xom-interest AT lists.ibiblio.org
Subject: XOM API for Processing XML with Java
List archive
- From: "Jim Stuttard" <jim.stuttard AT btinternet.com>
- To: "John Cowan" <jcowan AT reutershealth.com>
- Cc: xom-interest AT lists.ibiblio.org
- Subject: Re: [XOM-interest] serializing XOM objects
- Date: Wed, 01 Jun 2005 12:17:53 +0100
On Tue, 31 May 2005 09:21:58 -0400, John Cowan <jcowan AT reutershealth.com> wrote:
produce two implementations at the sameIn couldn't agree more. If you don't try reversing any values in your first cut design then you haven't explored the problem. The permutations of interfaces and implementations are too complex to keep in your head at once so you might as well notate them as candidate implementation classes while your thinking about final designs. Two implementations also provide validation of algorithms and implementations and one can be seen as a test for the other. I've always been criticised for doing this as a waste of time but it hasn't stopped me since my C days.
time. This trick was used when developing the IBM 360 architecture,
which was designed to be executable on a whole variety of systems with
different price-performance. The IETF's requirement for two independent
interoperating implementations of any protocol to be standardized is
probably related.
So thanks for the IBM360 ref.
From now on I think I'll call this Xtreme Redundancy.
Cheers
Jim
--
Using Opera's revolutionary e-mail client: http://www.opera.com/m2/
-
Re: [XOM-interest] serializing XOM objects,
Jim Stuttard, 06/01/2005
- <Possible follow-up(s)>
- Re: [XOM-interest] serializing XOM objects, Michael Abato, 06/01/2005
-
Re: [XOM-interest] serializing XOM objects,
Steve Loughran, 06/01/2005
- Re: [XOM-interest] serializing XOM objects, Luca Passani, 06/01/2005
Archive powered by MHonArc 2.6.24.