Web services: Beyond the hype

Bob Sutor is not one of those rah-rah types who easily gush about the transforming power of the Internet. But when it comes to Web services, he is a man on a mission.

Working behind the scenes to rally industry support, Sutor simply believes Web services constitute a revolutionary advance that will fundamentally change the way companies conduct business over the Web. The potential is so great, he believes, that common self-interest will in the end trump narrow jealousies.

You may have heard all of this before, but Sutor is not some dot-com Gen Xer. As IBM's director of e-business standards strategy, Sutor is in a position to make things happen in the arcane alphabet soup world of Internet protocols.

He still has a tough sell ahead but Sutor, a 43-year-old former mathematician, has already collaborated with Microsoft and others to fashion the underlying plumbing necessary to turn his vision of Web services into reality. Recounting to a listener what's at stake, Sutor makes it sound simple: Web services will engender more efficient ways for businesses to build software and let companies more easily conduct transactions with others. For consumers, he envisions a future where people can access their personal information online on any device--and do everything from shopping and banking to checking their e-mail or calendar.

Analysts describe Sutor as a key player in the Web services field, one who also has the necessary political skills to maneuver within the Byzantine world of industry standards and win the support from rival software makers such as Sun Microsystems, Oracle and BEA Systems.

"Bob is very politically astute and knows how to manage the ins and outs of relationships in standards bodies and between vendors," said Uttam Narsu, an analyst with Giga Information Group. "Bob is quite vital and has lots of experience with standards bodies and is good at encouraging the back-channel communications that are necessary for getting the bodies to work in harmony."

Sutor, who received a doctorate in mathematics at Princeton University, taught college-level math before joining IBM in his early 20s. As a member of IBM's research staff, he worked on mathematical and Internet publishing projects before joining the company's software group in 1999.

CNET News.com recently spoke with Sutor about the future of Web services, IBM's relationship with Microsoft, and the newly launched Web Services Interoperability Organization.

Q: Will Web services be more important for businesses or consumers?
A: With emerging technologies, it's always business. When something new comes out, people don't trust it...until they're confident enough that there's industry acceptance. For consumers, it's just plumbing. When people use the Web to buy and sell and fill out forms, they don't know what the underlying technology is. That's how it should be. What we're trying to do is simplify, so that when you have businesses talking to businesses, it's almost as easy as when people use the Web, viewing and buying things. And the notion is based on standards.

Are Web services some sort of Holy Grail for software?
The most compelling thing as a technologist is making the world appear more homogenous, trying to cut through all the different ways that people implement (computing) systems. A company could say, "I will only use things from this one vendor," and then through a merger or acquisition, they quickly blend and have different software and hardware. We have to get to a point where people can build systems the way they want to build them but still talk to the rest of the world.

Is that the appeal of Web services for you?

What appeals to me about Web services is there's elegance to it. If I were a business, I frankly want this to work. I want to buy something from you. I don't want to forever be down in the details on how to hook up our systems to do it. This is about e-business, of how we are going to bring businesses together. It's hard to argue that it isn't a good idea.

IBM and Microsoft have been the pioneers in creating the Web services specifications--SOAP, WSDL and UDDI--which are the emerging standards behind the industry shift toward Web services. SOAP, for example, is a communications protocol that allows companies to link and conduct e-business. How did the relationship come about? Did IBM approach Microsoft or vice versa?

"We have a very good opportunity to learn from our previous integration work inside the enterprise and do the right thing for the Internet."

It was gradual. It came about in organizations like the W3C (World Wide Web Consortium). We were working on a number of Web-based standards. What happened was Microsoft published SOAP version 1.0 in November 1999. We generally agreed with the need for something like that. We had a conundrum. We probably assumed that there was no way Microsoft would change this. Should we create something else and compete? We had backroom conversations in some of the standards meetings. We negotiated with them, telling them we think we should do it this way, and we sorted that out.

April 2000 was the turning point, when we co-authored SOAP version 1.1. It showed that, yes, we could work together and reach agreements. We still very much come from different perspectives on building these (software) platforms. When we produce things together, it's always negotiated, just like the work with others in standards organizations. We surprised each other.

Can rival software makers really cooperate and make Web services truly compatible? How can they avoid creating proprietary lock-ins to their specific technology?
This sort of question is very interesting because it seems to answer itself: You avoid proprietary lock-ins by cooperating to make sure Web services are interoperable. This is what we are trying to do with WS-I, the Web Services Interoperability Organization that we helped found with Microsoft, Oracle, HP, SAP, Intel, Fujitsu, Accenture, BEA and others in early February. This level of cooperation is not new, so I'm always surprised when people think it is and that it couldn't possibly work.

Remember that most of the Internet standards were created by similar cooperation in the IETF (Internet Engineering Task Force) and the W3C in the 1990s. Does the Web work? Is there interoperability for pulling down and viewing Web pages? We are trying to do the same for the standards that affect how applications talk to applications via the Internet. This requires patience and cooperation, an understanding that it will take time, and a good sense of optimism that it can--and should--happen.

When will Web services become a reality? At what stage is the industry at in building the standards needed to turn this into a reality?
We've gotten the initial buzz. With any new technology, there is a lot of education. We are in the early-adopters phase. We've gotten those who love new technology and try anything new. And we'll start seeing companies put Web services into production, where it starts snowballing. We see a lot of software development tools coming out this year, so now, it will become much more real for developers.

What are the roadblocks? You can do a lot now with Web services, but there still seems to be a lot of work left.

"When people use the Web to buy and sell and fill out forms, they don't know what the underlying technology is. That's how it should be."

There are major efforts under way in security to ensure that we have the standards we need. Since one way of looking at Web services is as "middleware for the Internet," you should expect to soon see Web services standards efforts in the areas of reliable messaging, transactions and work flow.

Provisioning (or the way companies introduce services) will be very important--as will systems management for Web services. The industry work in WS-I will help define this road map, and the standards organizations like the W3C and OASIS will make it real. These standards are layered in various levels of dependencies; until we create the lower levels, we cannot do the upper ones. You can expect to see the standards work that started in 2000 continue for another couple of years. This will allow us to build the specifications in a coherent way while having time to educate developers and the executives making investment decisions about how and when to use Web services technologies to the best benefit for their companies.

There were things like TCPIP that are now simply part of the fabric of the Internet. We don't talk about it anymore. In the same way, in five years, we won't talk about Web services. It will be the way we build programs, and we'll be talking about higher-level aspects and the interesting things coming out.

What's driving you? Why do you groove on Web services so much?
Perhaps because of my mathematical background, I'm attracted to the notion that we could develop a well-designed and elegant set of standards that can be plugged together in a variety of ways to solve a broad range of business problems. Once you get past the rhetoric from companies seeking to define Web services in their own terms or to better benefit their own (software) platforms, there is really substantial agreement on the goals for Web services and even much of the technology. We have a very good opportunity to learn from our previous integration work inside the enterprise and do the right thing for the Internet. I'm especially excited about how naturally Web services technologies fit with the emerging grid architecture and, I believe, how it will complement the new wireless architectures being developed.