On the Criteria To Be Used in Decomposing Systems into Modules

"On the Criteria To Be Used in Decomposing Systems into Modules"Information Hiding에 관해 얘기할 때 항상 인용되는 David Parnas의 paper다. 전에 읽을 때는 그저 Information Hiding의 개념을 처음으로 도입한 논문이구나 하고 생각했는데, Original wiki의 Information Hiding 엔트리를 살펴보다가 링크가 걸린 "Abstraction, Encapsulation, and Information Hiding"이란 글에서 다음과 같은 내용을 발견하였다.

"Hiding information", in and of itself, was not new. For that matter, the isolation of difficult and/or likely-to-change design decisions in modules was also not new. (Dijkstra had done this earlier in his implementation of the "THE"-Multiprogramming System.) The significance of Parnas’s 1972 article on software module specification lay in two areas:

– His advocation and specification of the (then innovative) technique of basing system modularization on design decisions. (You would have to say that the article presented a significantly different view of Dijkstra’s "levels of abstraction" approach.)

– His use of the term "information hiding". Virtually every article which mentions the topic traces its origin to [Parnas, 1972b].

Obviously, Parnas did not say all information hiding is good, nor did he say that all information hiding techniques are equally useful. He was identifying a particularly pragmatic approach to information hiding.

댓글 달기

이메일 주소는 공개되지 않습니다.

이 사이트는 스팸을 줄이는 아키스밋을 사용합니다. 댓글이 어떻게 처리되는지 알아보십시오.