diff --git a/GL/draw.c b/GL/draw.c
index 02a4c10..cc9a744 100644
--- a/GL/draw.c
+++ b/GL/draw.c
@@ -35,6 +35,26 @@ void initAttributePointers() {
     VERTEX_POINTER.stride = 0;
     VERTEX_POINTER.type = GL_FLOAT;
     VERTEX_POINTER.size = 4;
+
+    DIFFUSE_POINTER.ptr = NULL;
+    DIFFUSE_POINTER.stride = 0;
+    DIFFUSE_POINTER.type = GL_FLOAT;
+    DIFFUSE_POINTER.size = 4;
+
+    UV_POINTER.ptr = NULL;
+    UV_POINTER.stride = 0;
+    UV_POINTER.type = GL_FLOAT;
+    UV_POINTER.size = 4;
+
+    ST_POINTER.ptr = NULL;
+    ST_POINTER.stride = 0;
+    ST_POINTER.type = GL_FLOAT;
+    ST_POINTER.size = 4;
+
+    NORMAL_POINTER.ptr = NULL;
+    NORMAL_POINTER.stride = 0;
+    NORMAL_POINTER.type = GL_FLOAT;
+    NORMAL_POINTER.size = 3;
 }
 
 static GLuint byte_size(GLenum type) {
@@ -46,6 +66,7 @@ static GLuint byte_size(GLenum type) {
     case GL_INT: return sizeof(GLint);
     case GL_UNSIGNED_INT: return sizeof(GLuint);
     case GL_DOUBLE: return sizeof(GLdouble);
+    case GL_FLOAT:
     default: return sizeof(GLfloat);
     }
 }
@@ -90,7 +111,7 @@ static void _parseColour(uint32* out, const GLubyte* in, GLint size, GLenum type
 }
 
 static void _parseFloats(GLfloat* out, const GLubyte* in, GLint size, GLenum type) {
-    GLubyte i = 0;
+    GLubyte i;
 
     switch(type) {
     case GL_SHORT: {
@@ -107,8 +128,10 @@ static void _parseFloats(GLfloat* out, const GLubyte* in, GLint size, GLenum typ
     } break;
     case GL_FLOAT:
     case GL_DOUBLE:  /* Double == Float */
-    default:
-        for(i = 0; i < size; ++i) out[i] = ((GLfloat*) in)[i];
+        default: {
+            const GLfloat* ptr = (const GLfloat*) in;
+            for(i = 0; i < size; ++i) out[i] = ptr[i];
+        }
     }
 }
 
@@ -123,27 +146,52 @@ static void _parseIndex(GLshort* out, const GLubyte* in, GLenum type) {
     }
 }
 
-/*
-inline void mat_trans_float3_nodiv(const GLfloat* mat, GLfloat* v) {
-    GLfloat x = v[0] * mat[0] + v[1] * mat[4] + v[2] * mat[8] + mat[12];
-    GLfloat y = v[0] * mat[1] + v[1] * mat[5] + v[2] * mat[9] + mat[13];
-    GLfloat z = v[0] * mat[2] + v[1] * mat[6] + v[2] * mat[10] + mat[14];
 
-    v[0] = x;
-    v[1] = y;
-    v[2] = z;
-} */
+/* There was a bug in this macro that shipped with Kos
+ * which has now been fixed. But just in case...
+ */
+#undef mat_trans_single3_nodiv
+#define mat_trans_single3_nodiv(x, y, z) { \
+    register float __x __asm__("fr12") = (x); \
+    register float __y __asm__("fr13") = (y); \
+    register float __z __asm__("fr14") = (z); \
+    __asm__ __volatile__( \
+                          "fldi1 fr15\n" \
+                          "ftrv  xmtrx, fv12\n" \
+                          : "=f" (__x), "=f" (__y), "=f" (__z) \
+                          : "0" (__x), "1" (__y), "2" (__z) \
+                          : "fr15"); \
+    x = __x; y = __y; z = __z; \
+}
 
