Re: Proposal: 6NF

vldm10 wrote:

> Above definition is good for the operations on one set. For example we
> can use it for one set of the attribute's values in a database.
> We can define operation on more then one set: it is mapping from set S1
> x, ... ,x Sn to S.
> For example: multiplying a vector by scalar.
> We can define a conditional operation, which is mapping from a subset
> of S1 x, ..., x Sn to S. For example the conditional operation is x -
> y, where x, y are natural numbers is the conditional operation. Here
> the condition is x < y.
>
> All these definition are based on the set concept.
> However objects on which we are applying the operations (including the
> resulting objects) not necessary need to be gathered in the sets. They
> can be characterized by the properties. For example "to be a
> point", "to be a natural number" or "to be a set".
> (The property "to be a set" is non-collectivizing).
> Hope, that this is more light on the operations.
Agreed. Shortcuts should not be taken to define the undefinable. Description certainly does not constitute mathematical proof....In a word, non determinism of set operations outputs should be accepted *de facto*.

Still idiots as Jan hidders and Keith Duggar keep creating artificial correlation between
operated values, domains and operators. Reminds me of series searches for functional expression... I just can't believe some of these people teach math!!!

> Vladimir Odrljin
Received on Fri Oct 20 2006 - 22:56:24 CEST