Graphics Reference
In-Depth Information
semantics go beyond a mere rendered image, but for which an image is one pos-
sible view of the AM.
In our simple WPF-2D clock application, the AM contained only the current
time of day, but it could be extended to include alarm-related data (date, choice
of alarm sound, enabling of “snooze control,” etc.) or support for multiple time
zones. Note that the AM need not include any inherent geometry; for example,
there's nothing geometric about the time of day, and the rotation-based analog
display of clock time is simply one way to display that data.
Most applications have a heterogeneous AM containing both nongeometric
and geometric data, and the latter can be further subdivided into
abstract geo-
metric
(not in a form ready for the IM layer) versus
ready for rendering
(in a
form ready for the IM layer, e.g., geometry in the form of a triangle mesh).
Consider the breakdown of the AM of a chess application.
• Nongeometric data would include
-
Current board location of each piece (i.e., the square on which it
resides)
-
Record of each move since the game started, to facilitate export of a
game “transcript”
-
Chess strategy data used by the game to plan its moves
-
Duration of the game in progress, the player whose turn it is, the amount
of time left for her move, etc.
• Abstract geometric data might include
-
Mathematically defined shapes of the pieces (which must be converted
into meshes in order to be made ready for rendering)
-
Motion paths, specified as cubic Bézier curves, to support animation of
the movement of pieces from square to square
• The ready-for-IM geometric data might include
-
Geometry and materials for rendering the chessboard itself
-
Camera definitions for several points of view (if the UI allows the user
to choose from several POVs, e.g., directly overhead, POV from seated
avatar, etc.)
-
Modeling transforms for the pieces (e.g., if the user is able to control
the 3D positions and orientations of the pieces beyond their abstract
locations on specific squares)
Now consider the highly complex AM for a CAD/CAM representation of a
jet airliner, consisting of millions of components, each including geometric, spa-
tial layout, and connectivity/joint data; behavioral data used in aircraft-operation
simulation; part numbers, costs, and supplier IDs used in procurement; mainte-
nance/repair instructions or cautions; and much more. In addition, each compo-
nent “lives” in several organizational systems for the purpose of searching and
filtering; for example, a spatial organizational system might separate components
into regions (e.g., cockpit, main cabin), but a functional one might separate com-
ponents into distinct systems such as electrical or hydraulic.
These databases thus act as a confluence of many types of data, used and
manipulated by a large variety of different systems and applications, of which only
a fraction are “computer graphics” programs.
