Designing software for producibility

This article addresses the questions, what is software design quality and how can measurement help to improve it? There are two principal quality factors: traceability and producibility. The former is the usual notion of software quality as conformance to require0ments. The latter relates to the difficulty of implementing the proposed design. Poor producibility means increased probability of error, lower productivity, and higher maintenance costs during the rest of the product's life cycle. Measurement can help to improve design quality in several ways. It establishes a basis for evaluating design practices and guiding design process improvements. Specifically, measuring complexity provides criteria for making decisions about design alternatives that maximize producibility. This concept of quality and the applications for measurement to software design are illustrated with data from actual projects.