Well now you've gone and started another holy war!  While I agree with Date's findings for the most part, I'm not quite prepared to accept his complete prohibition on NULLs,...

Don't do it!  This is a train wreck waiting to happen.  This is essentially the same thing that I refered to as a MUCK table in this article:

http://www.sqlservercentral.com/columnists/dpeterson/lookuptablemadness.asp

Also see...

Ahh, I wondered if that was you.  It must be rough out there in paradise!

While the MUCK table might be properly normalized when viewed in isolation, the database that utilizes a MUCK cannot possibly be properly normalized.

Good articles on database design are always needed. ...

Actually, I made a typo.  I should have said 3NF since 1NF says nothing about key dependence...  Sorry for any confusion. 

So while Troy's point about 1NF is valid (assuming...

To add to what Joe said:

Use of MUCK tables violates either the Information Principle or 1NF--or both.  It is common to see a MUCK table defined with like so

CREATE TABLE...

I think you are minimizing the risks of a MUCK table and maximizing the percieved benefits...

The risk of drift is real, but using MUCKs does nothing to address it.  Using...

STOP, LOOK, and LISTEN.  This is a terrible way to design a database.  Follow the link above and look at my article here:

http://www.sqlservercentral.com/columnists/dpeterson/lookuptablemadness.asp

and look at the responses to this...

It isn't possible to use these abominations properly. 

I don't see how defining views can possibly help this situation.  Views can't be referenced by FK constraints.  Using such a method throws you immediately back to triggers or UDF's...

Not unless you are willing to allow them to alter the table (by adding a FK constraint).  This can get pretty messy when it comes time to roll out upgrades...

Customer configurable attributes can be a huge problem.  The fact that commercial software so often can't properly define the requirements for all of their potential customers is one of the...

So, you lost the log file(s) but the data files were fine?  Then you restored the log file and attempted to attach the database?

If by restored, you mean that you...

Ahh, I didn't specifically address the old UDF enforced integrity in my article, but the same argument that I used against triggers applies here.  Using the MUCK approach, your referencing...

Yes and no...  This is generally the right idea, but it won't do much good if your various meta-luns are all ultimately created from the same raid arrays (which is...