Abstract
Several issues related to solid modeling are examined. Fundamental background definitions and concepts are presented, concentrating especially on dual-representation solid-modeling environments. An ideal architecture is hypothesized and it is shown that the current state of the art leads to one of two compromise architectures in practice. Also studied are lower level development strategies, including schemes that trade-off, to varying extents, the goal of minimizing flexibility in geometric forms as provided to the designer. The need for software architectures that promote development of application functions has led to the development of standard modeling-system interfaces at the procedure call level. Several open issues are identified. Some prevent the complete realization of the ideal architectures, whereas others arise as areas potentially benefiting from a flexible dual-representation environment.

This publication has 28 references indexed in Scilit: