Fact or Fiction???

It's true.

Several factors.

1) Most of the Windows developers at Microsoft are C++ types. If you've been coding in C++ for 15 to 25 years, are you going to be more productive in that language, or a new language that is only a few years old?
2) C++ code can still be more performant and made to run in less memory with less resources than C#. There's a big caveat there: as long as you're a top-notch C++ programmer. If you aren't all that great then you'd be better off sticking with C# or VB.NET.
3) There's a rule in software development at Microsoft: take no dependencies if you want to ship on time. The Longhorn team made a decision to take no dependencies on other teams. To do so would mean waiting for technology to be finished. Remember, .NET 2.0 isn't yet finished. To build Longhorn on top of .NET 2.0 would mean building a house on a sand foundation. Not good. How about building on top of 1.1? That wouldn't be good either because Windows developers need many of the features that are in 2.0.
4) The "don't rewrite from scratch" Joel Spolsky rule. It's a good rule to follow if you want to ship anytime soon. Since existing versions of Windows were built in C++, rewriting pieces of them in C# or VB.NET would mean breaking this rule and potentially lengthening development time.

I should say "it's mostly true." Since I haven't seen every single piece of Longhorn yet (and most of it is still under development). It's quite possible that there's some .NET code that'll ship in Longhorn somewhere. Just not in major pieces.

Yes, Aero is the new UI for Longhorn.

Most of Aero is not built on top of Avalon, as far as I understand.

Yes, it's safe to say that the part of Avalon that is used in Longhorn (or that Aero uses) is not managed. Correct.

>Does that mean WinFX is dead? Indigo?

NOOOOOOO!!

WinFX and Indigo are completely alive and are, at least in part, built on managed code.

But, WinFX and Indigo have been decoupled from Longhorn. How do you know that?

They run on Windows XP and Windows Server 2003.

We're talking about code that's only making it into Windows, not about packs or applications or other things that run on top of it.

Buzza,

Avalon is no longer a core feature of Longhorn. Hope that helps. Back when that article was written it was, now it's not.

Avalon is, again, decoupled from Longhorn. I am running Avalon on Windows XP here.

>BTW, Scoble's spin is probably factually true and pretty good.

Damn, why is everything I write "spin." I'm being dreadfully honest in this thread -- to the point that I'm sure I'll get a call from someone in PR tomorrow.

This isn't spin. It's from hundreds of interviews (including one with Jim Allchin, the guy who runs Longhorn) and working with developers who know what's going on.

>Robert Hess was announcing it like it was set in stone back then bringing on the project leaders and talking about it with some degree of confidence.

Yup, I was pretty confident back then too.

But then we got to the PDC and developers and CTOs said "nice, but ain't gonna use any of this stuff until at least 2010."

Think about it, how long will it be before Longhorn gets any kind of massive adoption? Based on adoption curves of XP, it'll be years, even after Longhorn ships.

So, the decision was made to go and completely rethink Longhorn and the strategy around Avalon/WinFX/WinFS/Indigo.

That's why today those things will run on Windows XP.

Yes, it takes us a while, but we do listen to customers when they tell us we're smoking crack.

>So no winfs, avalon, indigo its not based on .net 2.0, am I missing something, whats new in Longhorn.

Lots of stuff, but if I told you then I'd really have a ton of PR folks to deal with when I woke up in the morning.

I didn't say that Avalon and Indigo aren't based on .NET 2.0. Just that they aren't a core part of Longhorn anymore.

Let's just say the PDC this year is gonna be pretty interesting again. http://msdn.microsoft.com/events/pdc/

RHM, the problem is, we've only shared what we've cut. We haven't shared much of what's new.