Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!uwm.edu!rpi!uupsi!grebyn!dlawson
From: dlawson@grebyn.com (Drew Lawson)
Newsgroups: comp.theory
Subject: Re: Why not multiple out parameters? [again]
Message-ID: <21838@grebyn.com>
Date: 6 Sep 90 11:56:58 GMT
References: <1990Aug28.203643.11214@zaphod.mps.ohio-state.edu> <4019@rex.cs.tulane.edu> <2501@l.cc.purdue.edu>
Reply-To: dlawson@grebyn.UUCP (Drew Lawson)
Organization: Grebyn Timesharing, Vienna, VA, USA
Lines: 25

In article <2501@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>> 
>> If a module's different kinds of results
>> do _not_ form a conceptual unit,
>> then the design is kludgy and should be redone.

>Kludgy it is not.  For example, a procedure can produce a result and a
>domain indicator, such as the truth of a condition.  Or an integer
>quotient and a floating remainder can be produced.


All of these examples show an output of a single conceptual object.
Returning a quotient and remainder is simply returning the result
of a division.  This does not complicate things.

The problem is procedures which return (set VAR parameters, etc.)
a quotient, remainder and the size of the largest available memory
block.  This is three data items representing two unrelated values.


-- 
+--------------------------------------------------------------------+
| Is life an illusion?                          | Drew Lawson        |
| Or does it just seem that way?                | dlawson@grebyn.com |
+--------------------------------------------------------------------+