Designing a
Method Network (Hierarchy)
As you progress from one level of detail to the
next, you should find the sub-processes required to perform the
overall tasks for which they are assigned. At each level of
analysis, you should perform an evaluation to identify duplicate
process behaviors under different higher-level processes.
At this stage of analysis, duplicated processes
are not eliminated or consolidated because further analysis at
greater granularity might reveal subtle differences in the processes
after all. What is done is that the two (or more) processes are
labeled to identify their system duplicity and the analysis
proceeds. Ultimately, you'll want to proceed until decomposition is
no longer possible. At that point, the effort turns to organizing
the defined processes. In which case, you start by grouping the
processes into the three categories named earlier: independent, base
level, and aggregate.
The next step is to revisit the duplicated
processes and consolidate them. This is not to say that you will
create one big package and put redundant sub-processes together.
Instead, the effort is focused on organization. Later, redundant
processes will be grouped categorically by function. These steps
have proven to be a most effective means of code organization in
terms of maintenance and deployment.
