Is G J Chaitin's "A toy version of the LISP language" Report RC 5924,
IBM Watson Research Center, available for download anywhere?
I believe that he wrote (a version of) it in C--and it's that that I'm
looking for.
Tia
--
G.C.
Note ANTI, SPAM and invalid to be removed if you're e-mailing me.
George Cox wrote:
> Is G J Chaitin's "A toy version of the LISP language" Report RC 5924,
> IBM Watson Research Center, available for download anywhere?
>
> I believe that he wrote (a version of) it in C--and it's that that I'm
> looking for.
The first link on this page:
http://www.cs.umaine.edu/~chaitin/unknowable/lisp.html
is the word "C" which links to the C source for Chaitin's interpreter. It's
a 950-line program. Naturally, given its size, it's a very limited dialect,
doesn't follow any particular standards, and introduces some conventions all
its own ("M-expressions", which aren't quite what McCarthy originally
designed) - but of course the important point is that it works for the
examples which Chaitin provides. There are also versions of the interpreter
in Java and for Mathematica.
If you plan to use it for anything other than Chaitin's examples, you'll
need to peruse Chaitin to find out more about its syntax, which *requires*
that you omit parentheses whenever possible at the top level, and adds
parens based on a function's known numbers of arguments, so you end up with
something a bit like REBOL. This:
define foo lambda (x) * + x 1 2
...means (defun foo (x) (* (+ x 1) 2)), and you can't cheat and include the
parens anyway, because it'll come out wrong.
Anton
P.S. Is Chaitin's Lisp a Lisp? ;)
Anton van Straaten wrote:
>
> George Cox wrote:
> > Is G J Chaitin's "A toy version of the LISP language" Report RC 5924,
> > IBM Watson Research Center, available for download anywhere?
> >
> > I believe that he wrote (a version of) it in C--and it's that that I'm
> > looking for.
>
> The first link on this page:
> http://www.cs.umaine.edu/~chaitin/unknowable/lisp.html
>
> is the word "C" which links to the C source for Chaitin's interpreter. It's
> a 950-line program. ...
Thanks for that. I had a reply from Chaitin as well by e-mail.
--
G.C.
Note ANTI, SPAM and invalid to be removed if you're e-mailing me.