W3C home > Mailing lists > Public > public-sws-ig@w3.org > May 2004

Re: Superfluous attributes in OWL-S?

From: David Martin <martin@AI.SRI.COM>
Date: Mon, 10 May 2004 12:57:46 -0700
Message-ID: <409FDEBA.7020403@ai.sri.com>
To: Daniel Elenius <daele@ida.liu.se>
Cc: public-sws-ig@w3.org

Hi Daniel -

It's a fair question.  I don't know of anyone who claims that these 
properties are crucial for any reason, and I'm not aware (offhand) that 
they've been used in any significant ways.

I should note that these properties were not conceived for the purpose 
of providing a unique identifier, but more for something human-readable.

When you say "instance names", do you mean the rdf:ID?  If so, I don't 
think of that as a suitable replacement for a "name" property, in 
general.  I imagine the rdf:ID will be automatically generated in many 
situations, and often may not be human-oriented (readable).  Also, it 
may be desirable to ensure that rdf:ID is unique.  So I wouldn't want 
that to be the only way to give a name to something.  Also, I don't see 
why a process shouldn't have its own distinct name (that is, different 
from a related Profile).

I think probably the best argument for eliminating these properties is 
that rdfs:label can be used for this purpose.


Daniel Elenius wrote:
> Hi.
> What's the point of these datatype properties:
> process:name (in Process.owl)
> profile:serviceName (in Profile.owl)
> They are both defined to range over xsd:string. It seems to me that the 
> instance names could be used instead. At least process:name seems 
> entirely superfluous. Perhaps serviceName is reasonable if a service is 
> to have a "public name" shown for purposes of discovery, that is 
> different from the name of the Profile instance.
> Regards,
Received on Monday, 10 May 2004 15:57:04 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 17:32:45 UTC