Web services could topple if we don't address internal issues surrounding standards People who think the world of effortless Web services and XML communication are just around the corner are confused, thinking that having a clear view of the future means it’s a short distance to get there. The clear view is of the world in which Web services expose each of the company’s applications. UDDI (Universal Description, Discovery, and Integration) can be used to discover how a company will talk to partners. Then XML standards defined for each application and industry will make connecting one company to another more like configuring a router than like the heavyweight systems integration currently required. On the Web services front, the news is good for the XML visionaries. The Web services architecture is getting better, and gaping holes in matters such as security and change control are being addressed. UDDI is not widely used to find what integrations are possible, because services must first exist before they are discovered. Still, UDDI is a simple directory service and will work as promised — when the time comes. The not-so-short distance to overcome involves the standards. I have been involved in standards-setting in several industries, and it is a slow process for a number of reasons. First, standards-setting is almost never someone’s day job. Even if a trade group is trying to jump-start an effort by staffing the effort, most experts from the industry are working on standards at quarter-time or less. The standards efforts grind along first to the data dictionary stages, then to the creation of comprehensive database or XML schemas, then to the creation of special purpose schemas. Lots of arguments delay the process. After a standard is announced it must be adopted, another slow process because the benefits of adoption are not distributed equally. Despite the benefit to an industry of having standards in the long term, not all companies are ready for them. Some companies have better infrastructure and more complex supply chains and will benefit more in the short term than will other companies. Those who have little to gain will not move quickly. This further lengthens the time needed to get a critical mass of adoption that forces the entire industry along. Then, of course, there are those pesky vendors who just love to argue about gaining control of infrastructure that will bring billions in profits. Why is there a Liberty Alliance and a Passport initiative? Can’t we all just get along? Apparently not. And until agreement comes on such fundamental issues as a universal identity, progress on universal interoperability will be slow. Those who have been at this stuff for years, such as Oasis and Rosetta.net, use years as a unit of measure for their progress as well. The bright side for shareholders is that companies will not wait for standards to achieve interoperability with their most important partners. They will create standards bilaterally to get the business value as fast as possible. If this is done with good engineering practices, companies will get the benefit now and reduce the cost of change later. They will also bring field experience to standards efforts, which will only make the standards better. But this will create a huge number of overlapping and incompatible standards — a Tower of Babel that will be with us for a long time. Software Development