Re: looking for a Borland metaprogramming hint

From: fernando_cacciola (fcacciola_at_[hidden])
Date: 2001-12-05 16:03:41

--- In boost_at_y..., "David Abrahams" <david.abrahams_at_r...> wrote:
> 
> ----- Original Message -----
> From: "Fernando Cacciola" <fcacciola_at_g...>
> 
> > 1) The partial specialization code is ill-formed.
> >
> > Look here:
> >
> >   (A) template <class F> struct arg_tuple_size;
> >   (B) template <class R, class A0> struct arg_tuple_size<R (A0::*)
()> ....
> >
> >   (B) is a specialization of (A), but it is *adding* a second 
template
> > argument.
> 
> ALL of my compilers which support partial specialization disagree 
with you,
> including comeau which you can try for yourself at
> www.comeaucomputing.com/tryitout.

Which can only mean that I am totally wrong!!!
Sorry about that. 
I would have bet that was illegal.

BTW. Borland seems to agree with me. I caught that because of
the compiler errors it gave...

Fernando Cacciola