[foaf-dev] proposal for FOAF 1.0

Etan Wexler ewexler at stickdog.com
Sun Jun 13 18:25:26 CEST 2010


Dan Brickley (in
<http://lists.foaf-project.org/pipermail/foaf-dev/2010-June/010294.html>),
asking about my proposal for a FOAF 1.0 that would define only one
property:

> What would it be useful for?

Well, <http://xmlns.com/foaf/v#a> covers a strict superset of the
cases that <http://xmlns.com/foaf/0.1/knows> covers. If you think that
<http://xmlns.com/foaf/0.1/knows> is useful, then you should see that
<http://xmlns.com/foaf/v#a> is useful. Setting domain and range aside,
the current FOAF specification has very slightly more normative
content about <http://xmlns.com/foaf/0.1/knows> as the specification
that I have proposed has about <http://xmlns.com/foaf/v#a>.

I intentionally omitted domain and range from the specification of
<http://xmlns.com/foaf/v#a>. I expect that, if FOAF becomes truly
popular, document authors will publish assertions that something that
is not a person knows another thing and assertions that something
knows another thing that is not a person. I expect, for example, that
one would find an assertion that somebody knows some group. The
current FOAF vocabulary and specification decree that the assertion
implies that the group is a person. The lay author, though not
intending to imply that the group is a person, has too little
familiarity with the semantics of RDFS to understand the problem. No
amount of tutorial effort will correct the mass of lay authors.

Asserting (and specifying) that the domain of the acquaintanceship
relation is the set of persons raises deep, intractable arguments
about personhood. I have, therefore, avoided specifying that the
domain of <http://xmlns.com/foaf/v#a> is the set of persons. It’s
reasonable to assert (and specify) that the domain of the
acquaintanceship relation is the set of agents. I have avoided
specifying that the domain of <http://xmlns.com/foaf/v#a> is the set
of agents because I want to avoid the selection of a particular term
for the set of agents, coining a term for the set of agents, and
bloating the specification with synonymous domains.

-- 
Do not include my address in public replies. I will read public
replies on the list.


More information about the foaf-dev mailing list