* [2010-11-12 15:33:01 +0100] Henry Story <henry.story@bblfish.net> écrit: ] I'd start differently. Start with the social web, and simple terms such ] as foaf and sioc. The build up meanings from the ground up, piece by ] piece by introducing value at each point in the game. FOAF avoided a minefield by using foaf:knows instead of e.g foaf:friend. Still, what exactly it means that a foaf:knows b is kept deliberately vague. It probably has as many interpretations as there are FOAF profiles. Maybe there is some basic consensus about the meaning which is the intersection of all (non-pathological) interpretations. But chosing the appropriate interpretation depends very much on the context or purpose of the communication or task at hand. In your slides I think you have implicitly assumed a context which has something to do with very basic questions of identity -- this is useful but is hardly the only context in which foaf:knows links between people can be considered and it isn't at all clear if the assumptions you make will hold in other contexts. ] Global naming is going to be useful, but by taking such a big ] problem, the linked data community is just confronting many big problems ] simultaneously, which is why it can seem intractable. The network effect ] will end up working itself out. This seems very hand-wavy to me. I agree that global naming is useful. But sorting out the myriad interpretations of these global names is a hard problem that I don't think is going to just work itself out. Cheers, -w -- William Waites http://eris.okfn.org/ww/foaf#i 9C7E F636 52F6 1004 E40A E565 98E3 BBF3 8320 7664
Received on Sunday, 14 November 2010 12:34:15 UTC