NumberXaero

Read up on orthographic projections, the whole point is that things wont change in size, the far and near plane values are probably culling the triangle after you move back.

I dont see the directional light being given a direction vector, and I dont see anywhere where youve given the point light a position. Did you set either?

Theres not a lot to learn about xml, but a good starting place would be the tinyxml lib/docs
http://www.grinninglizard.com/tinyxml/

void EnableOpenGL(HWND hwnd,HDC * hdc,HGLRC * hrc); <-- BIG L


end of file

void EnableOpenGl(HWND hwnd,HDC * hdc,HGLRC * hrc) <-- little l

Are you updating your projection matrix?

Where is "bulletIterator" defined? I dont see it, I see "bulletIT", is bulletIterator a typo? or another variable that shouldnt be there?
I would call the line "(*bossIT).hit();" before doing any list clean up just so that the logic of your if statement collision check is done before things start getting deleted and moved around. erase removes the current element and returns an iterator to the next, which is end() if you erase the last element.

   for (std::vector<bullet>::iterator bulletIT = aBullet.begin(); bulletIT != aBullet.end(); ) // Loops around each bullet
   {
	if((*bossIT).xPosition() + 32 >= (*bulletIT).xPosition() && (*bossIT).xPosition() <= (*bulletIT).xPosition() + 10 && // Detects Collision
		 (*bossIT).yPosition() + 32 >= (*bulletIT).yPosition() && (*bossIT).yPosition() <= (*bulletIT).yPosition() + 10)
	{
		 (*bossIT).hit();
	 bulletIT = aBullet.erase(bulletIT);   // update iterator with next if collision
	}
	else
	   ++ bulletIT;   // no collision go to next bullet
   }

In my 3D studio max exporter I have a flag that makes Y up because Max's up is +Z. What the flag does is applies a +90 rotation about the +X axis (right handed) to the object (or models skeleton root) transform, this rotates Max's +Y (my Max forward convention) to align with opengl +Y up, and swings the Max -Z (down) up to become the opengl forward (-Z is forward in opengl for me). If the objects transform is translated away from the origin you may have to swap its z and y, and then negate the new z.

There are many places where you can solve it. Flag it on export if you can. Flag it on the model when loading. Manually rotate the object before exporting. I wouldnt mess with opengl, Id rather tweak the data. I do it this way because it doesnt have an impact on animations, its just an additional transform at the root so that objects are not on their side (or face down on floor). Overall its just a pain.

I think maya is Y up by default but can be changed, Max doesnt allow it to be changed, in case you do import into Maya, initially you might have the same problem as with opengl.

The \Samples and \Documentation folders are pretty good places to start, what are looking for?

You should be able to just apply a scale to the root joints local transform and have it propagate down through all the child joints world transform calculations. Then just use each joints world transform with the bind pose matrices as usual.

I guess you are referring to the snowman example, where you can hit the snowmans nose with a snowball and the nose falls to the ground.
In that example they build the snowman, and set the nose shapes filter data with a custom detachable flag. In the contact report callback, when nose shape is hit by a snowball (also has a custom flag) the nose shape is saved in a list for detaching. At some other point at the end of the frame, the list of detachables is processed, each nose shape (created initially as part of the compound dynamic snowman?) is recreated as dynamic actor by itself.
The nose material, render model and transform are copied to the new dynamic actor. The dynamic nose is then dropped into the scene with a small force applied while the static nose is deleted.
So the original compound nose shape is detached from the snowman, and copied to a dynamic actor and allowed to move freely in the scene.

What physx version are you using?

If you have camera information of some sort that this frustum belongs to, you could use the forward, up, and right vectors to get 3 of the normals, negate them to get the other 3.
Then use these 6 normals with the view position and the l, r, t, b, n, f values to compute positions on the planes.

Whats the viewport set to?

Try a Quaternion from axis angle. You have the angle, you know you want to rotate about the x axis.

angleRad *= 0.5
q.w = cos(angleRad);
q.v = nrmAxis * sin(angleRad)

then combine this quaternion with the existing one, and use the result.

The info you need seems to be there

              	// transform matrix
        *TM_ROW0 0.6726   0.7400	-0.0000
        *TM_ROW1 -0.4469  0.4062	0.7971
    	*TM_ROW2 0.5898   -0.5361   0.6039
    	*TM_ROW3 546.8461 -133.5005 375.1562 	// translation

        *TM_POS 546.8461	-133.5005   375.1562   // translation pulled from matrix

        *TM_ROTAXIS -0.7092 -0.3138 -0.6314   // resulting axis from convert matrix to axis/angle
        *TM_ROTANGLE 1.2224       			// angle part of axis/angle conversion (in radians it seems)

    	*TM_SCALE 1.0000	1.0000  1.0000   	// no scale (probably because its a camera)

you could take this info and convert it to angles if you want, look up matrix to yaw pitch roll.

Need to see more code to know exactly why, but the place you are running that loop, does not have "inventory" in its scope, so it cant "see it" and therefore cant print its contents.