copies of NT and W2000
servers. That's an 8 to 1 over SUN and almost 4 to 1 over all UNIX
vendors. This doesn't include Linux, as nobody can figure out exactly how
many Linux copies are actually running.

More importantly, 60 % of Sun's sales revenue came from 10 % of the
units at the high end. People I know who worked at Sun in the mid 1990s
say that most of Sun's profits come from the high end units.

( July - September 2000 Sun unit sales - from SUN
publicity release 12/18/00 )

$ 750 Million - 723 units at $1.03 Million average
each
$ 902 Million - 8,110 units at $111 K average each
$1,179 Million - 88,108 units at $ 13K average each

That's $2,831 in server hardware sales for SUN +
$2,200 in software and accessories for a total $ 5.0 Billion for 3rd
quarter 2000.

Intel servers running Linux or W2000 generally cost 1/2 to 1/3 the
cost of a UNIX or Solaris box. If Microsoft shows that clusters of
inexpensive servers can replace the top end UNIX units, it will be a
disaster for Sun. Half their revenue will go away and all their profits.
That is why Microsoft is a threat, not really a competitor at the high
end.

Scott McNealy himself has said that Java is a calling card, a way to get
into customer's doors to talk abut servers. Sun can only maintain its
lock on Java by casting somebody as an "evil empire" that Sun must protect
against. By waving the Microsoft flag, Sun has been able to delay
releasing Java to the public domain. If you read the San Jose Mercury
News, (http://www.BayArea.com) you will see the large and real contempt
Silicon valley types have for software engineers in the rest of the world.
They are the "smartest" the "best" and the "brightest". Anyone who proves
otherwise is subject to attack.

(2) Languages and Copyrights
In the 50 years of computer science, no major language except Java
has been copyrighted. Where would the industry be if Bell Labs had
copyrighted C (1978) and C++ ( 1984), and sued Hewlett Packard for
creating STL. What if Bell Labs sued anyone who built a run-time library
that didn't pass Bell Labs tests. What if McCarthy ( MIT 1959 ) had
patented Lisp, and sued to stop new implementations. Lots of what we
know about recursion, denotational semantics, and dynamic binding comes
from Lisp.

All languages are built from the ideas of prior engineers. Bridge
designs and automobile designs are based on imitating prior engineering
success. The same holds true for programming languages. Quotes from
History of Programming Languages ( Thomas Bergin ACM Press, 1991
0-201-89502-1) tell most of the story about Lisp, ALGOL 68, C, and C++.

Here is part of the Lisp story by Guy Steele :
"Moreover in the 1960s ,IBM and MIT disputed who had
invented core memory and IBM insisted on enforcing its patents against
MIT. MIT responded by declining to use IBM equipment as extensively as it
had in the past. This provided further impetus to use DEC equipment
instead, particularly for Lisp and AI ................ ( page 248)

Because Sussman had been studying ALGOL [Naur 68], he
suggested starting with a lexically scoped dialect of Lisp. .... It
appeared that such a mechanism would be needed anyway for keeping track of
acquaintances for actors. Lexical scoping allowed actors and functions to
be created by almost identical mechanisms. " (pg 249 )

Dennis M. Ritchie talking about C
"The scheme and type composition adopted by C owes
considerable debt to ALGOL 68,although it did not, perhaps, emerge in a
form which ALGOL adherents would approve. The central notion I captured
from ALGOL was a type structure based on atomic types ( including
structures ), composed into arrays, pointers ( references ) and functions
( procedures ). ALGOL 68's concept of unions and casts also had an
influence that appeared later " ( page 679 )

Bjarne Stroustrup on C++
" One common question at C with Classes presentations was "Why use
C? Why didn't you build on say, Pascal? One version of my answer is
found in Stroustrup [ 1986b]." ........................

" At the time I considered Modula-2, Ada, Smalltalk, Mesa, and Clu
as alternatives to C and as sources for ideas for C++ [ Stroustrup 1984b],
so there was no shortage of inspiration. However only C, Simula, ALGOL 68,
and in one case BCPL left noticeable traces in C++ as released in 1985.
Simula gave the ability to declare variables anywhere in a block (Section
15.3.3.1) and BCPL gave // comments (Section 15.3.3.1)" ( page 711 )

The quotes above highlight three points
(a) all languages are built from the successes of prior art. Part
of an engineers' responsibility is to not re-invent the wheel. We build on
the backs of others, whether for airplane wing design or for programming
language design.
(b) Copyrights on languages are a bad idea. None of the
advancements above would have occurred if each language had been
copyrighted.

(c) SUN needs a copyright on Java to get in the door of its
customers. It needs to sell those big servers. SUN used the "Microsoft
Bully flag" to keep control of Java.

Now that Microsoft is gone, do you think SUN will release JAVA to the
public domain?

bob furick

---
Ice Z - Straight Outta Redmond