Re: Dawn doesn't like 1NF

"Laconic2" <laconic2_at_comcast.net> wrote in message news:usqdnWvN0NSkvezcRVn-rQ_at_comcast.com...
>
> "Marshall Spight" <mspight_at_dnai.com> wrote in message
> news:qp8cd.374924$mD.201649_at_attbi_s02...
> > >
> > > I am not sure you got my point. In order to carry out the process you
> > > outlined, you have to know what the logical subelements ARE. For a
> > > UDT, only U knows what the logical sublements are.
> >
> > Why should it be the case that only the user knows what the logical
> > subcomponents are?
> >
> > [This is the fundamental question at hand; everything else I've
> > written below is just a distraction from this essential question.]
>
> Because the user defined the datatype. The meaning of equality is type
> specific. If the system didn't know the datatype to begin with, then it
> doesn't understand equality for that type.

I guess I don't understand this perspective. Perhaps an example is in order. What advantage is there in keeping the system in the dark about the logical components of a data type? It seems like a big disadvantage to me.

Marshall Received on Sat Oct 16 2004 - 18:44:15 CEST