Microsoft's Ballmer: .NET is about integration

From...

By Michael Vizard and Mark Jones

(IDG) -- As part of an ambitious effort to create an architecture that fosters data andapplication integration, Microsoft has laid out a broad foundation based on XML technologies that will be marketed under the name of Microsoft.NET. In an interview with InfoWorld Editor in Chief Michael Vizard and West Coast News Editor Mark Jones, Microsoft CEO Steve Ballmer, who will be a keynote speaker at the InfoWorld CTO Forum this week, talks about how he sees this "bet-the-company" strategy paying off for Microsoft customers and its industry allies.

Why should corporations pay any attention to Microsoft.NET today?

Ballmer: There's a ton of information that is essentially locked in back-office systems today. We want to help [companies] bring that information together in new applications. We want to help them expose the information to the consumer. The way we would propose doing that is to essentially wrap it via XML and then build next-generation applications that pull things together using the XML infrastructure. This is about enterprise application integration. This is about business-to-business. This is about unlocking, getting knowledge of back-office systems to front office.

What's the core business model behind Microsoft.NET?

Ballmer: We will build software, servers, and tools that have .NET-and XML-platform capability built-in, and we will sell those as we sell software today. We will also have a set of services that you should think of as sort of customer-facing as opposed to developer-facing. These will be advanced services for consumers and knowledge workers that use an XML data store that the user has running on the Internet. These additional services on top of that somebody might subscribe to as part of Windows or on top of Windows or on top of Office, etc. We will also charge developers some [sort of] fixed fee to use our services per year because there's real operational costs in serving a developer. But we don't have any model under consideration that calls for transaction fees and that sort of thing.

IDG.net INFOCENTER

Related IDG.net Stories

Features

Visit an IDG site

IDG.net search

The first iteration of this new model would appear to be HailStorm. Are other versions planned for different customer segments?

Ballmer: HailStorm, as announced, is very end-user focused. The schema that we talked about are very much oriented toward the end-user. There could well end up being other schema that we introduce targeted at other audiences. Certainly, we will have a set of schema that we target at the small-business customer, for example. There's no announced plan and I'm not trying to announce any plans now. But certainly as you think of people who try to build b-to-b scenarios, there will be some schema to help integrate the world for end-users that is standardized and available in the cloud.

Who will deliver these services? Microsoft, system integrators or telecommunications providers?

Ballmer: Our basic business is software. That implies the need for great partnerships and relationships between us and the telcos. They all understand that IP will be important for delivery. We understand that they will be essential companies and that our business interests are intertwined, and so we spend more and more time with those guys. Not figuring out how to fight, because we're not going to fight. But figuring out how to really partner to offer consumers the richest set of sort of IP data and telecommunication services.

As this scenario plays out, what will be the impact on IT people? Will they become general contractors who integrate diverse sets of services over the Internet?

Ballmer: In about 10 years, yes. Over the next few years, no. What it really means to most IT departments is better development tools, better servers, better everything -- the ability to build applications that are easily integrated with other applications. I don't think the basic model will change all that very much over the next three years. What will change over the next three years is that people will build and deploy applications that integrate much more seamlessly together using .NET and using XML as the interchange.

Given the momentum behind Java among developers and corporate customers, where will the momentum for your approach come from?

Ballmer: We'll have the classic issue where the guys who can be quickest to jump on the new phenomena are not always the guys who have the greatest investment in today's approach. We don't have a choice at Microsoft -- we've got to bet on the future because we're a platform vendor in some ways. We can't be the platform vendor and not make your leading-edge applications keep up with your platform. You'd look ridiculous. But a lot of the guys who've been early adapters of new technologies have tended to be smaller companies, startup companies looking to break out of the box. I think we've also seen a lot of interest among the big established ISVs. I mean Java is nothing really. At the end of the day all of these guys have to embrace XML. Java is inadequate, and the way that applications will be extended will be by responding to XML messages. It won't be by sending somebody a Java program. The way applications will be built will be loosely coupled, not tightly coupled, which is what today's Java programming model encourages, as opposed to today's XML Internet-style model, which is a sort of a more loosely coupled model. That doesn't mean that people won't be able to use Java the programming language, or [use] Java skills in the XML world. But Java 2 Enterprise Edition [J2EE] and the Java operating system basically will be a throwaway. People are going to move in an evolutionary fashion.

What are the major points of difference between you and Sun about the role of XML?

Ballmer: XML is a message format, but it also implies a programming model. I don't send you a Java program that you run. I send you an XML message and you send me back an XML message. Yes, it's a data exchange format, but it is also the backbone for the way you write loosely coupled applications that extend one another and complement one another and work together. I don't think Sun gets that, frankly. Or maybe they do get it but strategically it is inopportune for them to get it.

What's your take on the ongoing war of words between Oracle and IBM over databases?

