Re: Mixing OO and DB

From: David Cressey <>
Date: Mon, 10 Mar 2008 12:49:20 GMT
Message-ID: <k7aBj.1528$Zo3.1034_at_trndny05>

"mAsterdam" <> wrote in message news:47d4e742$0$14360$
> [snip denormalization vs materialized views]
> Brian Selzer wrote:
> > Patrick May wrote:
> >> Brian Selzer writes:
> >>> Patrick May wrote:
> >> ... decoupling the application logic from the database schema.
> >
> > I don't agree with this. You're equating the database schema with the
> > database implementation. The schema specifies what information is to be
> > can be recorded. As such the schema is an integral part of the
> > specification, and it cannot be decoupled, but that doesn't mean that
> > database implementation cannot. The schema does not specify how
> > is physically recorded, nor does it specify the process by which the
> > recording takes place.
> The failure to make this distinction (what vs how with regard to
> data) would score high on a hypothetical top ten of 'misconceptions
> to get rid of a.s.a.p. - for DB beginners with an OO background'.

It's unnecessary to single out OO programmers for this comment. Back in the mid 1980s, I taught Rdb/VMS courses to DEC customers and software specialists (after learning it myself). The programmers I taught were generally from a COBOL, BASIC, FORTRAN, or C background, and not from an OO background. The distinction between "what" and "how" was just as novel to them as it is today for the OO programmer.

The "what" versus "how" distinction can be generalized from database query languages to cover just about any aspect of computing. Even the difference batween machine language and assembler can be discussed in terms of "what" vs "how", although not completely. Received on Mon Mar 10 2008 - 13:49:20 CET

Original text of this message