Re: [stack] impure concatenativity: let without translation

On Wed, Mar 19, 2008 at 11:58 AM, John Nowak wrote:

>
> On Mar 19, 2008, at 11:58 AM, Joe Bowbeer wrote:
>
> > On Wed, Mar 19, 2008 at 1:58 AM, John Nowak wrote:
> >
> >> Just moments ago, I gave this example of the quadratic formula:
> >>
> >>  (define (quad a b c)
> >>    (let ((d b 4 a c * * -)
> >>          (r d 0 >= [d sqrt] ["quad: sqrt of negative" error] if)
> >>          (x 2 a *)
> >>          (p b r - x /)
> >>          (n b r + x /))
> >>      p n))
> >>
> >>
> >
> > Is quad invoked postfix or prefix?
>
> Postfix, as are *all* procedures. Only special forms are infix.
> Perhaps this is a better declaration form:
>
>     (define (a b c quad)
>        ...)
>
> - John
>

Better.  Better yet is:

   (...
      (a b c quad) define)

This maintains a consistent right-to-left reading direction.  If postfix is
the "right thing," there's no such thing as too much, right?

But if that's even better, then why not go right-to-left all the way?

   (...
      (c b a dauq) enifed)

Naive question: Is there any reason why concatenative languages are
(uniformly) postfix?  Would prefix work just as well, pushing elements onto
the stack from right to left?

--Joe


[Non-text portions of this message have been removed]