Tests

[timschmidt]

I broke a ton of tests when I built out the geo 2D subsystem. The tests have largely been thrown together over time and could really use some care and attention. Is anyone interested in improving the tests?

[winksaville]

I saw this and thought of a "hard" test for is_manifold. Create a rectangular surface centered on the XY plane with one edge a sin wave in the Z plane the other edges would all be straight lines. You then extrude that wavy plane in the Z direction to create a solid. I'm guessing would be many triangles at the straight edge opposite the sin wave.

The fundamental operation is creating the weird curvy 2D plane, can csgrs do that?

[winksaville]

Thoughts?

[timschmidt]

csgrs should be able to do it. You would generate the line as a 2D geometry then extrude it.

I'm just thinking more about granular function-level tests.