Introduction: The widespread adoption of the IEEE/ISO standard for binary floating point arithmetic has brought a measure of order to the chaos of technical computation. Persistent complaints about calculator design, spreadsheet implementations, and about computational software generally demonstrate that general computation suffers from the lack of similar support. Composite arithmetic addresses these broader problem.
However, composite arithmetic could also be a basis for improving technical programs. This paper contrasts the use of floating point arithmetic and composite arithmetic in technical computation.
For the entire collection see [Zbl 0963.00027].