Understanding Structural Complexity Evolution: A Quantitative Analysis

Background: An increase in structural complexity makes the source code of software projects more difficult to understand, and consequently more difficult and expensive to maintain and evolve. Knowing the factors that influence structural complexity helps developers to avoid the effects of higher levels of structural complexity on the maintainability of their projects. Aims: This paper investigates factors that might influence the evolution of structural complexity. Method: We analyzed the source code repositories of 5 free/open source software projects, with commits as experimental units. For each commit we measured the structural complexity variation it caused, the experience of the developer who made the commit, the size variation caused by the commit, and the change diffusion of the commit. Commits that increased structural complexity were analyzed separately from commits that decreased structural complexity, since they represent activities of distinct natures. Results: Change diffusion was the most influential among the factors studied, followed by size variation and developer experience, system growth was not necessarily associated with complexity increase, all the factors we studied influenced at least two projects, different projects were affected by different factors, and the factors that influenced the increase in structural complexity were usually not the same that influenced the decrease. Conclusions: All the factors explored in this study should be taken into consideration when analysing structural complexity evolution. However, they do not fully explain the structural complexity evolution in the studied projects: this suggests that qualitative studies are needed in order to better understand structural complexity evolution and identify other factors that must be included in future quantitative analysis.

[1]  Audris Mockus,et al.  Developer fluency: achieving true mastery in software projects , 2010, FSE '10.

[2]  Joseph K. Kearney,et al.  Problems with software complexity measurement , 1985, CSC '85.

[3]  David Lorge Parnas,et al.  Software aging , 1994, Proceedings of 16th International Conference on Software Engineering.

[4]  Fabio Kon,et al.  A Study of the Relationships between Source Code Metrics and Attractiveness in Free Software Projects , 2010, 2010 Brazilian Symposium on Software Engineering.

[5]  Audris Mockus,et al.  Predicting risk of software changes , 2000, Bell Labs Technical Journal.

[6]  Jose Luis Roca An entropy-based method for computing software structural complexity , 1996 .

[7]  Guy A. Schiavone,et al.  The effects of development team skill on software product quality , 2006, SOEN.

[8]  Chris F. Kemerer,et al.  A Metrics Suite for Object Oriented Design , 2015, IEEE Trans. Software Eng..

[9]  Vishal Midha,et al.  Does Complexity Matter? The Impact of Change in Structural Complexity on Software Maintenance and New Developers' Contributions in Open Source Software , 2008, ICIS.

[10]  Akito Monden,et al.  An analysis of developer metrics for fault prediction , 2010, PROMISE '10.

[11]  Anas N. Al-Rabadi,et al.  A comparison of modified reconstructability analysis and Ashenhurst‐Curtis decomposition of Boolean functions , 2004 .

[12]  Katherine J. Stewart,et al.  Exploring Complexity in Open Source Software: Evolutionary Patterns, Antecedents, and Outcomes , 2010, 2010 43rd Hawaii International Conference on System Sciences.

[13]  Min Zhang,et al.  Performance Comparison of Software Complexity Metrics in an Open Source Project , 2007, EuroSPI.

[14]  Audris Mockus,et al.  Expertise Browser: a quantitative approach to identifying expertise , 2002, Proceedings of the 24th International Conference on Software Engineering. ICSE 2002.

[15]  Dewayne E. Perry,et al.  Metrics and laws of software evolution-the nineties view , 1997, Proceedings Fourth International Software Metrics Symposium.

[16]  Martin Hitz,et al.  Measuring coupling and cohesion in object-oriented systems , 1995 .

[17]  Katherine J. Stewart,et al.  Opportunities and Challenges Applying Functional Data Analysis to the Study of Open Source Software Evolution , 2006 .

[18]  Vineet Sinha,et al.  Using dependency models to manage complex software architecture , 2005, OOPSLA '05.

[19]  Chris F. Kemerer,et al.  How software process automation affects software evolution: a longitudinal empirical analysis , 2007, J. Softw. Maintenance Res. Pract..

[20]  Paul Clements,et al.  Software architecture in practice , 1999, SEI series in software engineering.

[21]  Claes Wohlin,et al.  Experimentation in software engineering: an introduction , 2000 .

[22]  Keith H. Bennett,et al.  Software maintenance and evolution: a roadmap , 2000, ICSE '00.

[23]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[24]  Robert C. Martin Agile Software Development, Principles, Patterns, and Practices , 2002 .

[25]  Odd Steen,et al.  Practical knowledge and its importance for software product quality , 2007, Inf. Softw. Technol..

[26]  James E. Tomayko,et al.  The structural complexity of software an experimental test , 2005, IEEE Transactions on Software Engineering.

[27]  Rajiv D. Banker,et al.  A model to evaluate variables impacting the productivity of software maintenance projects , 1991 .

[28]  Ahmed E. Hassan,et al.  Predicting faults using the complexity of code changes , 2009, 2009 IEEE 31st International Conference on Software Engineering.

[29]  Nicholas A. Kraft,et al.  Cyclomatic Complexity and Lines of Code: Empirical Evidence of a Stable Linear Relationship , 2009, J. Softw. Eng. Appl..

[30]  Raghvinder S. Sangwan,et al.  Structural Epochs in the Complexity of Software over Time , 2008, IEEE Software.