OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.


Help: OASIS Mailing Lists Help | MarkMail Help



   Re: [xml-dev] Re: XML and the Relational Model

[ Lists Home | Date Index | Thread Index ]

On Tue, 2003-08-26 at 13:37, Mike Champion wrote:

> Is there any chance that someone could elaborate? 
> I've heard that there are lots of these anecdotes, but
> I don't know any details. 
> Certainly a straightforward normalization of a serious
> document-oriented schema such as DocBook would lead to
> a nightmarish number of tables, but I'm not clear
> offhand on why normalizing just the metadata would be
> such a problem.  

If the metadata is simply, author(s) and issue, which might be all
Nature wants, clearly it's not a problem.

If you're trying to trace the development of a corpus, where essays
become chapters of books, works are reprinted under different titles,
authors change their names (and pseudonyms are exposed), journals change
their names (and issue numbering/naming schemes), essays that became
chapters of books are also reissued in second editions (as essays) with
different co-authors....

The first such project I worked on was just that, with the added fillip
of being about science fiction, so you got to track characters, series
(story cycles), universes and ghu knows what else.

The second was 15 years later and involved goverment documents. It was
attempting to bring a massive semi-structured database off of a y2k
vulnerable system. The whole project was botched from top to bottom, but
that didn't make attempting to move from semi-structured to relational
any less of a nightmare.



News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS