Mathematical Software: The Next Generation?
Mike Dewar · 2001
For more than ten years, there have been regular predictions of radical changes in the design of mathematical software. Conventional programming languages in general and Fortran in particular were dead, and interactive packages would only survive if they adopted open protocols and became more component-based. These predictions have so far failed to come true and the market for libraries such as NAG on the one hand, and packages such as Matlab, Maple and Mathematica on the other, continues to expand. At the same time mathematical and statistical technology has become more pervasive, often embedded inside application software packages which allow people who have no formal mathematical training to perform sophisticated computations which a few years ago would have required the services of a specialist.