Some remarks on Newton–Sabatier methods