Object constrained to ground where other is pointing

[ourfriendbrutus]

So I want to have one object point at a spot on the ground, or any input Y value, and and a second object be point constrained to the spot the first object is pointing. A simple point constraint at the end of, say, a joint, won't work because the top object won't stay perpendicular to the object on the ground. Does anyone know of a good solution to this? 

[joojaa]

is the ground flat?

[ourfriendbrutus]

Assume the ground is flat, yes. But a way to change the ground level to something other than world 0 would be good. 

[pilou--2]

I would create a polyPlane (subdivisions width and height at 2 - we'll call it targetPlane for the exemple) which gives you a plane with a vertex on its center. 

select the ground and shift+select the targetPlane open transfertAttribute options.

leave everything disabled except for the vertex position. Sample space is World and search method is closest along Normal.

transfer.

Parent the targetPlane to the first object. 

et voilà ! 

now you get a projected plane depending on its vertex normal and the transformation of your first object. You just have to take the position of the vertex on the center of the plane.

 alternative : you could have a targetPlane with one polygon and make a rivet. probably a more efficient technique bye the way.

[joojaa]

I was under the impression that the pointing locator keeps changing position of the point. So that he is asking the same in reverse. Like the rayintersection fo a point ray and object. (ray intersect node is fine but what if it has to be deployable).

[ourfriendbrutus]

@PilouFace, that works great. I had no idea that Transfer Attributes worked like that; it's certainly a solution that's workable. I will give it a try with a rivet contraint also. Is there anyway to accomplish the same thing, but to get a coordinate from a value or something that's theoretically infinite, and not going to have an edge or end, like a plane? 

@Joojaa, raycasting would work too. Would that be significantly slower than the offset rotation of the Pilou Face's method? 

Thank you both for the replies!