Re: What's right with FORTH and wrong with c.l.f.

On May 5, 10:41 pm, John Passaniti <n...@[EMAIL PROTECTED]
> wrote:
> Aleksej Saushev wrote:
> >> And all the Forths adapted for Windows and *nix hosted platforms
> >> do feature easy mechanisms for calling C or other subroutines,
> >> so their users have the same access to libraries as anyone else.
>
> > I doubt that either.
>
> > It is a miracle, if you transparently create NUL-terminated strings
> > from pointer-counter pairs, and get the latter from the former at
> > the same ease. And I don't believe in such miracles.
>
> A subroutine call is a miracle?  For years, I've never had any problem
> passing string literals between programming languages.  All you have to
> do is choose a representation that makes everyone happy:
>
>         [count] [text] [0]
>
> When I want to pass this string to a language like C, I point to [text]
> and the null at the end keeps the usual library routines happy-- it
> doesn't see the leading count.  When I want to pass this string to a
> language like Forth, I either point to [text] and dereference [count],
> or I point to [count] as appropriate.  It doesn't see the null
terminator.

Exactly how Win32forth does it. Some string operators also append
silent nulls; there are very few cases where an explicit +null is
required, but +null is there if needed.

Without call support, most hosted Forths would be blind and dumb.

--
Regards
Alex McDonald