Why Build an MDA Process?

The purpose of defining an MDA process is to determine the domains and bridges, as expressed by models and mapping functions with their associated marks. For each model, we will follow a process that establishes how to go about constructing a good model, reviewing it, testing it, and so forth.

Similarly, the tasks of building mapping functions and marking models have separate processes associated with them. These processes can be determined before or after the overall MDA process is established. The MDA process determines only to which models and which mapping functions each individual process applies.

Note that careful selection of the models, combined with the splitting apart of complex mapping chains into multiple hops, increases the reusability, portability, and longevity of both models and mapping chains for use in future projects. More immediately, you need a planned process for the project at hand, in which all of the players know which models to build and which can be reused, which mapping functions to build and which can be reused, and how to make use of legacy code.