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 NAA12584 for caml-redistribution; Mon, 9 Feb 1998 13:50:08 +0100 (MET)
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id RAA26310 for <caml-list@pauillac.inria.fr>; Thu, 5 Feb 1998 17:05:37 +0100 (MET)
Received: from krilin.dc.fi.udc.es (mail@krilin.dc.fi.udc.es [193.144.51.196])
	by nez-perce.inria.fr (8.8.7/8.8.5) with SMTP id RAA13390
	for <caml-list@inria.fr>; Thu, 5 Feb 1998 17:05:34 +0100 (MET)
Received: from quintela by krilin.dc.fi.udc.es with local (Exim 1.62 #1)
	id 0y0Tnn-0000Im-00 (Debian); Thu, 5 Feb 1998 17:05:23 +0100
To: Pierre CASTERAN <Pierre.Casteran@labri.u-bordeaux.fr>
Cc: caml-list@inria.fr
Subject: Re: Weak types ?
References: <199802031546.QAA21882@scrasmeustache.labri.u-bordeaux.fr>
X-Url: http://www.dc.fi.udc.es/~quintela
From: "Juan J. Quintela" <quintela@dc.fi.udc.es>
Date: 05 Feb 1998 17:05:23 +0100
In-Reply-To: Pierre CASTERAN's message of Tue, 3 Feb 1998 16:46:22 +0100 (MET)
Message-ID: <ytthg6eavgc.fsf@krilin.dc.fi.udc.es>
Sender: weis

Pierre CASTERAN <Pierre.Casteran@labri.u-bordeaux.fr> wrote:

> let toto =
>   let id x = x
>   in id id;;
> (*
>  val toto : '_a -> '_a = <fun>
>              ^      ^
> 
>              why not " 'a " ?  
>              
> *)

Hi,
	I really dont remember the reason, but I remember one solution,
to put explicity the varibles in the binbidng, i. e.:

# let toto y = 
    let id x = x
    in (id id) y;;
val toto : 'a -> 'a = <fun>

Now it works with the normal values:

# toto 5;;
- : int = 5
# toto true;;
- : bool = true

But the identity function returns again the extrange type variables:
# toto (fun z -> z);;
- : '_a -> '_a = <fun>


You can solve this again using the same technique:

# let f y = (toto (fun z -> z)) y;;
val f : 'a -> 'a = <fun>

 
I remember some mail (I think in this list) about this matter. It turs
to be problem with the eta reduction but I don't remind the details, I
remember the solution, to put the varibles in the bindings.

I hope this helps.

Cheers, Juan.

PS. Someone knows the details for this behaviour?