Information Technology Reference
In-Depth Information
Step 2: The next step is to design a class diagram. Determine the attributes and the
methods of each class.
Step 3: Dynamic modeling is the final step. In this step the actions of all classes
and subclasses should be described in a state chart.
You may be wondering what is meant by ''iterative and incremental''. What
iterative means is that this will continue over and over again in the same basic
form, but with slight modifications to the inputs each time. Incremental in this case
means that the whole project may not need tackled at the same time. It may be
necessary to get a general overview of the system and then break the architecture
down into smaller units or subsystems. Putting these terms together shows that this
process is not a big bang approach, where you go from nothing to something with
little in between. In fact with this approach it is the sheer opposite.
This architecture is very much akin to city planning and development. When a
city is being planned the planners begin by zoning the land they have to work with
to determine what function it will serve, such as residential, industrial, commer-
cial, etc. This is the same as creating a functional model where actors are used to
showing the interaction with the system. City planners use maps and zoning code
to show how people will interact in that part of the city. Second, city planners
would need to attract business, or display some reason for people and their money
to come to where ever it is this new city will be. The layout of each business and
the resulting development (for example the homes to house employees of a large
factory) can be seen as the classes and their relations to one another. Finally,
highways, utilities and services must be in place. In order to do this it needs to be
understood what, when and how each entity needs to function. This is closely akin
to a state chart where the systems function is laid out for closer examination. This
example is very reliant on the fact that the method is incremental and iterative. If
the planners had to fill every inch of the city instantly at its conception they would
never succeed, but since details can be created and modified over time it becomes
much easier to manage the development process.
7.7.1 Data Design
Data hiding or encapsulation has been the concern of object-oriented design since
its inception. Code reuse creates an abundance of easily accessible code entities for
easy and tested use. The next step in code reuse is a common data representation
format. Abstract Data Types (ADTs) are the answer to this problem. The modules
resulting from the procedural programming proceeding OOD give no hint on how
to implement them as meaningful components (Elienes
1995
).
For instance, without ADT's a programmer may produce several separate
pieces of code, individually and invariably differently from each other or they
could create the least common denominator of the data type and allow for specific
types to be applied dynamically. Assume this programmer is creating a queue, a
