Scheme and Functional Programming 

Abstract: This paper describes an approach for compiling Scheme's tail calls and first-class continuations to JavaScript, a dynamic language without those features. Our approach is based on the use of a simple custom virtual machine intermediate representation that is translated to JavaScript. We compare this approach, which is used by the Gambit-JS compiler, to the Replay-C algorithm, used by Scheme2JS (a derivative of Bigloo), and Cheney on the MTA, used by Spock (a derivative of Chicken). We analyse the performance of the three systems with a set of benchmark programs on recent versions of four popular JavaScript VMs (V8, SpiderMonkey, Nitro and Chakra). On the benchmark programs, all systems perform best when executed with V8 and our approach is consistently faster than the others on all VMs. For some VMs and benchmarks our approach is moderately faster than the others (below a factor of 2), but in some cases there is a very large performance gap (with Nitro there is a slowdown of up to 3 orders of magnitude for Scheme2JS, and up to 2 orders of magnitude for Spock).

Abstract: Many languages include a syntax for declaring programmer-defined structured data types, i.e., structs or records. R6RS supports syntactic record definitions but also allows records to be defined procedurally, i.e., via a set of run-time operations. Indeed, the procedural interface is considered to be the primitive interface, and the syntactic interface is designed to be macro expandable into code that uses the procedural interface. Run-time creation of record types has a potentially significant impact. In particular, record creation, field access, and field mutation cannot generally be open coded, as it can be with syntactically specified records. Often, however, the shape of a record type can be determined statically, and in such a case, performance equivalent to that of syntactically specified record types can be attained. This paper describes an efficient run-time implementation of procedural record types, discusses its overhead, and describes a set of compiler optimizations that eliminate the overhead when record-type information can be determined statically. The optimizations improve the performance of a set of representative benchmark programs by over 20% on average.

Abstract: Quasiquotation in Scheme is nearly ideal for implementing programs that generate other programs. These programs lack only the ability to generate fresh bound identifiers, as required to make such code-manipulating programs hygienic, but any Scheme programmer knows how to provide this ability using gensym.In this tutorial we investigate hygienic quasiquotation in Scheme and in languages influenced by Scheme. Stepping back from implementation issues, we first identify the source of the freshness condition in the semantics of a hygienic quasiquotation facility. We then show how gensym is needed to break a meta-circularity in interpreters and compilers for hygienic quasiquotation. Finally, following our recent work, we present a type system for hygienic quasiquotation that supports evaluation under dynamic λ-abstraction, manipulation of open code, a first-class eval function, and mutable state.This tutorial outlines Scheme programs implementing an interpreter, a compiler, and a macro for hygienic quasiquotation.

Abstract: Self-reconfigurable, modular robots are distributed mechatronic devices that can change their physical shape; modules are programmed individually but must coordinate across the robot. Programming modular robots is difficult due to the complexity of programming a distributed embedded system with a dynamically evolving topology. We are currently experimenting with programming-language abstractions to help overcome these difficulties. This tutorial describes a few such experiments using Scheme to control simulated modular robots.