From mboxrd@z Thu Jan  1 00:00:00 1970
Received: (from weis@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id SAA31431 for caml-redistribution; Tue, 14 Dec 1999 18:26:29 +0100 (MET)
Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id JAA20683 for <caml-list@pauillac.inria.fr>; Tue, 14 Dec 1999 09:20:39 +0100 (MET)
Received: from pauillac.inria.fr (pauillac.inria.fr [128.93.11.35])
	by concorde.inria.fr (8.8.7/8.8.7) with ESMTP id JAA13668;
	Tue, 14 Dec 1999 09:20:37 +0100 (MET)
Received: (from xleroy@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id JAA24437; Tue, 14 Dec 1999 09:20:35 +0100 (MET)
Message-ID: <19991214092035.33700@pauillac.inria.fr>
Date: Tue, 14 Dec 1999 09:20:35 +0100
From: Xavier Leroy <Xavier.Leroy@inria.fr>
To: Norman Ramsey <nr@eecs.harvard.edu>, caml-list@inria.fr
Cc: nr@labrador.eecs.harvard.edu
Subject: Re: OCaml and tail recursion
References: <199912131727.MAA20344@labrador.eecs.harvard.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mutt 0.89.1
In-Reply-To: <199912131727.MAA20344@labrador.eecs.harvard.edu>; from Norman Ramsey on Mon, Dec 13, 1999 at 12:27:37PM -0500
Sender: weis

> I have just completed my first nontrival Caml program (an implementation
> of the rsync algorithm) and I am distressed about the treatment of
> tail calls.  My code has to go through files one character at a time,
> and as an SML programmer from way back, I wrote the code using three
> mutually recursive functions that make tail calls to each other.
> Imagine my surprise when I started getting errors with stack overflow!

I'd like to see the code, since ocamlc performs tail calls, of course.
(ocamlopt too, although on some platforms the tail-call optimization
doesn't apply to calls with many arguments, e.g. 8 or more.  But
ocamlc has no such limitation.)

- Xavier Leroy