That's fine with me. I never really wanted Windows to be Windows if that meant unreliability and performance problems and viruses and fences everywhere that prevent true integration.
But he seems to think that the complexity of Windows provides some kind of barrier to entry in the desktop market. If so, he's living in a dream world. All he has to do is look at Apple's OS X. It is, in the opinion of many, an operating system superior functionality and usability to Windows. Microsoft isn't the only OS company in the galaxy. Indeed, the thing that has saved Microsoft from serious competition in the OS space is not the complexity of Windows, it's the personality quirks of Steve Jobs.
I suspect Mr. Hilf could run the numbers better than I can, but I've long thought that if Steve Jobs would just tweak OS X to run garden-variety Windows apps on garden-variety PCs and price it to be competitive with Windows XP, by this time next year 12 percent of the PCs in America would have moved off Windows. Look at the dent Firefox has put in Internet Explorer.
OS X is very much an "inside the box" product. If you begin to think outside the box, Mr. Hilf's paycheck is in even bigger jeopardy. As software becomes a Web service, the integration is going to take place farther and farther away from the desktop, and the need for complexity in the OS is going to diminish. This is at least part of the reason for the flurry of excitement last year when it looked for a day or two as if Google was going to bring out its own, presumably Linux-based, desktop operating system. Google is the company that is demonstrably thinking farther outside the box than any other software company on the planet. Its approach to software is exactly what Hilf speaks so scornfully of -- and what has made the Internet such a success: "small pieces loosely joined," to use David Weinberger's book title.
I grant you that the reasons for complexity change over time. The complexities of Windows originally solved the problem of providing a standard -- and therefore learnable -- graphical user interface. Today the complexities of Windows revolve around solving the problems of security. But the very complexity of Windows seems to make it part of the problem, not part of the solution. Virtualization seems to be a promising avenue for protecting the desktop, and as soon as Microsoft gets Vista out the door, I suspect it will devote a great deal of attention to virtualization. But if Microsoft really thinks it can hide behind the complexity of its operating systems, sooner or later it will find it's not hiding from anything. Integration, like the graphical user interface, won't be the problem anymore -- not because.Linux solved it, even, but just because the world moved on and Microsoft was left inside its box.