By now you have heard the news: WinFX will be shipping on XP, WS03 in order to make both Longhorn and WinFX ship with higher quality and sooner. If you haven’t already, check out JimAll’s Channel9 interview. Longhorn will contain core-OS level features focusing on the basics: security, reliability, manageability and performance (and cool new look-and-feel feaures). While WinFX will be shipped with Longhorn and available for XP SP2 and WS03 in order to expose the developer platform to a broader reach. Of course Longhorn will ship the .NET Framework 2.0 (Whidbey) redist and WinFX, but there will not be lots of interdependencies dependencies between WinFX and Longhorn. Along those lines, WinFS, the storage engine for Windows, will not be on this train. It will likely be in beta when LH RTMs.
At PDC2003 Jim Allchin announced WinFX. At the time he made a passing comment that really stuck with me. I imagine that most folks missed it or quickly forgot it. But he clearly said that this is an unprecedented early-look at a major OS release and that some of what you are hearing today will NOT make it into the final product. Turns out he knew what he was talking about (a bazillion years shipping Windows releases will do that for you I guess ;-)). While it is true that some of what we talked about at that PDC will not make it into the LH product, I still strongly believe in the vision we laid out and the company is still executing on that vision, it is just going to take us a couple of hops to get all the way there.
I think this is a great move for a couple of reasons:
1. Reach. We are now able to bring the greatness of WinFX to a much broader set of customers, making WinFX an even more compelling development platform.
2. Timing. This plan optimizes for shipping sooner. And that is goodness for the entire ecosystem. There is really only one way to make a software project ship sooner: Simplify. And simplify is what we did with Longhorn and WinFX. We simplified the features, simplified the dependencies, we simplified the product. As an aside, adding resources rarely makes the project more simple and therefore often does not help you ship sooner.
As far as the stuff I deal with on a day-to-day biases things will get a lot easier for me. We have a more scoped set of APIs in WinFX so I can focus better on cleaning those APIs up and making them shine.
What do you think? Good move or not?
If you had $100 to spend today on everything you ever heard would be in “WinFX” or “Longhorn” where would you spend it? This is exactly the question we are asking internally now as we crisp up our definition of WinFX over the next few months. Your feedback would be very valuable.