Ballmer: If you take out mainframe databases, which is a very specialized market, we're the number two database vendor in the world. We've got a billion-dollar plus database business. I know who's number two and it's not IBM. Oracle's the lead dog in the race. But the world is moving to PC architecture servers. The real competition and the battle there is clearly Microsoft and Oracle. The data says we just passed them, again on the Windows platform. Going forward, XML's got to be part of the storage strategy and it's got to be part of the programming strategy. You're going to want to be able to write Dot-Net programs and run them, either outside the database or inside the database; your choice.

Microsoft recently acquired Ncompass, a provider of content management software. What role will this technology play?

Ballmer: If you take a look at the kinds of applications that our enterprise customers are trying to build, we're finding that they want to not only have to help to manage and build the code, but manage and produce the content. After looking at that long enough and saying we're one tool short to serve these customers so we went ahead and bought NCompass.

Microsoft has spent a ton of time and energy fighting the browser wars. But give the limited user interface of the browser, was this a war worth winning?

Ballmer: We haven't married the best of local execution with the browser. We're starting to do more of that now. But today, you pick either a fairly rudimentary UI or you pick browser UI. We've started to break down that wall. If you take a look at the way Windows XP or Office XP works you have browser embedded in application or an application embedded in the browser. We're starting to knit together to the point where you can't tell sort of where local code versus browser-based things begin and end. That's the right way to go. The browser is not the end all. After you get out a few years, when we have the really complete implementation, you won't be able to tell the difference. The technology won't go away, but it will become something bigger.

Why are there so many hooks between Microsoft Office XP and MSN?

Ballmer: It adds value. You get the name of a place and you find out how to get there. It means a map and we have a preferred mapping solution. But all smart tags are replaceable. If somebody wants to install a smart tag that takes them to their own favorite place, they can do it. There's nothing in the system that sort of stops them from doing things the way they want to do them. Not surprising, we point you to maps that are up on MSN. We take you to data sources that we can support and where we make some money. But that doesn't stop other people from customizing that to the way they want it.

How will wireless networking be ultimately married with handheld devices?

Ballmer: People are going to have different form factor devices. You will have a small device that you carry. You will have probably a device that's sits near your television set. You may well have something that we think of as more like a PC that you carry around. We'll think of those things over time as working not only in the home, but outside the home. You could have pocket-sized device that has telephone type services. When you walk outside the building, it loses the 802.11 connection or Blue Tooth connection in your house, and it starts connecting you up to the 2-1/2 G or 3G network. You get to work and it roams you to the 802.11 network at work. That's a very conceivable scenario. If you take a look even at what's built inside of Windows XP, we've built the software that reconfigures the network stacks and keeps you connected or re-connects you when you when you get to another 802.11 spot.

What other areas need work?

Ballmer: There's a whole slew of things. We've redesigned the server-programming model pretty well. And you'll hear some announcements about that over the next couple of weeks. There's still some work to do on the client side programming model. The file system's got to know about XML documents. Smart Tags are just the start of how we train the user interface to be XML information aware. There's a lot to do in all parts of the system. Caching, naming, proxy, I mean everything. The system infrastructure will get re-plumbed over time.

So what has Microsoft as a company learned about the needs of corporate customers over the last five years?

Ballmer: It was a different day and age, and I think we've learned a lot of lessons, People told us we didn't realize that all we did there was getting into the starting line.I think at this stage, we have a good sense of interoperability issues. We've had a constant problem in our enterprise strategy on how do we provide interoperability with other systems. XML is the standard. It's not controlled by one company, not by us, nor anybody else. And in some sense, it's the perfect answer to the problem of how to solve enterprise interoperability. We also continue to improve our products. We think we've really got security. Today we actually have very good answers on scaling up and scaling out. I think we have a lot of experience at this stage in the enterprise issues. Is our execution all, everything that it needs to be for the future? No. But it's not because at this stage we have a fundamental lack of understanding of the enterprise.

Why does Microsoft carry so many different products in its portfolio? Wouldn't it be easier for customers if there were fewer products to try and figure out?

Ballmer: What you wind up then is some products where some of the capabilities that just get ignored. The customers don't know it's there, the sales force doesn't sell it. You built something brilliant and it would almost get better exposure if it was a separate product. So we've gone from the company who tried to put everything into a very few boxes to a company that now tries to put things into more boxes. But you could envision a world in which we had fewer products, in which some of the concepts we have today get better integrated. I expect to see some of that over time.

What will all is product integration mean to companies that integrate applications for a living?

Ballmer: Service vendors will focus more and more of their time and energy on real unique value-added. It just means the money that went into that wouldn't evaporate, but that money could be spent on real enterprise application integration and really on new applications. There's a pretty insatiable appetite for better IT solutions. You don't want people sort of doing low-level systems configuration and network configuration that software can do.

How big a challenges does Web services present for security?

Ballmer: I see it as an area that needs work, there's no doubt about it. If you want the world to move around XML documents, there really needs to be a standard way to encrypt XML information.

We're working with the W3C and some other companies to create standards around how you encrypt XML information. Work certainly needs to be done so that this revolution, if you will, has the security that people will expect.