Re: [Gretl-devel] problems with automatic matrix-to-scalar conversion