Internet Link Exchange
Member of the Internet Link Exchange

The 'Futzing Factor' And The Network Computer

March 19, 1997

Ok, we've read all the hype about the Oracle Network Computer, Sun JavaStation, IBM NetStation and Microsoft NetPC, but what will really happen in corporate America when all these devices are readily available and proven alternatives to the fully equipped desktop PC?

Here's my prediction: not as much as any of these interested parties seem to think. The basic idea of the "net computer"--I'll use this as the generic term-- is to provide access to computer applications via a GUI without incurring the high cost of operation of today's PC.

Since that cost comes from the combination of the PC's complexity and variability (no two seem to be the same in all respects, including installed hardware devices and software), Larry Ellison's original idea was to simplify the end-user device to the point where any operational decisions could be made centrally and costs could be contained.

So let's say that, a year from now, enough net computers of some variety have been in the field for long enough so that some research organization can go around computing the cost of ownership and tell us all the sad news. It's still going to be expensive to run distributed client/server operations with net computers, because net computers themselves are still only part of the operation- -they all have to connect to their servers, which will connect to other servers, which will connect to other servers, and so on ad infinitum. But it will probably be less expensive than running those operations with full-fledged PCs on end-user desktops.

And why is that? It's the "futzing factor," a term popularized in the latest Gartner Group survey of PC cost of ownership. Futzing around with all the options that PCs allow us to install and modify (again, both hardware and software) costs a lot--in fact, in my estimation it costs more in lost productivity than all the surfing of all the nonbusiness-related Web sites that you might look at in a typical day.

Just think about how much time it took to make the screen look the way you really wanted it, the network card to hum, the modem to respond, the hard drive to be organized and all the rest. This first generation of network computers will take a lot of those options away, and that will be the most important thing that they do to reduce total cost of operation of a client/server network.

My prediction is that net computers in their various forms will be wildly popular with the IS groups in most corporations because they bring a lot of issues back into the IS shop to be resolved. They also will be immensely popular with the typical "I just want to get my job done" end users (assuming, of course, that IS can keep all those servers up and connected). But they'll be definitely unpopular with the "I just got the latest gizmo on my desktop PC" hacker.

You can be sure that most large organizations will have a combination of PCs and net computers around the shop in the not-too-distant future. There aren't many that will make the commitment to throw all those Pentiums out the Window and replace them with, say, Sun's JavaStations running HotJava or IBM's NetStations, running whatever.

As to the popularity of the NetPC running a scaled-down Windows 95, it will depend on just how many options Microsoft leaves in--and how much the company allows us hackers to keep on futzing.

Jerrold M. Grochow is CTO at American Management Systems, an international consulting and systems development company. He can be reached at

Source: PC Week

Copyright © 1996 NCNS News. All rights reserved.