Yag does a great job of explaining to those who don't get it why not all MS products are using DotNet as a native framework.
If you read all the comments, you'll see that part of the issue has to deal with the components that have not yet been put into DotNet (Uniscribe, and features for multi-lingual text etc).
Yag's basic premise though is that no one should expect every single MS app to flip over to DotNet overnight - there's a lot of work involved in moving it over and as the DotNet framework matures, more features get added in.
How this will apply to the entire Avalon scenario is up for grabs. It reminds me a little of how DNA has matured into DotNet and also how FoxPro has matured into Europa. The basic MS approach of development frameworks and almost any new technology (at least in my view) is:
Phase 1: Tell Everyone How they Should Be Doing It (VFP example: VFP 3 / Win Example: COM and Windows DNA)
Result: Some people get it, some people don't.
Phase 2: Show Everyone How To Do It (Ex: VFP 6 with the FFCs / DotNet v 1.0 / COM + )
Result: More light bulbs go off and more people get it
Phase 3: Write Tools so people can do it easier (VFP 8 / Whidbey)
Result: More applications start using it across the board.
Phase 4: Continue to mature the toolset (VFP 9/ _____________ )
Result: More applications use it than don't
Phase 5: Come up with a new idea (or rename the older one) and go back to Phase 1
Seriously though, that is the basic path that these things take.
But at any step in the road, people get frustrated with waiting for Phase 3 and jump off the wagon. At other times, you may think you're ready for step 3 but you took a wrong course and have to go back to step 2 to get it right.
In some scenarios, MS jumped the gun and did a lot of steps 1 and 2 in secret (umm, research) and then introduced a product like MS Bob only to be assailed by critics and the public.
In the development world, MS has been a little more open in this context, perhaps to avoid the surprises or headaches of newer technology. Some succeed, some fail but each one builds on each other.
Just because MS had a monopoly in DOS didn't mean it would naturally succeed with Windows 9x. Likewise with XP and future versions.
Many of the technologies found in DotNet 1.0 needed to mature to be "production-ready" but just like SpaceShipOne is simply the first version of what will be more tourist trips into space, the conceptual foundation of managed code is the one that MS has built its future on.
Now it's a matter of getting to phase 4 and ensuring you don't lose any more developers on the way.
Microsoft and .NET (by yag)