Implementation Of Business Policies Using Object Oriented Methodologies And Design Patterns

In the early problem-solution era of software programming, functional decompositions were mainly used to design and implement software solution. In functional decompositions, functions and data are introduced as two separate entities during the design phase, and are followed as such in the implementation phase. In general, separation of function and data in a software program causes tight coupling between the two. Tight coupling means that a change in a data may require multiple changes to the design and code through the system. Also, tight coupling of data and function adversely impacts the cost of software maintenance. For example, the year 2000 phenomena (Y2K) broke many software program logics that involved date arithmetic operations (subtracting year portion of the date), processing the year portion of a date produced erroneous results. Conceptually the correction is minimal; “just change the year from two digits to four digits”, however we now know that correcting the date error cost the business communities billions of dollar. The correction included fixing the software so that the year portion of a date is represented as four digits attribute versus two digits programmed earlier. The major cost was due to “tight coupling”; date was tightly coupled with all functions using the date; as such, changing the date required making several changes throughout the software systems.