> -----Original Message-----
> From: Ronald Bourret [mailto:]
> Sent: Friday, November 09, 2001 4:35 PM
> To:
> Subject: Re: [xml-dev] Caught napping!
>
>
>
> First, it's important to remember that native XML databases model XML
> *documents*, not the data in those documents.
That's an extremely good point ...
>
> Second, every native XML database defines two different models for an
> XML document.
I'd argue that a "native" XML database system presents a logical model that
is "XML" and hides the physical model, if it is different; that's the
differentiator between "native" and "XML-enabled" database systems. To just
mention the examples I know best, the mapping of XML onto Tamino's internal
data structures is more or less invisible to the user (except for indexing,
which can't be expressed in XML-ese); the mapping of XML onto Oracle 9i
tables is explicitly the responsibility of the database designer.
>
> In the long run, I think you will still see
> a variety of physical models, but very few logical models, with all
serious
> players standardizing on XQuery's model. Similarly, I predict that an
> ODBC-like interface (with XQuery playing the role of SQL) will be
> standardized within a couple of years (XML:DB with XPath is
> here today).
I sure hope so! It would be interesting to hear from the MS and Oracle
people on this point ... y'all seem to be holding your XQuery and/or SQLX
cards kinda close to the vest <grin>
