Actually it's simpler than that....

Quite simply I would have a relation for each unique type of screek. From what you have said that...

Jeff Moden (8/1/2010)

---

Maybe I have your intention wrong but look again. Tom said that a NULL could not be used to explain WHY data was missing (or unknown my...

Jeff Moden (8/1/2010)

---

Oh jeez! C'mon and get off your high horse, David. What's next? A lecture for hours to come on what you think the semantical differences...

Tom.Thomson (7/29/2010)

---

If anyone attempts to use SQL's single null to indicate why data is missing they are being very stupid. Codd explicitly required NULL (and later the A_NULL mark)...

I read what you wrote a few times but...

Jeff Moden (7/28/2010)

---

Since Codd rather well endorsed NULLs, why are they considered such an irritant even by some of those that appear to follow his every word with nearly religious...

I am using the term "view" in the same sense that Codd used it (at least in the RM/V2 book). View meaning derived relation.

My point was simply that in Codd's...

Tom.Thomson (7/25/2010)

---

Note that the guaranteed-access feature represents an associative-addressing scheme that is unique to the relational model. It does not depend at all on the usual computer-oriented addressing. Moreover, like...

doobya (7/24/2010)

---

plenty of validation at the application level

plenty more at...

Ensuring the right results from your data is more important to most business users than either simplicity or performance. So if you sacrifice accuracy in the name of simplicity or...

KIM: Good morning, friend. How are you today?

EDGAR: Very well thanks. But what about you? If you don't mind me asking, why are you walking in that very strange and...

This is a hopeless task unless the data is never going to change. Otherwise the results of your static analysis could prove invalid the moment the data first gets updated....

It is included in the disc version of the ACM SIGMOD anthology, although I don't think the text is available online.

http://dblp.uni-trier.de/db/conf/icde/Codd86.html

That's the rebublished version of the Computerworld article.

The 13 "rules"...

You would actually need two CHECK constraints because this is effectively a constraint on both tables whereas each CHECK constraint can apply only to one table.