Talk:Euclidean distance
|
Is that formula for approximating the distance using only integers wrong? Surely if dx and dy are the same (eg, at a 45-degree angle) then dx + dy - 2×min(dx,dy) will always be 0, and the approximation will always be 100% error?