Re: Database design, Keys and some other things
Date: 27 Sep 2005 07:47:32 -0700
Message-ID: <1127832452.044079.260610_at_z14g2000cwz.googlegroups.com>
David Cressey wrote:
> "Marshall Spight" <marshall.spight_at_gmail.com> wrote in message
> news:1127799407.754373.267390_at_g49g2000cwa.googlegroups.com...
>
> > How can you derive age if birthdate isn't an attribute of the
> > person? Where would you get the birthdate from?
>
> The decomposition of the universe into entities and attributes is somewhat
> subjective. (See the discussion of "nominalism" elsewhere in this thread).
>
> The modeling of attributes is somewhat situation dependent, much as we would
> like to pretend that it is not.
>
> If I'm creating a data model for a personnel system, I'm going to model
> "date of birth" as an attribute of a person.
>
> If I'm creating a data model for a maternity ward, I'm going to model "date
> of birth" as an attribute of a birth.
> The person(s) born would inherit the attribute due to their relationship to
> the birth.
>
> It's a different model for a different situation.
Agreed on all counts.
What I was tring to get at was the contradiction between saying "birthdate is not a attribute of person" and "age is a derived attribute of person from birthdate and now()." But I fully agree that "The modeling of attributes is somewhat situation dependent."
Marshall Received on Tue Sep 27 2005 - 16:47:32 CEST