Re: 1 NF

>> the relation that contains the combinations of parts ever shipped in one or more shipments, a set of sets if you will. <<

This is not 1NF in most cases be a single attribute because each part has meaning in the data model. YOu teally do want to look at them. To make 1NF atomic you would need a data model in which the lowst unit is the shipment -- a container level model.

>> If it is a single attribute relation, then it may be hard to display without some device such as a "generated key". <<

You would usually have a shipment number that might be derived from source, destination, etc.

>> If certain relations we can imagine are not possible in Codd's RM without an arbitrary number of attributes then I'd think that his theory is broken. <<

The theory is still good; modeling is hard.

And not everything fits into RM. Documents and text searching require semantics and return a fuzzy revelance match number. Pictures are even worse. A human can look at a political cartoon, a symbol (Doonsberry's floating war helment) and a photogrpah and know that they are all George Buch. There is no way to write a join for that kind of matching. Fractals curves also have problems, since there is an infinite distance between any two points and you cannot determine if a point is inaide or outside of a closed fractal curve. Received on Thu Mar 01 2007 - 14:30:00 CET