Peter Bowditch's Web Site

Advertising policy

The way we were...

This article appeared in The Sydney Business Review on 15 September, 1995

This column appears on the 29th anniversary of my induction into the computer business. The launch of Windows 95 has made me think about how much has changed over those years, and also how little has changed.

There must be some people alive who have not heard some of the hype about Windows 95. Much of the hysterical and ill-informed opposition to the product seems to be based on a fear that one company will dominate the personal computer industry. Too late! Two companies have had almost total control since 1981 - Microsoft and Intel. Market domination has been a fact since commercial computers were first produced. In 1966, the second-largest computer company in the world, Honeywell, had 5% of the market. The largest then, IBM, is still the largest player in the game.

Another worry for the pessimists is that a single operating system will define the computer industry. Too late! An operating system creates the environment in which software runs, and sensible software manufacturers create their products for the biggest markets. The industry has been dominated for years by three strains of evolution - DOS/Windows, UNIX, and IBM's OS-360 and its descendants. There have been other systems which sold well (DEC's VMS, Apple's Mac, etc) but they have all been marginal and confined to a single manufacturer each. DEC and the other surviving minicomputer makers are now embracing UNIX, Apple is allowing (too late) other people to use its operating system while simultaneously and desperately telling customers that their machines can run MS-DOS and Windows. There have been clones of IBM's mainframes since the 1970s (Hitachi, National Semiconductor, Amdahl) claiming compatibility with IBM's system software.

Then there are the people who complain that they may have to update their hardware and learn some new things. I can remember managers who whinged about the retraining costs for programmers when hard disks replaced tapes, about the costs of providing terminals for programmers when multiuser online systems came along, and about how the proliferation of personal computers was a threat to discipline. I'm sorry, but they just don't make 78rpm record players or 80286 processors any more, which is why record companies make CDs and software companies write for the Pentium. By the way, the realistic minimum system for Windows 95 (486DX2-66, 8Mb RAM, CD-ROM) is also the minimum for most games these days.

There are always some software writers who complain when any new technology comes along. They did it when IBM introduced the 360 architecture, they do it each time a new COBOL standard comes out. Usually, their software is badly written, or they believe that nothing will change, or they are just too lazy to learn something new. I have two accountant clients using different tax preparation software packages written for DOS. The 1995 versions of these things require so much memory that they can't be run from Windows, and one won't run if any network software is loaded. Windows 95 will in fact give these kludges a reprieve, but DOS has gone and so have software developers who think that DOS is all that they or their customers need.

So if some things never change, what does change? Engineering, manufacturing, economics and what's possible! The price and power of today's personal computers was unimaginable just a few years ago. The range of affordable software running on these machines is immense and getting better. In 1983 the construction company I worked for paid $180,000 for a project management program to do what Microsoft Project or CA SuperProject can now do for around $1000. In 1988 I was using a statistics package which cost Macquarie University $100,000 to do functions which are now included in Microsoft Excel's $600 price. The computer on my lap has more memory and disk capacity than the total of all machines in the first three mainframe installations in which I worked. The principles of computer logic, program design and so on have changed little. The power to do things has changed a lot.

The challenge is to use this power effectively. Meeting this challenge requires the ability to accept that today's problems are not tomorrow's and that change is inevitable. Bob Dylan had some advice for those who want time to stand still: "don't stand in the doorways, don't block up the halls".

Aahh! The computer business. Chaos one day, different the next. You wouldn't be dead for quids.

Copyright 1998- Peter Bowditch

Logos and trademarks belong to whoever owns them