登入選單
返回Google圖書搜尋
Why is it Difficult to Program in Von Neumann Languages?
註釋Often the assertion is made that programs in von Neumann languages like Fortran, Pascal, or C are difficult to synthesize, analyze, modify, or extend. While the assertion appears to be correct, usual arguments given in this support are not very convincing. This Note focuses on one part of this assertion, namely, that programs in von Neumann languages are difficult to synthesize, and develops a new argument in this support. Frequently it is necessary to express in a programming language statements of the form 'each of functions f1 ..., fn are to be computed for the same entity.' The presence of destructive assignment in von Neumann languages gives rise to the possibility that programs modify their inputs. This makes it difficult to express statements of the above form in them. Applicative languages such as pure Lisp or pure Prolog do not contain destructive assignment, so such statements can be expressed very easily in them.