Results of measuring the performance of the storage management subsystem in an implementation of SNOBOL4 are described. By instrumenting the storage management system, data concerning the size, lifetime, and use of storage blocks were collected. These data, like those obtained from conventional time measurement techniques, were used to locate program inefficiencies. In addition, these measurements uncovered some deficiencies in the storage management system, and provided the basis upon which to judge the heuristics used in the garbage collector.

This paper concems the representation and processing of data resulting from the measurement of programs written in high-level programming languages. The value of various "perspectives" of measurement results, especially in languages having many levels of activity, is illustrated. A representation of measurement data based on ideas from relational data bases is presented, in which perspectives can be represented as hierarchically-organized permuted projections of measurement data. Applications described include measurement of Snobol4 and Fortran programs.

This paper describes a graduate Computer Science course in which all aspects of the programming process are examined.