In Ina Fried's article Microsoft's 'big bang' could be its last, I couldn't disagree more with Gartner fellow Tom Bittman's comment that Windows XP (or any other Microsoft product) is "stuck in the weeds..."
Windows XP SP2 is NOT the same product today that Windows XP was in 2000. It is dramatically improved -- more stable, more security conscious, more reliable -- though not generally more feature-rich.
The problem for Microsoft is not that Vista will come out five years after XP was first released. The problem is that Microsoft does not make any money off of interim releases (Service Packs), so it must make up those upgrade costs on pricey retail sales and on OEM sales of new systems. The enterprise balked at Software Assurance and customers constantly complain about the high cost of upgrades.
But what about the competition?
Like Windows XP, the last four releases of Mac OSX (that's 10.1, 10.2, 10.3, and now 10.4) were virtually identical (feature-wise) to the original and yet EACH required a new license -- and, in many cases, applications software upgrades. In contrast, over that same period, Windows XP Service Pack 1 and SP2 were released (along with a long list of security and performance updates) and all were free. No one complained about the cost of the Mac OS X upgrades -- but Microsoft didn't lose any market share to Apple during that period either.
Yes, Microsoft would benefit financially from fewer "service packs" and more frequent "featureless upgrades" and the associated license fees but, in the end, the consumer and the enterprise would spend more money and the enterprise would be even more reticent to upgrade between releases.
So what about the Linux factor?
Well, I am not so sure that Linux is any more feature-rich today than it was when Windows XP shipped either, but I do know that the retail prices of fully supported Linux servers and clients are about the same as those from Microsoft So where are those disgruntled Microsoft customers going to go to get all those new features?
It would certainly be to Microsoft's advantage to be able to deliver more features at a faster pace in order to generate customer interest in upgrading (and thus spending more money), but perhaps right now Microsoft needs to be paying more attention to those markets where they fall short -- not on the desktop but in the server room.
If Microsoft wants to stay on top, they need to be able to compete for the lucrative server business. They don't need to add features. They need to improve performance so they can truly compete with UNIX on the high-end of the TCO scale and with Linux on the low-end of the TCO scale.
Ease of use goes a long way when dealing with the consumer -- but price-performance is a far more valuable commodity to the enterprise than low upfront costs or pretty new bells and whistles.