dcs program

Programming Languages

The group is investigating factors associated with programming language design and implementation, in collaboration with IBM. Currently two themes are active.

Rexx

Rexx is an acronym which stands for "Restructured Extended Executor Language". Rexx is a procedural language that allows programs and algorithms to be written in a clear and structured way. Rexx doesn't look that different from any other procedural language. Here's a simple Rexx program:

/* Count some numbers */
say "Counting..."
do i = 1 to 10
  say "Number" i
end

What makes Rexx different from most other languages is that it is also designed to be used as a macro language by arbitrary application programs. The idea is that application developers don't have to design their own macro languages and interpreters. Instead they use Rexx as the macro language and support the Rexx programming interface. NetRexx is a programming and scripting language which has been designed to be a simple, effective, and complete alternative to the Java language. With NetRexx, you can create applications and applets for the Java environment faster and more easily than by programming in Java - and you can compile or interpret them, as appropriate. IBM Object Rexx is an object-oriented programmming language suited for beginners as well as experienced OO programmers. It is upward compatible with previous versions of classic Rexx and provides an easy migration path to the world of objects. Because it can be used with Rexx conventional programming, Object Rexx protects your investment in existing Rexx program code. It provides many programming interfaces to existing applications, such as DB2, C, and C++ applications.

Decimal Arithmetic

Most computers today support binary floating-point in hardware. While suitable for many purposes, binary floating-point should not be used for financial, commercial, and user-centric applications because the decimal data used in these applications cannot be represented exactly using binary floating-point. The problems of binary floating-point can be avoided by using base 10 (decimal) exponents and preserving the precision of numbers. We are investigating a decimal arithmetic which achieves the necessary accuracy while conforming to the relevant ANSI and IEEE standards for integer, decimal fixed-point, and decimal floating-point arithmetic.