|
|||
|
|||
Sun's Role in the Java World
JavaOne, June 12, 2003
Java is different things to different people—a language, a platform, a community. To Jonathan Schwartz, Sun's executive vice president of software, it is the core of the company's competitive software strategy. Java Pro caught up with Schwartz before JavaOne to learn how Java plays into Sun's strategies for integration, Web services, and Project Orion. Q: As we head into JavaOne in 2003, where does Sun stand in the marketplace against its biggest competitors, both within and outside the Java community? Let's start with Microsoft. Schwartz: We are number-two to Microsoft and we try harder when it comes to delivering software products into the world. The community of people who support Sun is a couple orders of magnitude larger than the community of people who support Microsoft, given that the number of companies that support Microsoft is about one. Our alternative to the Microsoft monopoly, the Java platform, is all about interoperability and portability, which are anathema to Microsoft's competitive strategy. Some people say that we're anti-Microsoft, and they are right. We believe in open systems, interoperability, portable code, and portable data—and these are all things that Microsoft has a hard time with. Q: How about IBM? Schwartz: If Microsoft is a one-choice company, then IBM is the all-choices company. When you think about the number of operating systems that company now supports, it's staggering. IBM has a conflict of interest because it wants to be all things to all people. Its real agenda is to amplify complexity so it can send in a bunch of Global Services consultants to make it all work. Q: What about BEA? BEA is a great company. It is a partner, and while we have some spats with them in public, our folks work together very well. We need to do what we can to help BEA because the resources required to be a player in the future will be broader than most companies expect. I think the real platform competitors in this space are going to be Sun, Microsoft, and IBM. Q: And how does Sun compete? Schwartz: We offer choice in terms of platforms for running our platform. The Java platform runs on all of those different platforms, so we don't need to deliver an operating system to you, nor do we need to add to the complexity by scaring you with a message around incompatibility or scaring you that we're going to "end-of-life" the boat you've been riding. Q: There was a perception that Java was slow to embrace Web services. Has that changed? How will Sun get in the driver's seat on future standards? Schwartz: Java is the dominant language through which Web services are enabled. With J2EE 1.4, we're actually the first to deliver WS-I basic profile support in the platform, far ahead of anybody. You have to implement Web services in something and Java is the best place to do it. Many more developers are using Java to do Web-service development than any other platform. By comparison, there are no real .NET Web services in deployment at a level that would suggest Microsoft's platform message was successful. You can use its technology, but it's not really going out and deploying. Regarding the higher-order standards, we are focused on the pragmatic things. When you see all the game-playing around reliable transactions and reliable messaging and security and choreography, that attests to the fact that you need some kind of orchestrated process and body to ensure that one standard exists, and not one product from one vendor. Q: Where are these higher-level standards going to be resolved, if at all? Schwartz: I think they ought to be resolved in something like OASIS and not in a smoke-filled room. The biggest impediment for users in building out business systems is not knowing the way forward based on either the platforms they own or their desire to escape a platform they own. When Microsoft is telling you one thing, IBM is telling you another, and Sun is telling you yet another, that doesn't help you move forward. So unless and until all of the members get together and agree on what the standards ought to be, the reality is we're going to be using the standards we already have. Q: But even within Java's own standards body, the Java Community Process, there has not been much movement on workflow technology, for example. BEA has offered an interesting idea with XMLBeans, but it hasn't gone anywhere. Schwartz: Sun gets only one vote in the JCP. If BEA wants to start a JSR in XMLBeans, it doesn't need Sun's permission to do it. [BEA] can just grab five of its partners and make it happen. I think we are interested in wanting something like this, but the issue is ensuring that the standard that BEA would like to espouse is one that will be available to everyone, and not be specific to BEA. I think what BEA wants are fundamental changes to the Java language itself, and there are about 740 other people we need to talk to about it, given that the principal objective of Java is compatibility. The ugly thing about democracies is that they can take some time to make decisions because you want to ensure that you bring everybody along. Unlike some other bodies, the JCP is open to 100 percent of the people on the planet. Q: One way to add workflow extensions to Java is through the JCP. Another way, since you own the technology, is to do it yourself. Schwartz: We can't do that because we have a community to bring along. To have a specialized Java that works just for Sun wouldn't be very useful for the world, nor for BEA. What is important for the community is to ensure that we have one Java, not one that BEA can barely afford to support. Q: A lot of the competition in the industry is over integration. What do you see as Sun's value proposition or advantages in this area? Schwartz: Everything that we've ever done is about interoperability and open standards. Our fundamental belief as a company is that open standards yield a rising tide that lifts all boats. So whether it is NSF or HTTP or Java or Web-service standards, interoperability enables a greater competitive terrain in which more businesses will want to connect their systems. They'll want to do it using standards and basic infrastructure, and that's exactly what Sun provides. Our competitive advantage is that in Java, we have the largest developer community in the world. Despite IBM's belief that people are going to implement Web services with COBOL, we believe that you're going to implement Web services with either Java or .NET, and obviously we have an enormous developer community that looks to Sun for guidance and leadership around what happens next with Java. Q: The J2EE market used to be centered on sales of application servers, but that has become more of a commodity product. What's the current status of your play in this game? Schwartz: We have unit-volume leadership but not revenue leadership on our app server. Our answer to that is our Orion strategy, which aggregates all of the application platform components into a delivery vehicle that we can deliver on x86 and 64-bit systems. We are beginning to see significant customers unhooking $20,000-per-CPU app servers. Anyone who is looking at signing up for a $10 million app-server deal is beginning to look at the one that comes bundled in Solaris. With Project Orion, you get the app server, directory, identity, portal server, all of the Web-services developer profiles, and a Solaris platform—probably about $30,000 worth of software for about a thousand bucks. Q: What are you excited about at JavaOne that is going to make the industry take notice? Schwartz: There are 100 million Java-enabled phones in the world—a hundred million. There were 120 million PCs shipped last year. We are on target to out-ship PCs. Right now, 100 million consumers are walking around with a little bit of Java in their pockets. When you talk about integration, this is a big part of it. Integration is a challenge when you are trying to connect business systems. We expect to connect consumer systems running on mobile devices, as well as enterprise systems running on servers. About the Author |
