In my office, I have some pictures of personal significance hanging on
the walls, cabinets, computers, etc., but the technically interesting
illustrations are all in books, papers, magazines, etc. stored in
filing cabinets, book shelves, and desk drawers. I think the designer
needs the freedom to define a coordinate system (and scale) for the
scene *and* a coordinate system (and scale) for each object in the
scene, and they'll probably differ.
When I'm walking through my office, I mentally use the "walls and
furniture navigation mode." When I'm sitting in my office reading a
technical journal, I use "journal navigation mode." While I'm reading
that journal, if I want to look up a definition in a medical
dictionary, I switch to another mode, then switch back to the journal.
Each of these cognitive modes might have a different coordinate system
attached to the VRML scene which lets me navigate and find the
information I seek. I believe that the scene designer should be able
to make an artistic decision by specifying the coordinate system and
scale appropriate to the subject and context at hand, but the user (via
the browser) should be able to override that artistic decision.
Movement mode within the scene should likewise be specified by the
scene designer, with optional user override. This allows the human
view, the gods-eye view, the fly-on-the-wall view, the kangaroo view,
-- Ken Jenks, NASA/JSC/SD5, Space Biomedical Research Institute
firstname.lastname@example.org (713) 483-4368
"One who conquers others is great.
One who conquers himself is mighty."
-- Lao Tze