I am learning WebGL by "tinkering" with various tutorials (both math and code), the spec, my own imagination, etc.
Currently, I am working deeply through Gregg Tavares' WebGL Fundamentals Tutorial. It is awesome, and below you will find some of my experiments based on his work.
The Orthographic 3D example uses 4x4 matrices to implement 3 dimensional transforms. It applies x, y and z axis rotation each with a separate matrix, along with x, y and z translation using a single matrix and x, y and z scaling using a single matrix to apply the transformations.
It is orthographic because it does not use a perspective matrix to give the 3D image perspective. Perspective would cause the parts of the image that are further down the line of sight to appear smaller.