Mining Existing Assets for Software Product Lines

Publisher Website
Google Scholar
Abstract
Mining of existing assets offers an organization the potential to leverage all, or part, of its cumulative system investments, and thus represents a critical practice area in implementing a software product line. However, there are significant risks in achieving success because of the poorly documented and maintained state of many existing systems and the fact that many systems were initially developed for different paradigms than current distributed, Web-oriented, object-oriented approaches. Four basic steps are required to successfully mine assets: (1) preliminary information gathering, (2) making decisions on whether to mine assets and which type of overall strategy to use, (3) obtaining detailed technical understanding of existing software assets, and (4) rehabilitation of assets. This note outlines basic considerations for each of these steps. It outlines typical information to collect before an analysis. It then outlines a model for making decisions on mining legacy assets, and discusses the technical understanding of assets and the rehabilitation of assets. Because of its importance as a strategy for product lines, architecture reconstruction is discussed, as it is supported by an automated tool set known as the Dali workbench.

This publication has 0 references indexed in Scilit: