yensao,

Basically, the relational database model is to extract the business data and
store it in rows and columns of related tables. You have to extract the
data from the objects in the client to store them in the database. On the
other hand, you have to reassemble the data into objects in the client.

The object database model is to store the objects in the database as
objects. There is no extracting data from or reassembling data to objects.

Why hasn't the world converted to object databases? Because of the
reliability, speed and size constraints of most object database systems.

Hope that helps!

Regards,

Randy

Randy A. Ynchausti wrote:
Doesn't that depend on the complexity of the data?
The more complex data, the better performance with object databases.
That is the impression I have after reading some material about
different DBMS.

About size constraints, I have visited a page
<http://www.slac.stanford.edu/BFROOT/www/Public/Computing/Databases/>
where they use an OODBMS to handle a very large amount of data:
<qoute>
This database is arguably the largest in the world: as of Sat Dec 6
00:01:11 2003, over 895.0 TB has been stored in 847149 files.
</qoute>

The biggest RDBMS I have seen was in
<http://www.wintercorp.com/VLDB/2003_TopTen_Survey/TopTenWinners.asp>
and it had a size just over 29.2 TB.

Anyone that knows about a larger RDBMS?

--

/Anders

"Randy A. Ynchausti" <randy_ynchausti (AT) msn (DOT) com> wrote

The Relational Model is the direct application of predicate logic to
the data management field.

Nonsenses, you are regurgiting mindless marketing crap of the worst
class.

The Object Database Model does not exist. The called OODBMSes use the
network model.

The world was converted to object-SQL databases like Oracle and DB2.
The called OODBMSes were a fiasco.

It didn't.

Regards
Alfredo

Alfredo

Thanks for your response. It gave me a good chuckle.

Regards,

Randy

"Randy A. Ynchausti" <randy_ynchausti (AT) msn (DOT) com> wrote

If only one could find humour in the nonsense you wrote...

Bob,

Sorry, you don't score as high as Alfredo because of lack of originality.
Thanks for playing though!

Regards,

Randy

Bob,

[snip]

Sorry to disappoint both you and Alfredo with the discussion here. It seems
that you have cornered the intelligent response market.

Maybe developing the first intelligent, adaptive AI-based real-time process
control software that use an object-oriented database system
(www.kscape.com) in real-time makes me still as green as they come?

Having a 600 GB object-oriented database (GemStone) in production for a
large shipping company over the past 10 years; connected to a Sybase
database with hundreds of tables and millions of records; connected to an
Oracle relational database with hundreds more tables and millions more
records doesn't count for much with you two.

I am waiting to hear from you both what experience I still lack and how this
endeavor is going to crash to the ground in a heap of flames. And more
importantly why the basic descriptions of the relational model and
object-oriented model for databases that have been offered here can be so
evil.

Anxiously listening as always,

Randy

"Randy A. Ynchausti" <randy_ynchausti (AT) msn (DOT) com> wrote

You might know something abouty process control software, but you don't have
a clue about data management. Your posts reveal profound ignorance of the
subject.

Bob,

Thanks for not disappointing me.

Warmest regards,

Randy

"Randy A. Ynchausti" <randy_ynchausti (AT) msn (DOT) com> wrote

What you wrote was plain nonsense. That's all.

You clearly lack basic theoretical knowledge about data management.

It is not evil, it is completely wrong.

If you want to learn you have a good starting point here:

http://www.dbdebunk.com

In the books section.


Regards
Alfredo