Recursive Programming

ously, and the available memory space is therefore used rather uneconomically. Furthermore—and this is a more serious objection—it is then impossible to call in a subroutine while one or more previous activations of the same subroutine have not yet come to an end, without losing the possibility of finishing them off properly later on. We intend to describe the principles of a program structure for which these two objections no longer hold. In the first place we sought a means of removing the second restriction, for this essentially restricts the admissable structure of the program; hence the name “Recursive Programming”. More efficient use of the memory as regards the internal working spaces of subroutines is a secondary con­ sequence not without significance. The solution can be applied under perfectly general conditions, e.g. in the structure of an object program to be delivered by an ALGOL 60 compiler. The fact that the proposed methods tend to be rather time consuming on an average present day computer, may give a hint in which direction future design might go. Recursive Programming