Re: relations aren't types?
Date: Tue, 30 Dec 2003 01:53:37 -0500
"Marshall Spight" <mspight_at_dnai.com> wrote in message
> "Lauri Pietarinen" <lauri.pietarinen_at_atbusiness.com> wrote in message
> > My understanding is that in the scope of relational databases,
> > atomicity is defined in terms of whether the relational operators can
> > "see" the value
> > or not without the help of "non-relational", or scalar operators.
> Interesting. I wonder if that's definition Bob is using, thus causing
> our disconnect about atomicity. Where did you get this definition,
> may I ask?
I don't make any distinction whether the operations are "non-relational". If all the dbms knows about the contents of attributes is a set of operations that apply to them as values in toto, then the values are atomic.
Thus a string type might have an operation: MID(String s,Integer start,Integer length): String
From the logical perspective of the dbms, the MID operation operates on the whole of the string value, s, and just as it operates on the whole of the integer values, start and length.
Conceptually and physically, the MID operation may operate on components of the String value, but logically it operates on the whole thing.
When dealing with relation valued attributes, the relational model treats relational operations applied to those attributes just as it does any other domain operation.
In my mind, the only operation that treads near a gray area is the ungroup operation. Received on Tue Dec 30 2003 - 07:53:37 CET