Gilpin: Users Face Inevitable Tension Between Short- And Long-Term Goals

Over the next five years, companies that have previously built trading rooms with lower-level middleware technologies like messaging and Object Request Brokers (ORBs) will increasingly utilize those and other low-level middleware in the context of higher-level products, like application servers, that are easier to use, configure and manage.

Not only is the industry of middleware vendors driven to provide higher-value solutions that offer more strategic positioning possibilities, but those that use these technologies are highly motivated to migrate incrementally toward such higher-value solutions. That is because the cost to maintain their own middle-office infrastructure on top of these lower-level technologies is simply too high compared to using more commercialized solutions.

The trade-off that will balance this trend on Wall Street is the fact that trading rooms require special features that are not as widespread in higher-level products, like IP multicasting and object caching support.

So, only those vendors that have these special features in their higher-level products will be able to attract Wall Street customers to move away from their in-house infrastructure.

The biggest challenge will be the inevitable tension between short-term and long-term goals, acting out in the context of strategic middleware application platforms, as IT managers have to reconcile the need to address Y2K and EMU issues while simultaneously making bet-the-business choices of architecture and platform that will make or break their ability to respond to the business challenges of the near future, such as Internet applications, globalization and business-to-business integration.

Toward a greater usage of distributed-component technology and platforms. The primary business driver for this will be the need for greater flexibility in IT infrastructure, rather than reuse, although the economics of reuse via acquisition of products in a componentized form will continue to be attractive.

The tremendous investment in existing systems and technology, primarily based on architectures that are relational-database-centric. Many are extremely reluctant to migrate any of these applications or data to a new platform, without a really strong justification to do so, and they have not yet seen that case made in a truly compelling fashion. I believe it will be possible to build such a case over the next couple of years, but that has not yet been done except in some specific contexts.