The article discusses the importance of software reliability and the lack of a universally accepted measure for it. It explores reasons for this and suggests a solution. The paper covers the software life cycle, fail-safe programming, and types of virus programs that disrupt systems. It also presents models for debugging and operating software. The main focus is on creating reliable software and estimating its reliability during normal use.