-void transformToEyeSpace(GLfloat* point) {
+#undef mat_trans_normal3
+#define mat_trans_normal3(x, y, z) { \
+    register float __x __asm__("fr8") = (x); \
+    register float __y __asm__("fr9") = (y); \
+    register float __z __asm__("fr10") = (z); \
+    __asm__ __volatile__( \
+                          "fldi0 fr11\n" \
+                          "ftrv  xmtrx, fv8\n" \
+                          : "=f" (__x), "=f" (__y), "=f" (__z) \
+                          : "0" (__x), "1" (__y), "2" (__z) \
+                          : "fr11"); \
+    x = __x; y = __y; z = __z; \
+}
+
+
+
+inline void transformToEyeSpace(GLfloat* point) {
     _matrixLoadModelView();
     mat_trans_single3_nodiv(point[0], point[1], point[2]);
 }
 
-void transformNormalToEyeSpace(GLfloat* normal) {
+
+inline void transformNormalToEyeSpace(GLfloat* normal) {
     _matrixLoadNormal();
     mat_trans_normal3(normal[0], normal[1], normal[2]);
 }
 
+
 static void submitVertices(GLenum mode, GLsizei first, GLsizei count, GLenum type, const GLvoid* indices) {
     static GLfloat normal[3] = {0.0f, 0.0f, -1.0f};
     static GLfloat eye_P[3];
@@ -242,7 +290,7 @@ static void submitVertices(GLenum mode, GLsizei first, GLsizei count, GLenum typ
 
             GLubyte j;
             for(j = 0; j < MAX_LIGHTS; ++j) {
-                if(isLightEnabled(i)) {
+                if(isLightEnabled(j)) {
                     calculateLightingContribution(j, eye_P, eye_N, to_add);
 
                     contribution[0] += to_add[0];
diff --git a/GL/immediate.c b/GL/immediate.c
index eced866..ab29025 100644
--- a/GL/immediate.c
+++ b/GL/immediate.c
@@ -89,6 +89,14 @@ void APIENTRY glVertex3fv(const GLfloat* v) {
     glVertex3f(v[0], v[1], v[2]);
 }
 
+void APIENTRY glVertex2f(GLfloat x, GLfloat y) {
+    glVertex3f(x, y, 0.0f);
+}
+
+void APIENTRY glVertex2fv(const GLfloat* v) {
+    glVertex2f(v[0], v[1]);
+}
+
 void APIENTRY glVertex4f(GLfloat x, GLfloat y, GLfloat z, GLfloat w) {
     glVertex3f(x, y, z);
 }
diff --git a/GL/lighting.c b/GL/lighting.c
index 064b21c..eb95b17 100644
--- a/GL/lighting.c
+++ b/GL/lighting.c
@@ -21,9 +21,6 @@ void initLights() {
     memcpy(MATERIAL.specular, ZERO, sizeof(GLfloat) * 4);
     memcpy(MATERIAL.emissive, ZERO, sizeof(GLfloat) * 4);
     MATERIAL.exponent = 0.0f;
-    MATERIAL.ambient_color_index = 0.0f;
-    MATERIAL.diffuse_color_index = 1.0f;
-    MATERIAL.specular_color_index = 1.0f;
 
     GLubyte i;
     for(i = 0; i < MAX_LIGHTS; ++i) {
@@ -231,36 +228,52 @@ static float FPOW(float b, float p) {
 void calculateLightingContribution(const GLint light, const GLfloat* pos, const GLfloat* normal, GLfloat* colour) {
     LightSource* l = &LIGHTS[light];
 
-    struct vec3f L, N, V;
+    struct vec3f L = {
+        l->position[0] - pos[0],
+        l->position[1] - pos[1],
+        l->position[2] - pos[2]
+    };
 
-    L.x = l->position[0] - pos[0];
-    L.y = l->position[1] - pos[1];
-    L.z = l->position[2] - pos[2];
+    struct vec3f N = {
+        normal[0],
+        normal[1],
+        normal[2]
+    };
 
-    N.x = normal[0];
-    N.y = normal[1];
-    N.z = normal[2];
+    struct vec3f V = {
+        pos[0],
+        pos[1],
+        pos[2]
+    };
 
-    V.x = -pos[0];
-    V.y = -pos[1];
-    V.z = -pos[2];
+    GLfloat d;
+    vec3f_length(V.x, V.y, V.z, d);
 
     vec3f_normalize(L.x, L.y, L.z);
     vec3f_normalize(V.x, V.y, V.z);
 
