Chapter 4: Discovering Web Services

Overview

In the previous two chapters, we looked at two-thirds of the infrastructure that makes Web services possible. You now know how to communicate with Web services, and you know how to describe Web services. The piece of the puzzle that’s missing is how to actually find the Web services you want to use.

Most Web services are still discovered by traditional means—you’re told the address of the service by e-mail or on a Web page, and you navigate to it directly, creating the proxy class as required. Nothing is wrong with this method of discovering Web services, and indeed both Amazon (http: //www.amazon.com/webservices) and Google (http://www.google.com/apis) use it to promote the use of their Web services.

However, what we really need is a method that standardizes the process of discovering Web services and allows searches for Web services that perform particular tasks. For example, we might want to find a Web service that performs Internet searches; without such a standard, we might not know that Google provides such a service.

Two technologies currently allow this method of searching: Microsoft Discovery (DISCO) and Universal Description, Discovery, and Integration (UDDI). We’ll look at both of these in turn. A third possibility, WS-Inspection, is on the horizon; it sits on top of the UDDI infrastructure and allows Web services to be described in a more technology-agnostic way. At the moment, this specification has no real implementation. We’ll look at WS-Inspection in more detail when we cover Global XML Web Services Architecture (GXA) in Chapter 12.