vtkQuad is defined by the four points (0,1,2,3) in counterclockwise order
I am wondering what “in counterclockwise order” formally/rigorously means.
Indeed, when we look at the associated figure, it is obvious that we see 0 1 2 3 vertices of the quad in counterclockwise order.
However, what if I go behind (and not front of) my screen. If I could see my screen from behind, the same exact quad vertices would now be clockwise ordered. And of course, this quad would still be valid…
Then, I cannot understand what really mans this constraint especially regarding vtkTriangle definition which does not state anything about node ordering.
Just to build on what @toddy said, the point ordering is critical as you know.
When you look at the surface, if the normal is pointing directly towards you then ordering has to be anticlockwise.
I’ll be putting up a change to LinearCellDemo and the Python version. I noticed that the ordering in the triangle strip example there is in fact clockwise! I will also add a few (commented out) lines that, if uncommented, will show the back face.
This was the clarification I wanted to get, thanks. The vtkQuad documentation does not state anything about the normal pointing towards me or not.
Should I report a bug or do you think that the documentation is enough for the VTK documentation lamda readers?
I guess this rule is generalized to all polygons in VTK?