Moving from data architecture to event architecture in the EMS environment

Information models and databases are at the center of the design of most current software systems. The job done by the active portions, the applications, is to transform the data as needed. If the applications can, they work in a strictly stateless, data transforming paradigm, because this creates the simplest sort of reliable system. Our EMS systems are complex, real-time systems that cannot adopt this simplest paradigm, but they nevertheless follow a data-centric architecture. This is about to change. "Message buses" or "integration buses" or "event buses" are different names for the ability to connect applications with high speed flexible messaging. This technology is now ready for prime time and the question is - how are we going to use it? Are we going to take advantage of the active voice communication that events can provide? Or, are we going to stay predominantly in the familiar data orientation? The situation is somewhat analogous to the now-familiar experience that procedural programmers, when given object oriented languages, often did not go through the paradigm-shift and produced procedural code written in object languages. In this paper, we focus on the advantages that can be achieved in the temporal domain - a long-standing trouble-spot in EMS design.