>>From a point of view of someone designing a VR room or building it
>>would make a lot of sense to use a coordinate system that is natural
>>for architectural plans, i.e. to have X and Y in the plane of the plan
>>and Z upwards in the vertical direction. For a right handed coordinate
>>system that gives:
>>
>> X increasing to the right
>> Y increasing into the screen
>> Z increasing up the screen.
Dave Raggett's point is well taken. I won't argue about what x, y, or z
should indicate. Whatever people agree on. However, rotations of objects
for placement in a virtual world where one prefers +z as up would be an
easily implemented translation.
Using this system, every 3D object would have a center based on it's local
coordinate position and scale. This would allow a Graphical Interface
system to make use of `place the center of the object here` routines
to aid in rapid placement of objects. Scaling could be automatic to
the scale of the environment or the scale of the newly placed object.
A total environment could then be treated as an object in and of itself
and made available via a VRML transaction request.
I think it is obvious to me that VRML can only be a *transaction*
language. That is to say, that building a Virtual Reality environment
will best be done using a CAD-like environment, not a scripting
environment. Sure, everything will get translated into VRML but the
actual placement and creative process will most likely be through
a Graphical, perhaps Virtual also, User Interface.
HTML, I'm not belittling it by saying this, is really a very
sophisticated form of punctuation that is appropriate to character
based information. I think file formats such as OOGL have much more
bearing on the issue of VRML than perhaps people realize.
My idea is to allow for a multi-object *combination* that could then be
treated, by other VR environments, as one object itself. Something like:
I have a VR house, I need a really cool garage complete with classic
Corvette. I found this one on the net. I'll stick my new garage centered
right here next to the house.
Movement, such as boat navigation, is an entirely different issue and could
make use of a standard Cartesian coordinate system to make it's calculations.
>> The IEEE Distributed Interactive Simulation (DIS) protocol standard has
>> strictly defined coordinate system that many virtual worlds &
>> autonomous agents comply with. It works well. Information on DIS & protocol
>> reference manuals are available from the Institute for Simulation &
>> Training at the University of Central Florida.
I always prefer to stick to standards whenever possible. Is there further
information on how to obtain the above info, perhaps via the Internet?
Thanks for your time.
Bob.
-- Robert K. Foster Analyst / Designer / Programmer Mid-Michigan MRI, Inc., Lansing, MI, USA rkf@rad.msu.edu