Reference from Day 1

Where we left off from Day 2

Fixing up the end caps.

To do this, I first made a quick hemisphere using some boolean operators. It looks gnarly, but it’s really just a bunch of low-level math and transformations.

Luckily, this is easily condensed with a node group!

So this is great and all, but when we try to instance it on the endpoints, it doesn’t quite rotate the right way.

Using this set of nodes, I was able to get the left cap working. It uses the positions of of the bezier curves’ handle and end point to calculate the direction it should point in. However, this only works for one cap. The other one is inverted!

Left Cap

Right Cap

Unfortunately, this means we’re going to have to split it up to two instances on points. No big deal.

All I had to do was make two separate groups with slight alterations. (Switching the endpoint selection and the curve handle positions)

And it works!*

• Not really. Using Align Euler to Vector, we get the cap pointing in the right direction, however not rotated correctly on the local x axis. This leads to sliding as seen below.

To fix this, I plan on separating the components of the rotation vector, and instead finding some other way to control the local x axis rotation. I imagine this will involve finding the rotation of the profile circle at the end.