> The camera is essential. When you first enter a scene you must look at it
> from some particular viewpoint. If there is no camera in the scene, the browser
> must make something up. Sometimes this is ok, but many times it is not. Consider
> walking into a room. If there is no camera, the browser could easily decide to look
> at the whole room --- from the outside. Obviously the browser will provide
> interactive control of the camera after the scene is entered.
All right, that's a good point. It'd be nice, though, if the spec would
give an indication that the camera position is not to be considered
binding for static vrml files (ones where the/a server is not actively
involved in updating the scene in response to user actions, and the
client has to do everything itself).
> o Is the primary use of VRML ever going to be as static .wrl files?
> It will certainly be one important use. Consider airline companies looking
> at the design of a new Beoing aircraft (although one can imagin wanting
> to see the flaps move I suppose).
If I were looking at aircraft designs, I'd like to get them in a
heavy-duty CAD format of some sort. I don't think engineering
visualization was ever seriously a purpose of VRML, and if it was it
shouldn't have been. No one format can realistically cover all the
extremes of 3d modelling and imaging applications, and to try would only
make it either virtually unimplementable or a woeful failure at its goals.
-- James Deikun, University of Pittsburgh #include <std_disclaimer.h>