I got tired of botching the 4-bearing method — both the stationary and the moving variant. With the vanilla ±several-degree spread, a rounded midpoint, and bearings printed as whole integers, every plot I ran came out wrong by enough to matter. So I made t
