Media Type review request for application/dialog-info+xml

[Mark Baker]

On Tue, Jan 25, 2005 at 02:39:17AM +0100, Bjoern Hoehrmann wrote:
> * Martin Duerst wrote:
> >At 04:30 05/01/25, Bjoern Hoehrmann wrote:
> > >* Allison Mankin wrote:
> > >>The Transport Area requests a Media Type review for the proposed
> > >>new type application/dialog-info+xml, intended for the IETF tree, and specified
> > >>http://www.ietf.org/internet-drafts/draft-ietf-sipping-dialog-package-05.txt
> > >>(Section 8.1).
> > >
> > >Why does it list .xml as file extension?
> >
> >Is there anything wrong with using a .xml extension? If yes, what?
> 
> Well, e.g. RFC 3236 states:
> 
> [...]
>   It is not recommended that the ".xml" extension (defined in
>   [XMLMIME]) be used, as web servers may be configured to
>   distribute such content as type "text/xml" or
>   "application/xml".  [XMLMIME] discusses the unreliability of
>   this approach in section 3.  Of course, should the author
>   desire this behaviour, then the ".xml" extension can be used.
> [...]

Yes, I was surprised I managed to get away with including that, given
the penchant for the use of */xml types at the time (and to this day,
it seems).  The relevant paragraph from section 3 is this;

  An XML document labeled as text/xml or application/xml might contain
  namespace declarations, stylesheet-linking processing instructions
  (PIs), schema information, or other declarations that might be used
  to suggest how the document is to be processed.  For example, a
  document might have the XHTML namespace and a reference to a CSS
  stylesheet.  Such a document might be handled by applications that
  would use this information to dispatch the document for appropriate
  processing.

Emphasis on "might".

Said another way, there is no specification which licenses a recipient
of a */xml-described (say) XHTML document, to infer that the sender
intended the message to convey XHTML semantics.  Therefore, using a
.xml extension provides different semantics than using the .dif
extension.  Unless this is desirable - which I seriously doubt - I
agree that .xml should be excluded.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca