Does code decay? Assessing the evidence from change management data

Abstract

AbstractÐA central feature of the evolution of large software systems is that changeÐwhich is necessary to add new functionality, accommodate new hardware, and repair faultsÐbecomes increasingly difficult over time. In this paper, we approach this phenomenon, which we term code decay, scientifically and statistically. We define code decay and propose a number of measurements (code decay indices) on software and on the organizations that produce it, that serve as symptoms, risk factors, and predictors of decay. Using an unusually rich data set (the fifteen-plus year change history of the millions of lines of software for a telephone switching system), we find mixed, but on the whole persuasive, statistical evidence of code decay, which is corroborated by developers of the code. Suggestive indications that perfective maintenance can retard code decay are also discussed.

Combining Single-Version and Evolutionary Dependencies for Software-Change Prediction

“…Such changes can lead to potential side effects and/or violations of the underlying assumptions. Software-change prediction methodologies that provide all the entities that need to be appropriately co-changed are important for sustained evolution of a software system [5,14]. Two broad groups of methodologies are described in the literature for supporting software changes.…”

Successful Reuse of Software Components: A Report from the Open Source Perspective

“…With reference to software decay, past SE literature has firmly established that software architectures and the associated code degrade over time [13], and that the pressure on software systems to evolve in order not to become obsolete plays a major role in their evolution [23]. As a result, software systems have the progressive tendency to loose their original structure, which makes it difficult to understand and further maintain them [33].…”

Is It Dangerous to Use Version Control Histories to Study Source Code Evolution?

“…Eick et al [13] identified specific indicators for code decay by conducting a study on a large (∼100,000,000 LOC) real time software for telephone systems. These indicators were based on a combination of metrics such as number of lines changed, commit size, number of files affected by a commit, duration of a change, and the number of developers contributing to a file.…”