Talk a little more about the incubation of new technologies, rather than just having integrated innovation, going forward. Let me give you an example: WPF, Windows Presentation Foundation, we have it in the marketplace, but the Windows shell doesnt use it yet. And thats OK.Its a good example of incubate or incubate and innovate than opposed to innovating and integrating all in one big bang. Take what were doing with advancing storage. You know were taking the WinFS and some of those principles and well have a release of SQL Server that includes those technologies. We can bring it to the file system and worry about the shell and changing all the applications in the second phase. So, in some sense, you could say it is what we did with .Net itself, which first shipped as an application runtime with developer tools and now its migrating to be system services, and I think there is good learning in that for us. But Windows PowerShell is an example of a new technology that you managed, relatively successfully, to incubate and integrate successfully. How do you decide which new technology goes on which track? Its all judgment at the end of the day about what is manageable, technological risk and where you go beyond. I mean, it is not like we are not going to do any integrated innovation, its just that we are thoughtful and we have to learn from experience where we are crossing the line beyond the state-of-the-art innovations coming together. If someone came out and said that Boeing was going to build a new airplane under a new design process, with jet engines that have never been used on any aircraft, with a new material for the shell that has never been used before in flight, Ill bet that is a project that takes longer than saying we are going to reuse more components from the past and only innovate in a few of the key component areas. It is the equivalent of that analysis. Next Page: Service provider.
If we tried to hold Windows until the shell could rewrite, then were back on very long cycle times.