[foaf-dev] foaf:mbox but not IFP?

[Dan Brickley]

On 04/01/2008, Alexandre Passant <alex at passant.org> wrote:
> On Jan 3, 2008 5:16 AM, Dave Brondsema <dave at brondsema.net> wrote:
> >
> > What should I use instead of foaf:mbox when the property should NOT be
> > an Inverse Functional Property?  Example: I'm starting an address book
> > (and more) webapp for my family and my Aunt and Uncle both use the same
> > email address.
> >
> > http://xmlns.com/foaf/spec/#term_mbox says "Note that there are many
> > mailboxes (eg. shared ones) which are not the foaf:mbox of anyone" but
> > it doesn't say what they should be.
>
> Why not using it for a foaf:Group which includes your Aunt and Uncle ?
> So, the mbox will be an IFP for a group consisting of both of them.
> It will be OK, since the domain of this property is a foaf:Agent, not
> a foaf:Person, unless "ie. an Internet mailbox associated with exactly
> one owner, the first owner of this mailbox." means that the owner
> should be only a single person ( but in that case the domain should be
> changed in the specs).
> IMHO, using a group makes sense.

Yup, I was about to suggest Group, which you can just write as
<Agent>, since all groups are agents. There are some properties which
are Person-only (see http://xmlns.com/foaf/spec/images/foafspec.jpg
(unless you've tabulator installed, sorry)) but in general that should
work.

It is possible to be fancy and enumerate the contents of the group,
but for a simple addressbook, just writing "Agent" with name, mbox and
other applicable properties should make sense.

We currently have :gender as a property of agents, for example (this
allows non-human agents like pets or fictional beings to have gender).
I'd suggest most groups probably don't have a gender (this is noted in
the spec already :)

cheers,

Dan