DanBri's Bus and the Fate of FOAF
Without exception, whenever I have been involved in a discussion about URI persistence, the question arises: "What happens to FOAF if Dan Brickley goes under a bus?" It's always the same man that gets talked about and it's always the same bus.
Here is Dan, here is a bus.
The question is not about the fate of one person: for the avoidance of doubt, if DanBri really were to go under a bus it would be an absolute tragedy and I would be among the countless mourners. The recent suicide of Aaron Swartz gives a poignant counterpoint to any apparent flippancy. But that's not what we're talking about here.
The issue is this: along with Libby Miller, Dan Brickley invented the Freind of a Friend vocabulary (FOAF), a vocabulary that has been massively implemented and that is seen as one of the basic vocabularies of the (Semantic) Web. It's hosted via a service that Dan pays for and to which he and Libby have access. That potentially amounts to a 'single point of failure' — a weak point. What if Dan and Libby are no longer around to look after the domain? How can anyone be sure that the FOAF spec and schema will still be available on the Web in a year's time, or twenty years' time?
And are instiutional online assets any more stable?
Well, yes, they are. And FOAF is stable too. And yes it will be there in twenty years' time, but that's because people have thought about it, planned for it, and made commitments. Key point: DCMI has access to the DNS for xmlns.org as well as Dan and Libby.
It's that kind of discussion that is at the heart of the study on persistent URIs I did last year that was published in the European Commission's Joinup platform and that I recently re-published on this site as a Web page.
Dan Brickley is unlikely to go under a bus I'm pleased to say, but if he did, FOAF would be OK.