Geometry Transformation Pipeline
To create an image of a 3D scene, OpenGL has to go through several
steps, starting with a set of vertices that define a form, and ending
with many forms placed together in a scene that has a viewpoint, width,
height and depth and appears on a computer screen.

These steps, called the "Transformation Pipeline" map the geometry from one
"space" to another.  For instance several 3D models, each centered at the origin,
may be mapped by glTranslate() into a "world space" where each model has a
position relative to the world space origin.

Below is a summary of the steps in the Transformation Pipeline.


1) Make the geometry
    define shapes
    build geometry in the "model space" (or "object space")
    vertices are typically centered at the origin

    functions:
        glBegin()
        glVertex()
        glEnd()


2) Position the geometry
    define camera and geometry positions
    place geometry in the "world space" and the "eye space"
    vertices have new coordinates relative to the world origin
    vertices are positioned relative to the eye (gluLookAt)
    the MODELVIEW matrix

    functions:
        gluLookat()
        glTranslate()
        glRotate()
        glScale()


3) Project the geometry
    define camera lens
    place geometry in the viewing volume aka the "clip space"
    the PROJECTION matrix

    functions:
        gluPerspective()
        glFrustum()
        glOrtho()


4) position the rendering on screen
    define the shape and position of the rendering
    place geometry in the "screen space"
    the Viewport is not a matrix (array of four numbers: x,y,w,h)

    functions:
        glViewport()