Publications about Dylan#
- LLVM Code Generation for Open Dylan (by Peter Housel at ELS 2020 pdf bib slides video)
The Open Dylan compiler, DFMC, was originally designed in the 1990s to compile Dylan language code targeting the 32-bit Intel x86 platform, or other platforms via portable C. As platforms have evolved since, this approach has been unable to provide efficient code generation for a broader range of target platforms, or to adequately support tools such as debuggers, profilers, and code coverage analyzers.
Developing a code generator for Open Dylan that uses the LLVM compiler infrastructure is enabling us to support these goals and modernize our implementation. This work describes the design decisions and engineering trade-offs that have influenced the implementation of the LLVM back-end and its associated run-time support.
- Extending Dylan’s type system for better type inference and error detection (by Hannes Mehnert at ILC 2010 pdf bib)
Whereas dynamic typing enables rapid prototyping and easy experimentation, static typing provides early error detection and better compile time optimization. Gradual typing provides the best of both worlds. This paper shows how to define and implement gradual typing in Dylan, traditionally a dynamically typed language. Dylan poses several special challenges for gradual typing, such as multiple return values, variable-arity methods and generic functions (multiple dispatch).
In this paper Dylan is extended with function types and parametric polymorphism. We implemented the type system and a unification-based type inference algorithm in the mainstream Dylan compiler. As case study we use the Dylan standard library (roughly 32000 lines of code), which witnesses that the implementation generates faster code with fewer errors. Some previously undiscovered errors in the Dylan library were revealed.
- Automatically generated type-safe GTK+ binding for Dylan (by Hannes Mehnert at ILC 2009 pdf bib)
We present an automated way to get language bindings for GTK+ for Dylan, an object-oriented functional programming language related to Lisp. Dylan supports multiple inheritance, polymorphism, multiple dispatch, keyword arguments, pattern-based syntax extension macros, and many other features. The generated binding is type-safe, no up- and downcasts are needed.
- A domain-specific language for manipulation of binary data in Dylan (by Hannes Mehnert and Andreas Bogk at ILC 2007 pdf bib)
We present a domain specific language for manipulation of binary data, or structured byte sequences, as they appear in everyday applications such as networking or graphics file manipulation. Our DSL is implemented as an extension of the Dylan language, making use of the macro facility. Dylan macros, unlike Common Lisp macros, are implemented as pattern matches and substitutions on the parse tree, and we will show the strengths and limits of this approach for the given problem.
- Secure Networking (by Andreas Bogk and Hannes Mehnert at 23rd Chaos Communication Congress pdf)
We noticed a lack of a decent secure framework for handling network packets, we have designed and implemented major parts of a TCP/IP stack in the high level programming language “Dylan”, focusing on security, performance and code reuse. This paper describes the binary-data library, as well as network night vision.
- Efficient Compression of Generic Function Dispatch Tables (by Eric Kidd - Technical Report Dartmouth College 2001 pdf bib)
A generic function is similar to an overloaded operator, but provides a way to select an appropriate behavior at run-time instead of compile-time. Dujardin and colleagues have proposed an algorithm for building and compressing generic function dispatch tables.
We present several modifications to their algorithm, including an improvement to Pseudo-Closest-Poles and two new algorithms for compressing pole tables. The two new compression algorithms are simple and fast, and one produces smaller output than the original.
- Partial Dispatch: Optimizing Dynamically-Dispatched Multimethod Calls with Compile-Time Types and Runtime Feedback (by Jonathan Bachrach and Glenn Burke - Technical Report 2000 pdf bib)
We presented an approach to gaining back complete class hierarchy information by delaying the construction of dispatch caches until the whole class hierarchy is available at run- time. Run-time call-site caches can then be constructed as specialized decision trees built from disjointness and concrete- subtype operations on actual arguments combined with compile-time inferred types injected into the run-time. Unnecessary decision steps can be avoided and often run-time dispatch can be completely eliminated. We consider this to be a nice half-way house between full static compilation and dynamic compilation which mitigates the runtime expense of separately compiled components while satisfying our implementation constraints of code shareable components, multi-threaded runtime, incremental development, “pay as you go philosophy”, and interoperability with standard tools.
- D-Expressions: Lisp Power, Dylan Style (by Jonathan Bachrach and Keith Playford - Technical Report 1999 pdf bib)
This paper aims to demonstrate that it is possible for a language with a rich, conventional syntax to provide Lisp-style macro power and simplicity. We describe a macro system and syntax manipulation toolkit designed for the Dylan programming language that meets, and in some areas exceeds, this standard. The debt to Lisp is great, however, since although Dylan has a conventional algebraic syntax, the approach taken to describe and represent that syntax is distinctly Lisp-like in philosophy.
- Modern languages and Microsoft’s component object model (by David N. Gray, John Hotchkiss, Seth Laforge, Andrew Shalit and Toby Weinberg - Communications of the ACM May 1998 bib)
As the computer industry’s reliance on component software increases, it becomes increasingly difficult to integrate complex component systems. Modern development environments take a variety of approaches to simplifying this programming problem. This article describes how three programming language implementations provide access to Microsoft’s Component Object Model (COM), and how they attempt to reduce the complexity of writing and using COM objects.
- A Monotonic Superclass Linearization for Dylan (by Kim Barrett and Bob Cassels and Paul Haahr and David A. Moon and Keith Playford and P. Tucker Withington at OOPSLA 1996 PDF bib)
Object-oriented languages with multiple inheritance and automatic conflict resolution typically use a linearization of superclasses to determine which version of a property to inherit when several superclasses provide definitions. Recent work has defined several desirable characteristics for linearizations, the most important being monotonicity, which prohibits inherited properties from skipping over direct superclasses. Combined with Dylan’s sealing mechanism, a monotonic linearization enables some compile-time method selection that would otherwise be impossible in the absence of a closed-world assumption.The Dylan linearization is monotonic, easily described, strictly observes local precedence order, and produces the same ordering as CLOS when that is monotonic. We present an implementation based on merging and a survey of class heterarchies from several large programs, analyzing where commonly used linearizations differ.
- Discovering the way programmers think about new programming environments (by Joseph Dumas and Paige Parsons - Communications of the ACM June 1995 bib)
An interesting usability study of a prototype development environment for the Dylan programming language is presented here. This study’s purpose is to determine just how close the prototype is to developers. New approaches to source code organization and to the relationship between the environment and the application being developed are introduced. An asessment of how effectively the prototype conveys these innovations to Dylan developers is also given, followed by some proposed improvements.