RE: XML design of Database

Subject: RE: XML design of Database
From: "Vun Kannon, David" <dvunkannon@xxxxxxxx>
Date: Tue, 30 May 2000 09:45:31 -0400
Andreas' presentation is clear on some things, not on others.

XML as normalised relational tables means table per element, not table per
nesting level. Passing from XPath to SQL on such structures should be
mechanical and optimisable. This representation also lets your XML
interoperate directly with everything else in the database.
Andreas's description of Tamino is not as direct as his description of the
other approaches. From what he does say, it sounds like Tamino uses a
network/hierarchical DB (Software AG owns a few of those, I believe).

-----Original Message-----
From: Schoedl, Andreas [mailto:Andreas.Schoedl@xxxxxxxxxxxxxx]
Sent: Tuesday, May 30, 2000 4:53 AM
To: 'xsl-list@xxxxxxxxxxxxxxxx'
Subject: RE: XML design of Database


Do you mean a DB that stores XML?

If yes, here's an overview!

[Caveat: I am employee with a major database system developer, Software AG.
Take this as an technical overview over the mechanisms employed today to
store XML to databases, not as an objective report for any one of them.]

Some R-DB-vendors store XML-data normalized in a host of db-tables. The
deeper the XML hierarchie, the more tables to resolve the relations are
necessary. Storing and retrieving gets slower and slower, because more and
more tables have to be accessed.

Some R-DB-verndors store one XML document with all <tags/> as one long line
of characters (CLOB). As you might guess, searching with XPath-Expressions
on that will take a good full-text search module within that DB-system.

Some OO-DB vendors create an object hierarchie that matches the document
hierarchie, transfer the document's data to new object instances and store
them. This is already much nearer to the meaning of XML.

There's one DB that stores XML natively, it is my employer's, so I'm
certainly not neutral on that:
The DB is called Tamino. When you want to store an XML document to it, you
first have to load the doctype for this particular "class" of documents.
With loading this doctype, the database internally creates a hierarchical
structure for all declared tags. (Similar to what you might see when opening
a DTD or XML file within a tree-view of a graphical editor.) From now on,
all documents matching this doctype are stored in this structure. Since this
structure mirrors the tree-construct XPath is working on, the query language
of Tamino is XPath-like with some additions from XML QL drafts. This all
works without transforming XML to objects or relations and, thus, is quite

I won't take this further, since it tends to degrade in SPAM.

Andreas Schödl

> -----Original Message-----
> From: Sydney [mailto:moisi_fr@xxxxxxxx]
> Sent: Monday, May 29, 2000 6:48 PM
> To: XSL-List@xxxxxxxxxxxxxxxx
> Subject: XML design of Database
> Hi
> I want to know if a DB made in XML is:
> a) an object DB
> b) a relational DB
> c) an object relational DB
> I know this question is a little bit curious.
> Excuse me if this question is not related to all
> XSL-List thread(off-topic).
> Thanks
> Sydney
> ___________________________________________________________
> Do You Yahoo!?
> Achetez, vendez! À votre prix! Sur
>  XSL-List info and archive:

 XSL-List info and archive:
The information in this email is confidential and may be legally privileged.
It is intended solely for the addressee. Access to this email by anyone else
is unauthorized.

If you are not the intended recipient, any disclosure, copying, distribution
or any action taken or omitted to be taken in reliance on it, is prohibited
and may be unlawful. When addressed to our clients any opinions or advice
contained in this email are subject to the terms and conditions expressed in
the governing KPMG client engagement letter.

 XSL-List info and archive:

Current Thread