From owner-sc22wg5@dkuug.dk  Fri Jul 11 16:29:42 2003
Received: (from majordom@localhost)
	by dkuug.dk (8.12.8p1/8.9.2) id h6BETgl8067909
	for sc22wg5-domo; Fri, 11 Jul 2003 16:29:42 +0200 (CEST)
	(envelope-from owner-sc22wg5@dkuug.dk)
X-Authentication-Warning: ptah.dkuug.dk: majordom set sender to owner-sc22wg5@dkuug.dk using -f
Received: from Mail.Math.Princeton.EDU (mail.math.Princeton.EDU [128.112.18.14])
	by dkuug.dk (8.12.8p1/8.9.2) with ESMTP id h6BETVEc067904
	for <sc22wg5@dkuug.dk>; Fri, 11 Jul 2003 16:29:38 +0200 (CEST)
	(envelope-from adonev@Math.Princeton.EDU)
Received: from atom.Princeton.EDU (atom.Princeton.EDU [128.112.143.9])
	(authenticated bits=0)
	by Mail.Math.Princeton.EDU (8.12.9/8.12.9) with ESMTP id h6BETO66009559
	(version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NOT)
	for <sc22wg5@dkuug.dk>; Fri, 11 Jul 2003 10:29:25 -0400
From: Aleksandar Donev <adonev@Math.Princeton.EDU>
Organization: Princeton University
To: WG5 <sc22wg5@dkuug.dk>
Subject: Re: (SC22WG5.2868) Interfaces
Date: Fri, 11 Jul 2003 10:29:24 -0400
User-Agent: KMail/1.5
References: <200307101854.h6AIsF5s062025@dkuug.dk> <200307110951.h6B9pBQm066608@dkuug.dk>
In-Reply-To: <200307110951.h6B9pBQm066608@dkuug.dk>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200307111029.24376.adonev@math.princeton.edu>
Sender: owner-sc22wg5@dkuug.dk
Precedence: bulk

John Reid wrote:

> Please let's require complete agreement in all interfaces of a separate
> procedure. We are trying to keep this simple, aren't we?
I am in support of always requiring *completely* matching characterstics. 

But I do believe there is benefit in *not* requiring to repeat the 
redeclaration, but rather allow:

SEPARATE PROCEDURE Name
	! No dummy argument/result/attribute redeclaration allowed
	! No FUNCTION or SUBROUTINE
END PROCEDURE

The reason being, as I explained earlier, that there are developers who would 
be programming both the module and submodule in sync (Ctrl-? on the keyboard 
to switch from the interface to the implementation) and requiring redundant 
cut-and-paste is not nice. Otherwise I will end up using a shared include 
file and macros that generate the interface and use these both in the module 
and the submodule. But why if I don't have to?

Aleksandar