Critique: DSPL/AIR-CYL
Brown, Chandrasekaran ^*

      The main discussion in this paper is around understanding design, or pieces thereof. The breakdown of classes of design and assumptions made seemed intuitive. So far this presents design, or rather possible perspectives of design which are poorly understood, far better than the other system papers we've covered. This is not meant to be critical of the other papers, but rather call out that they were mostly concerned with the particular method(s) used by the system rather than with an overall design philosophy and a way to represent this in a way that a computer system could take advantage of. Then again, this very paper is referenced in PRIDE, DPMED and Dominic.

      While the example used, namely design of air cylinders, demonstrates well the complexity of the problem this system and methodology can be used on, it was lengthy. The paper did a good job of keeping focus on design process but one can only hold an interest for air cylinder dynamics and intricacies for so long.

      Where this paper mentions performance degradation I find the statement "...that as designs get to be only just class 3 the performance of the system will degrade..." unclear. Is this meant to say that attempts by systems to do class 1 design and class 2 design would be more successful than class 3 as the number of dependencies grow?