The evolution of customer middleware requirements

In the past, commercial client/server systems were typically limited to configurations employing a single relational database management system (DBMS). Moreover, both client and server components were either designed by a single vendor, or constructed using one of the many application programming interfaces (API) available, restricting that application for use with the single DBMS for which that API was designed. Since then, customer expectations have grown to include connectivity and interoperability requirements for the heterogeneous mixture of applications and database products which comprise their enterprise: a diverse set of APIs, operating platforms, filetypes, and network operating systems. Middleware products have emerged to provide interoperability for these components via “gateways”, which map the APIs, data manipulation languages (DML), and database wire protocols of these diverse components. In addition to basic interoperability, the ease-of-use and performance characteristics of the middleware model have begun to change. The distributed database, constructed from middleware technologies, is seen as the means by which the distributed, heterogeneous enterprise computing environment can be integrated as a homogeneous unit