Design Development Paradigm

This section will cover the design methodology of the PPF project and various design considerations.

The iterative method will be the basis for design of the Parallel Particle Framework. We choose the iterative method for the following reasons:

• I simple skeleton of the PPF can be developed easily.
• The nature of the problem requires all basic components to exist before the PPF can be tested.
• Correctness and accuracy are vital to the success of the PPF and therefore the ability to test the PPF must be fundamental to the design paradigm chosen.
• Some of the features of the PPF are not crucial to its functionality. In the advent that the project deadline approaches before the PPF is completed, prototypes of the least important modules are already incorporated into the PPF.

The following list provides a basic set of guidelines to be used as a preliminary set of iterative steps during the design process:

• Create independent workspace with working polymesh code. DONE!
• Develope fake physics package to randomly move particles around a problem mesh. DONE!
• Write a driver file to model a package using the PPF, use the physics package as the base for the phsyics.
• Create a skeleton of the PPF, include all classes and methods, however they will have no functionality to begin with.
• Work on the PPF_Logic methods first. This will require simultaneous development of the PPF_Parallel_Wrapper. These modules must be working correctly before proceeding.
• Fill in the PPF_Statistics and PPF_Timer methods.