Atom namespace, qname-uri-qname roundtripping

[Benjamin Nowack <bnowack@xxxxxxxxxxxxxxx>]

Hi,

as the atom namespace is declared as "http://www.w3.org/2005/Atom";
(i.e. without a trailing separator such as "/" or "#"), term URIs
from atom documents are expanded by xml parsers to
   - http://www.w3.org/2005/Atomfeed
   - http://www.w3.org/2005/Atomtitle
etc.
This makes it difficult for automated qname generation after
atom docs have been processed by generic aggregators which
mix terms from different vocabularies. It would be nice if
atom adopted good practice from rdf vocabularies such as
dc, cc, rss 1.0, etc and used a trailing separator, so that
re-serialized data doesn't end up as
   <Atomfeed xmlns="http://www.w3.org/2005/";>
which will then break simple atom parsers that ignore
xml namespaces.

(sorry if this has been discussed before, but the ML archive
didn't seem to be searchable..)

cheers,
benjamin

--
Benjamin Nowack

Kruppstr. 100
45145 Essen, Germany
http://www.bnode.org/