Edit: Sorry not sure what happened to the title, it got shortened after I submitted the post. Can't seem to change it at this point. It was originally “Help orienting a polygon towards a camera”

I'm following Bobby Anguelov's tutorial on getting a sprite shader set up using a geometry shader approach. I'm using it to draw lines on screen, however I'm having a few issues modifying to fit my needs. First off, this is the shader

[maxvertexcount(6)]
void gs_line(line InVertex input[2], inout TriangleStream<OutVertex> output)
{

	float3 inVertex1Pos = input[0].pos.xyz;
	float3 inVertex2Pos = input[1].pos.xyz;

	//get the vector that the line represents
	float3 lineVec = normalize(inVertex1Pos - inVertex2Pos);

	//get the vector looking at camera
	float3 lineCenter = (inVertex1Pos + inVertex2Pos) / 2.0f;
	float3 towardsCamVec = normalize(CameraPosition - lineCenter);

	//get the vector that is the up direction from our view point (the normal of the plane made by line and towardsCamVec)
	float3 localUpVec = normalize(cross(lineVec, towardsCamVec));

	//compose our line out of two triangles
	OutVertex v[6];
	v[0].pos.xyz = inVertex1Pos + LineThickness * localUpVec;
	v[0].color = input[0].color;
	v[1].pos.xyz = inVertex2Pos + LineThickness * localUpVec;
	v[1].color = input[1].color;
	v[2].pos.xyz = inVertex1Pos - LineThickness * localUpVec;
	v[2].color = input[0].color;
	v[3].pos.xyz = inVertex1Pos - LineThickness * localUpVec;
	v[3].color = input[0].color;
	v[4].pos.xyz = inVertex2Pos + LineThickness * localUpVec;
	v[4].color = input[1].color;
	v[5].pos.xyz = inVertex2Pos - LineThickness * localUpVec;
	v[5].color = input[1].color;


	//move to clip space 
	for (int i = 0; i < 6; i++)
	{
		v[i].pos.w = 1.0f;

		v[i].pos = mul(v[i].pos, WorldTranslationMatrix);
		v[i].pos = mul(v[i].pos, View);
		v[i].pos = mul(v[i].pos, Projection);
	}

	//output all the vertices upstream
	output.Append(v[0]);
	output.Append(v[1]);
	output.Append(v[2]);
	output.RestartStrip();
	output.Append(v[3]);
	output.Append(v[4]);
	output.Append(v[5]);
	output.RestartStrip();
}

I have two issues with this code. One is that there is a part that I'm not 100% clear on and the second is that I'm having trouble modifying the code to fit my needs. Starting with the first issue, the code above works well, however what I don't get is this - I orient the two triangles towards the camera before I translate the sprite. Meaning that after I translation, shouldn't the sprite now be facing the old direction of where the camera was before it was translated? In other words shouldn't the translation mess up the sprite effect, since it's done after the fact that I oriented it? Shouldn't I first translate it and then orient the sprite from the new location. However it's not the case, the code works as expected and the sprite always remains facing the camera. Why is that?

The second issue I have is I'm trying to add rotations where I can specify a rotation matrix to rotate the sprite around something but for some reason it's messing up the sprite effect. The triangles become misaligned and stop facing the camera.

I just add a rotation multiplication before the translation like this

//move to clip space 
	for (int i = 0; i < 6; i++)
	{
		v[i].pos.w = 1.0f;


		v[i].pos = mul(v[i].pos, WorldRotationMatrix);
		v[i].pos = mul(v[i].pos, WorldTranslationMatrix);
		v[i].pos = mul(v[i].pos, View);
		v[i].pos = mul(v[i].pos, Projection);
	}

How can I accomplish this?

P.S If you need I will post some screenshots of the artifact I'm getting but when I add rotations the sprite basically looks like a flat 2d polygon that I can walk around, the camera orientation gets messed up and the sprite effect disappears (meaning instead of having constant thickness when I walk around it, the line gets thinner when I view it from the side and wider when the polygon is viewed from the front).

Any help is appreciated with either issue. Thanks!