Re: Development Trends in Web and Oracle

From: DA Morgan <damorgan_at_x.washington.edu>
Date: Sun, 13 Mar 2005 20:06:06 -0800
Message-ID: <1110772972.912694@yasure>

Galen Boyer wrote:

> On Sat, 12 Mar 2005, andy_at_andyh.co.uk wrote:
> 
>>On Sat, 12 Mar 2005 21:21:03 -0000, "Hexathioorthooxalate"

>> If you've got an XML Schema isn't it possible to automatically derive
>> a relational schema that can store the XML? (Or at least, one
>> possible relational schema - there are likely various ways to
>> represent an XML Schema in relational form, but given consistent
>> rules it should be possible to deterministically come up with the
>> same relational schema for a given XML Schema?). If you're already
>> inside a relational database, and it's mapping XML into a relational
>> schema, then you could have the option of issuing an XPath query
>> which the server translates internally to SQL against the data; or
>> you could just use SQL against the relational schema itself.

> 
> 
> I believe I'm with Hexathioorthooxalate on this point.  What's the
> reason to take the xml data and get it to a relational form, if the only
> use for the data is to pass it back to the application in xml form?

If your caveat was valid I'd agree too. The question at hand though is one that relates to the job that is supposed to be performed by the database architect working with the application architect.

If both are competent then they should agree to use XML as it was designed to be used and an appropriate use of XML is not to "pass it back to the application."

If data is being exchanged between applications/systems I would agree that it is appropriate. But that is a far cry from intentionally designing an application that EXPECTS XML as its input which is just another version of "my only tool is a hammer."

-- 
Daniel A. Morgan
University of Washington
damorgan_at_x.washington.edu
(replace 'x' with 'u' to respond)

Received on Sun Mar 13 2005 - 22:06:06 CST