Re-implementing a legacy system

Harry Sneed*, Chris Verhoef

*Corresponding author for this work

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

Re-implementation is one of the alternatives to migrate a legacy software system next to conversion, wrapping and redevelopment. It is a compromise solution between automated conversion and complete redevelopment. The technical architecture can be revised and the code replaced, but the functional architecture – the use cases remains as it was. The challenge of this approach is to preserve the functionality while changing the technical implementation. This approach is taken when conversion is not feasible and redevelopment is too expensive or too great a risk. It entails more than a 1:1 transformation but less than a total rewrite. The same components remain with different contents. In this paper the case for reimplementation is presented and the process described. The tools required to support the process are identified and their use illustrated. Finally, two industrial case studies are presented, one with a VisualAge/ PL/I-DB2 system and one with a COBOL-IMS application.

Original languageEnglish
Pages (from-to)162-184
Number of pages23
JournalJournal of Systems and Software
Volume155
DOIs
Publication statusPublished - 1 Sep 2019

Keywords

  • Code refactoring
  • Code rewriting
  • Data renaming
  • Reverse engineering
  • Software migration
  • Software re-implementation

Fingerprint

Dive into the research topics of 'Re-implementing a legacy system'. Together they form a unique fingerprint.

Cite this