in reply to Re (tilly) 1: Run-time generation of magic constantin thread Run-time generation of magic constant