Re: Requesting Annotea annotations and bodies from a single query
The TAG are having that argument right now. The trouble is the notion
of URL is overloaded, and people get confused between different usages.
Example: does <URL:http://www.webthing.com/~nick/> identify a subject
that is vaguely human? A browser pointing at that URL will retrieve
a series of bytes, and render them as a page about that subject.
OTOH, RDF uses URL explicitly as identifiers (the "words" of the
semantic web), and detaches them entirely from any content that
might be retrieved by a GET on a URL. This in turn leads to confusion
when Annotea substitutes XPointers - which *do* depend on a GET
to have meaning.
> In your
> situation, each user's browser points them to different places (their
> own computer), making it not universal.
Not a problem. A URL doesn't have to point to anything at all.
The file:// protocol is perfectly valid. Don't forget, the
content of an http:// URL can be multivalued too. Try changing
your language preferences in your browser, and you'll find
different content at some familiar sites!