Web services: Time for a reality check?

As with most things in life, the reality lies somewhere in between.

In fact, this is one of the most significant--yet surprisingly practical--innovations to hit the information technology industry in recent years. But do critics who accuse the IT industry of hyping Web services have a point? Yes.

In the scramble to help customers implement a decade's worth of expensive technology, many vendors and standard-setting bodies view Web services as a sweeping new enterprise IT model. They like it because it allows them to build applications using services scattered all over the Internet (owned and operated by different providers) with components they can assemble and discard as needed.

But while this may be the dawn of a exciting new era, most of the hard work remains unfinished. The reality is that major Web services vendors have only agreed on three loosely defined standards: Simple Object Access Protocol (SOAP), Universal Description, Discovery and Integration (UDDI) and Web Services Description Language (WSDL).

Even more disconcerting is that almost none of the ongoing work will standardize the transformation of data from one form to another. Nor does a consensus exist on the interfaces for the services, or even how to architect them.

Against such a half-constructed backdrop, why do I still believe Web services are destined to become one of the most significant innovations in IT? One reason is that Web services allow managers to start with the technologies they already have on hand. A company may find it worthwhile to buy a really good implementation or toolkit, but such an investment is optional.

Also, Web services offer IT a way to finally resolve the painful problem of application integration. That makes it a supremely valuable tool for businesses implementing Web services in areas like order management, shared forecasts, or connection of internal systems.

For example, consider the hypothetical case of a buyer wanting to inquire about the status of an order. The seller's manufacturing system would need a SOAP interface. Then the buyer's purchasing system could "call" the seller's system over the Internet using Extensible Markup Language (XML) and ask, "When will my order ship?"

The payoff is a significant enhancement in the value of a company's investment in enterprise application integration software. Linking enterprise systems to systems or applications that otherwise would have been left out of the loop, a company can cut the overall cost of integration by 50 or 75 percent. But why stop there? What if all of a company's systems were linked and could then be linked to its most important customers and suppliers?

Given the fractured state of Web services, this might sound like more of the kind of hype the critics disdain. But remember that the true promise of Web services is in reducing costs, not eliminating them. Companies in the retail and banking sectors have already implemented Web services behind the firewall using private networks, SSL (Secure Sockets Layer) and password protection in order to address security concerns. While not visionary, these implementations are proving both economical and effective.

The promise of Web services is exciting, but it will likely take years to make good on it. In the meantime, the companies reaping the early fruits are those with well-developed--and realistic--expectations.

Are you optimistic or pessimistic about the future of Web services? TalkBack below.

John Radko is chief architect for "href="http://www.gegxs.com/index.jsp">GE Global eXchange Services.