This paper proposes a model which explains the process of learning about computation in terms of well-accepted software engineering concepts, and argues that our approach to understanding how problem-solving skills are acquired is an innovation over well-accepted learning theories and models. It examines how it all students make sense of computational processes; by reporting on experimental observations that have been made with school children, and with university undergraduates. We observed little difference between children and adults with regard to how they learn about computation, and suggest that the strong similarities are due to a common set of problem-solving techniques which are fundamental to all problem based learning, in general, and learning about computation, in particular. To conclude, we demonstrate that our model --- based on software engineering concepts --- is useful when reasoning about the relationship between problem solving and learning to program.