Assessing technical debt by identifying design flaws in software systems

Abstract

Tough time-to-market constraints and unanticipated integration or evolution issues lead to design tradeoffs that usually cause flaws in the structure of a software system. Thus, maintenance costs grow significantly. The impact of these design decisions, which provide short-term benefits at the expense of the system's design integrity, is usually referred to as technical debt. In this paper, I propose a novel framework for assessing technical debt using a technique for detecting design flaws, i.e., specific violations of well-established design principles and rules. To make the framework comprehensive and balanced, it is built on top of a set of metrics-based detection rules for well-known design flaws that cover all of the major aspects of design such as coupling, complexity, and encapsulation. I demonstrate the effectiveness of the framework by assessing the evolution of technical debt symptoms over a total of 63 releases of two popular Eclipse® projects. The case study shows how the framework can detect debt symptoms and past refactoring actions. The experiment also reveals that in the absence of such a framework, restructuring actions are not always coherent and systematic, not even when performed by very experienced developers.