Re: OO and relation "impedance mismatch"

From: Kenneth Downs <firstinit.lastname_at_lastnameplusfam.net>
Date: Tue, 05 Oct 2004 19:43:18 -0400
Message-ID: <1jbvjc.nh5.ln_at_mercury.downsfam.net>


Laconic2 wrote:

>
> "Kenneth Downs" <firstinit.lastname_at_lastnameplusfam.net> wrote in message
> news:fr2vjc.8q4.ln_at_mercury.downsfam.net...
>

>> We may have different definitions of data dictionary.  My definition of
>> dd is something designed by me and populated by db analysts/designers
>> that is
>> used to build databases.  Neither SQL Server nor DB/2 uses the term that
>> I have ever seen, but I think I've heard people use the term "data
>> dictionary" in connection with Oracle to mean the vendor-supplied
>> description of an existing database.

>
> AFAIK, your definition of "data dictionary" is the classic definition of
> a
> "passive" or "off line" data dictionary.

Yes

>
> Oracle uses the term "data dictionary" to refer to the tables in the
> SYSTEM
> schema. These tables contain metadata that describes the schema objects
> (tables, indexes, views, and the like) and database objects (tablespaces
> and the like).

This is a bummer. I am doomed to endless confusion when talking to Oracle people, cuz I'm not willing to stop using the term data dictionary.

>
> If you know the definitions for the SYSTEM schema, then you can bootstrap
> your way up to knowing (except for semantics) the content of any Oracle
> database. some of this is done automatically by tools that can reverse
> engineer an existing Oracle database back to a conceptual data model, and
> then forward engineer for implementation on a different database.

SQL Server and DB/2 both have nifty and very complete system tables that you can use to find out anything. But they are of course proprietary and different. But because they describe ultimately the same thing, they can be recast to layouts that are cross-platform useful.

>
> For many years, the SQL standard described the language itself, but
> didn't
> have a tight definition for what the metadata ought to look like. Then,
> in
> SQL-92 (I think), they added standard metadata definitions to the
> standard.
> (If I'm wrong, Joe Celko will surely jump in). Most vendors like Oracle
> did
> not go back and redefine their own system tables. Instead, they just
> added views that make the metadata appear to conform to the standard.
>
> HTH.
Yeah, in SQL Server they have something called the INFORMATION_SCHEMA tables that provide a description of the systems which AFAIK are MS's offering that conforms to the standard. But they have weird gaps that made them that made them not useful in my prior pass at this project, I was better off with the proprietary tables.

-- 
Kenneth Downs
Use first initial plus last name at last name plus literal "fam.net" to
email me
Received on Wed Oct 06 2004 - 01:43:18 CEST

Original text of this message