JDJ: Let's start off with XML in relation to IBM, Bob. Can you tell us why XML is important to the market and to IBM?
Sutor: XML provides a standardized, flexible and powerful method for exchanging data among many different platforms and applications across the Internet. IBM is a leader in database and transaction systems, and XML will be the standard mechanism for interoperability among data warehouses, repositories and Web applications. Now IBM, of course, is a very big business, so in addition to the services and products we provide for our customers, we're using XML to make ourselves a first-class e-business. It's basically just a great way of representing data. I usually like to think about what XML is about using four "I"s:
I think of XML first of all being about
Information, perhaps messages (exchanged between applications or platforms) or Web documents.
Interoperability comes next: getting applications that previously had proprietary file formats to open up and talk to each other. This extends the life of both the applications and the data.
Integration: using this data coming from lots of different sources together to create new products that you never really imagined before.
Independence: I mentioned platform independence, but also device independence. People are using PalmPilots, browsers and ATMs for financial applications, and XML allows you to present this information to all these different devices.
JDJ: What's IBM's role in helping to develop XML?
Sutor: Two ways: first of all, in the W3C, which we think of as the core horizontal standards organization. We're involved with most of the key XML working groups such as XML Schema. I myself, for example, was one of the authors of the Document Object Model. On the industry side, traditionally we've been involved with a number of the industry-specific consortiums such as the OAG and the OMG, and recently such things as the Open Travel Alliance and the Financial Products Markup Language with one of our customers, J.P. Morgan. Really recently, one I'm personally very excited about because it happened at OASIS is XML.org. This is a brand new organization that has come together to act as an umbrella for developing standards.
Sagar: Do you feel that at some point XML may replace EDI?
Sutor: I expect that there will be a transition from using traditional EDI formats to more open XML standards that are being developed. In the meanwhile, there are a lot of people doing conversions between the old formats and the new ones. One example is the Financial Information Exchange. They've come up with very nice methods for preserving the old-style methods while moving to a new XML format.
Sagar: Some people are seeing it as coming in and solving all the data-formatting problems. And you know, being in the industry long enough, nothing really solves all the problems. There's a lot of hype associated with XML and then there's some reality. Would you elaborate on what you think may be the direction it's going to take and maybe what problems it's not going to solve?
Sutor: It's an enabling technology, so it doesn't solve the problems by itself. It allows you to develop new markup languages. But you have to create those correctly and they should be done in a vendor-neutral way. For example, there's a fair amount of data modeling that should be done to produce a good XML DTD or schema. If it's done badly, you're going to have a bad product. A lot of what we're doing is introducing people first of all to the technology so they're aware of it, and then showing them how they can really use XML in the best possible way to integrate it with their products. From a developer's point of view, getting started with XML is very easy. If you go back to your EDI question, parsing out that EDI information and getting the information can be tricky because it's all encoded very tightly, although that might be efficient for moving the data around. XML is simple to read and it's easy to write new applications. There are parsers such as our XML4J that are available for free right now so the barrier for entry is very low. I think what's really generating a lot of the excitement about it is that there are so many tools available, such as those on our alphaWorks site. So the developers are very excited, but there's real work to be done by industry groups to create high-quality DTDs.
Sagar: One of the things you mentioned earlier was that XML is now going out across the wireless, PalmPilots; there's WML; there's 3XML, WIDL. One of the dangers I see coming up is that there will be so many vertical definitions of XML in industry that it's going to be a hard thing to actually keep a tight rein on what is going on. Is OASIS planning to help solve such problems? Do they have some strategy as to how to control what becomes standardized?
Sutor: Sure, the strategy is summed up by XML.org. We see XML.org as being in many ways complementary to W3C but for industry-specific standards. This means that XML.org can act as an umbrella for bringing together people to work on standards in industries. Thus there would be less need to set up all the ".org" XML Web sites we've been seeing. They could work within XML.org and have, for example, a human resources XML working area and write their standards there. It's actually a little trickier because there are many fine consortiums already out there that have developed standards. We are developing liaison relationships with many consortiums, but since things are being worked out, I don't want to talk about the details right now. If there's another organization that we can look at and say, "You have really done this in the right way. You have brought in enough people in the industry. People are agreeing to this and it has real, practical applications and implementations are happening," then in some sense we would like to liaison with them and be able to recommend their work.
JDJ: Getting back to open standards. Why should developers care about open standards?
Sutor: I've been a developer for a long time and I know how easy it is to create proprietary data formats. XML provides a really easy way of providing data in a format that can be changed easily and reused by other applications. Open standards are good because of the interoperability and integration aspects I mentioned before. In the long run standards lower the support costs because fewer technologies are involved. The data and the application will both be useful for a much longer time and the data will not be tied to particular platforms. This is important because 70% of the Fortune 1000 companies use three or more server platforms. Just as Java means portable programs, XML means portable data.
Sagar: Since this interview is coming out in JDJ, I'd like your opinion on the relationship between XML and Java. Do they play well together? And what is the role that OASIS or XML.org will play in defining the standards for the Java industry?
Sutor: XML.org isn't going to concern itself initially with looking at Java issues. It will keep itself to XML and work on the industry standard. That is already a big enough problem. Getting back to the first question, Java and XML are complementary. There are a lot of good tools for using them together. If you go to IBM's alphaWorks site, there are now 26 technologies available for doing XML, most of them in Java. Basically, Java is a great tool for dealing with XML.
Sagar: Why is that?
Sutor: You want to use them in pretty much the same sort of environment. The same things that make Java so attractive: the independence of platforms and the interoperability.
Sagar: Let's talk about the organization of OASIS. There are different industries that are involved in it. How do you make sure it stays unbiased and doesn't have more representation from one industry to the other? I'm not talking specifically about IBM. I'm just talking about whoever gets in. What's the way you define committee selections and so forth?
Sutor: We're still working on the process. In fact, we're looking at what happens in other organizations and we'll be building our process from that. You also have to remember that XML.org is part of a much larger OASIS, which already has about 90 numbers. We already have a lot of input. XML.org is not independent of OASIS, but rather an activity within it. So in addition to the partners in XML.org, we've got companies in OASIS such as Microsoft that have not signed up for XML.org, but they can comment on the standards process we are developing. Of course, we would welcome their support for XML.org and their joining the XML.org steering committee.
Sagar: What is the exact relationship between OASIS and XML.org? Is XML.org just a Web site for OASIS or is it a totally separate entity?
Sutor: No, it's not a separate entity at all. We refer to it as an initiative, which means a project that is going on within OASIS. So today, basically, it's an organization that operates within OASIS. It doesn't have separate bylaws, so ultimately the OASIS board of directors makes the final decisions and has the fiscal responsibility. XML.org was formed initially by nine partners who put up $100,000 (large companies) or $25,000 (small companies) each to get this thing started. It's much more specific than the general OASIS. These people have helped to get it kick-started and so we can build the necessary infrastructure.
JDJ: I'm assuming we can go to XML.org to learn more.
Sutor: That's correct and every day the site will grow to become a real portal for learning more about using XML in industry.