It was bound to happen. After years of nearly empty rhetoric, commercial enterprises are finally beginning to use what has come to be called client-server computing technology. Naturally, a reaction has set in. Glass house information processing executives, aided and abetted (if indeed the whole thing was not indeed at their instigation) by vendors with a lot at stake in a vision of computing’s future that looks suspiciously like computing’s past, are beginning to talk about a return to limited function terminals… without fully understanding the implications of their words. We don’t doubt that these information processing executives have accurately perceived a problem: the loss of control over data that can occur in a distributed, decentralised workstation environment. But by aligning themselves with advocates of a renaissance of the dumb tube they are making a big mistake. They are asking to pay for something that is going to be costly and useless: dumb tubes with mice. The emerging vision is being shaped by companies like Oracle Corp and IBM Corp. These two vendors have a lot in common these days. In the first place, they are justifiably nervous about the immense power of Microsoft Corp and Intel Corp. And that’s just one of their problems. Their second big issue is one that gives them something in common with Microsoft – a belated appreciation of the appeal of Internet technology – and something else that again gives them nightmares starring Bill Gates – they may secretly believe that Microsoft will do a better job of co-opting the Internet than they will. A third problem that these vendors would like to solve in a real hurry is the threat that Lilliputian machines pose to Brobdingnagian ones. That last item is also on the fear list of a vast army of computing executives whose careers have grown though an attachment to mainframes, like ivy on a stone building.
Drastic overreaction
Everyone knows that there will always be settings in which end users do not need a fully functional desktop computers. And there are plenty of cases in which local accumulations of data and local applications can create problems for enterprise computing. But throwing away the personal computers is a drastic overreaction. The right answer is to tame the desktops and preserve every possible option to make end users more productive. Chances are, some of the good software will require smart terminals. And if it turns out that all you need on the desktop is an applet or two dozen, well, they will be available for personal computers as well as so-called Internet terminals. No software vendor is going to succeed at productivity applications without going after the market segments that will always buy smart workstations, such as users on the road, personnel in small branch offices, high-level managerial and clerical employees and everyone else who needs a more versatile computing tool kit than a terminal and server set-up can economically provide. And there is one other thing about a personal computer that is not going to be true of a snazzy but brainless terminal: a personal computer can allow an end user to ride out the storm when a network or server or central system crashes. – Hesh Wiener
From the March 1996 edition of Infoperspectives published by Technology News of America Co. All rights reserved.
