Describing Web Services

< BACK  NEXT >

[oR]

The technology of Web services can be broken down into four separate areas, each addressing a particular aspect of the problem. Those areas are as follows:

Web services rely on four fundamental technologies

• Describing information sent over the network: Invoking a remote operation commonly involves passing in parameters and getting back some kind of result. With Web services, this information is described using the Extensible Markup Language (XML). Universally accepted as the modern lingua franca for describing data, XML allows a variety of information to be described and exchanged.

• Defining Web service capabilities: Some mechanism must exist to allow the provider of a Web service to specify the technical details of exactly what services are offered. As with other types of services, it makes sense to group related operations into interfaces and then provide some way to describe each of those operations. For Web services, this can be done using the Web Services Description Language (WSDL).^[1] Each WSDL-defined interface contains one or more operations, and WSDL itself is defined using XML.

  ^[1] Commonly pronounced wizdel.

• Accessing Web services: Once an interface has been defined, clients must use some protocol to invoke the operations in that interface. There's no single required protocol (in fact, WSDL explicitly allows specifying different protocols for invoking the operations in an interface). The most important choice, however, is the Simple Object Access Protocol (SOAP). SOAP provides a way to identify which operation to invoke, to convey that operation's inputs as XML-defined data, and to return any outputs, also as XML-defined data. SOAP itself defines only a simple envelope for conveying this information, one that can be carried in various ways. For example, SOAP calls can be carried on the Hypertext Transfer Protocol (HTTP), on a messaging technology such as the Simple Mail Transfer Protocol (SMTP), or in other ways.

• Finding Web services: For developers to create clients that use Web services, there must be some way for those developers to learn what services are available, what they provide, and what their interfaces look like. Given the existence of the Internet, it makes sense to create a standard registry for storing and accessing this kind of information. This is exactly what's done by the technology defined for Universal Description, Discovery, and Integration (UDDI). Using UDDI, providers of Web services can advertise their offerings in a standard way, allowing clients to learn what services each provider offers and letting creators of client software learn what they need to know to build those clients.

Each of these technologies was created by groups of vendors and users working together. XML, for instance, was created by a large group working under the auspices of the World Wide Web Consortium (W3C), while WSDL was created primarily by Microsoft and IBM. SOAP comes from a group somewhere in between in size, with Microsoft, IBM, UserLand Software, DevelopMentor, and several other organizations playing a role. UDDI was originally developed by Microsoft, IBM, and Ariba, although many more organizations have since joined the effort. The key point to notice about the origin of these Web services technologies is that none of them is a single-vendor solution. Instead, Web services based on XML, WSDL, SOAP, and UDDI can be used across virtually all platforms, languages, and object models.

All of these technologies have broad vendor support

< BACK  NEXT >