-    GLfloat LdotN;
-    vec3f_dot(L.x, L.y, L.z, N.x, N.y, N.z, LdotN);
+    GLfloat NdotL;
+    vec3f_dot(N.x, N.y, N.z, L.x, L.y, L.z, NdotL);
 
-    GLfloat f = (LdotN < 0) ? 0 : 1;
+    GLfloat f = (NdotL < 0) ? 0 : 1;
 
     GLfloat VdotN;
     vec3f_dot(V.x, V.y, V.z, N.x, N.y, N.z, VdotN);
 
-    GLfloat VdotR = VdotN - LdotN;
+    GLfloat VdotR = VdotN - NdotL;
     GLfloat specularPower = FPOW(VdotR > 0 ? VdotR : 0, MATERIAL.exponent);
 
-    colour[0] = l->ambient[0] * MATERIAL.ambient[0] + f * (l->diffuse[0] * MATERIAL.diffuse[0] * LdotN + l->specular[0] * MATERIAL.specular[0] * specularPower);
-    colour[1] = l->ambient[1] * MATERIAL.ambient[1] + f * (l->diffuse[1] * MATERIAL.diffuse[1] * LdotN + l->specular[1] * MATERIAL.specular[1] * specularPower);
-    colour[2] = l->ambient[2] * MATERIAL.ambient[2] + f * (l->diffuse[2] * MATERIAL.diffuse[2] * LdotN + l->specular[2] * MATERIAL.specular[2] * specularPower);
+    GLfloat att = (l->position[3] == 0) ? 1.0f : (
+        l->constant_attenuation / (1.0f + l->linear_attenuation * d) * (1.0f + l->quadratic_attenuation * d * d)
+    );
+
+    colour[0] = att * l->ambient[0] * MATERIAL.ambient[0] + f * (l->diffuse[0] * MATERIAL.diffuse[0] * NdotL + l->specular[0] * MATERIAL.specular[0] * specularPower);
+    colour[1] = att * l->ambient[1] * MATERIAL.ambient[1] + f * (l->diffuse[1] * MATERIAL.diffuse[1] * NdotL + l->specular[1] * MATERIAL.specular[1] * specularPower);
+    colour[2] = att * l->ambient[2] * MATERIAL.ambient[2] + f * (l->diffuse[2] * MATERIAL.diffuse[2] * NdotL + l->specular[2] * MATERIAL.specular[2] * specularPower);
     colour[3] = MATERIAL.diffuse[3];
+
+    if(colour[0] > 1.0f) colour[0] = 1.0f;
+    if(colour[1] > 1.0f) colour[1] = 1.0f;
+    if(colour[2] > 1.0f) colour[2] = 1.0f;
+    if(colour[3] > 1.0f) colour[3] = 1.0f;
 }
diff --git a/GL/matrix.c b/GL/matrix.c
index e9e4703..0c36219 100644
--- a/GL/matrix.c
+++ b/GL/matrix.c
@@ -85,8 +85,8 @@ static void transpose(GLfloat* m) {
 
 static void recalculateNormalMatrix() {
     memcpy(NORMAL_MATRIX, stack_top(MATRIX_STACKS + (GL_MODELVIEW & 0xF)), sizeof(matrix_t));
-    transpose((GLfloat*) NORMAL_MATRIX);
     inverse((GLfloat*) NORMAL_MATRIX);
+    transpose((GLfloat*) NORMAL_MATRIX);
 }
 
 void APIENTRY glMatrixMode(GLenum mode) {
@@ -271,6 +271,10 @@ void glMultTransposeMatrixf(const GLfloat *m) {
     mat_load(stack_top(MATRIX_STACKS + MATRIX_IDX));
     mat_apply(&ml);
     mat_store(stack_top(MATRIX_STACKS + MATRIX_IDX));
+
+    if(MATRIX_MODE == GL_MODELVIEW) {
+        recalculateNormalMatrix();
+    }
 }
 
