Guy Kewney

Apple Computer is a big company and, if you sold software on commission, you would be in a very good mood the day you signed up the chief buyer of Apple as a customer.

But think about Estrella Van Damm, a sole trader consultant in the tiny Spanish town of Orgiva, whose business struggles to run on an outdated Pentium II with a dodgy keyboard.

You wouldn't waste the cost of a stamp to send this business a brochure. Yet the two companies are, potentially, the same.

Look back a few years, and you'll find that Apple started out using the Intel 8080 processor. Or at least, that was the plan. But the founders of Apple had cashflow problems.

They couldn't afford the Intel chip, and Intel wasn't prepared to do a deal. So they used an alternative chip, based on the rival Motorola architecture.

So much for nostalgia - but there's an important lesson for Microsoft here. At the end of 2004, Windows NT4 dies, and we've just had Windows 98 taken off death row for a short time. Microsoft is becoming quite nannyish about people working with out-of-date systems.

"So why would Microsoft invest in supporting those people?" inquired a consultant when I raised this subject. And, he suggested, Microsoft should regard these customers as "natural customers for Linux".

This reaction was widespread. When I queried the wisdom of this approach, they seemed to think I was advocating that Microsoft provide service to old software users as a form of charity. Absolutely not! But Microsoft does have a real problem here: a huge installed base that it simply can't cast off.

Businesses that are refreshing their hardware this year will throw out Pentium III machines that smaller companies can use to run Windows 2000. But running Windows 2000 on 'inherited' software is becoming difficult.

And when it comes to taking over machines with Windows XP, Microsoft's hard line on registration and updating means that Van Damm Consulting really will have to use Linux.

More to the point, when this small company stops working out of the garage and starts buying a dozen machines and a proper server, it will be a Linux user.

And the next generation of super-corporations will employ Linux-trained, Linux-equipped executives who think Windows is strange. They'll have relationships with Linux developers, and find Microsoft licensing confusing.

Fortunately, Microsoft's own executives don't seem to be suffering from the same myopia as some of its consultants. Senior executives at Redmond see the solution as the development of a mini-footprint version of Windows, based on Windows Mobile.

Such a platform could fulfil the functions of Windows 3.1 and Windows 9x that second-generation owners of hand-me-down PCs use them for today, but would be inherently more stable.

Disadvantages? Well, the margin for Microsoft on a PC is already smaller than you might think. A small-footprint ARM-based PC, clocked at 200-400MHz, can't be sold for much less than a full-spec Pentium 4.

But there are savings, and the more savings there are, the more Microsoft's share has to be cut.

I think that the 'developing world PC' is a potential money-maker, and that Microsoft will start promoting it shortly. And once that bandwagon starts rolling, the kiss of death for Windows 98 will be final and swift.

Like this story? Spread the news by clicking below:

 del.icio.us     Digg this     reddit!