Re: why arcs never be merged?

At 12:02 PM 11/29/02 -0800, JeffZhang wrote:

>Dear all,
>   When two graphs merged, arcs are never merged,why? What the difference 
> between a single
>assertion and duplicated assertions?

Do you mean the distinction between these graphs:

    ex:foo ex:property ex:bar .

and

    ex:foo ex:property ex:bar .
    ex:foo ex:property ex:bar .

?

Semantically, they are the same in that each entails the other.  How they 
are treated syntactically is rather up to an application.


Or do you mean:

    ex:foo ex:property ex:bar .
    ex:foo ex:property ex:bif .

?

These are two distinct statements -- what would it mean to merge the 
arcs?  Each arc is, in effect, the combination of subject, object and 
predicate:  no arc exists separately from a statement.  The URI that labels 
an arc may, however, be a separate node in the graph, and the presence of 
multiple arcs with a given URI don't mean that the URI node can appear 
multiple times.

A slightly different way of looking at this is that the property URI is a 
kind of type label, and an arc is an instance of that type.

>   The literals,which were not merged in previous specificaions, are 
> merged in this version.

That's down to a semantic change in the nature of literals, a big 
discussion quite different from the matter of arcs.

>Would the same change happen to arcs?

There is no discussion that I've heard that calls to change the role of arcs.

#g


-------------------
Graham Klyne
<GK@NineByNine.org>