Re: Date is Incomplete - database application software and database theory

From: Eric Kaun <>
Date: Wed, 19 May 2004 20:30:18 GMT
Message-ID: <utPqc.308$>

"mountain man" <> wrote in message news:c5Fqc.47229$
> "Alfredo Novoa" <> wrote in message
> > On Mon, 17 May 2004 13:56:53 GMT, "mountain man"
> > <> wrote:
> >
> > >> Applications depend on the data management approach
> > >> but the contrary is not true.
> > >
> > >The modern database systems management paradigm turns
> > >a blind eye to the applications environment
> >
> > As it should be. Theory should not be limited by the current
> > technology. Theory is the basis for future technology.
> The theory needs to address current technology.

That's absolutely silly. Certainly new technologies can proceed without a formal basis, and decent ones can even suggest a new formal basis as a subject of theoretical study, but this is purely suggestion. To suggest that theory address all current technologies is nonsense - the cart leading the horse, the tail wagging the dog. Not everything that exists in reality is worthy of theoretical study.

> Current
> technology allows application data objects, in fact promotes
> these, but the RM theory does not address this.

Not true at all, though again your terminology confuses me: what is an "application object"? Relational was invented to benefit applications.

> Current technology is the stepping stone to future technology.

Not always - theory is also a stepping-stone. Not everything that exists today will survive, nor should it.

> > >It's easy to point a finger in the other direction, but in
> > >the final analysis examination of contributing causes to
> > >the current state of the nation includes the "paradigm"
> > >and its incompleteness.
> >
> > Only in your personal analysis, and I am afraid it is very wrong.


> Well I have not yet received any form of logical reason why
> the analysis is wrong, so I am unlikely to alter it.

You've offered no analysis. All you've said is "the relational model doesn't address X [where X is something I still have trouble understanding], therefore it is incomplete." No analysis at all. It would help if you'd offer something resembling a beginning of such theory; not all the details, but something that suggests you understand what the relational model does say. I simply don't know what you're looking for.

> Your opinion does not reflect the increasing trends
> of investment in these machines since 1979.


> And as I have pointed out before, the current data management
> theory is not modern, it is decades old. It reflects a practice
> that went out of date 25 years ago.

Out of date implies obsolete, which implies that something newer and better exists. What is that something?

> > Nonsenses. The division between applications and DBMS's is a perfecty
> > valid principle now.


> This assertion fails to understand the nature of RDBMS stored
> procedures: application system components totally defined within
> the RDBMS.

Simply put, those components are not application components. They're database components, which (like other database components) are used by applications.

> > It is as valid as the ancient "divide and
> > conquer" principle, and a derivation of it.
> Yeah, it is barbaric.

So Big Ball Of Mud is more civilized?

> > What don't you like in such division?
> It is out of date.

The number 0 is also old, though not as old as the number 1. I suggest we retire it as well.

  • erk
Received on Wed May 19 2004 - 22:30:18 CEST

Original text of this message