View article

A key problem in view-based software development is merging a set of disparate views into a single seamless view. To merge a set of views, we need to know how they are related. In this extended abstract, we discuss the methodological aspects of describing the relationships between views. We argue that view relationships should be treated as first-class artifacts in the merge problem and propose a general approach to view merging based on this argument. We illustrate the usefulness of our approach by instantiating it to the state-machine modelling domain and developing a flexible tool for merging state-machines.