Re: [PD] nqpoly4 with selector abstraction