[rdfweb-dev] Friendster

Jim Ley jim at j...
Thu Dec 19 11:17:32 UTC 2002

"Danny Ayers" <danny666 at v...>
> >I think FOAF is viral already - it just lacks the strong incentives to
> >drive uptake. It's viral in the sense that to get listed in someone
> >else's FOAF file you really ought to create your own, thereby doubling
> >the number of FOAF files. But what's the incentive to get listed in a
> >friend's FOAF file? What benefit do I derive from being there? With
> >the Hotmail viral thing you got an email service, with FOAF you
> Good point, I can only think of the foafnaut.

Given that interests are still rather limited in nature despite the
number of people, common interest searches wouldn't give you much
currently, there's sufficient information with codepiction, and knows to
do the foafnaut interface successfully, but interest based searches there
still isn't the data. We got lucky with foaf:knows in that people
authored it without having the tools that consumed it (probably due to
Leigh's foaf-a-matic to encourage them...)

http://jibbering.com/foaf/interest.1 gives us the current list of
interests, and to be honest I don't think it tells us a lot yet, I don't
think any "friend discovery" is going to come from it...

> One other aspect of the virulence could (has been?) extended, the
> of knows: fields - so if John says he knows Mary, Mary's foaf can
easily be
> updated with a reciprocal knows:

That doesn't follow though, just because I know someone it doesn't mean
they know me, back to Tim Mansfield's pub analogy, I call the people in
the pub on the friday night my friends, but they call me the odd stranger
who's always hanging around...

> (btw, I still haven't got any friends in my foaf data but that's not
down to
> the technology

If you're in the London, I'll buy you a beer and call you "friend"...


More information about the foaf-dev mailing list