On the impact of aspectual decompositions on design stability: an empirical study

  • Authors:

  • Phil Greenwood;Thiago Bartolomei;Eduardo Figueiredo;Marcos Dosea;Alessandro Garcia;Nelio Cacho;Cláudio Sant'Anna;Sergio Soares;Paulo Borba;Uirá Kulesza;Awais Rashid

  • Affiliations:

  • Lancaster University, United Kingdom;Kiel University of Applied Sciences, Germany;Lancaster University, United Kingdom;Federal University of Pernambuco, Brazil;Lancaster University, United Kingdom;Lancaster University, United Kingdom;Lancaster University, United Kingdom and Pontifical Catholic University of Rio, Brazil;Pernambuco State University, Brazil;Federal University of Pernambuco, Brazil;Pontifical Catholic University of Rio, Brazil;Lancaster University, United Kingdom

  • Venue:
  • ECOOP'07 Proceedings of the 21st European conference on Object-Oriented Programming
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Although one of the main promises of aspect-oriented (AO) programming techniques is to promote better software changeability than objectoriented (OO) techniques, there is no empirical evidence on their efficacy to prolong design stability in realistic development scenarios. For instance, no investigation has been performed on the effectiveness of AO decompositions to sustain overall system modularity and minimize manifestation of ripple-effects in the presence of heterogeneous changes. This paper reports a quantitative case study that evolves a real-life application to assess various facets of design stability of OO and AO implementations. Our evaluation focused upon a number of system changes that are typically performed during software maintenance tasks. They ranged from successive re-factorings to more broadly-scoped software increments relative to both crosscutting and non-crosscutting concerns. The study included an analysis of the application in terms of modularity, change propagation, concern interaction, identification of ripple-effects and adherence to well-known design principles.