Adam Wulkiewicz
Adam Wulkiewicz
The result of set operation for boxes would have to be something like a Collection of Boxes, a MultiBox or something like a BoxifiedPolygon. In cartesian it indeed might as...
I think it won't be much more complex in spherical, you only have to take winding and poles into account. The boxes are still the same in the coordinates space,...
> I will look into it after finishing the implementation for cartesian coordinates. Ok, though it's possible that noone will find a use for it. It might be more beneficial...
Wow, that's a lot of thoughts. :) > other algorithms such as area (I'd need to be able to make it return floating point results for integer coordinates, though) The...
@barendgehrels
`difference()` cases currently failing validity test are: - intersect_holes_intersect_and_disjoint - intersect_holes_intersect_and_touch - intersect_holes_intersect The problematic fragment is the same in all of them.
Yes, there was always a plan but there was not enough free time. The serialization support is still experimental and conditionally enabled with a macro definition because it's implemented in...
> Is there a way to reconstruct the exact tree from the public API? No, there is no public API for this. And yes, the placement of elements in the...
> How would you implement it outside the rtree? It depends on the internals of the rtree which I don't see a public API exposing. There is non-documented, internal API...
Thanks for the explanation. I'll treat this issue as a feature request but I don't know when I'll look into this.