Influence of the theory of algorithms on algorithmic practice

Abstract

Nowadays the majority of algorithms explicitly constructed and used by mankind in his activity consists of computer programs (see [Knuth 74], [Knuth 74a]). The scale of programming activity can be partly realized by examining the problems created by management in that field, see [Brooks 75]. So programming can be identified with algorithmic practice and theoretical programming in a broad sense — with theory of algorithms (for example Gödel’s incompleteness theorem may be considered as a theorem of theoretical programming, cf. [Gluš 79]). However the term “theoretical programming” is commonly used in a different sense: for a part of algorithm theory concentrated around the relation between a program (as a syntactic, uninterpreted object) and its semantic meaning (value). Naturally it is typical for theoretical programming to be interested in topics motivated by practice and not considered by classical theory of algorithms, such as parallel programming (see [Kot 74]) or data structures (see [Scott 70]). Nevertheless the “common part” of theoretical programming and the theory of algorithms is large, and the classical theory of algorithms has definitely influenced programming practice. This influence was ideological in nature, however, because theorems can not all be used for practical purposes. Here we will try to trace this influence by discussing related results of algorithm theory.