Re: collection access: inferred property vs. syntax

On Wed, Feb 02, 2005 at 05:16:06PM +0900, Yoshio FUKUSHIGE wrote:
> 
> >So if the server understands predicates that an intermediate might not
> >understand, it needs to break the data down to a universally
> >understood form:
> >
> > Eric sharesOfficeWith _:list1.
> > _:list1 owl:member Philippe.
> > _:list1 owl:member David.
> >
> >(which loses the closed aspect to the list).
> 
> Well, then in the case you cited, we will get another graph:
> 
>  Eric sharesOfficeWith _:list2.
> _:list2 owl:member Michiko.
> _:list2 owl:member Yoshio.
> 
> #Let's suppose they are all your colleagues in Keio.
> 
> So, what I want to say is that what you get (i.e. solutions above)
> is different from the following and will cause no problem.
> 
> Eric sharesOfficeWith _list.
>  _list owl:member Philippe.
>  _list owl:member David.
>  _list owl:member Michiko.
>  _list owl:member Yoshio.

  {?x sharesOfficeWith ?l.
   ?l owl:member (?m1, ?m2). } => { ?m1 sharesOfficeWith ?m2 }

You're right. So I guess the only issue I remain concearned about is
the loss of boundries on the list.

Yoshio (verbally) explored the idea that custodians of owl:member
statements are resposible for maintaining completeness. Given that
the query

  SELECT *
  WHERE { ( Eric sharesOfficeWith ?l ) ( ?l owl:member ?who ) }

can't distinguish this graph with a closed list:
  
  Eric sharesOfficeWith _:list1.
  _:list1 rdf:first Philippe.
  _:list1 rdf:rest  _:list2.
  _:list2 rdf:first David.
  _:list2 rdf:rest  rdf:nil.

from this list with a repeated property:

  Eric sharesOfficeWith _:list1.
  _:list1 owl:member Philippe.
  _:list1 owl:member David.

we can shift the burden of closure onto the database and protocol by
saying:

  Any custodian of owl:member statements must not capriciously removed
  some of them without removing all of them and the arc that references
  the list.

  Any list enumeration that is, or may have been, terminated before
  the list was competed must behave as if some of the owl:member arcs
  had been removed. That is, it must removed the remaining owl:member
  arcs and the referring arc.


I think this conveys Yoshio's thoughts. Corrections welcome.
-- 
-eric

office: +81.466.49.1170 W3C, Keio Research Institute at SFC,
                        Shonan Fujisawa Campus, Keio University,
                        5322 Endo, Fujisawa, Kanagawa 252-8520
                        JAPAN
        +1.617.258.5741 NE43-344, MIT, Cambridge, MA 02144 USA
cell:   +81.90.6533.3882

(eric@w3.org)
Feel free to forward this message to any list for any purpose other than
email address distribution.