A ton of catch phrases here but... SUNW Reply # of 11411
Sun Microsystems
The Wall Street Journal -- October 26, 1998
Manager's Journal: The Computer of the Future
----
By David Gelernter
Your job if you are an academic technology researcher is to invent ideas, dress them up snugly and send them out into the world. You can cheer them on, but in the end they have to make it on their
own. Your reward if they do is to be acclaimed a "visionary" and issued a temporary license topontificate.
Recently some of our ideas have made it, and I've got my license and the clock is running, so here are my predictions -- about the world beyond Microsoft Windows, the future of the Web, the prognosis
for academic technology research in general.
My pontification credentials: When Sun Microsystems announced a bold new technology strategy this summer, it was big news (because, for one, the anti-Microsoft forces seem to be rallying around Sun).
It was also a gratifying moment for our Yale research group. Sun's new strategy is called Jini -- an everybody-rhumba approach to networking in which all sorts of electronic devices can pool their
efforts and strike up liaisons as soon as they are plugged in, without worrying about compatibility or formal introductions. Jini is built using a system called JavaSpaces, a contraction of "Java" and "tuple
spaces." Java (the chart-topping hit programming language) was partly inspired by my 1991 book "Mirror Worlds"; Nicholas Carriero and I invented tuple spaces in the early 1980s. Jini uses some
other ideas from "Mirror Worlds" also.
A tuple space is a sort of shared bulletin board that allows computers to communicate by posting notes. A mirror world is a "live" image on a computer screen of your organization (company, school,
city); it can be tuned in over a network, viewed and "entered" from afar. Some people claim that the 1991 book anticipated the World Wide Web, which took off in 1994.
A New Haven company called Scientific Computing Associates foresightedly built the first industrial-strength version of tuple spaces in the mid-1980s. Today SCA has customers world-wide,
from oil and pharmaceutical companies to microelectronics and financial services. But only recently, as computer communication has emerged as one of technology's most important topics, has the
computer industry itself begun to chug relentlessly in its direction. IBM, among other companies, now has its own version of tuple spaces.
Prediction-wise I am hardly batting a thousand. We've launched plenty of other ideas (more radical than these) that are still wandering in the wilderness, ultimate fate undetermined. So, do
my predictions come true? Maybe. Here goes:
The Microsoft Windows view of computing (formerly known as the Apple Macintosh view) is obsolete. It dates from a long-dead era when most people owned only a relative handful of computer
files, since they hadn't been computer users for very long; when e-mail and the Internet were largely unknown; when memory and computing power were expensive. In the New View, your files will be
stored not on some particular computer but afloat in cyberspace, so you can tune them in from any network-connected machine: at home or work, in a hotel room or a phone booth.
Filenames and directories are gone: You're relieved of filing and organizing duties, except when you actually want them.
Everything is laid out on a single browsable timeline, and it's the system's job to get you what you want. You tell it to fetch "my last letter to Schwartz," or "everything dealing with
zeppelins"; or you
can cruise the cyber-landscape, browse from 10,000 feet and set down where you like.
Your files
and e-mails stretch into the past, your calendar and appointments into the future. The
system can
squish any group of documents into an "executive summary" at the click of a mouse. But
unlike most
of today's systems, it won't have a million options and switches and "features"; your mother, or even a
CEO, will be able to master the controls in five minutes.
We're building such a system, called Lifestreams. (I have a commercial interest in this prediction; the
system is being developed by Mirror Worlds Technologies, a company that is partly mine.) But Lifestreams is only one example. The winning New View might (for all I know) be manufactured by
the National Biscuit Co., or Microsoft. In any case, its characteristics will be similar to what I've described.
The real "Internet era" will begin when two things happen. First, the Internet's most important mission will be to feed you information about yourself -- to store your own files and cyberstuff, not show you
other people's Web sites. The Lifestreams system uses the Internet this way; other systems will too.
(To a limited extent, some already do.) Phone, television, video-and-music libraries and other communication services will all come together on the Internet ultimately -- but it won't work unless
the second condition is met: Your postmodern computer, TV and phone must all be operated (like a Buick, Jaguar and Hyundai) using basically the same set of controls. Lifestreams can serve as such a
"universal dashboard." Other systems will too.
And a new kind of network is about to emerge alongside the Internet: "cyberzones" --
currently a research topic, not a product. A cyberzone is associated with a particular place -- say, an airport gate. When you enter the zone with your cellphone or laptop, you can tune in and pick
up information. (When is the flight due? What's the weather en route?) Classrooms can be cyberzones; also office buildings, libraries, supermarkets, stadiums and lots of other places. And a
group of laptops or cellphones can create a zone from scratch by themselves, on the spur of the moment.
When businessmen gather around a conference table, each one pushes a button on his laptop and an ad hoc cyberzone leaps into existence.
All these potential advances have grown out of academic technology research. Our ideas have made money, increased productivity, created jobs and knowledge. But will universities
continue to attract enough good young minds to their computer science and engineering faculties?
A few years ago, Lifestreams had matured enough to spawn a serious project. We asked ourselves:
Should it be a university project (we raise grant money, build the software at Yale, distribute it free)?
Or do we start a new company? I favored the academic approach, being an academic.
But we decided to try both and see what happened. We described the project to what seemed like the right federal funding agency, and also to prospective private investors. Result: We couldn't get the funding agency to invite us to Washington for a visit, but prospective investors came to New Haven to see us.
Nowadays, private capital often seems a lot more enterprising than federal research agencies.
The new company is bubbling along beautifully, thanks to my inspired former graduate students Eric Freeman, Scott Fertig and Susanne Hupfer. As for us profs, the university has made us
a generous offer. It can't touch industrial salaries; it can't underwrite the people and equipment we need for software research. But it can allow us to be entrepreneurs in our spare time. Like
thousands of other professors, I've taken the plunge.
The problem is, we didn't come to the university seeking the freedom to be entrepreneurs. We came seeking the freedom not to be. We are good at thinking, writing, making things; no good at raising
money, overseeing budgets, placing conference calls.
It's lucky that the university has an ace up its sleeve: absolute intellectual freedom; the liberty to roam anywhere. No privilege is more precious. I'm grateful for it every day. I owe a lot, also,
to a few federal funding officers who did support our research back in the old days. I myself have no grounds for complaint -- but I wonder about the future. My younger, harder-headed colleagues
look at me; they look at my counterpart at (say) Sun Microsystems; they can't figure out why the hell they'd want
my job and not his. If they have to be businessmen anyway, why bother with students?
If I were running a university, I'd think it over.
---
Mr. Gelernter is a professor of computer science at Yale and author, most recently, of
"Machine
Beauty: Elegance and the Heart of Technology" (Basic, 1998).
Powered by Quote Agent® and News Agent® from IDD Information Services
Copyright © 1998 Dow Jones & Company, Inc. All Rights Reserved.
|
