Some remarks on the Dirichlet problem on infinite trees