My initial idea for this implementation:
- if acute triangle -> split into 4 triangles with vertices on midpoints of the 3 edges of the larger triangle
- if obtuse triangle -> split into 2 triangles with vertex at the midpt of the largest base edge
- repeat until all triangles are smaller than a certain area or edges fulfil a certain minimal length
the geometry will be used for small area terrain animations/deformations. would like to know if there're any pitfalls or even simpler alternatives.