Fun_People Archive
3 Feb
Nothing at all to do with Monica Lewinsky.


Content-Type: text/plain
Mime-Version: 1.0 (NeXT Mail 3.3 v118.2)
From: Peter Langston <psl>
Date: Tue,  3 Feb 98 17:32:44 -0800
To: Fun_People
Precedence: bulk
Subject: Nothing at all to do with Monica Lewinsky.

Forwarded-by: Nev Dull <nev@bostic.com>
Forwarded-by: "Joel N. Weber II" <devnull@gnu.org>
http://www.pbs.org/cringely/text.html

                              [LINK]

                         February 2, 1998
                      PBS Online I, Cringely
            "I think, therefore, I think." Volume 1.44
                           I, Cringely

May the source be with you: two important events that have nothing
                at all to do with Monica Lewinsky
                       by Robert X. Cringely
  ______________________________________________________________

Nearly 20 years ago, I worked for awhile in the White House. The
President back then was Jimmy Carter, former Democratic governor of
a southern state who had a pretty and forceful wife, a young
daughter, an embarrassing brother, and who thought that the answer
to just about any problem could be found in some new government
policy. Sound familiar? Carter had admitted to Playboy his "lust in
the heart" for other women, but most people focused on his
born-again Christianity. In truth, Carter's personality had
elements of both qualities (after all, he was a politician). The
most telling story I recall was how the Secret Service tried to
please Carter by removing all the booze from Air Force One only to
have the Commander-in-Chief ask for a highball during his first
foreign trip as President.

It's not Monica Lewinsky, I know, but that highball anecdote at
least shows that things are generally not as they seem.

Last week in the world I work in, the apolitical, often asexual,
but definitely cut-throat world of computers, two important things
happened that were also not as they seemed. Compaq bought Digital
Equipment Corp. and Netscape started giving away its Web browsing
software for free.

The DEC sale says much more about Compaq than it does about
Digital, a company that had been adrift for several years. Sure,
Digital already had all the elements for fashioning its own
salvation and renewed vigor, but like Dorothy, it had always been
that way. Those ruby slippers hadn't done anything to increase
shareholder value in at least a decade, so there was no compelling
reason they were likely to get tapping now. So the merger comes not
so much from Digital's failure as Compaq's success: DEC could have
limped along for years but Compaq needed it NOW.

And why did Compaq need DEC so much it would throw almost $10
billion (nearly $5 billion in cash!) into buying what most other
computer companies just would have seen as a pack of troubles?
Because Compaq CEO Eckhard Pfeiffer is a man with a mission, to run
the largest computer company in the world. Pfeiffer's target is
IBM.

To compete with IBM, Compaq has to be a truly global company with a
complete product line. Starting with a PC hardware company,
Pfeiffer added networking, mainframes (first from Tandem and now
DEC), minicomputers, workstations, microprocessors, service, and
software. The parts of IBM he specifically avoided were chip-making
factories, which are hideously expensive and almost instantly
obsolete, and basic research. Compaq will buy any new technologies
it needs, thank you.

The result is a lean and mean vertically-integrated company that
can meet or beat IBM at every turn. DEC fills some important gaps
for Compaq, notably in microprocessors, workstations, and software.
Tandem Computer, which Compaq bought last year, uses MIPS R10000
processors in its Nonstop systems and that had to change. Expect
Tandem to move to DEC Alpha chips. Digital workstations give Compaq
a one-up on IBM and allow the company to at least appear as a blip
on Sun's radar screen. Sun is, of course, another Pfeiffer target.

The Alpha chips and Digital's discount Pentium deal with Intel were
attractive, too, because now Compaq has even more leverage with
Intel. Not only can they, as the largest user of microprocessors in
the world, get the best prices, but the Alpha CPUs offer to Compaq
the same leverage they gave to DEC -- the ability to tell Intel to
go to hell and roll their own if necessary. It will never become
necessary.

But the most interesting part of the deal is the software expertise
it brings to Compaq. Minicomputer and mainframe companies, unless
they are building clones of some other system, have to be
responsible for their own system software, yet Compaq had always
left that to Microsoft. No more. Digital came with abundant
software talent. All those engineers who left Digital for Microsoft
to do Windows NT left plenty of their friends behind.

So let's say Compaq is successful and turns itself into a leaner
and meaner IBM, who does that threaten besides IBM? Microsoft.
Suddenly Compaq has its own operating systems, and good ones at
that. Sure, Windows 95 and NT are bread and butter, but doesn't
Pfeiffer have some room left for cake? Microsoft's biggest single
customer will not be so dependent as before.

And then there is Netscape, which a few days ago announced that its
Navigator and Communicator browsers would be available for free. On
the surface this looks like Netscape just had to meet Microsoft's
price point for Internet Explorer (free), but the announcement is
much more insidious than that. In the area of browsers, at least,
Netscape has made a master stroke that even Microsoft may not be
able to overcome.

How do little software companies co-opt big software companies? One
method invented in 1982 by Sun Microsystems is to freely license
source code. Sun came from the Berkeley Unix community, where Unix
source code was licensed to universities for a flat fee. An entire
generation of programmers grew up doing regular recompilations of
their favorite operating system. It's hard to imagine now how
amazing that was, because just about all the OSes before Unix were
strictly commercial with the source code hidden in corporate
vaults. Microsoft source code has never been seem by anyone who
didn't work for Microsoft or IBM. But Sun would sell a license for
anything, everything, and in doing so made itself a powerhouse in
workstation software. The hardware was just carried along by the
code.

Then came the Open Software Foundation, the Free Software
Foundation, and products like the GNU compilers and, of course,
Linux. Suddenly not only was source code available, it was free!
The success of Linux, FreeBSD and several other important-yet-free
operating systems is a result of these movements. Thousands of very
capable programmers are working hard to improve this code, for
free, for fun. It's revolutionary. and it's what Netscape is
proposing to do with its Web browsers.

For now the browsers are free, but with Version 5, Netscape says
the source code will be free, too. Why? Well, it's a little like
Tom Sawyer painting that fence: giving away the source code will
encourage those thousands of programmers to donate their efforts to
improving Netscape's product. That's more programmers than Netscape
could ever afford to hire and more programmers than even Microsoft
would devote to one project. It should result not only in Netscape
regaining profitability, but in Netscape browsers retaining market
dominance. And one thing's for sure: Microsoft won't follow suit.
When Bill Gates first heard about the idea of giving away Internet
Explorer, he called it Communism. Well, Netscape's move IS
Communism and Microsoft can't do the same without shaking the very
foundation of its corporate structure.

If it's such a great idea, why wait until Communicator 5? Because
lots of other companies own code that's in Netscape browsers and
that code has to be either licensed under the same terms or
removed. And there's also the issue of expurgating those
embarrassing comments in the source code that programmers never
thought would be read by anyone outside Netscape.

So there you have it, two important blows against Microsoft.
They're no Monica Lewinsky, but they're a start.


prev [=] prev © 1998 Peter Langston []