Chair: Thomas Ball, Bell Laboratories, Lucent Technologies
Tim Bray, Textuality
Charles Consel, Irisa/University of Rennes
Mary Fernandez, AT&T Labs-Research
Paul Hudak, Yale University
James R. Larus, Microsoft Research
Doug Lea, State University of New York at Oswego
Jay Lepreau, University of Utah
Brad A. Myers, Human-Computer Interaction Institute, Carnegie Mellon University
Todd Proebsting, Microsoft Research
David S. Rosenblum, University of California, Irvine
Michael Schwartzbach, University of Aarhus

Invited Talks Coordinator
Carlos Puchol, Bell Laboratories, Lucent Technologies

Letter from Thomas Ball, Program Chair

Dear Colleagues,

It is my pleasure to invite you to attend the Domain-Specific Languages Conference, the second sponsored by USENIX, again in cooperation with the ACM Special Interest Groups on Programming Languages and Software Engineering. I am confident of a repetition of the very high quality of interaction that occurred at the first DSL Conference in 1997. I have no doubt the quality of the presentations will be just as outstanding.

Domain-specific languages have had substantial impact on how software is created, maintained, and modified. Prototypical examples of DSLs are YACC, SQL, spreadsheets, and HTML. These DSLs exemplify many of the unique attributes of the DSL approach:

• DSLs automatically provide programmers with guarantees of correctness, performance, and security that are simply unachievable with general-purpose languages such as C, C++, and Java (think of YACC and SQL).
• DSLs allow non-programmers to program (think of spreadsheets).
• By providing high-level abstractions tailored to the problem domain, DSLs allow programmers with general skills to program in a new domain without having to know platform details (think of HTML and Web services).

DSL'99 advances the practice of DSL design, DSL implementation, and software engineering by:

• providing examples of successful DSLs;
• highlighting the spectrum of benefits DSLs provide (e.g., compile-time guarantees of behavior, improved program performance);
• uncovering design principles and methodologies for creating DSLs;
• explicating design techniques and tools for working with DSLs throughout the software engineering lifecycle;
• providing a framework within which language designers from different domains can easily communicate; and
• facilitating a community that will continue to study and refine the practice of software engineering through DSLs.

DSL'99 features refereed technical papers and invited talks, along with "hot research reviews." The conference offers technical papers on various approaches to DSL construction, on new DSLs for problem domains such as specifying hardware circuits and robot control protocols, and on the creation of data-intensive Web sites and collaborative applications. Brad Myers from the Human-Computer Interaction Institute (CMU) kicks off the program with his keynote address. He describes the results of empirical studies designed to discover the most natural programming paradigms for nonprofessional programmers. Peter Lee, from Carnegie Mellon University and Cedilla Systems Inc., speaks on "Language Technology for Performance and Security," and Philip Wadler, from Bell Labs, provides a tantalizing glimpse of "The Next 700 Markup Languages." Our hot research reviews look at DSLs for programming active networks (Carl Gunter, University of Pennsylvania) and explore how to design and create DSLs using program specialization (Charles Consel, Irisa/University of Rennes).

I look forward to seeing you October 3-5, 1999, in beautiful downtown Austin, Texas.

Sincerely,

Thomas Ball
Bell Laboratories, Lucent Technologies
For the Program Committee