Andrea Montanari wrote:

I think Ed was referring to duplication of data, not duplication of
rows. Since your pkey is not part of the data you've modeled,
duplication is still possible.

Anyway, this is one of the holy wars I like most (way above vim vs
emacs), and I read many threads - and occasional swearwords and death
threats (*) - on the subject. It has been entertaining and instructive.

I've come to the conclusion that we, as species, have evolved in an
environment with a profusion of surrogate keys, for millions of years...
I'd need to research more on the matter, and publish a paper somewhere.


(*) I'm not making this up.

Andrea Montanari wrote:

I think Ed was referring to duplication of data, not duplication of
rows. Since your pkey is not part of the data you've modeled,
duplication is still possible.

Anyway, this is one of the holy wars I like most (way above vim vs
emacs), and I read many threads - and occasional swearwords and death
threats (*) - on the subject. It has been entertaining and instructive.

I've come to the conclusion that we, as species, have evolved in an
environment with a profusion of surrogate keys, for millions of years...
I'd need to research more on the matter, and publish a paper somewhere.


(*) I'm not making this up.