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 XAA25247 for caml-redistribution; Mon, 2 Jun 1997 23:49:46 +0200 (MET DST)
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 XAA25170 for <caml-list@pauillac.inria.fr>; Mon, 2 Jun 1997 23:42:00 +0200 (MET DST)
Received: from pauillac.inria.fr (pauillac.inria.fr [128.93.11.35]) by concorde.inria.fr (8.8.5/8.7.3) with ESMTP id XAA28753; Mon, 2 Jun 1997 23:41:53 +0200 (MET DST)
Received: (from weis@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id XAA25165; Mon, 2 Jun 1997 23:41:53 +0200 (MET DST)
From: Pierre Weis <Pierre.Weis@inria.fr>
Message-Id: <199706022141.XAA25165@pauillac.inria.fr>
Subject: Re: Instruction return
In-Reply-To: <199706021620.JAA20114@blaze.accessone.com> from Dwight VandenBerghe at "Jun 2, 97 09:23:19 am"
To: dwight@pentasoft.com (Dwight VandenBerghe)
Date: Mon, 2 Jun 1997 23:41:52 +0200 (MET DST)
Cc: caml-list@inria.fr
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: weis

> This is one of the two main control problems with functional programming,
> IMHO.
> The solution I use is:
[...]
Dwight's solution uses the a user defined exception Found_it instead of the
predefined exception Exit, with a loop or an iteration functional.

> The other control problem is when you want the equivalent of:
> 
>     x = foo(a);
>     bar();
>     return x;
> 
> If you try 
> 
>     let x = foo(a) in
>     bar();
>     x
> 
> then bar gets done first, instead of second.

That's not true: the language semantics ensures that foo (a) is
evaluated before the sequence bar (); x. (This is intuitively evident
since the value of x may be required to evaluate the "in" part.) 

(Try for instance:
 let foo x = print_string "Hello "; print_string x; x;;
 let bar () = print_string " world!";;
And then
# let x = foo ("you") in bar(); x;;
Hello you world!- : string = "you")

>  i use this instead:
> 
>     let f x = bar(); x in	
>     in f (foo a)

This is provably equivalent, although slightly offuscating.

(Theory meets practice in this case, since we get:
# let f x = bar(); x in f (foo "you");;
Hello you world!- : string = "you")

> I haven't found any other serious problems with Caml's logic flow.  These
> two just take some getting used to.

Good. Especially if we discarde the second problem :)

Pierre Weis

INRIA, Projet Cristal, Pierre.Weis@inria.fr, http://pauillac.inria.fr/~weis/