Fodder

• Hypertext web hypothesis: it's useful to have short strings stand for text (and graphics)
• Semweb Hypothesis: it's useful to relate new terms to older, more ubiqutously known ones formally. See also: TimBL's Total Cost of Ontologies (TCO) argument in his ISWC 2005 presentation

some conjectures, or perhaps open problems

• Only URIs that connect to communications protocols are worthwhile; URNs don't sufficiently facilitate consensus of meaning. Re "How can I mint a URI that refers uniquely to myself, and maintain control of it?" you can only control the meaning of a URI inasmuch as the world finds it useful.
• FOL is insufficiently expressive; we need PRA
• FOL is overly constraining; we need to drop LEM

Points to make:

• scaling down is as important as scaling up; I shouldn't have to make an international treaty to make a lunch appointment with a colleague.
• tagging is valuable
• URI authority comes from consent of the governed
• Dns -> ip mapping :: uri->meaning mapping

cases to explore:

• follow-your nose, self-describing web. start with text+a programming language or two, add another language
• more on Grice's maxims and inform/request(command)/ask
• cyc:subabstrac, time context
• Global namespace per village. Lunch w sandro ant on the dog story