Re: [xml-dev] XML Database Decision Tree?

[Ronald Bourret <rpbourret@rpbourret.com>]

Dan Weinreb wrote:

> But what do you mean, in this context, by "succeed"?  If the native
> XML DBMS's succeed in supplanting the relational DBMS's, and taking
> over their role as the "database of record" for the classic kind of
> key corporate data that has to be accessed by many different groups,
> then yes, the DBA's would then become the lords of the native XML
> DBMS.

That is the role I am talking about.

> However, I don't think that's the kind of "success" that native XML
> DBMS's are currently looking for.  I think it would be very hard for
> them to succeed at that.  Why should people currently using RDBMS's
> for their classic key corporate data have any interest in converting
> to XML as a data model?  After all, the relational model basically
> serves their needs pretty well.

I'm not 100% sure of this. When I ask native XML database vendors about
their target markets, and whether they include the kinds of
transactional data currently stored in relational databases, all of them
say they are not interested in data that fits relational databases
extremely well.

They then get this faraway look in their eyes and hint that, should
their database outperform an RDBMS on certain transactional data, and
should people decide to use their native XML database as the database of
record, well, then, they're not about to dissuade them. Witness, for
example, Mike Champion's comments about bill-of-material data.

I'm not saying that native XML database vendors definitely are going
after the kind of data that might interest a relational DBA, but it
certainly seems like a realistic possibility.

And if a native XML database becomes the database of record for
semi-structured data (as opposed to a virtual database that gathers this
data from multiple back ends), then I'm sure DBAs will become
interested.

>  There is not currently an industry
> crisis among RDBMS users caused by deficiencies in the relational
> model.  Furthermore, changing the data model in such systems from
> relational to anything else would entail a huge amount of work and
> risk.  A vast superstructure has been built around these relational
> databases: report generators, reports written for those report
> generators, all kinds of tools and add-ons from the relational
> database vendors, the fact that lots of people have been trained in
> concepts of third normal form, the SQL language, and so on and so
> forth.

Agreed. This is why the native XML database vendors say they would be
foolish to go after strongly relational data.

> When we use the phrase "native XML DBMS", I think we are using DBMS in
> the broad sense.  Of course the broad sense is a lot less specific,
> since it doesn't say which of the services we're talking about.  So
> it's hard to generalize about native XML DBMS's.  In order to get
> specific and get down to the real facts, you have to talk about use
> cases: what is the native XML DBMS for in this case, and what are its
> responsibilities in this case, etc.

In many cases, I agree. A number of the native XML database people I
have talked to view their database more as middleware -- that is, part
of the application -- than as the database of record. This is certainly
the case in application integration scenarios.

-- Ron