Software Stability:

Recovering General Patterns of Business

The software stability approach is required to balance the seemingly contradictory goals of stability over the software lifecycle with the need for adaptability, extensibility and interoperability. This workshop paper addresses the issue of how software stability can be achieved over time by outlining an approach to evolving General Business Patterns (GBPs) from the empirical data contained within legacy systems. GBPs are patterns of business objects that are (directionally) stable across contexts of use. The work explains, via a small worked example, how stability is achieved via a process of ‘sophistication’. The outcome of the process demonstrates how the balance that stability seeks can be achieved.


UML 2003, 6th International Conference on the Unified Modeling Language, 20 October 2003, San Francisco, California, USA


Mark Lycett (Brunel University)
Chris Partridge (BORO Solutions, Brunel University)
Sergio de Cesare (Brunel University)