To me, it looks like FOAF is an excellent example of the point
Janes was trying to make : that application ahould precede
specification.
Syndication started with some companies who had an application :
they wanted to pipe data from one machine to another. And the
specification followed from that.
FOAF started as a specification and the hope that people would
then go and develop applications that used it.
Ok, actually the FOAF case is a bit more complex. It started
with a misreading. It looked like there was an application : YASNS
such as Friendster, Orkut and Tribe were demonstrating it; and that
what was needed was a decentralized, supplier-independent
version.
That's a worthy goal. But in retrospect, it's not clear a) what
the true application and value of the YASNS is, or b) whether FOAF
actually captures it.
My hunch is that the original buzz about YASNS was about
"collecting" people and showing them off … to other people.
But, by definition, FOAF is designed for machine-to-machine
communication, and showing off to a machine doesn't have the same
attraction. Especially when there aren't any programs which can
turn that information into something for people to read. (Unless
you are using the very alpha PeopleAggregator)
Beyond this, Friendster et. al. gave a web-presence to users who
might not have had one. Whereas FOAF needs you to have a web-page
before you can play. (Unless you are using the very alpha
PeopleAggregator)
But if you already have a web-presence, there's another
well-established way to show-off a collection of people to other
people : the "blogroll" convention. It's much easier to add a link
to them there than it is to construct a FOAF record, even with
FOAF-o-matic.
The other excitement of the YASNS was to do with *membership*.
Of exclusive invites. Of who would accept you as a friend. FOAF,
has all the representational power (and more) of the YASNS'
databases, without any of the actual experience, application or
functionality these services provide. That's why it's
languishing.