[Gretl-users] Re: cannot forecast with Nonlinear Least Squares