What is the reasoning behind the .Net platform strategy?

There are a few dreams that are pretty important as driving factors behind .Net. The first dream is that everything can and will be a Web Service. Every piece of software, hardware, and the network will be programmable through XML interfaces. This doesn’t mean that you have to go back and re-architect and re-do everything, you can put XML interfaces on top of legacy systems, but moving forward almost all software will be written with native XML messaging interfaces — not just software, but hardware and networks as well.

Dream number two is that anyone can write these XML Web Services in whatever language.

The third dream is that whenever two XML Web Services meet, they should automatically work together. Today the core problem for anyone that writes software that talks to more than one system is that it’s incredibly hard to inter-operate and integrate.

The fourth dream is that the user should always be in control.

Dream five is to take a real shot at distributed computing. Why put everything on a server and access pictures of it through HTML? Why not have XML all the way up to a Smart Client, servers talking to servers in a programmatic way? That can really take a bite out of the dream of distributed computing.

Those are the dreams. We are executing on that through .Net.

So as you evangelize the .Net platform, how is the competitive landscape evolving, in terms of what companies like Sun and BEA are also developing in the realm of Web Services?

When you look at Sun, they made three promises — one was a simple productive language called Java, two was ‘Write Once, Run Anywhere,’ and three was everything integrating very easily.

In terms of the first promise they delivered. We are happy for that, Java is one of the languages supported on the .Net platform, but so are 23 others. In terms of Write Once, Run Anywhere they just failed. For example, nobody writes a piece of code that runs on multiple application servers. Write Once, Run Anywhere was more of an insurance policy for people rather than any concrete technical dream. In terms of integration, systems still don’t inter-operate.

So they really failed dramatically on Write Once, Run Anywhere and integration. And unfortunately for Sun, Web Services solves those problems. And they can’t move fast enough on the Web Services front. They recognize it dramatically dilutes the Java franchise but they can’t move fast enough.

You have to ask, will Sun survive? They have a battle on the SPARC architecture with Intel, they’ve got battles on their servers with Compaq, IBM, HP, and Dell, and they’ve got a battle on the software front with Microsoft. It’s the last of the vertically integrated companies.

Ultimately if this .Net vision comes true we have a situation when everything comes through Microsoft. How do you address the age-old question of a monopoly?

That criticism is a last resort when you can’t compete. Go back to the dreams that I laid out. This vision of .Net and Web Services is independent of Windows and independent of our architecture.

That’s the core value we think. We have to out-execute on windows, but it in no way pre-supposes that these Web services are running on Windows. This is a vision that is not in any way dependent on Windows, but we’re confident that we are going to out-execute. And when a competitor says it’s a lock-in you have to wonder if it’s a last resort.

So the core is independent of other Microsoft products?

An XML Web Service does not depend on anything except for the XML stack. We have to implement that using .Net. We are betting on the public standard.

Give some specifics about how the future of computing could operate as a result of these developments.

The single word I would use is distributed. It’s going to continue to be distributed in every sense of the word, and those distributed kinds of apps will drive interesting scenarios.

Let’s take the Internet in 2000, before XML became a buzzword. It was a very mainframe-centric way to think about life. You crunched a bunch of numbers and servers and you pushed the pictures out using HTML. Think about XML all the way to the client. Then you can have a two way browsing experience. You interact with the data, not just with the pictures. Say you’re looking at an SAP database with an Excel spreadsheet. When you have XML all the way to the server and the client you can interact with that as opposed to just seeing Web pages. That makes incredible scenarios possible. You use whatever processing power intelligence, cache and storage is there on the client and you take maximum advantage of what the user wants to do and what information data is on the backend.