Viewpoint: Is Microsoft backing the wrong OS?
Microsoft may be king of the software heap, but it risks losing its way on OS strategy
?Pile it high, sell it cheap? used to be the war cry of Tesco. And as supermarket philosophies go, it has been resoundingly successful ? elevating the retail giant to among the most profitable companies in the country.
Bill Gates might not exactly be in the baked bean business, but his approach isn?t massively removed from Tesco?s. Except he piles IT high and sells IT cheap. It?s a strategy that has equally catapulted Microsoft to the top. But as the Seattle software colossus moves from shrink-wrapped products to sales focused more on enterprise solutions, there are worrying signs that its operating system strategy has become hopelessly confused.
If the company?s hype is to be believed, Microsoft is rapidly moving towards the fusion of Windows 95 and NT, making NT a superset of Windows 95. For that to occur, a consistent set of APIs is necessary. Yet what is happening? The number of program interfaces unique to each operating system is growing by the week.
Moreover, Microsoft has always insisted that a key difference between the two is that NT is built on a micro kernel, whereas Windows 95 will never be, though one of the key reasons for a micro kernel is that it allows the operating system to be hardware independent. Yet now we learn NT will no longer run in future on Power PC or Mips platforms, effectively ruling out a major part of the Risc server and workstation market. Whether NT will continue to be supported on Digital?s Alpha chip also has to be in doubt, given that the excuse for axing Power PC and Mips was that neither shipped in sufficient volume. That leaves the prospect of NT running only on Intel ? dispensing at a stroke with what was meant to be one of the key differences between NT and Windows 95.
Not only that, but the two operating systems continue to have different interfaces, when the justification for supporting the same interfaces is now more compelling than ever. Whether it?s a strategy that pays dividends for Microsoft is not the point. Gates is big enough to look after himself. But for Vars, forced to choose which operating system to invest in developing, the choice can be critical to survival.
As Microsoft grows in stature, there are indications it is losing focus in other areas too. At one time it slated the mainframe as a technological behemoth, yet now it argues its Transaction Server is ideally positioned to play a key role in three-tier, client/server computing. It was once highly sceptical about the network-centric paradigm as posited by Oracle CEO Larry Ellison and built around the Web metaphor. But Microsoft is now hurtling at a furious pace towards this very technology.
The problem, of course, is that Microsoft is getting too big for its boot-ups. Whereas once it had clear business vision, it is now saddled with a plethora of applications and tools, not to mention the problem of supporting a huge legacy of Dos and Windows products. But perhaps the biggest danger of all is in Gates striving to be all things to all customers, be they standalone PC users or vast corporations.
History has already demonstrated how other IT giants like IBM and Digital paid the price for similarly attempting to be a one-stop shop, yet in the software sphere this appears to be the very same strategy that Gates is pursuing. Whether he too will end up hopelessly overstretched has yet to be determined. Either way, Gates needs the channel ? which is probably why in the past few years he has worked hard to build up one of the best Var networks in the business.
It is probably why, in the US at least, Microsoft is urging business affiliates to acknowledge the company as their preferred software supplier in the hope of keeping an edge and ? in an almost domino effect ? impose its products from on high at every conceivable opportunity. Having Microsoft as your best software pal doubtless brings with it huge marketing clout when pitching for a contract. But those who, having signed up for the program yet don?t include Microsoft in a solution sale, are then obliged to give the Seattle developer a hour-long ?courtesy? briefing on why.
No one can accuse Gates of not being pushy. Yet at the end of the day, if a reseller is to survive it means satisfying the customer. Microsoft products may be piled high, they may be cheap. But without clearer focus on operating system strategy and how Microsoft?s myriad other tools and products gel within a larger business enterprise, there is a risk of the whole lot tumbling down. Like poorly stacked cans of baked beans.