Re: Unknown SQL

Just saw something worth responding too - namely, Carl attempting to redefine relational terms. Could we be above that, please?

In comp.databases.theory Carl Rosenberger <carl_at_db4o.com> wrote:
: Your argumentation is like:
: "You need to understand C to work with your newsreader, since it is
: programmed in C."

His argument is also correct. Your system has no concepts aside from the Java concepts, ergo, users of your system need to know Java.

:> >If a programmer uses any of these JDK classes, we store them directly,
:> >without any detours through relational mappers to tables, which would be
:> >unsuitable.
:>
:> Of course such a wrapper would be unsuitable. Classes map to Domains and

: What do Domains map to?
: Relational databases map classes to tables, no matter how many unnecessary
: in-between layers you invent.

It is perhaps improper to *tell* relationists what their terms are. It's always easy to win a debate when you get to define both sides' arguments. Perhaps your should believe Bob (or Date) when they say classes map to domains -- they're not redefining your terms for you, are they?

:> >Can you program a parser with it, a text editor or a web-browser?

This is just wierd. Can you program a text editor with a flat file? A database does not deal with behavior - it deals with data (and validates its correctness via constraints). Whatever your model, OO, Relational, network, etc - this does not change.

: I am only pointing out that one paradigm is given:
: The programming language.

How unfortunate, given that there are two problem spaces - the storage of validated data and the sequential operation of a program.

: Lets take Java as a starting point.
: The programming language Java uses objects to represent data.

Uh, no. Objects represent (data + behavior). Received on Sun Jul 22 2001 - 01:28:39 CEST