Problems with hacd algorithm

First of all, it's very cool that you implemented this great algorithm!

However, it kinda fails when I try to decompose the following simple mesh:


`hacd` here doesn't result in (what I'd expect) 2 or 3 separate convex objects, but 28 single triangles.

I wasn't sure if it's a problem that my mesh has a unified index buffer and therefore duplicate vertices?
Anyway, the length of the resulting decomposition changes into 13 when calling `split_index_buffer` while recovering topology before, and in this case, at least the first element in the vec of trimeshes looks more or less correct:


But the remaining 12 also seem to be just triangles.

I called `hacd` with error = 0.03 and min_components = 0. Changing these values doesn't help.

Here an .obj representation of the above mesh (here with already splitted indices):

> g Mesh
v 0 0 0
v 1 0 0
v 1 0 1
v 0 0 1
v 0 1 0
v 1 1 0
v 1 1 1
v 0 1 1
v 2 1 0
v 2 1 1
v 2 0 1
v 2 0 0
v 1 1 2
v 0 1 2
v 0 0 2
v 1 0 2
vn 0 -1 0
vn 0 0 -1
vn -1 0 0
vn 0 1 0
vn 0 0 1
vn 1 0 0
f 1//1 2//1 3//1 4//1
f 5//2 6//2 2//2 1//2
f 8//3 5//3 1//3 4//3
f 5//4 8//4 7//4 6//4
f 12//2 2//2 6//2 9//2
f 9//4 6//4 7//4 10//4
f 10//5 7//5 3//5 11//5
f 11//1 3//1 2//1 12//1
f 9//6 10//6 11//6 12//6
f 16//6 3//6 7//6 13//6
f 13//4 7//4 8//4 14//4
f 14//3 8//3 4//3 15//3
f 15//1 4//1 3//1 16//1
f 13//5 14//5 15//5 16//5