Re: abnormal forms

From: Bob Badour <bbadour_at_pei.sympatico.ca>
Date: Sun, 16 Apr 2006 17:23:00 GMT
Message-ID: <U3v0g.60869$VV4.1129266@ursa-nb00s0.nbnet.nb.ca>

paul c wrote:

> x wrote:
>

>> "paul c" <toledobythesea_at_oohay.ac> wrote in message
>> news:QHa%f.6021$WI1.5342_at_pd7tw2no...
>>
>> ...
>>
>>> For example, (using ttm-style braces and for convenience omitting type
>>> names), is a SUPP{S#} relation logically the same value as a SUPP{{S#}}
>>> relation that is (somehow) constrained to have only one tuple?  (not to
>>> be confused with one that somehow allows multiple tuples).
>>
>> Maybe you should define an abnormal join first. :-)

>
> Not sure if you were kidding, but as opposed to defining abnormal join,
> I wonder if it isn't already defined. Regardless, one would have to
> decide how join could be involved since it must be involved. Just what
> that would mean is another question. if an sva and an rva in two
> otherwise different relations have the same name would the join (or
> <AND>) of the two contain an rva or an sva or both?

If you are talking about a normal join, I can see arguments for two different results: 1) An error for trying to evaluate equality of values with no common supertype. Or 2) A relation of cardinality zero with an attribute whose most specific type is the universal supertype.

> if you accept that certain rva's can't be expressed as sva's with the
> same cardinality there might be times when the decision needn't be made,
> such as in a constraint expression.

Asssuming that sva means an attribute with any type not a relation, an rva and an sva have different types. One is a relation and the other isn't. Thus one can never express an rva as an sva, and cardinality is meaningless for a generic sva. Received on Sun Apr 16 2006 - 12:23:00 CDT