David Burrows (4/27/2011)

---

If, as you state, the DDL in question does conform to Date's 1NF as in 'There's no top-to-bottom ordering to the rows' can you post example DDL that...

....

You have never read 1NF definition or even the Relation definition.

Example of 1NF definition, from the only book I have right now :

Database Systems - The complete Book Second Edition...

You'r so wrong about 1NF, I can't add anything.

........ I can't believe it...

YOU are the...

Obviously you don't know what a predicate is, or you'r just ignoring predicate variables. Nice 😉

You'r calling Occam's razor on the number of tables ? What a joke... Why are...

Well we have a problem here.

I think you didn't understand what 1NF is.

Let's take data example to make it clear !!

What you are citing are the properties a TABLE must...

tabraham 36699 (4/26/2011)

Ideally, Chris Date's vision of an OO-Relational DB will emerge, with, among other things, a formally defined range for each attribute and an explicit value representing UNKNOWN, with...

@tom.Thomson

Tom.Thomson (4/25/2011)

---

Assuming <S#> is a key, that's only slightly different from my version:

For supplier S# the information we have about his name is indicated by SNAME and the...

What the ??

What is violating 1NF, and what is your definition of 1NF ?

I give up, good luck guys 😉

@tom.Thomson

When you write Dace, I guess you mean Date.

I actually own almost every books Chris published. He wrote hundred of pages on NULL topic.

Maybe I'll come later and...

I don't know Dace's work, but maybe you could explain it for me,...

Tom.Thomson (4/24/2011)

---

This of course relies on using the original definitions of the normal forms, not the perverted versions with an explicit forbidding of NULL added long after the definition was...

I have no issues with you address solution but it requires an extra join and it can degrade performance.

If the objective is perfect normalization then go with an extra table...

My demonstration was maybe too short, but this was my point.

NULL impacts nomalization...

Tom.Thomson (3/7/2011)

---

But Heath's theorem isn't the point...

Tom.Thomson (1/26/2011)

---

David Portas (1/22/2010)

---

Irrespective of that, the Normal Forms, upto and including 6NF don't leave any room for the possibility of nulls. Codd himself says so: "The concepts and rules...