Reusing proofs of program correctness in ENCOMPASS
Robert B. Terwilliger · 1990
If a program's proof of correctness can be reused, higher quality may be achieved with reasonable cost. Unfortunately, reusing proofs of program correctness is difficult. The approach being taken toward this problem in the ENCOMPASS project is explored. Specifically, the author presents examples of three types of proof reuse: instantiating (reusing) a parameterized component and its proof, reusing a development step with its proof, and reusing a provably correct program schema. The author believes that, although program verification will, in general, remain expensive, the reuse of verified components may become practical through the use of such methods.>