Increasing Fault Detection Effectiveness Using Layered Program Auralization

This paper presents a new approach to using music for debugging computer code, layered program auralization. We use layers of musical structure to represent the state and behavior of a computer program while it is running, taking advantage of metaphorical relationships between musical structure and programming constructs. This paper makes three contributions to the debugging and program auralization literature. We use cadences, recognizable patterns of chords that usually dictate an ending, to metaphorically represent nesting depth and hierarchical control structures. Auralizations, in our system, use more than one perceptual mapping at a time. Further, we decompose our auralizations into units, which we call layers. This decomposition is similar to using classes in object oriented programming. We see debugging as a time consuming, difficult, task, and present a solution where music is played to the programmer during program execution. Our goal is to increase debugging effectiveness and to improve a programmer’s comprehension of the runtime behavior of computer programs.