In: html/boost_proto/user_s_guide/extending_proto.html there's: <--------- cut here ----------- namespace boost { namespace proto { // A specialization of generate<> that causes expression in the // calculator_domain to be wrapped in the calculator<> wrapper. template< typename Expr > struct generate< calculator_domain, Expr > { typedef calculator< Expr > type; static type make( Expr const & expr ) { return type( expr ); } }; }} This instructs Proto to wrap all expressions in the calculator_domain in the calculator<> wrapper.

--------- cut here ----------- OTOH, in:

libs/xpressive/proto/example/calc3.cpp there's: <--------- cut here ----------- // Tell proto how to generate expressions in the calculator_domain struct calculator_domain : proto::domain<proto::generator<calculator_expression> > {};

--------- cut here -----------

which suggests to me that there's 2 different ways to wrap expressions in a particular domain. Is that right? If so, what are the advantages of one over the other method?