• DocumentCode
    2024389
  • Title

    Butterflies: a visual approach to characterize packages

  • Author

    Ducasse, Stéphane ; Lanza, Michele ; Ponisio, Laura

  • Author_Institution
    Software Composition Group, Bern Univ.
  • fYear
    2005
  • fDate
    1-1 Sept. 2005
  • Lastpage
    7
  • Abstract
    Understanding sets of classes, or packages, is an important activity in the development and reengineering of large object-oriented systems. Packages represent the coarse grained structure of an application. They are artefacts to deploy and structure software, and therefore more than a simple generalization of classes. The relationships between packages and their contained classes are key in the decomposition of an application and its (re)-modularisation. However, it is difficult to quickly grasp the structure of a package and to understand how a package interacts with the rest of the system. We tackle this problem using butterfly visualizations, i.e., dedicated radar charts built from simple package metrics based on a language-independent meta-model. We illustrate our approach on two applications and show how we can retrieve the relevant characteristics of packages
  • Keywords
    object-oriented programming; program visualisation; reverse engineering; software metrics; software packages; systems re-engineering; butterfly visualization; class generalization; language-independent metamodel; object-oriented systems; package metrics; program understanding; radar charts; reverse engineering; software package characterization; software visualization; system development; system reengineering; Application software; Informatics; Java; Packaging; Radar; Reverse engineering; Software maintenance; Software packages; Software systems; Visualization; Program understanding; package metrics; reverse engineering; software visualization;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Software Metrics, 2005. 11th IEEE International Symposium
  • Conference_Location
    Como
  • ISSN
    1530-1435
  • Print_ISBN
    0-7695-2371-4
  • Type

    conf

  • DOI
    10.1109/METRICS.2005.15
  • Filename
    1509285