Appeal No. 97-2477 Application 08/212,908 Moreover, what is more important is what the appellant’s specification regards as a behavioral description as opposed to a structural description. Also, for the purpose evidently intended by the examiner, it is inappropriate not to include these references in the stated ground of rejection. See, e.g., In re Hoch, 428 F.2d 1341, 1342 n.3, 166 USPQ 406, 407 n.3 (CCPA 1970). Accordingly, these references have not been considered. As a further backup position, the examiner cited to Watkins. (Answer at page 8). Specifically, the examiner referred to column 2, lines 46-63 of Watkins which simply describe in general terms that simulation may be performed at several different levels, i.e., the component-level model which describe the exact behavior of a specific component such as a gate or a transistor, the high level behavioral model which provides a logical or mathematical equation or set of equations describing the behavior of the component, viewed as a “black box”, and circuit-level models which comprise a plurality of component-level or behavioral-level models. This does not demonstrate that Dunn’s specific gate icons and logic circuit icons are behavioral descriptions. 10Page: Previous 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 NextLast modified: November 3, 2007