Mapping to a globe is a lot easier than mapping to a flat projection. Latitude and longitude are essentially spherical coordinates (http://en.wikipedia.org/wiki/Spherical_coordinate_system). Once you convert them to radians, you can use some really simple math to extract cartesian (x,y,z) coords.

In my case I pass a 2D vector of lat,lon to two different functions, first to convert to spherical coords then to convert to cartesian:

Vec3D latToSphere(Vec2D ll, float r) {

Vec3D v = new Vec3D();
ll = toRadians(ll.x, ll.y);
v.x = ll.x;
v.y = ll.y – PI/2;
v.z = r;

return(v);
};

Vec3D sphereToCart(Vec3D ll) {

Vec3D v = new Vec3D();
v.x = ll.z * cos(ll.y) * sin(ll.x);
v.y = ll.z * sin(ll.y) * sin(ll.x);
v.z = ll.z * cos(ll.x);

return(v);

};

float degreesToRadians(float d) {
return((d/180) * PI);
};

This gives me a 3D vector with the X,Y,Z coords. Both this vector and the spherical 3D vector are really useful since I can use some of the built-in methods of the Vec3D Class (thanks, toxi!) to do a lot of the math required to make the pretty ribbons, calculate distances, etc.

Toxi’s geomutils are really really handy for this and a whole pile of other things: http://code.google.com/p/toxiclibs/

I’ve used the metacarta api myself, but never poked around processing yet (although ProcessingJS is pretty wonderful).. something I’ve added to my todo list.

I’m really curious as to how you mapped a lat/long to an image of the globe. Did you use the Google Earth API? I know you’ve talked about the Wolfram projections… but can you be a bit more specific?

cheers!
john
@worldlyjohn