Rich Salz and I created a QName to URI mapping IETF WD, it's at http://www.ietf.org/internet-drafts/draft-rsalz-qname-urn-00.txt. It puts a solution on the table to the vexatious Qname to URI problem. I think it's very clever to put the Namespace name before the namespace URI, to avoid all those darned separator problems that URI+Name concatentation algorithms have. This is my very first IETF Working Draft, and I'll be interested in following and learning the IETF process.
An XML QName, like then content of foo
<foo xmlns:xkms="http://ww.w3.org/2002/xkms#">xkms:Reject</foo>
becomes a URN like:
urn:qname:xkms:Reject:http://w3.org/2002/xkms#
The alternatives mostly involve concatenating the Name at the end, ala
http://ww.w3.org/2002/xkms#Reject,
but what if the QName is:
<foo xmlns:xkms="http://ww.w3.org/2002/xkms">xkms:Reject</foo>
?
I have some small doubts because of the issue of the permanence of URNs and the transience of URLs.
I should also say that I really had wanted a QName to URI binding that could make dereferenceable URIs out of the Qname, as I listed in my older listing of qname/uri techniques. In my *ideal* world, the web community would have either:
1) A single canonical way of binding Qnames and URIs,
2) A mechanism for specifying an algorithm for the particular qname/uri binding.
This spec is a start on both fronts because it could be the canonical way (though I doubt it) and you know the algorithm from the uri scheme.
At any rate, it's a possible step forward to help with the Qname/URI mapping problem.
