| Previous slide | Next slide | Back to the first slide | View Graphic Version |
The deepest and most serious technology gap in what we are proposing is developing a language for describing the wide variety of different physical mechanisms that may arise in a design project. Each component may be analyzed at a variety of different levels of detail—sometimes beams are viewed as rigid lumped parameter objects, sometimes they are continuous and must be analyzed with finite element methods. Sometimes the mechanical aspects are important, other times the electro-magnetic, or chemical aspects need to be included. How does one combine all of these different ways of looking at a physical object into a coherent, computationally effective structure? When one combines representations of different parts, how does one extract the relevant descriptive information in form that can be combined to produce an efficient simulator.
At a more technical level there are many issues that still need to be addressed with respect to compound document technology. While it appears to be relatively straightforward, although tedious to wrap the active model structure in OLE, OLE was developed business collaboration in mind. This places different demands on the underlying memory system, synchronization structures and exception mechanisms than arise in scientific and engineering computations. Furthermore, the Web based extensions, ActiveX and ActiveMovie are still very young. Although promising, there is still a great deal of uncertainty on issues like: How do changes propagate among shared/linked objects, especially when separated by the Internet? How do these issues change, when mobile computers enter the picture, and parts of the knowledge web, become disconnected?
Finally, we expect that new design management policies will be needed to take full advantage of the increased level of communication and sharing made possible by the active model approach.