There is a definite business appeal to cheap Network PCs, but the concept is not flawless

There is a definite business appeal to cheap Network PCs, but the concept is not flawless

As noted last week, we're on the verge of another generation of operating systems--the basic software for personal computers--with new versions being beta-tested by Apple, IBM and Microsoft. All provide more features, more power and more stability, and all demand more hardware. But if proponents of the so-called Network Computer have their way, users may be able to get by with much less while still doing more.

Oracle is a company in the business of selling high-end database software to big companies. Typically, these companies keep their databases on big, old-style mainframes, and Oracle CEO Larry Ellison claims to hate personal computers, saying he doesn't like having to drive to the store to "buy bytes." Instead, he has proposed the Network Computer--a sort of scaled-down PC that gets its software and stores its data over a network. The network might be corporate, or it might be the biggest network of all, the Internet.

Ellison suggests that a properly designed Network Computer could sell for under $500. This has been scoffed at: personal computers were designed in reaction to so-called dumb terminals, which are screens plugged into a big central computer and only able to run software approved by the high priests of corporate computing. Many suspect that Ellison's Net PC is an attempt to return to 1970s-style computing.

Still, companies including IBM and Sun Microsystems are backing Oracle in developing specs for the Net PC. They envision a minimally powered box, with 4 megs of RAM and a CPU equivalent to Intel's 486, which was popular a couple of years ago. No hard drive is needed, and home users could keep costs down by plugging into the TV set.

The computer would boot up quickly, and connect to the network or the Internet, and run a Web browser. Using Java software, it could run a wide range of software (yet to be developed), with the programs as well as the user's data residing in the network.

Oracle et al don't expect to be selling the hardware--they're developing the standard, and releasing it to the industry. They expect that regardless of who makes the hardware or which company's CPU is used, common standards equal easy connectivity. Their hopes are to develop two main markets: places where the high cost of owning a personal computer has been a barrier, like schools, libraries, kiosks in malls, and homes, that could get Internet connectivity with cheap Net PCs, and the business market. The business market might well be the bigger of the two, not only replacing a $2,000 PC with a $500 terminal, but allowing companies to minimize training and support costs that typically run several times the purchase price of the actual computer.

With more and more companies investigating Intranets, internal networks sharing information through common, easy-to-use Web browsers, a low-priced, low-maintenance system based on Network PCs might make a lot of sense.

Critics point out a few weaknesses in this strategy. Such terminals may be usable in a corporate network, where high-speed connections are already in place, at least once Java programs are developed to allow users to carry out common tasks like word-processing. But are users going to want to run applications over the Internet? Not if they have to rely on today's low-speed connections. And are they going to want to store sensitive data over an unsecured Internet connection? Will they want to give up the independence and features of a full-fledged PC for a low-priced network machine?

In some cases, it may depend on what the price spread turns out to be between low-end, full-function PCs and Larry Ellison's dream machine. Now if they can get its price down to around $200 or so, the Net PC might make more sense. And there's no reason they can't: beneath the covers of all those Sega and Nintendo game machines lie real computers, being sold at a profit at about that price point.

If Ellison gets his way, this merger of the Internet and cheap hardware could make those new operating systems from Apple, IBM, and Microsoft seem big, bloated and largely unnecessary. Maybe less really will be more.