Open, stable autonomic adaptation: a grand challenge

Treating autonomic systems as solving individual, isolated problems misses the opportunity to consider adaptation as an object of study in its own right. We propose a grand challenge of supporting adaptation in an open and stable manner, embracing the evolution of system goals and constraints over time without sacrificing our ability to describe, implement and analyse autonomic solutions. We suggest a possible approach using topology, and comment on the impact and problems of this novel approach.