News

IBM's Web Services: A Horse with No Name?

• By Scott Bekker
• March 21, 2001

Joseph McKendrick

With all the hype being generated around Web services - generally defined as application-to-application interaction over the Internet - why has IBM, the world's largest computer company, been so quiet on the subject?

Some analysts agree that IBM has had the right ingredients in its mix for some time, but more work needs to be done to connect the dots and make the Web concept workable. This is the case among all major vendors touting Web services, states Yafim Natis, analyst at GartnerGroup, who recently co-authored an analysis of IBM's Web services strategy. He notes that IT executives, "should take IBM's Web-services initiative seriously, but much work remains to be done before any vendor can claim mainstream acceptance of its variant of the Web-service architecture."

IBM executives have been on the road to articulate IBM's Web-services stance and to counter the blizzard of announcements around .NET, Sun's ONE, and Hewlett-Packard's e-Services and e-Speak. Big Blue's message: Make no mistake about it, IBM is in this game, too. "What you haven't heard from IBM is some kind of new name associated with this effort," says Steve Holbrook, Technology Evangelist for Web Services at IBM. "We haven't seen a need at this point to rename or coalesce the effort."

IBM's Web-services strategy revolves around a number of different initiatives, Holbrook explains. For one, IBM has been actively promoting and adopting standards such as SOAP, among others. Also, IBM has been aggressively positioning its WebSphere application server and VisualAge toolkits as key pieces of middleware and development platform for e-business. "WebSphere, VisualAge for Java, and WebSphere Studio are the areas where we're focusing to begin with, but Web services and the underlying protocol will extend to our whole product portfolio," Holbrook says.

Another component of its Web-services development support strategy is the previewing of close to 40 percent of its bleeding-edge technologies on its publicly accessible alphaWorks Web site. Once a technology has cleared a standards body, IBM seeks to get user input as quickly as possible, Holbrook explains. AlphaWorks includes IBM's Web Services Toolkit 2.1, which provides a runtime environment as well as examples to design and execute Web services.

IBM's goal is to extend the life of its legacy systems and applications, Gartner's Natis says. Thus, many of IBM's Web-services tools are designed to "wrap or modernize existing mainframe or other applications, alongside Java support for Web services," he explains. "Microsoft, looking to modernize the overall Windows environment, will primarily provide tools for building new service-oriented applications in Windows alongside interoperability with preexisting COM+ applications. While .NET is a new foundation architecture and a revolution for Microsoft, Web services represents a technology layer for IBM aimed mainly at enabling incremental evolution of enterprises' existing business-application systems."

IBM's Holbrook offers his own competitive distinction between IBM's and Microsoft's Web-services approaches: IBM is embracing open-source software. The company is investing $1 billion to support Linux across all its platforms, and WebSphere is built on top of Apache - an open source Web server. "WebSphere is our centerpiece strategy," he points out. "Everything's coming together under the WebSphere umbrella. And underneath that is the Apache Web server."

Holbrook hopes that IBM's heavy investment in open-source software will attract developers. "IBM is embracing developer mindshare developing in the open-source space. We want to make money off solutions built on top of the infrastructure, not the infrastructure itself."

IBM will begin rolling out Web-services-enabled products within the next few months. Rapid adoption is likely among Big Blue's loyal core of customers, Natis predicts. "IBM's customers will experience lower entry barriers to adopting IBM's Web services than other customers adopting the radical and innovative approach promoted by Microsoft," he explains.

Microsoft's .NET strategy is likely to "initially be hampered by the higher cost of entry caused by the need for users to fully adopt the completely new .NET stack before experimenting with Microsoft's Web services." Natis thinks Microsoft will catch up over the next two years as it enables encapsulation of proprietary Windows applications into Web services.