Minimizing the maximum lateness in a single-machine scheduling problem with the normal time-dependent and job-dependent learning effect.