Until recently, most specifications for computer systems were prepared in the form of written documents — similar in a way to the written specifications prepared by architects for new buildings. Architects, however, also rely on the use of drawings to convey information about proposed developments whereas computer systems have been traditionally specified simply by the use of written statements (supplemented perhaps by flow diagrams).
In the case of architecture, a drawing does more than just convey detailed information. It provides a model which is used by the designer to test the viability of his or her proposals. That is, the architect can visualise whether something is going to work or not. Imagine the carbuncles that we might have to suffer if architects were not able to create pictures of their schemes first!
In much the same way, computer systems designers require methods for specifying new systems which enable them, first, to examine fully the logic and suitability of any proposals, and second, to convey a user’s requirements in the most precise way possible. To meet these requirements various structured methodologies have been developed, which result in system specifications consisting of written and diagrammatic information — analogous to the specifications and drawings used in construction and engineering. An example of such an approach is Structured Systems Analysis and Design Methodology (SSADM).
SSADM consists of a number of interrelated techniques, such as entity-relationship modelling and data-flow diagrams, which enable the process of systems design to be tackled in a much more logical and systematic manner than using traditional methodologies.
This particular approach is well suited to the complexity inherent in property systems. For example, the data structure of a property system must be able to cope with the often complex relationships between property interests, sites, buildings, units and controlling bodies.
The recently published CIPFA guide on asset registers for local authorities illustrates the need for such a structured approach. Within the guide the three main entities identified for data-storage purposes are site, establishment (eg school) and building. However, the use of entity-relationship modelling reveals that property (ie interests in property) is the most important entity about which information needs to be stored, and should form the core of any administration system.
In this way, SSADM may be used to create data structures that are both logically correct and which provide the foundation for systems that may be readily adapted over time, to suit the changing requirements of users. To continue the analogy used earlier, it is no use building a structure on shaky foundations!
