The class comments for java.awt.GraphicsConfiguration don't properly describe
the way that GraphicsConfigurations are implemented on win32.

"The GraphicsConfiguration class describes the characteristics of a graphics
destination such as a printer or monitor.
There can be many GraphicsConfiguration objects associated with a single
graphics device. For example, on X11
windowing systems, each visual is a different GraphicsConfiguration. On PCs and
Macintoshes, the different screen
resolution/color resolution combinations would be different
GraphicsConfiguration objects."

On Windows, we decided (for 1.3) that GraphicsConfigurations should reflect
available PixelFormats. These have nothing to do with screen resolution, and
only apply to the current color depth, not all available color depths.

On X, GraphicsConfigurations are used to draw using different visuals, and on
Windows different GCs can be used to get different graphics capabilities
(double-buffering, compatibility with OpenGL, etc). This is done by using the
Canvas(GraphicsConfiguration) constructor. I don't, however, think that one
could pass in a GraphicsConfiguration representing (as documented) a resolution
and color depth different from the current display's settings and have it work.

It should be documentated that GraphicsConfigurations represent different drawing capabilities in
the current settings of a display device.