[XML4LIB] re: databases or xml

Kevin S. Clarke ksclarke at stanford.edu
Thu Feb 7 16:10:49 EST 2002


On Thu, 2002-02-07 at 12:29, Stirnaman, Jason wrote:
> I'm just a newbie, but heregoes:  Is it correct to say that xml and
> databases should not be mutually exclusive?

I'd say this is true.  I do not think the question is really XML or a
database.  The question is really (IMO) what type of database or
retrieval system best meets your information needs.  If as Elizabeth
Shaw said your information is highly structured then maybe a relational
database is best.  In this case, you are probably going to create a
highly structured schema for your system.  Using XML here gives you
flexibility and the separation of content/presentation, as Jason notes.

The first post said, essentially, that presentation can be done with
other tools that allowing dynamic information to be presented based on
the structured data in the database.  This is true.  XML, I think,
generalizes this process so that not so many database specific programs
have to be written to do this.  There is a lot more customized
programming, in my opinion, to generate a dynamic presentation from raw
database-specific data than there is from an XML stream that comes out
of a database (even a relational one).  In my opinion, XML provides a
higher level (or more abstract level) for working with a library's
dynamic data than does a customized program (SSI/servlet/etc.) that
accomplishes the same thing.

There are also the advantages that Nathan Williams mentions... Creating
a custom schema for your known highly structured data in a relational
database is fine, but what if you want to receive data, in various
structures, from other places.  Using a native XML database will give
you the advantage of not having to create all the custom tables in a
relational database to handle each new type of data structure.  There is
a great deal of flexibility in being able to accept a semistructed
document and serve it from a database almost immediately on receiving
it.

That my 2cents at least :-)

Kevin


-- 
Kevin S. Clarke (ksclarke at stanford.edu)
Digital Information Systems Developer
Lane Medical Library, Stanford University

"Pick a job you love, and you will never 
 have to work a day in your life."

                             --Confucius



More information about the xml4lib mailing list