Re: Can we solve this -- NFNF and non-1NF at Loggerheads

"Dan" <guntermann_at_verizon.com> wrote in message news:gpWNd.15063$uc.3993_at_trnddc08...
>
> "Alan" <not.me_at_rcn.com> wrote in message
> news:36qgunF4ut60pU1_at_individual.net...
> >
> > "Dan" <guntermann_at_verizon.net> wrote in message
> > news:1107817027.196841.324050_at_g14g2000cwa.googlegroups.com...
> [snip]
> >
> > They do exist as encapsultaed data elements, but you are comingling the
> > Relational Model and an implementation.
>
> I did say we were supposing, didn't I?
>
> When modeling the data (in the
> > Relational Model world), one sticks to the rules of normalization.
>
> [snip]
> > Just because you can model something relationally, and then go off of
that
> > model during implementation, it does not make the non-relational items
> > relational. You no longer have a normalized database, that's all. It's
not
> > good or bad, it's just (hopefully) appropriate for what you need to do.
>
> I think we'll have to recognize and respect the idea that we have a
> different perspective on this. My belief is that if the RM is unaware of
> the contents or internal (particularly logical) representation of some
type,
> it is outside the bounds of the "data model" and normalization rules. If
I
> understand you, you are of the opinion that this encapsulation is
> irrevalent, and all internal logical structure is part of the data model.
I
> think we can agree to disagree on this point.
>
> - Dan
>
>

I think we actually agree more than it seems, and we just have a different way to explain the idea. In any case, I agree to agree or disagree, whichever it is :) Received on Tue Feb 08 2005 - 15:11:17 CET