Re: Concrete datatype use case

At 06:39 PM 10/24/01 +0100, Brian McBride wrote:
>I would agree that having two size properties doesn't change the meaning 
>in this case, but if I'm paying extra money to store or process them, then 
>it matters to me.

Ah, implementation details ;-)  In such a case I think that you would pay 
for a superior implementation that is able to recognize the equivalence 
(through knowledge of the data typing) and only store the information once 
-- possibly selecting a canonical form -- on the basis that this entails 
all the other forms that might be used, assuming the data typing is the 
same in all interpretations considered to be models.

#g


------------------------------------------------------------
Graham Klyne                    MIMEsweeper Group
Strategic Research              <http://www.mimesweeper.com>
<Graham.Klyne@MIMEsweeper.com>
------------------------------------------------------------

Received on Wednesday, 24 October 2001 18:56:09 UTC