While I do no agree with some of the opening statements, the authors make some good points about functional reasoning in general. In particular, the problem of the term function not having a standard definition conflicts with function being the most important concept in determining the basic characteristics of a design. From the numerous CAD occurrences I assumed a predominantly mechanical design view of functional reasoning.

      I think that an argument against functional design playing a central role in ensuring design quality and innovativeness. The functions that are provided are usually specified. How they are provided is central. Maybe this just a case of misleading presentation instead of being a point to be argued. I had trouble accepting the explanation for the trouble representing functions that do not transform something. The example given is that of a fixture. It can be said that a fixture does perform transformation. It takes all external forces that would otherwise move the fixed object and redirects them to some more immovable structure. This can be further illustrated by exerting some force beyond the capacity of the fixture to redirect force and the fixture breaks.

      When introducing existing research a list of focus points is given. I thought it odd that function representation was not in this list, especially since the first example system given is Functional Representation. could the authors have meant the first item in the list, "definition of function", to mean representation?

      I know it should come as no surprise by now, but the authors' work is given over four times more space in the article than the average of the other systems included in the survey.

      "Practicing designers using the FBS modeler typically report that..." "None of these comments is specific to the FBS modeler..." Now doesn't that seem a little fishy? Also, one of the comments, "Lack of experience generally makes explicitly describing functions difficult." is too vague. Lack of experience with what? It could be the design system, the domain, design in general, etc., all of which would make things difficult.

      Despite all of this I found this survey article very informative and interesting. There are a good number of systems and approaches described, and in enough details that some comparison can be made. While some space could have been saved by omitting some of the authors' own work, that work did illustrate some interesting points, which the authors were not shy in sharing.