RE: (whimsy) Bug-free software

["David Crocker" <dcrocker () eschertech com>]

Actually, there is often a very big difference between whether the failure
occurs at the requirements level or at the implementation.

The difference is that the end-user may have access to the requirements document
(he/she may have helped to develop it) and can check it against his actual
requirements (assuming he knows what they are!). But even if the end user has
access to the implementation detail, he is unlikely to understand it, much less
be able to determine whether it is correct.

David Crocker
Escher Technologies Ltd.
www.eschertech.com
Tel. +44(0)1252 336565  Fax +44(0)1252 320954


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of Terrence Enger
Sent: 06 February 2004 05:30
To: der Mouse; [EMAIL PROTECTED]
Subject: Re: [SC-L] (whimsy) Bug-free software


t 19:10 2004-02-05 -0500, der Mouse wrote:
> Now of course you might be talking about higher-level requirements
> documents, in which case you have something of a point - but when a
> malfunction occurs, whether the bug is in the code written to implement
> the requirements document or in the requirements document itself is a
> distinction of little import to the end-user with the problem; all it
> affects is whom to blame, and that much only at levels at which there
> is a distinction between the sources of the two.

As Dick Francis put it in another context: It does not
matter to a bettor whether the jockey fell of the horse, or
the horse fell and took the jockey down with him.  The
bettor loses either way.  Still, it makes a world of
difference to the jockey.