From owner-sc22wg5  Thu Jul 26 06:55:45 2001
Received: from imo-m06.mx.aol.com (imo-m06.mx.aol.com [64.12.136.161])
	by dkuug.dk (8.9.2/8.9.2) with ESMTP id GAA15839
	for <SC22WG5@dkuug.dk>; Thu, 26 Jul 2001 06:55:45 +0200 (CEST)
	(envelope-from Wclodius@aol.com)
From: Wclodius@aol.com
Received: from Wclodius@aol.com
	by imo-m06.mx.aol.com (mail_out_v31.9.) id m.7f.17bb2384 (4197);
	Thu, 26 Jul 2001 00:55:02 -0400 (EDT)
Message-ID: <7f.17bb2384.2890fca6@aol.com>
Date: Thu, 26 Jul 2001 00:55:02 EDT
Subject: Re: (SC22WG5.2153) Dynamic arrays and interfacing with C
To: jkr@rl.ac.uk, SC22WG5@dkuug.dk
MIME-Version: 1.0
Content-Type: text/plain; charset="US-ASCII"
Content-Transfer-Encoding: 7bit
X-Mailer: AOL 4.0 for Mac - Post-GM sub 146


In a message dated 7/24/01 3:00:40 AM, Richard Maine writes:

>That idea was basically to
>have an intrinsic that generated a Fortran pointer from an address
>and a mold.  See paper J3/00-168.  One could argue spelling trivia,
>but the idea seems pretty simple and workable to me.
>
I am surprised that there were significant objections considering the strong 
semantic and implementation similarities to the RESHAPE intrinsic. About the 
only problems I can see is that the Fortran compiler's memory management 
might have to be more strongly coupled to its associated C compiler, but I am 
not even certain that is required. Would there be any need for a capability 
to explicitly delete the dope vector that is implicitly created by such a 
facility? I am not particularly enthusiastic with the name suggested in 
J3/00-168, and I am not enthusiastic about overloading RESHAPE, as C scalars 
and pointers are described as if they are different types unlike Fortran's 
arrays and scalars. Perhaps MAP_C_POINTER or MOLD_C_POINTER, or MAP_C_MEMORY?