The perturbed Korteweg–de Vries equation considered anew