Re: The stupidest design I ever saw
Date: Tue, 11 Apr 2006 18:09:10 +0300
Message-ID: <e1ggpn$ftb$1_at_emma.aioe.org>
"Neo" <neo55592_at_hotmail.com> wrote in message
news:1144767112.909515.213850_at_i40g2000cwc.googlegroups.com...
> > > > > "Formal Concept Analysis in Information Science" article by Uta
Priss
> > > > > http://upriss.org.uk/papers/arist.pdf
> > > Now a friendly alien from Mars (named bill) arrives on Earth. He is
capable of using his tools to dignose health problems and write
prescriptions to cure various ailments. This alien is a doctor but he is not
a human. To avoid such dilemmas start with a non-hierarcal arrangement of
classes as shown below. (It is difficult for you and most other people to
realize that there is an indirect, but no direct, relationship between
classes human and doctor, which can be derived from the data below).
> .
> > Then we need to redefine our terms. If we had said that doctor is-a
human, then we cannot say this alien is a doctor until we redefine that is-a
relationship.
.
> Perhaps you are one of the people for whom it is difficult to realize
> that there is an indirect/derivable, but no direct, relationship
> between classes human and doctor. Perhaps you prefer to setup initial
> data structures that will allow future dilemmas so that you can
> restructure shemas, migrate data and update related
> scripts/queries/code later.
What is wrong with deleting the is-a(doctor,human) from the is-a relation
and adding the is-a(alien,doctor) ?
Why is is-a relation an indirect relation ?
Received on Tue Apr 11 2006 - 17:09:10 CEST