> >Surely something simpler could be imagined, ~r<n> for example, which when
> >immediately preceding a value-requiring escape code requests that it
> >retrieve the <n>'th argument.
>
> That doesn't work if the value-requiring escade code can take a
> parameter, which in Common Lisp (and existing Scheme implementations)
> is almost all of them.
Why not?  The proposal I made was that ~r<n> precedes a full escape code, ie
~r3~a.  It seems crufty as well that the ~xxxxxx@* specifies the number as
the character succeding the escape tilde.  That violates the idea that
the character following the escape is some sort of command identifier.

>
> In any case, I don't think "~rNa" is enough of an improvement
> over "~xxxxxx@*~a" to justifying inventing a new convention.  Yes,
> it might have been better if Common Lisp made '@' ("absolute
> goto") the default, but they didn't.
I'm not so sure we should be writing this SRFI to be compatible with
Common Lisp or Kawa.  No offsense, but if we have the ability to
improve, why not?

	Scott