Approaches to System Design
There are three fundamentally different approaches to defining a data processing system: data-driven, output-driven, or process-driven. Each of these is discussed below.
Using a data-driven approach, you first define all the data fields for an application. Next, you specify the relationships between the data fields and then develop supporting processes. Proponents of this approach believe that if they can define all data-related information in advance, processes to support the entry, update, review, and retrieval of data will flow naturally as a by-product of data definition.
In an output-driven design, you first define all the required outputs (both screen displays and printed reports) and, then, you develop the data structures, and input and update processes that you need to derive them. If the desired outputs can be defined, they can work 'backwards' through the system design to define all the supporting structures that allow the system to produce the specified outputs.
Using a process-driven approach, you define each step in the work process (such as shipping finished goods, preparing invoices, etc.) and develop parallel steps in the application system to support the normal work flow. Proponents of this approach feel that a computerized application should be developed to reflect the day-to-day business operation of a company or organization. If the processes are well-defined, the details of data entry and retrieval can be logically defined as a secondary task.
APPX supports any of the three approaches that are described above. After you have developed the preliminary definition of a proposed application, however, there is a preferred sequence of steps to make the building of an application in APPX more straightforward. |