Re: Another view on analysis and ER

David Cressey wrote:

>
>"rpost" <rpost_at_pcwin518.campus.tue.nl> wrote in message
>news:c1f35$4766eca5$839b4533$16718_at_news1.tudelft.nl...
>> David Cressey wrote:
>>
>> [...]
>>
>> >The question of whether it's "the same entity" or "a different entity" is
>> >also moot.
>>
>> Until you consider other relations referring to the same entity.
>> Their attributes may or may not have to be updated as well.
>
>Are they really "referring to the same entity"?

Yes. (What else do you expect me to answer?)

>Or are they "referring to
>data values that describe and/or identify some entity"?

Nope. That's the whole point of this discussion: sometimes we're able to able to state that two references refer, in the given context, to the same thing (i.e. that changes to that thing affect both occurrences) without having said how the thing is identified.

>Can an entity be
>identified without data?

Regarding references from the model to the outside world, and assuming that all access to the data is in terms of the query (and update) language): no.

Regarding references within the model (if the model allows them): yes.

-- 
Reinier