Re: [Boost-users] tsp 2-approximation using triangle inequality