Hi Peter and all
Hard to get time these days for checking this board..Just got an interesting article in my mail box that I want to pass along a propos the current discussion on where do we go from here...(Bill Joy is in the NOVL BOD.)
Regards
Victor
THE ANGLER
Bill Joy on Sun's long road from Java
to Jini -- and where Microsoft comes in.
By Anthony B. Perkins
The Red Herring magazine
January 1999
If we were charged with inventing the computer business
all over again, Bill Joy would certainly be one of our
starting five players. (The others would probably be Bill
Gates, mainly because you wouldn't want to compete
against him; Steve Jobs, for his marketing and product
design savvy; Andy Bechtolsheim, for his hardware
architecture and networking expertise; and Bill
Campbell, who would be our inspirational leader and
chief baby-sitter.) Mr. Joy's past achievements --
including inventing Berkeley Unix and cofounding Sun
Microsystems -- would have been enough to land him on
the team. But his most recent efforts as the spiritual guru
of the Java and Jini movements make him arguably our
most valuable player. John Doerr, the Kleiner Perkins
Caufield & Byers partner who first sponsored his firm's
investment in Sun and sits on its board, concurred with
this assessment recently: "Bill Joy is rarely wrong," he
told a packed house at a November Churchill Club
meeting in Silicon Valley.
We first talked with Bill back in the summer of 1995 (see
"Software's Other Bill"). Netscape had just gone public
(Nasdaq: (NSCP), and the word Java was beginning to
mean more than coffee in Silicon Valley. To Bill, the PC
didn't represent a particularly great use of technology. He
dismissed it as a relatively inexpensive device that "you
can buy ... in parking lots across the country." Besides,
he said, "Microsoft owns all the application categories,
anyway, and they have no real interest in doing much
innovation, so the whole PC space is just kind of boring."
There was a raging debate going on then over what the
dominant home computing device would be. Bill argued
that no single device would prevail: "To me, the PC, the
TV, and the telephone are three different things -- they
offer three different solutions." This was our belief at the
time, too, and we stand by it today (see "Ten Trends for
the Post-PC World"). Likewise, we have always
believed, as Bill has, that the infamous network computer
proposed by Oracle would end up being not one
product, but many products, whose value would derive
from their affordability and their connection to a universal
network (see "The Strange Evolution of the NC").
The emphasis of Bill's sales pitch for Java, however, has
changed quite a bit over the years. At our interview, he
explained that the core technology for Java was really the
outgrowth of his realization four or five years prior that
consumer electronics needed better software. So his
team folded a Sparcstation into a little box with a touch
screen, animated interface, and wireless link. The box
had spread spectrum capability and a PCMCIA card:
basically it was the prototype for a fully animated,
cartoonlike remote control for the home. It was "way
cool," according to Bill, but it never took flight, because
the communications, game, and computer companies all
started crashing into each other, and no one firm really
owned enough of the standards to bring it all together.
"The whole thing became one big gridlock," as Bill put it.
The next thing Sun tried to do was angle Java toward the
TV set-top box. And, as it turned out, a cheap enough
solution couldn't be designed for this application either.
What next? When Sun saw the Internet coming in the
spring of 1994, it decided to focus solely on applying
Java to the Net. "Supercharging browsers became our
new mission," Bill said that sunny afternoon.
Well, as we all know, the third attempted implementation
of Java didn't really work out either. To date, Java does
nothing for your browser except slow it down, and it has
yet to become the platform of choice for embedded
systems. Java has, however, become an important
programming language that makes developers more
productive and a crucial middleware architecture that
helps tie together disparate systems across
heterogeneous networks. The surge in the adoption of
Java applications for servers and middleware,
particularly for very large and small companies, has
surprised even the most optimistic pundits. By 2002,
according to the International Data Corporation, the size
of the Java market should exceed $2 billion, with nearly
1.5 million programmers coding in the language.
Bill isn't conceding the consumer market, however. In
November at NDA, the Red Herring's conference for
CEOs, he made a presentation about Jini, Sun's
Java-based networking technology. Jini, Bill said,
promises to let devices ranging from PCs to smart cards
discover other devices on the same network, without the
need for configuration, drivers, or directories. Cool stuff
-- in principle. And, following the examples of Linus
Torvalds with his operating system and Netscape with its
browser, Sun plans to open up Jini for modification and
expansion by outside developers. The licensing model,
which developers call "community source," means that
Jini can be tweaked, hacked, and hopefully improved,
for free. Developers of commercial applications using Jini
will have to pay royalties to Sun, but Bill has said he
expects those royalties to be no more than $1 or $2 per
consumer device sold.
Red Herring editor Jason Pontin and I caught Bill in the
lobby after his talk and pressed him for more details on
his current vision for computing. "The PC of the next
wave will be a combination organizer, pager, and cellular
phone -- a pocket-size personal communicator that's
always on and always hooked to the Net," he said. And
did he believe that this next wave would crash down on
Microsoft? "That's why Gates is worried about Symbian
[the new operating system for handsets that the leading
players like Motorola, Ericsson, and Nokia are rallying
around]. You combine this threat with the undisputed
success of the PalmPilot -- this must be scaring the heck
out of Microsoft," he answered. Bill Joy's theme for the
future, he explained, is "simplicity for the consumer."
But hadn't Java failed to catch on for the client side?
"Well, there have been some strong forces trying to
prevent it," he claimed. His real answer came next.
"When the tools and next generation of virtual machines
come out, which will improve performance and
compatibility -- combined with the fact that you can
program in Java four times as fast as you can in C --
even the big guys will be forced to convert," he said.
Finally, we asked Bill what market share he estimated
that Java could earn in its mighty battle with Microsoft
for the client. "I think that we can get somewhere
between 30 and 70 percent and that some kind of
Coke-Pepsi balance can be struck in the industry," he
said. "You know, Windows could run Java, and Java
could run Windows, and we could compete in the way
that we're supposed to." And if Sun were successful in
this goal? "We'd then have to operate under some of the
same constraints that Microsoft, as a dominant player,
should be operating under already."
|
