This wonderful paper proposes special constraints for software patterns so that they can be processed to become executable software components. The design representation of patterns can help make them controllable software artifacts. The challenge is to define constraints so that the pattern does not become trivial. Štěpán admits that the “exact form of variability modeling is one of [his] challenges.”
Though this paper is a work in progress, it does provide a good tutorial of the state of patterns and pattern design. It is easy to read and insightful. I especially liked the contribution in section 4, where the author lays out a procedure that allows a pattern to become represented as a connector.
Recursion, especially across different instances of the pattern’s execution, needs to be treated in the feasibility examples of section 5. These examples are at quite a low level. Once we understand the constraints, we must apply them to higher-level patterns.
This is a worthwhile paper, and the research it introduces is worth following.