Complex number functions for std.math

[Anders F Björklund]

James Dunne wrote:

> Here's what I've devised for a floating-point type system similar to D's:

I think you should also include half (16-bit) and quad (128-bit) floats.

--anders