 /* Set the GL viewport */
diff --git a/GL/private.h b/GL/private.h
index 6f081cd..b64c192 100644
--- a/GL/private.h
+++ b/GL/private.h
@@ -42,27 +42,24 @@ typedef struct {
 } TextureObject;
 
 typedef struct {
+    GLfloat emissive[4];
     GLfloat ambient[4];
     GLfloat diffuse[4];
-    GLfloat specular[4];
-    GLfloat emissive[4];
+    GLfloat specular[4];    
     GLfloat exponent;
-    GLfloat ambient_color_index;
-    GLfloat diffuse_color_index;
-    GLfloat specular_color_index;
 } Material;
 
 typedef struct {
-    GLfloat ambient[4];
-    GLfloat diffuse[4];
-    GLfloat specular[4];
     GLfloat position[4];
     GLfloat spot_direction[3];
-    GLfloat spot_exponent;
     GLfloat spot_cutoff;
     GLfloat constant_attenuation;
     GLfloat linear_attenuation;
     GLfloat quadratic_attenuation;
+    GLfloat spot_exponent;
+    GLfloat diffuse[4];
+    GLfloat specular[4];
+    GLfloat ambient[4];
 } LightSource;
 
 
diff --git a/gl-api.c b/gl-api.c
index 852d42d..a62e717 100644
--- a/gl-api.c
+++ b/gl-api.c
@@ -251,14 +251,6 @@ void APIENTRY glColor4ub(GLubyte r, GLubyte  g, GLubyte b, GLubyte a) {
 
 //== Vertex Position Submission Functions ==//
 
-void APIENTRY glVertex2f(GLfloat x, GLfloat y) {
-    return _glKosVertex3ft(x, y, 0.0f);
-}
-
-void APIENTRY glVertex2fv(const GLfloat *xy) {
-    return _glKosVertex3ft(xy[0], xy[1], 0.0f);
-}
-
 void APIENTRY glRectf(GLfloat x1, GLfloat y1, GLfloat x2, GLfloat y2) {
     pvr_vertex_t *v = _glKosVertexBufPointer();
 
@@ -289,36 +281,6 @@ void APIENTRY glRectiv(const GLint *v1, const GLint *v2) {
     return glRectfv((const GLfloat *)v1, (const GLfloat *)v2);
 }
 
-void APIENTRY glKosVertex2f(GLfloat x, GLfloat y) {
-    pvr_vertex_t *v = _glKosVertexBufPointer();
-
-    v->x = x;
-    v->y = y;
-    v->z = 10;
-    v->u = GL_KOS_VERTEX_UV[0];
-    v->v = GL_KOS_VERTEX_UV[1];
-    v->argb  = GL_KOS_VERTEX_COLOR;
-
-    _glKosVertexBufIncrement();
-
-    ++GL_KOS_VERTEX_COUNT;
-}
-
-void APIENTRY glKosVertex2fv(const GLfloat *xy) {
-    pvr_vertex_t *v = _glKosVertexBufPointer();
-
-    v->x = xy[0];
-    v->y = xy[1];
-    v->z = 10;
-    v->u = GL_KOS_VERTEX_UV[0];
-    v->v = GL_KOS_VERTEX_UV[1];
-    v->argb  = GL_KOS_VERTEX_COLOR;
-
-    _glKosVertexBufIncrement();
-
-    ++GL_KOS_VERTEX_COUNT;
-}
-
 //====================================================================================================//
 //== GL Begin / End ==//
 
diff --git a/include/gl.h b/include/gl.h
index 38e3caa..eb9dc46 100644
--- a/include/gl.h
+++ b/include/gl.h
@@ -436,10 +436,6 @@ GLAPI void APIENTRY glVertex2f(GLfloat x, GLfloat y);
 GLAPI void APIENTRY glVertex2fv(const GLfloat *xy);
 #define glVertex2iv glVertex2fv
 
-/* Non-Standard KOS Primitive 2D Submission.  This will perform no tranformations on the vertices. */
-GLAPI void APIENTRY glKosVertex2f(GLfloat x, GLfloat y);
-GLAPI void APIENTRY glKosVertex2fv(const GLfloat *xy);
-
 /* Primitive 3D Position Submission */
 GLAPI void APIENTRY glVertex3f(GLfloat, GLfloat, GLfloat);
 GLAPI void APIENTRY glVertex3fv(const GLfloat *);