Re: Can the value set of a subsetting association end be equal (not a proper subset) to the set of the subsetted association end?



Re: Can the value set of a subsetting association end be equal (not a proper subset) to the set of the subsetted association end?

From: Arend Rensink <rensink_at_cs.utwente.nl>
Date: Mon, 22 Oct 2007 12:16:36 +0200
Message-ID: <471C7884.20606@cs.utwente.nl>
Wendell Ocasio wrote:

> Is this an inadvertent inconsistency?  Are non-proper subsets (ie. the collections are equal) allowed.  And if so, what would be the rationale?  Why not use redefines instead?

This seems to me an inconsistency which should be resolved by allowing subsets to be equal, i.e., not requiring them to be proper subsets. The reason why believe (hope?) this to be the right interpretation is that the usefulness of the subsets attribute would be greatly diminished otherwise: typically the point of declaring a subset is not to express that there are fewer instances of something, but to express that the presence of one thing implies the presence of another.

Redefines is not typically an alternative: the way I read it, with redefines, the "upper" (redefined) feature may *not* occur on instances of the "lower" classifier, whereas for subsets, the presence of a "lower" (subsetting) feature implies the presence of the "upper" one.

Yours,

-- 
Arend Rensink                         http://www.cs.utwente.nl/~rensink
Department of Computer Science             mailto:rensink@cs.utwente.nl
University of Twente                               tel: +31 53 489 4862
P.O. Box 217, NL-7500 AE Enschede, Netherlands     fax: +31 53 489 3247
Received on Mon 22 Oct 2007 - 11:16:47 BST