This paper describes successive reformulations of skeletal-plan refinement as a problemsolving method. We argue that, whereas ideas derived from planning literature helped to determine the overall structure of the planning systems, domain-derived considerations and architectural framework in which the systems were implemented played important roles in these reformulations. We illustrate the argument by describing a new framework that integrates knowledge-based applications with a temporal data-abstraction and data-management system. In this framework, both applications and temporal-data mediators are encapsulated as Common Object Request Broker Architecture (CORBA) objects. The skeletal-plan refinement method itself is formulated as a collection of cooperating CORBA objects. We have found that we needed to reformulate the method ontology, mapping relations and control structure of the skeletal-planning problemsolving method in this framework. Our experience suggests that problem-solving methods are not necessarily fixed structures that can be plugged into arbitrary application environments, and that we need to develop a flexible configuration environment and expressive mapping formalisms to accommodate the requirements of application environments. These requirements include the ways data are made available and the ways software components interact with one another.