... unless somebody where to, you know, define some precedence rules... lets see what I can come up with in 30 seconds or less.
In the stark and desolate context of XML+namespaces, an unbound prefix is a fatal error. In the warm and forgiving context of HTML, perhaps an unbound prefix could signal... an IRI?
Even so, helpful conformance checkers might want to flag with a gentle warning cases when a CURIE is found where the prefix matches a URI scheme and the reference is other than a simple NCName. Or an IRI is found with a simple ihier-part that is a NCName and an scheme that isn't bound as a prefix. If either of these rules generate too many false positives or false negatives, then the rules get tweaked a bit and we iterate a few times.
That’s just off the top of my head. This problem is unfortunate, but solvable.
It would be nice to be able to describe a video in a Web page with a standard syntax, and independently of which player is used (native, Flash, Java, Silverlight...).
I am thinking about an extension of the <video> tag to at least provide a thumbnail and a description, as I do not think this is already possible in HTML5 (or is it?).
Thanks for making a precision on the abusive usage of “real world”. The set of initial conditions one puts on his/her system is part of the real world. Maybe the sentence would have been more meaningful in saying:
“The Web by its distributed nature creates issues which are far more difficult to solve than in a controlled environment.”