I guess the title begs the question, if this is the end of the beginning, is it the beginning of the end? Hardly. But it is time to close the book on the first phase of Web services - the beginning of the hype curve.

Almost a year ago we decided that Web services would receive enough attention that we should consider devoting an entire magazine to the topic. We started with two teaser issues, one of which came out this time last year, and the response was dramatic enough that it justified our existence. In fact, WSJ is one of the fastest-growing titles in the history of SYS-CON Media.

This really has nothing to do with the end of the beginning, but it sets the stage well. A year ago we were all discussing possibilities. Would .NET catch on? Would Java support Web services? We even wondered exactly what Web services was and who was going to use it.

Some of the initial thoughts were grandiose and off base, but in general we got it right. Many of the initial visionaries and evangelists of Web services pictured a world where people could look up any service they wanted and then simply invoke it - on the Web, across the Web. This vision was coupled with the ultimate scenario, one in which companies exposed their business processes to one another via a global directory service. While the possibility still exists, we've seen that in the near term, companies are largely content to use Web services to integrate their own internal processes first.

What's become apparent in the year since we started exploring this topic is that Web services has achieved critical mass. It has enough horsepower to be a viable player in the market.

How can we judge this? Largely by the fundamental shift in the thought processes of vendors and implementers, from the realm of trial applications and rough-hewn support fit only for hardcore developers to the domain of working implementations and smooth packages aimed at high-level developers and business users.

Adding WSDL, UDDI, and SOAP support was relatively easy for the J2EE vendors, and Microsoft built it on top of their platform. Even a large number of vendors who aren't in the application server business provide a Web services stack. At this point it's not a question of where to get a UDDI provider, but which UDDI provider to use.

There are already choices for Web services deployments - multiple choices, with increasingly large numbers of referenceable implementations. What really starts to point to the next step in the evolution of Web services is the rise of companies focusing on two entirely different aspects of provisioning: ease of use and management of service.

On the one hand we have a variety of vendors building workbenches, IDEs, suites - all designed to make it easier to develop and deploy a Web service without ever having to actually write WSDL. These are the tools that take Web services from the esoteric realm of the hardcore developer into the realm of corporate America. These are the tools we really need for Web services to become endemic.

On the other hand, there's management. Web services, like any other corporate asset, needs to be watched, managed, and measured. Service-level agreements are meaningless without tools to measure the ability of a service to comply with them. Much like application servers, Web services won't fully penetrate the corporate infrastructure until it can be managed by a corporate management application such as Tivoli, Unicenter, or OpenView. As in the tools domain, vendors are debuting products that make Web services an integral player within the management space.

So it's the end of the beginning. But not to worry, it's not the beginning of the end.