Re: musings

Mike O'Dell <> Fri, 17 May 1996 01:57 UTC

Received: from by IETF.CNRI.Reston.VA.US id aa29679; 16 May 96 21:57 EDT
Received: from CNRI.Reston.VA.US by IETF.CNRI.Reston.VA.US id aa29675; 16 May 96 21:57 EDT
Received: from by CNRI.Reston.VA.US id aa07766; 16 May 96 21:57 EDT
Received: by (5.65c/5.61+local-23) id <AA18776>; Thu, 16 May 1996 18:44:18 -0700
Received: from by (5.65c/5.61+local-23) id <AA18738>; Thu, 16 May 1996 18:44:12 -0700
Received: from rodan.UU.NET by (5.65c/5.61+local-23) id <AA29642>; Thu, 16 May 1996 18:44:10 -0700
Received: from localhost by rodan.UU.NET with SMTP (peer crosschecked as: mo@localhost) id QQaqag01205; Thu, 16 May 1996 21:44:09 -0400 (EDT)
Message-Id: <QQaqag01205.199605170144@rodan.UU.NET>
Subject: Re: musings
In-Reply-To: Your message of "Thu, 16 May 1996 11:54:57 PDT." <>
Date: Thu, 16 May 1996 21:44:09 -0400
Sender: ietf-archive-request@IETF.CNRI.Reston.VA.US
From: Mike O'Dell <>
Precedence: bulk

the concept of "interoperable implementations of a requirements
document" is curious to think about.  there is no such thing
as an "implementation" of a requirements spec. there can be an
implementation of a system which is believed to meet the requirements
in the spec, and that system could be interoperable with another
system which claimed to meet the requirements, but "interoperability"
has nothing to do with the spec in any real sense because that is
a characteristic of the protocols in the system.

so one can have good interoperability (PPP on cisco talks to PPP
on 3com talks to PPP on Bay talks to PPP on Ascend etc etc) without
any notion of whether the box completely satisfies the requirements.

bottom line: the notion of "two interoperable implementations" as applied to
requirements documents makes no sense in general.