Re: relations aren't types?

> I hear logical contradictions:
> People can choose to access the representation of a scalar value AND the
> value is scalar because its representation is not visible?
>
> People can use dates "as a scalar value" (or not) AND this scalar-ness
> is inherent in dates themselves?

This is akin to saying that if I index into a list, it is no longer a list. The value I get from the invocation is not a list (necessarily), it's a value of the type of the elements in the list, but this does not make the list value any less a list value. In the same way, just because I access some component of some representation of a scalar value, it doesn't make the scalar value itself any less a scalar value. I suppose we could say that the representations are just syntactic sugar. All we really need are a set of operators that can extract new values from a given value. The original value is still a scalar. The representation just allows us a simple way to define all the operators necessary for dealing with the representation. Received on Wed Jan 14 2004 - 07:11:20 CET