Reference from Day 1
The nasty snag from yesterday
As of last time, we tried to make the end cap point in the correct direction. While we were able to do so, it ended up sliding in relation to the rest of the mesh.
I’m going to try breaking this problem down into it’s most basic parts. Right now, we have a bezier curve. This curve (in blender) has tangents. To visualize it, I made some quick axes in blender, and instanced it on the points with a rotation set to the tangent.
Noticing that it seemed to position the axes just about perfectly (at least on one cap) first try, I reformatted the hemisphere to make it like this by default:
Aaaaaaaand…?
DRATS! Not quite it either, but we got something much easier to digest! (Damn this node sure is tricky to deal with)
How I see it, it looks like we need to also incorporate the normal information of the curve in order to make that local.
After watching a helpful tutorial by CG Cookie and fiddling around with the inputs, my hypothesis was correct.
It FINALLY WORKS!
HUZZAH!
On to some cleanup work. I made sure to get rid of the bottom face of the hemisphere generated in the mesh boolean.
And then I went and joined all the geometry together in the main node graph. Ah, the beauty of a finished network!
