C# and Longhorn

I want to learn C# so I can create some really awesome applications for Longhorn. I have a good knowledge of SQL which should give me a pretty good head start. How do you plan for a platform that is years away when you do not know what tools we be already available and where will there be a niche to fill. I cannot afford to attend PDC, so I will already be behind in the tool development. Regardless of Longhorn, learning some C# seems like a good idea. I am a Java guy, so it should not be too tough. So far Scoble has not been able to convince me why I should start planning something big for this product launch. I would rather use the best of what’s out there now, and wait until the Longhorn vapors start to solidify before acting on anything.