Re: Unique Keys

Mikito Harakiri wrote:

>

>> yes, but that is not a unique constraint.  Is there anyway at all to
>> imagine it as a declarative unique constraint?

Not a domain, but a datatype.

In another subthread of this thread I suggested a datatype interval, composed of an ordered pair of any two values of the same base type, (a,b) where a<=b. Equality could then be defined as:

where x = (a,b)
and   y = (c,d)

x = y iff (a <= c <= b) OR (a <= d <= b)

This allows the declarative Unique and RI constraints.

The actual purpose of the thread was to ask if anyone thought this possible in any form. All positive answers led towards a datatype, which was what I wanted to find out.

FWIW, this entire thing has only encouraged me to believe even more that many 'special' constraints can be reduced to unique and referential, with the possibility of a few carefully chosen new data types becoming an important part of the solution.

-- 
Kenneth Downs
<?php $sig_block="Variable scope? What's that?";?>