Re: Impossible Database Design?

mAsterdam wrote:
> Jay Dee wrote:
>

>> Regarding domains:  the question, "Is absorption a property of some value
>> in the domain?" must be answered depending on what the domain is and how
>> it can reasonably be represented.  

Absorption is a property of infinity:

   oo + 1 = oo
Adding something to the value doesn't change the value.

Some of the contributors to this thread seemed to desire a value which means "the end of an interval which has no end." My point is that DD+L describe a technique that accommodates domains having such a value; all their technique requires is a finite set of values and few operators having prescribed semantics.

In the case of time, oo might mark an instant which will never come:

        until t = oo { <do something>; t++ }

would run forever -- at least, as well as we can know forever.

There are plenty of untidy questions to be dealt with: "How much is one less than infinity?" "How can I represent values which can be incremented forever and not reach infinity?"

There are analogs in arithmetic: "Is any positive value divided by zero equal to infinity?" Perhaps, "Is any positive value divided by infinity zero?" Iverson used an arithmetic in which any value divided by itself is unity, and deftly handled all the "zero divided by zero" cases. It would also be useful when one asks, "How much is infinity divided by infinity?" (Remember the 'N / oo = 0' question!)

All these questions are left to the domain modeler. Of course, if you don't need an infinity, these questions go away... Received on Sat May 20 2006 - 13:41:46 CEST