For annotating it can be useful to see the cross-section of the mesh at the current point - that is the intersection in the z-x and z-y planes at the current point. This can be visualised as a line overlaid on the viewport, or as a small view embedded in the viewport.
This can be useful to aid the user in exploiting symmetries and points of maximum curvature on the mesh, especially useful in the case of faces.
However, this generally is only useful if the mesh has a sense of direction or 'front' and 'up'. Then, the mesh can be aligned, e.g. so
front --mappedto--> [0, 0, 1]
up --mappedto--> [0, 1, 0]
and now the z-x and z-y have special meaning.
In terms of implementation this thus involves two considerations:
Adding a concept of direction to meshes - see #13
Efficiently generating the visualisation of the isolines. This can be most easily solved by sampling from a depth or shape buffer - need to see how this can be accessed from THREE.
This is now much more feasible since #72. It's on the back burner as it's not something we've had requested in a while, but wouldn't be that hard to implement now.
For annotating it can be useful to see the cross-section of the mesh at the current point - that is the intersection in the
z-x
andz-y
planes at the current point. This can be visualised as a line overlaid on the viewport, or as a small view embedded in the viewport.This can be useful to aid the user in exploiting symmetries and points of maximum curvature on the mesh, especially useful in the case of faces.
However, this generally is only useful if the mesh has a sense of direction or 'front' and 'up'. Then, the mesh can be aligned, e.g. so
and now the
z-x
andz-y
have special meaning.In terms of implementation this thus involves two considerations: