There was quite a lot of reporting and commentary last week over Gartner analysts Michael Silver and Neal McDonald's recent presentation alleging that Windows is collapsing. I stayed out of it; I think my views are reasonably obvious at this point: Windows should collapse, but it probably won't. Mary Jo Foley has a post up detailing many of the reasons why.

To her grounded analysis of Microsoft's planning and strategy that make the demise of Windows unlikely, I would add that there is a significant part of the IT industry and of IT professionals who have a vested interest in continuing to promulgate Windows, no matter how unwieldy and poorly suited it becomes for business. There is an entire generation of techs and executives for whom Windows is all that they know and all that they want to know. legacies like that don't die out quickly or even collapse. People will point to IBM's position when Microsoft was coming to the forefront of the PC market, but even the PC revolution didn't really knock IBM's existing mainframe and mini-computer business out of existence. Most of the organizations that were using them continued to do so for some time, and many still do, even though they have been eclipsed by the proliferation of PCs on desktops. There's every reason to think that this will continue to be the case with Windows as well. The only thing which might reasonably eclipse it at this point is SaaS computing, but just as with desktop PCs and mainframes, there isn't really complete overlap between the two; just as legacy programs continue to require mainframes to run on, PCs are still the primary means for accessing SaaS applications.

Of course, SaaS minimizes the importance of the operating system and Apple and Linux proponents won't hesitate to point out that they can provide either more stable or much cheaper mechanisms for the same purpose. Be that as it may, those who are familiar with Windows are going to continue to use it even if it's poorly suited for the tasks at hand; it's what they know, and it's what they have, and there are friction costs for changing those things.

All the same, it is expensive and inefficient, and over time, business will surely trend toward things that are less expensive or more efficient because those that do will succeed while others fail. So I wouldn't say that Windows is going to collapse, but unless Microsoft can pull something out of its hat better than a glorified Vista update for Windows 7, it's surely going to begin to wither in the next five to ten years.