Cart
[Home] [By Thread] [By Date] [Recent Entries]
> -----Original Message----- > From: John Cowan [mailto:jcowan@r...] > Sent: Monday, January 27, 2003 1:41 PM > To: Jeff.Lowery@c... > Cc: jcowan@r...; tbray@t...; > xml-dev@l... > Subject: Re: many-to-many > JC: > Oh, a representation can itself be a resource and have its > own meta-rep. Yeah, that thought occurred to me 10 secs after I hit post. I blame Monday. > The trick of using b-nodes is good in this respect: an RDF > b-node represents > a non-retrievable (non-addressable) thing-in-the-world. I'm not fluent on RDF, but taking that statement at face value: is the existence of b-node an assertion of non-addressability, or just non-addressibility within a given context? Either seems problematic to me. But maybe this one's better left for the RDF list; I don't want to load xml-dev with idle RDF questions. You can help me to the proper Draft, if you want. Or the proper draft would be better; make it a pint. > The same way we agree on a quotation convention such that > "'Boston' has six > letters" is reasonable and so is "Boston has six million > inhabitants", but > not vice versa. Hmmm... I think this sidesteps the question nicely. :-) I don't think it's possible to come up with a general or multitude of conventions to disambiguate cross-context boundaries so easily. It comes back to needing contractual understanding between parties: a common perceptual framework. But here we are again at the beginning of this permathread, so maybe it's time to hope off the circle until the next tangent approaches. In 'bout three months, I reckon.
|
