Survey of established practices in the life cycle of domain-specific languages

Abstract

Domain-specific languages (DSLs) have demonstrated their usefulness within many domains such as finance, robotics, and telecommunication. This success has been exemplified by the publication of a wide range of articles regarding specific DSLs and their merits in terms of improved software quality, programmer efficiency, security, etc. However, there is little public information on what happens to these DSLs after they are developed and published. The lack of information makes it difficult for a DSL practitioner or tool creator to identify trends, current practices, and issues within the field. In this paper, we seek to establish the current state of a DSL's life cycle by analysing 30 questionnaire answers from DSL authors on the design and development, launch, evolution, and end of life of their DSL. On this empirical foundation, we make six recommendations to DSL practitioners, scholars, and tool creators on the subjects of user involvement in the design process, DSL evolution, and the end of life of DSLs.