Re: does a table always need a PK?

From: Ed prochak <ed.prochak_at_magicinterface.com>
Date: 29 Aug 2003 08:03:35 -0700
Message-ID: <4b5394b2.0308290703.1f0b1179_at_posting.google.com>

"Heikki Tuuri" <Heikki.Tuuri_at_innodb.com> wrote in message news:<q593b.818$n62.433_at_read3.inet.fi>...
> Christopher,
>
> "Christopher Browne" <cbbrowne_at_acm.org> kirjoitti viestissä
> news:bij2l3$9tgnj$3_at_ID-125932.news.uni-berlin.de...
> > Centuries ago, Nostradamus foresaw when "Bob Badour" <bbadour_at_golden.net>
> would write:
> > > Even still, I find it unforgivable for someone to implement a
> > > self-proclaimed rdbms and not have any familiarity with this book. It is
> > > unforgivable for someone to claim an opinion on what constitutes an
> rdbms
> > > and not even know what 4VL is.
> >
> > I haven't seen any opinion being expressed, on the "MySQL side," as to
> > what they think constitutes an RDBMS.
> >
> > Certainly nothing resembling the "Peano Axioms" that the MySQL guy
> > seems to think _everyone else_ should consider as the level of
> > discourse.
>
> I have said that the concept of an 'RDBMS' is vague. Lee Fesperman and Bob
> Badour say it is not vague. It is their job to provide the exact definition.
> In the threads with Mikito Harakiri and Jan Hidders you find some reasons
> why I am asking these questions.
>

I've been observing this thread and really you miss the point. Lee and Bob mentioned the paper. Bob gave a link to it. Reading it leads to this:

<begin quote>
The term relation is used here in its accepted mathematical sense. Given sets S1, S1, ···, Sn, (not necessarily distinct), R is a relation on these n sets if it is a set of n-tuples each of which has its first element from S1, its second element from S1, and so on. We shall refer to Sj as the jth domain of R. As defined above, R is said to have degree n. Relations of degree 1 are often called unary, degree 2 binary, degree 3 ternary, and degree n n-ary. For expository reasons, we shall frequently make use of an array representation of relations, but it must be remembered that this particular representation is not an essential part of the relational view being expounded. An array which represents an n-ary relation R has the following properties:

1.Each row represents an n-tuple of R. 
2. The ordering of rows is immaterial. 
3. All rows are distinct. 
4. The ordering of columns is significant - it corresponds to the

ordering S1, S1, ···, Sn of the domains on which R is defined (see, however, remarks below on domain-ordered and domain-unordered relations ).
5. The significance of each column is partially conveyed by labeling it with the name of the corresponding domain. <end quote>

So Codd's paper is not adhoc, but really mathematical. The properties listed look very axiomatic to me. And to get back to the topic - is a primary key needed? - I think property 3 says it all. In the extreme case, all colums together form the key, but usually you have something less. So a PK is required. The only significant space for discussion is how to identify that key. (since there are practical reasons for wanting a PK that is less than all the columns). But tables must have a PK for it to be a relational DB by Codd's Definition as quoted above.

I hope that answers at least this question.   Ed Received on Fri Aug 29 2003 - 17:03:35 CEST