Standard ML of New Jersey

Standard ML of New Jersey
Name	Standard ML of New Jersey
Paradigm	Functional, imperative, modular
Designer	Robin Milner, David MacQueen, Mads Tofte
Developer	University of Chicago?
First appeared	1987
Typing	Static, strong, inferred
Implemented in	C (programming language), Standard ML (programming language)

Standard ML of New Jersey

Standard ML of New Jersey is an implementation of the Standard ML (programming language) developed primarily by researchers in the 1980s and 1990s. It became prominent in academic and industrial settings for work on type systems associated with researchers from institutions like University of Edinburgh, University of Cambridge, Massachusetts Institute of Technology, and Bell Labs. The implementation influenced projects at organizations such as Microsoft Research, IBM Research, and AT&T Laboratories and featured in events like the ACM SIGPLAN conferences and the International Conference on Functional Programming.

History

The project originated from the formalization efforts by figures including Robin Milner, Edwin S. (Ed) Brinksma? and collaborators at places like University of Edinburgh and Royal Society groups, building on the earlier ML work from Cambridge University and INRIA. Releases in the late 1980s and early 1990s were announced alongside publications at Conference on LISP and Functional Programming, ACM SIGPLAN-SIGACT Symposium on Principles of Programming Languages, and workshops at European Symposium on Programming. The implementation matured during collaborations with researchers from Aarhus University, University of Oslo, Carnegie Mellon University, and industrial partners such as Sun Microsystems and Digital Equipment Corporation. Over time, it was maintained by principal contributors including Mads Tofte and David MacQueen, and it served as a testbed for type inference research presented at venues like International Colloquium on Automata, Languages and Programming.

Design and Implementation

The implementation emphasized a faithful rendition of the Standard ML (programming language) language semantics originally formalized in documents by researchers associated with University of Edinburgh and INRIA. The compiler architecture combined front-end parsing and elaboration with advanced type reconstruction influenced by work from Robin Milner and later formalizations presented at IEEE Symposium on Logic in Computer Science and ACM SIGPLAN Notices. The runtime incorporated a garbage-collected heap inspired by implementations at MIT and contemporaneous systems such as Squeak (software)? and CMU Common Lisp; the code generator produced efficient machine code comparable to compilers developed at Stanford University and UC Berkeley. Key contributors demonstrated the implementation at conferences like International Conference on Functional Languages and Computer Architecture.

Language Features

The implementation supported the full Standard ML (programming language) feature set including parametric polymorphism described by Robin Milner and algebraic datatypes discussed in papers presented at ACM Symposium on Theory of Computing and European Symposium on Programming. It implemented pattern matching techniques similar to those used in studies at University of Cambridge and module-level abstraction akin to signatures and structures explored by researchers at INRIA and University of Oxford. The type system incorporated principal types and inferred polymorphism related to work from University College London and Princeton University. The system also provided a foreign function interface linking to libraries from Sun Microsystems and integration approaches demonstrated at USENIX workshops.

Performance and Optimization

Optimizations targeted by the implementation drew on research from Carnegie Mellon University and Stanford University into closure conversion, deforestation, and strictness analysis published at venues like ACM SIGPLAN Conference on Programming Language Design and Implementation and International Conference on Functional Programming. The code generation benefited from register allocation strategies developed in collaborations with groups at University of Illinois Urbana-Champaign and University of California, Los Angeles, and memory management improved via garbage collection research presented at International Symposium on Memory Management and Workshop on ML and its Applications. Benchmark comparisons in papers often referenced implementations from Haskell (programming language) communities at University of Glasgow and compiler projects at Oxford University Computing Laboratory.

Tooling and Ecosystem

The implementation shipped with a REPL and interactive tools inspired by environments from MIT Scheme and editors like Emacs and Vi (text editor), with support in integrations demonstrated at SIGCSE and ECOOP sessions. Tooling included pretty-printers, debuggers, and build scripts comparable to tools developed at Microsoft Research and IBM Research. The ecosystem fostered libraries and bindings contributed by groups at University of Tokyo, ETH Zurich, and Technical University of Denmark and was used in coursework at institutions such as Harvard University and Yale University.

Influence and Legacy

The implementation had an outsized impact on subsequent language research and implementations at institutions like Microsoft Research (influencing languages such as F#), Bell Labs projects, and academic efforts at University of Cambridge and University of Edinburgh. Concepts refined through the project informed type-system design in languages like OCaml from INRIA, influenced compiler toolchains at University of Chicago and Carnegie Mellon University, and appeared in textbooks authored by academics at MIT Press and Cambridge University Press. The research lineage extended to graduate programs at Stanford University, Princeton University, and ETH Zurich and continued to be cited at conferences including ACM SIGPLAN and International Conference on Functional Programming.

Category:Programming languages