Re: Question on functional dependency
Date: Thu, 23 Sep 2004 14:49:34 -0400
Message-ID: <2tidnYbcIsJBh87cRVn-hg_at_comcast.com>
"Josh Hewitt" <lajos.nagy_at_gmail.com> wrote in message
news:1c92edeb.0409230805.6b331e9a_at_posting.google.com...
> For example, in a relation on employees, usually the functional
We all know about "assume", right?
We don't have to assume that this is true. It's one of the questions we can
ask during the analysis phase.
If the FD is true in "the real world under consideration", the SME will be
able to tell us.
If you approach an existing set of data, and begin trying to find how
normalized it is, it's a good, if somewhat back assward, way of finding out
what the FD's really are.
It's amazing how many times it's useful to do exactly this.
One of the other things you get when you ask a programmer to design a
database is, "we don't have to learn that much about the subject matter.
We'll leave that for version 2."
There ARE some programmers who can become excellent database designers,
after suitable deprogramming. I consider myself a case in point. What gets
most programmers in trouble on their first database design is that they
> dependency (FD) Emp# --> Dept# holds. Notice though, that we assume that
> this is true because of the meaning that we assign to the attributes of
> the relation: in the real world an employee is assigned to at most one
> department.