Meta-analysis and Reflection as System Development Strategies

Abstract

This paper is about strategies for developing new approaches to solving difficult problems in complex computing system design, based on a very strong use of abstraction and reflection. We address problems in the design and development of Constructed Complex Systems, which are large, heterogeneous, usually distributed, systems managed or mediated by computing systems. We call this strategy "meta-analysis", since it is as much about the processes of modeling and defining elements of these systems as it is about new kinds of elements themselves. For some years, we have had interesting successes in defining new approaches to these problems, and we have recently noticed that there is some commonality in the approaches. We define what meta-analysis is, show some examples from our earlier research, and explain why it is a meta-informatic approach. © Springer-Verlag 2004.