From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <virgile.prevosto@m4x.org>
Delivered-To: caml-list@yquem.inria.fr
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78])
	by yquem.inria.fr (Postfix) with ESMTP id 00E55BBFA
	for <caml-list@yquem.inria.fr>; Mon, 20 Jun 2005 17:59:46 +0200 (CEST)
Received: from interferon.mpi-sb.mpg.de (mpii.mpg.de [139.19.1.1])
	by nez-perce.inria.fr (8.13.0/8.13.0) with ESMTP id j5KFxjpV009827
	for <caml-list@yquem.inria.fr>; Mon, 20 Jun 2005 17:59:45 +0200
Received: from amavis by interferon.mpi-sb.mpg.de with scanned-ok (Exim 3.36 #1 (Debian))
	id 1DkOgb-0006SZ-00
	for <caml-list@yquem.inria.fr>; Mon, 20 Jun 2005 17:59:45 +0200
Received: from maildist ([139.19.90.238] helo=mpi-sb.mpg.de)
	by interferon.mpi-sb.mpg.de with esmtp (Exim 3.36 #1 (Debian))
	id 1DkOgZ-0006SG-00
	for <caml-list@yquem.inria.fr>; Mon, 20 Jun 2005 17:59:43 +0200
Received: from mpiat2305 (Debian-exim@mpiat2305 [139.19.20.94])
	by mpi-sb.mpg.de (8.12.10+Sun/8.12.10) with ESMTP id j5KFxho5001300
	for <caml-list@yquem.inria.fr>; Mon, 20 Jun 2005 17:59:43 +0200 (MEST)
Received: from localhost.mpi-sb.mpg.de
	([127.0.0.1] helo=mpiat2305 ident=prevosto)
	by mpiat2305 with esmtp (Exim 4.50)
	id 1DkOgZ-0000Eb-8N
	for caml-list@yquem.inria.fr; Mon, 20 Jun 2005 17:59:43 +0200
Date: Mon, 20 Jun 2005 17:59:42 +0200
From: Virgile Prevosto <virgile.prevosto@m4x.org>
To: caml-list@yquem.inria.fr
Subject: Re: [Caml-list] signature and 'a generic types
Message-ID: <20050620175942.2287baed@mpiat2305>
In-Reply-To: <C6B8033B-6F2E-490D-85BB-55B37407448E@m4x.org>
References: <C6B8033B-6F2E-490D-85BB-55B37407448E@m4x.org>
X-Mailer: Sylpheed-Claws 1.0.0 (GTK+ 1.2.10; i686-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable
X-Virus-Scanned: by AMaViS perl-11
X-Miltered: at nez-perce with ID 42B6E7F1.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)!
X-Spam: no; 0.00; caml-list:01 damien:01 sig:01 val:01 func:01 struct:01 func:01 val:01 sig:01 struct:01 pairs:01 foo:01 2005,:98 tutto:98 oggi:98 
X-Spam-Checker-Version: SpamAssassin 3.0.2 (2004-11-16) on yquem.inria.fr
X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=disabled 
	version=3.0.2
X-Spam-Level: 

Hi,
Le 06/20/2005, =E0 05:20:40 PM, Damien Bobillot a =E9crit:
> Look at the following code :
>=20
>      module Module : (sig val func : 'a -> 'b end) =3D
>      struct let func a =3D a end

>      Values do not match:
>        val func : 'a -> 'a
>      is not included in
>        val func : 'a -> 'b
>=20
> Why ?

because the implementation of func is less general than what the
signature expects: if this was accepted, then you could use Module.func
in a context where 'a and 'b would get different instantiations.

>=20
> I also try to use the reverse match :
>=20
>      module Module : (sig val func : 'a -> 'a end) =3D struct
>          let l =3D ref []
>          let func a =3D List.assoc a !l
>      end

This is not the "reverse" match, which would rather be
module Module: (sig val func: 'a -> 'a end) =3D=20
struct
  let l =3D []=20
  let func a =3D List.assoc a l
end;;

which is a perfectly well defined module. In your case, problems come
from the ref: the type of l (hence of func) is not really polymorphic:
it uses what is called weak type variables (that you recognize because
of the '_' prefix in '_a and '_b), that on the contrary to normal
type variables can not be generalized. Otherwise, it would be possible
to add to l pairs of different types, such as in the following example:
module Module =3D=20
struct
  let l =3D ref [] (* type: ('_a * '_b) list *)
  let func a =3D List.assoc a !l
  let _ =3D l:=3D (0, "foo")::!l=20
   (* after that '_a is bound to int and '_b to string. *)
  let _ =3D l:=3D ("bar", 1)::!l (* type error. *)
end
--=20
E tutto per oggi, a la prossima volta
Virgile