A note on “a uniformly valid asymptotic solution for d2y/dt2 + y = a + εy2”