An integrating role for web services

For once it hasn't been as aggressive as is usual for the IT industry, and while there have been the common groupings of companies supporting one or other alternative proposals, there has been a healthy level of cooperation. Even Microsoft and IBM are in agreement in most areas. The reason for this unusual stance is one of necessity; the Web is bringing together independent trading companies who have historically developed their own systems and applications with little regard to working with outsiders. Web Services must work across heterogeneous systems and it is in the commercial interests of all suppliers to work within the framework. I doubt it will ever be truly simple; there will always be some conflict of interest, e.g. .NET versus Java. We can live with one or two alternatives, indeed there are advantages in having some competition, but we don't want the dozens of alternatives and incompatibilities that IT has suffered from in the past.
However there is an interesting side effect emerging from Web Services developments, a very useful one. The original idea behind Web Services was to build a technical infrastructure in which providers of a variety of services could offer them to a variety of potential users. The target market was of course e-commerce over the Internet. In particular it was conceived with applications in mind that involve interactive users, the so called consume-to-business (C2B) applications.
The Web Services products are based on Internet standards such as XML, SOAP, etc., including the specific Web Services Description Language (WSDL). They define the open methods by which one application can find an appropriate service and request that service from another, as well as how a service can advertise its capabilities.. It exploits robust message passing technologies, obviously with the ability to function outside a firewall. A key objective of such standards was to encourage the development of a wealth of "building block" services for business.
A typical scenario would be for a typical e-commerce application to access a credit checking service and then to use relevant customer data to call on the services of a warehouse and a transport agency. Almost any business that services other businesses could get involved, e.g. banking, insurance, shipping, stock-holding, etc. In many ways this is the modern and more flexible development of Electronic Data Exchange EDI).
This is a good concept, one of the more practical "component" ideas. However after about four years of existence it hasn't yet become a major revenue earner in itself. This is not too surprising since it requires a lot more than implementing a new application, it requires some significant changes to business culture. It is obvious that it is going to take a long time for businesses to come aligned to the new trading and service methods of e-commerce. Until a significant percentage of user organisations and service providers are committed, it won't work. Thus the progress so far is in fact satisfactory and is progressing quite well. But now there is an increased interest because so many companies have got a number of semi-autonomous divisions within the group. They are not ignoring the outside world, but it is a lot easier to try and put one's own house in order first, and the Web Services technology, and the available software to implement it, is an attractive way of integrating all those internal systems. It is this that appears to be the biggest market at the moment.
Web Services techniques will not solve all application integration issues. The development of applications running on Application Servers such as Web Sphere, Cold Fusion, etc. are still the most important since they directly interconnect Web front ends with core business applications, including the major legacy applications. Most organisations have tackled the more important of these application integration problemsnow and are beginning to look towards the next level of integration, which is where the Web Services concept and products come into play. Any organisation adopting this route will be making a sound investment for the future because the next step after intra-company integration is business-to-business e-commerce and a serious need for Web Services. It will be far easier to add the new class of service if the internal integration efforts are based on a similar technology.

 
Martin Healey, pioneer development Intel-based computers en c/s-architecture. Director of a number of IT specialist companies and an Emeritus Professor of the University of Wales.