The future of software

Companies like British Telecommunications whose business is based on the performance of their software systems, should emphasize accessibility, adaptability, transparency, fail-safe operation, and a human face. Software Producing such software is difficult and involves high costs and risks. Adopting the most appropriate methods, technologies, and tools at just the right time is a major problem for the software industry. Recognition of the critical role played by software in so many aspects of society has therefore led us to pursue the following goals: • Forming a vision of software and software development based on the systematic use of expert judgement and peer review; • Establishing the need for a long-term software-development research agenda; and • Outlining a long-term research agenda that helps meet society's future needs for software that is reliable, adaptable, available when needed, and reasonably priced. Predicting the future is a popular pastime in many disciplines. In the field of software development, the February 1997 Communications offered many personal hopes and visions for the future of computer technology [1]. Authors expressed their personal views relating to particular technologies and applications , including databases [2], the Internet [3], and computational humanities [6]. In another personal view of future needs, Stuart Shapiro [5] examined a number of key software technology publications dating from as far back as the 1960s, concluding we need a more pluralistic approach to software engineering, Society is increasingly dependent on large and complex software systems. Indeed, if many of the current predictions about the Y2K computer problem turn out to be accurate, this dependency will be demonstrated dramatically on January 1, 2000. Users need software that meets stringent requirements, supports a range of interaction styles, can be produced quickly, and can be maintained to keep pace with the ever-increasing demand for functional-ity, quality, flexibility, and cost-effectiveness.