What are the important research issues in domain-specific languages?

Julia Lawall

Many of us have developed, used, or desired domain-specific languages (DSLs) in a variety of problem areas. Furthermore, the working group will be co-organizing a conference on DSLs next year and a workshop on DSLs will be co-located with GPCE 2008. The goal of this discussion is to consider what are problems that we as a community should be working on in this area. Some issues that could be considered are:

* If you were teaching a course on DSL's, what would you put in it?

* If you had a new PhD student who was interested in DSLs, what literature would you point him to?

* What topics would you hope to hear about at the DSL conference?

* What kinds of industrial collaborations are possible in this area, and how would such collaborations make a research contribution?

* How much can DSL design and development be automated?

* What tools could we envision to develop DSLs? What are the benefits and shortcomings of existing tools, such as DSLTools, JastAdd, and Stratego

* What are the killer apps for DSLs? Does the question make sense?

* What is the relationship between DSLs and modeling languages?

* Which metrics could be used to evaluate the benefits of a DSL for a given application domain.

* Does the syntax of a DSL really matter? Is XML, as eg used in a configuration file, a DSL?

(some questions contributed by Charles Consel and Laurent Reveillere)