Some Aspects of a Software Reliability Problem
This is an incremental review that discusses software reliability problems for researchers and practitioners in information systems, without introducing new methods or data.
The paper addresses the lack of a generally accepted measure for software reliability, analyzing reasons for this issue and proposing an approach to solve it, though no concrete results or numbers are provided.
Obviously, the dynamism of software reliability research has speeded up significantly in the last period, and we can state the fact that its intensity is approaching, and in some cases is ahead of the information systems hardware reliability research intensity. Reliability of software is much more important than its other characteristics, such as runtime, and although the absolute reliability of modern software is apparently unattainable, there is still no generally accepted measure of reliability of computer programs. The article analyzes the reasons for the situation and offers an approach to solving the problem. The article touches upon the issue of general characteristics of information systems software life cycle. Considered software application reliability questions and use of fail-safe ensuring programming. Also presented basic types of so-called virus programs that lead to abnormal functioning of information systems. Much attention is given to presenting some known models used for software debugging and operating. So, this review paper consists of four sections: information systems software process creation, reliability of information systems software, using of fail-safe programs and estimation of software reliability according the results of adjusting and normal operation.