W3C home > Mailing lists > Public > www-rdf-interest@w3.org > November 2001

Re: Domain/Range: conjuntion or disjuntion??

From: Peter F. Patel-Schneider <pfps@research.bell-labs.com>
Date: Wed, 21 Nov 2001 11:56:18 -0500
To: tarod@softhome.net
Cc: brian_mcbride@hp.com, www-rdf-interest@w3.org, jena-dev@yahoogroups.com
Message-Id: <20011121115618Z.pfps@research.bell-labs.com>
From: tarod@softhome.net
Subject: Re: Domain/Range: conjuntion or disjuntion??
Date: Wed, 21 Nov 2001 16:16:35 GMT

> 
>   Tricks almost never work as you hope, I agree with that. My point was,
> you can be sure that an instance of CarsAndMotos is instanceOf Car and
> instance of MotorCycle, but you can't not be sure that instance of Vehicle
> is instance of Car or instance of MotorCycle.
> 
>   So, with CarsAndMotos I can simulate the conjuntion view of domains, but
> with Vehicles I can't simulate the disjuntion view of domains (and ranges).
>   For DAML projects, with CarsAndMotos you can perform inference without
> any problem.
> 
>   Regards,
>          Marc

Yes, your trick is one-sided one way and the other trick is one-sided the
other way.  You claim that your trick captures everything you want, but
that does not necessarily mean that your trick does not lose anything.  

In particular, suppose that I want to say the following:

	If something is in the domain of a property then it is allowable to
	have a value for the property.

This is not captured by your trick.

peter
Received on Wednesday, 21 November 2001 11:56:29 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 7 December 2009 10:51:52 GMT