2.5 Rotations: As a experiment, enabling Axis-Angle for Bones

The support for this is really quite hacky, and I might disable this later if we cannot get some parts to work nicely. 

Some notes:
* This is currently stored in the same variable that quaternions are stored in, since they both have 4 components. However, in RNA, I've added 2 properties specially for this. 
* There are some shearing issues using certain axes - i.e. (1,1,0) - that will need to be checked on.
* Transform code is really quite temporary for this. Just a quick demo of what can be done...

1 files changed, 6 insertions, 2 deletions

diff --git a/source/blender/blenkernel/intern/armature.c b/source/blender/blenkernel/intern/armature.c
index 95c4c2966c6..0c18817f8be 100644
--- a/source/blender/blenkernel/intern/armature.c
+++ b/source/blender/blenkernel/intern/armature.c
@@ -1986,11 +1986,15 @@ void chan_calc_mat(bPoseChannel *chan)
 	/* get scaling matrix */
 	SizeToMat3(chan->size, smat);
 	
-	/* rotations may either be quats or eulers (no rotation modes for now...) */
+	/* rotations may either be quats, eulers (with various rotation orders), or axis-angle */
 	if (chan->rotmode > 0) {
-		/* euler rotations (will cause gimble lock... no rotation order to solve that yet) */
+		/* euler rotations (will cause gimble lock, but this can be alleviated a bit with rotation orders) */
 		EulOToMat3(chan->eul, chan->rotmode, rmat);
 	}
+	else if (chan->rotmode == PCHAN_ROT_AXISANGLE) {
+		/* axis-angle - stored in quaternion data, but not really that great for 3D-changing orientations */
+		VecRotToMat3(&chan->quat[1], chan->quat[0], rmat);
+	}
 	else {
 		/* quats are normalised before use to eliminate scaling issues */
 		NormalQuat(chan->quat);