Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!cmcl2!harvard!seismo!decuac!avolio
From: avolio@decuac.DEC.COM (Frederick M. Avolio)
Newsgroups: net.lang.lisp
Subject: Re: FranzLisp question
Message-ID: <933@decuac.DEC.COM>
Date: Wed, 21-May-86 15:12:24 EDT
Article-I.D.: decuac.933
Posted: Wed May 21 15:12:24 1986
Date-Received: Sat, 24-May-86 03:52:09 EDT
References: <1319@enea.UUCP>
Organization: ULTRIX Applications Center, MD
Lines: 46

In article <1319@enea.UUCP>, pesv@enea.UUCP (Peter Svensson) writes:
> Yesterday, I made a little program in Lisp. It was supposed to give n
> random-numbers to see how the (random x) function was doing. ...
> (def foo
>   (lambda (n)
>        (prog nil
> 	    back  (cond ((zerop n) (return))
> 			(t (princ (random 100))
> 			   (sub1 n)
> 			   (go back)))))
> 
> And all that happens is that for every n>0 the lisp starts to barf
> the most humungus bignum ever to be seen by mortal man.  ...  WHY?

Two things... First, it doesn't give you a real big number.  Just a
bunch of little numbers with no spaces between them (princ, you see
you need a terpri or something).  Also, as far as I can see this
should run forever.  And I bet it does, eh? Because (sub1 n) returns
one less than n but IT DOES NOT CHANGE N!  You want (setq n (sub1 n))
instead.  Liek so:

(def foo
  (lambda (n)
       (prog nil
	    back  (cond ((zerop n) (return))
			(t (princ (random 100))
			   (terpri)
			   (setq n (sub1 n))
			   (go back)))))

Or better....

(def foo
  (lambda (n x)
    (cond ((or (zerop n) (minusp n)) nil)              ;; never be trusting...
          (t (cons (random x) (foo (sub1 n) x))))))

-> (foo 7 100)
(90 75 84 81 74 99 52)
-> (foo 2 100)
(64 1)

-- 
Fred @ DEC Ultrix Applications Center
INET: avolio@decuac.DEC.COM				* Fight the Fight *
UUCP: {decvax,seismo,cbosgd}!decuac!avolio	       * Rescue the Unborn *