What Has Been Learned From Applying A Formal Process Model To A Real Process

As software processes are multi-faceted and have to cover all the scope of software development ranging from tool invocation to project management, many choices are left about the formalisms used to describe them. Some are focussed on their behavioral aspects and may find it convenient to use computational model based on firing expert system-like rules, Others are interested in their higher level structures and try to start with organizational and project management aspects. As is indicated in the solutions sent to the last year's example problem, a varieties of formalisms are actually studied including rule based, data flow, active database, state transition, pem-net and conventional programming languages. Appropriateness of these approaches have to be checked against actual process descriptions as it was the primary objective of our last year's trial. In this paper, we introduce what we have learned from describing an actual industrial process in our formalism HFSP and want to draw some conclusion about desirable properties of process formalisms.