Simon St.Laurent makes the same argument in a few places, but I'll select
the one near the end of his message.
Creating a first-level xml space tells generic XML processors that they can
do _something_ to the material contained in the file. Then the second level
can provide a more specific description useful for processors that don't
want to waste their time with the wrong XML information.
There are other ways of doing this without an XML top-level type. The
process in your MUA or other viewer that dispatches a document of a
particular type/subtype pair can be told to launch an XML viewer even if
neither the type or the subtype says "xml". In the case of
"text/x-xmlnews", if you don't have an "xmlnews" viewer, you can still tell
the MIME dispatcher to dispatch "text/x-xmlnews" to the XML parser.
Having a top-level "xml" type allows you to default all subtypes for which
there is not a known specialized viewer to the generic XML viewer, just
like the top-level "text" type can default to the text viewer. It has
generally been found in the text case not to be very useful for most
unrecognized text subtypes. In order for us to want to create an "xml" main
type, we'd have to be very sure that dispatching to a generic XML viewer is
useful, and is more useful than writing the object out to disk.
Going back to Larry's challenge for real-world examples, I'd like to see a
bunch of examples where launching the generic XML viewer is really useful.
What value would I get out of that for many XML types?
--Paul Hoffman, Director
--Internet Mail Consortium