Monday, October 4, 1999

8:45 am - 10:00 am Invited Talk

Language Technology for Performance and Security, or, Making Life Better, Not Just Easier

Peter Lee, Carnegie Mellon University and Cedilla Systems Incorporated


Modern languages strive to make programming easier. However, the real importance of modern language technology does not lie merely in ease of use. The same design principles that enable programs to be constructed more easily also lead to improved performance and safety. In order to illustrate this, I will give an introduction to proof-carrying code. PCC allows exceptionally high levels of performance and safety but for practical reasons depends critically on the design principles that underlie better languages for programmers.

Peter Lee is an Associate Professor of Computer Science at Carnegie Mellon University. His approach of applying theoretical ideas in programming language design to practical systems has led to numerous research contributions in the areas of programming language design, compiler technology, networking, and operating systems. Most recently, he has focused his attention on developing Proof-Carrying Code, a technique which uses program verification to enhance the performance and safety of mobile code. He is a principal investigator for the DARPA-sponsored Fox Project and is also the co-founder and president of Cedilla Systems Incorporated.

10:00 am - 10:30 am Break

10:30 am - 11:45 am DSLs and Monads

Session Chair: Paul Hudak, Yale University

DSL Implementation Using Staging and Monads
Tim Sheard, Zine-el-abidine Benaissa, and Emir Pasalic, Oregon Graduate Institute

Monadic Robotics
John Peterson, Yale University, and Greg Hager, The Johns Hopkins University

11:45 am - 1:30 pm Lunch (on your own)

1:30 pm - 3:00 pm Hot Research Review

Session Chair: Todd Proebsting, Microsoft Research

A Methodology for Designing Domain-Specific Languages Using Program Specialization

Charles Consel, Irisa/University of Rennes


Domain-specific languages are mainly being developed in isolation. Furthermore, the difficult task of designing, structuring, and implementing a DSL requires expertise in multiple areas. DSLs can succeed only if development methodologies and tools are made available.

In this talk I give an overview of a methodology for developing DSLs. I also demonstrate how program specialization can map DSL interpreters into efficient (possibly just-in-time) compilers. The presentation is illustrated by concrete examples.

Charles Consel is a professor of computer science at the University of Rennes Irisa/Inria. He leads the Compose group at Inria. His group studies partial evaluation, a program transformation approach aimed at specializing programs with respect to given execution contexts. The work has been carried out with a program specializer for C called Tempo. This system has been successfully used in various applications such as operating systems and scientific code. A complementary research project is domain-specific languages: a software development approach that provides high productivity, easy maintenance, and improved safety (without giving up perfor-mance, thanks to partial evaluation). His work on programming languages, software engineering, and operating systems has led to many publications in major conferences and journals (POPL, PLDI, OOPSLA, ASE, SOSP, TOPLAS, ACM Surveys).

3:00 pm - 3:30 pm Break

3:30 pm - 4:30 pm Embedded Languages

Session Chair: Michael Schwartzbach, University of Aarhus

Domain-Specific Embedded Compilers
Daan Leijen and Erik Meijer, University of Utrecht

Verischemelog: Verilog Embedded in Scheme
James Jennings and Eric Beuscher, Tulane University

6:00 pm - 7:00 pm Conference Reception

8:00 pm - 11:00 pm Birds-of-a-Feather Sessions

Sunday Technical Sessions

Tuesday Technical Sessions