A desiderata for refactoring-based software modularity improvement

  • Authors:

  • Girish Maskeri Rama

  • Affiliations:

  • Infosys Technologies Limited, Bangalore, India

  • Venue:
  • Proceedings of the 3rd India software engineering conference
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

There exists number of large business critical software systems of recent vintage that are becoming increasingly difficult to maintain. These systems, written in newer languages such as C and Java are fast becoming legacy and showing the same symptoms of modularity deterioration reminiscent of older legacy systems written in Cobol and PL1. However, this problem has not received much attention from the software modularization community. In this paper, we argue that the modularization needs of these relatively newer systems, which generally have some modular structure, is significantly different from the needs of older, mainly monolithic, legacy systems. We emphasize the need for incrementally improving modularity and propose a software refactoring based approach to solve this problem, thereby, uniting hitherto two disparate threads of research - software modularization and software refactoring. As part of this refactoring based modularity improvement approach, we identify a set of modularity smells and specify how to detect these smells in poorly modularized software systems. A validation of these modularity smells is carried out using several open source systems.