Ideally, we would use the PNS model as the default one: r-tree, geometry
library (could solve also a few problems in the ratnsnest algorithm),
and there you go - a big refactoring! It sounds too scary to me to do it
before removing the legacy canvas.
Ideally, we would use the PNS model as the default one: r-tree, geometry
library (could solve also a few problems in the ratnsnest algorithm),
and there you go - a big refactoring! It sounds too scary to me to do it
before removing the legacy canvas.