I greatly enjoyed Gero's introduction. The brief glimpses into design history was a refreshing change.

      It was interesting to to see designers be called change agents. "Change Agent" is a term I've only recently come in contact with and only in a business context. ( www.FastComapany.com - Search for Change Agent )

      I wonder why the last mentioned goal of design research is "the potential automation of some design tasks". It seems that many, if not all, design tasks could be sought to be automated.

      Some of the terminology and wording is confusing throughout the article. What a "goal variable" is wasn't very clear. It is odd to say an artifact produces functions. The examples given for B_s and B_e are very similar.

B_s: light flux transmitted, the ventilation rate, and the various solar gains
B_e: light transmitted, ventilation rates, and solar collection

      In Gero's gives a model of design, F -> D but says there is no direct transformation from F to D. Being the overly optimistic person that I am, I fail to see why this is true. A brief explanation would have been nice.

      Another example model of design, F -> S is given but it seems incomplete. shouldn't it be F -> S -> D. I ask since the manner in when it is presented, it is said that the suspect partial model is another model. This follows a description of structure (S) as it relates to design description (D).

      It is unclear how variables for which there is insufficient knowledge to produce specific values for are transformed into requirements.

Some random comments:
If only software had Hammurabi's Code for quality...
Gero's scopious vocabulary inspires admiration bordering veneration.