For Immediate Release
Today marked the long-awaited public release of the Aldor programming language compiler. The release is available from www.aldor.org, currently hosted at the University of Western Ontario (London, Canada).
Aldor was originally developed by a team under the direction of Stephen Watt at the IBM T.J. Watson Research Center at Yorktown Heights, New York. Aldor was available for some time as part of the AXIOM distribution from the Numerical Algorithms Group (NAG), a not-for-profit British corporation. Now a free version has been made available to the public from Aldor.org.
Aldor is a programming language designed to allow both efficient symbolic and numeric computation, and to allow mathematical ideas be expressed naturally. It achieves a powerful blend of functional and object-oriented styles by taking both types and functions as first-class values. Many awkward features of other languages such as C++ and GJ are no longer necessary, as their capabilities arise naturally through simple and uniform combinations of primitives in Aldor. Natural facilities allow programs in Aldor to be linked to libraries written in C++ or Fortran.
The INRIA connections to Aldor are longstanding. Both Stephen Watt (past director of INRIA'se Project SAFIR) and Manuel Bronstein (director of INRIA's Project CAFE) worked together at IBM in the project that gave rise to Axiom. Later, Aldor development continued at NAG under the aegis of the ESPRIT project FRISCO, in which INRIA was a partner. Most recently, INRIA has been a chief contributor to fundamental Aldor libraries, LibAldor and LibAlgebra. LibAldor is a low-level library making certain language features simple to use. LibAlgebra is provides a collection of fundamental mathematical components, including the objects of linear and polynomial algebra.
LibAlgebra, developed primarily by Manuel Bronstein (INRIA) and Marc Moreno Maza (U Lille/Western Ontario), is a robust and diverse library that serves as a base for the differential equation library, Summit, and the library for solving polynomial systems, Delta. Together, these applications comprise several hundred thousand lines of code depending on LibAlgebra.
Ongoing collaboration between INRIA and the University of Western Ontario promises to continue the opportunity for software libraries developed at INRIA to be the basis of serious work in computer algebra based on Aldor.
Ongoing research in programming languages and compilers by Stephen Watt and his group at the University of Western Ontario continues to advance Aldor's power and efficiency.