Re: A searchable datastructure for represeting attributes?

From: Niall Litchfield <>
Date: Fri, 1 Mar 2002 09:30:59 -0000
Message-ID: <3c7f4a5a$0$8506$>

wise words indeed.

Niall Litchfield
Oracle DBA
Audit Commission UK
Please include version and platform
and SQL where applicable
It makes life easier and increases the
likelihood of a good answer

"David Cressey" <> wrote in message

> Niall,
> > It looks to me as if an xml document system maybe the way to go since
> > allows documents (which I'll call here objects) to self describe. That
> > the object itself can tell the app how to read it. This object/attribute
> > approach does not really work very well with the relational model as
> others
> > have already said.
> You may well be right. XML could be the right tool, as long as it is used
> properly.
> But if people begin using xml in a way that blurs the distinction between
> metadata and data,
> they may end up with documents that are "self describing" only when the
> rules are supplemented by some kind of private understanding between the
> author and the reader. That necessity for a private understanding is what
> eventually causes a breakdown in intelligibility.
> But this issue is not well understood, even by many in the profession.
> I think it's the reason why so many non relational designs are stored in
> relational databases.
> --
> Regards,
> David Cressey
Received on Fri Mar 01 2002 - 10:30:59 CET

Original text of this message