[Gretl-devel] Re: Standard error of OLS residuals