[rdfweb-dev] advocating use of rdf:ID /rdf:aboutattributesonfoaf:Person tags

Bill Kearney wkearney99 at hotmail.com
Mon Aug 18 15:19:00 UTC 2003


> There's considerable value in having an email address, most people use
> there's daily, there's no value to me in having a URI

There's no current value?  I'd argue against this as quite a few folks are quite
merrily contributing their part in the form of weblogs.  An entity they're quite
a bit *more likely* to have greater dominion over than an e-mail address.  As in
joe at workerbee.corporate.example.com is very fragile if the company fires him.
Where as Joe owning his own domain, or paying for a hosted one, stands to be a
lot less fragile.  Certainly not permanent for all time but let's get real,
they're all faught with hassles (most of them quite similar).

> , in any case, remember
> the mbox is defined as the 1st user of that email address so even if it
> stops being usable, it's still fine as an identifier.  We could say the same
> with URI, however with URI's the controller of the domain is surely the
> person who defines what the URI means, so we would not be able to do it.

If the *person* owns the domain it's a lot MORE durable than an e-mail address
might be.  Remembering, of course, that owning a website URL doesn't always me
one has e-mail services on it.

If not, then the same arguments made about URI control or fragility are directly
applicable to e-mail as well.

> Consider  the URI  http://Jim.example/foaf.rdf#me as soon as the owner of
> jim.example changes who it points to (maybe they lost the domain) then all
> FOAF which points to that URI is pointing at a different person, however if
> they'd not used the URI, but had just used
> foaf:mbox="mailto:jim at jim.example" then the Person hasn't changed, and the
> new owner is specifically excluded from using that email address as a
> foaf:mbox.

Specifically excluded by whom?  Who's "in charge" of that idea?  It's just as
fragile as a URI and possibly worse.  A URI, if retrieved, has the hope of
containing /something/ that might help resolve the process.  An e-mail address
is likely to just be a black hole; a message sent to it might elicit no reply at
all.

I'm not arguing that one is better than the other.  I'm suggesting that
arguments against URI durability are weak, at best.

-Bill Kearney




More information about the foaf-dev mailing list