Answer by rcober

3d animation is complex - there is always confusion. A 3d model is a collection of vertices. Typically called a mesh. One approach to animate it is called skeletal animation. You create a 3d "skeleton" that consists of joints and bones. (The joints really just represent transformation and rotation matrices used to deform the vertices at render time) This skeleton is the rig. It can also contain ik controls to help pose it. The skeleton is posed in various keyframes to create animations. walking, running, jumping... This is more efficient than keyframing the entire list of verts (like old quake md2) because the gpu can be used to perform the bone transformation and rotations. The verts are already stored on the gpu in a vbo - this is the t-pose. The gpu shader is the "skinned mesh renderer component". In order for the skeleton to be used to deform the mesh, each bone must be associated with the vertices it affects in the mesh. This process is sometimes called "skinning a mesh". In Mecanim, you can take an animation built for a different mesh/skeleton and retarget it to any mesh. This relies on the fact that both skeleton rigs are humanoid and similar. This is very useful and innovative. Hope that helps clarify things for you. The reason the vocabulary is so varied is because different tools have evolved that call it different things and use different approaches.