Re: dbms/rdbms software & its environment

"mountain man" <hobbit_at_southern_seaweed.com.op> wrote in message news:Z9Inb.169368$bo1.106498_at_news-server.bigpond.net.au...
> There have been a few astute posts here and there
> to the effect that notwithstanding the benefit of the
> development of "the relational model" for databases,
> for the last 20 years database theory (a la Date for
> example) has remained database centric in its thinking.
>
> Is this a fair summarisation?

I don't think I understand this, so I can't really comment.

> Finally three questions:
>
> 1) To what extent is "an environment external to the
> database" referenced within database theory?

In my experience, slim to none. In practice, we see things like jdbc and odbc as APIs for the external world to access the dbms, but these are not theoretical concerns. Date makes passing mention of such things in TTM but they don't seem to interest him.

> 2) What extent of the user interface in any general database
> application software system and supporting dbms/rdbms
> (or depending on how your view it ... in any general
> dbms/rdbms software and supporting application software
> system) is external to the dbms/rdbms, and what is internal?

It's all external. The only UI elements I've seen in a database have been user-visible strings, and that usually (but not always!) makes for a hard-to-internationalize application.

> 3) In the above, "internal" means literally defined with the
> dbms/rdbms software as sql or contraints or triggers or
> indexation management etc. Considering this definition
> are there any examples of solutions in today's world
> whereby database application software systems are
> functioning close to 100% within the dbms/rdbms environment
> rather than distributed on the clients, or application servers?
> (ie: they are defined close to totally within the dbms/rdbms env)

Not that I'm aware of.

Marshall Received on Wed Oct 29 2003 - 08:03:18 CET