Re: One-To-One Relationships

On 1 déc, 20:26, JOG <j..._at_cs.nott.ac.uk> wrote:
> On Dec 1, 7:13 pm, Cimode <cim..._at_hotmail.com> wrote:
>
> > On 1 déc, 18:21, JOG <j..._at_cs.nott.ac.uk> wrote:> On Dec 1, 4:52 pm, rp..._at_pcwin518.campus.tue.nl (rpost) wrote:
>
> > > > snip
> > > > ER modelling allows me to assert the identifiability of a relationship
> > > > *without* reference to identifying attributes.
>
> > > This is a contradiction to your previous statement. A relationship can
> > > not be identifiable if it has no identity. Its an oxymoron.
>
> > Huhh...Isn't a name a way to identify an entity?
> > [Snipped rest because I agree with most part]
>
> A name is an attribute, just like any other attribute.
And does not a name is the best attribute to distinguish an entity from another entity ? Is it mandatory in ER to suppose that no attribute should have the role of distinguishing the entity? Does E/R modeling prevent the integration of metadata as being a part of entity definition ? I am asking because I am not familiar with rules of E/R modeling... Received on Sun Dec 02 2007 - 14:21:20 CET