On Tue, 13 Apr 1999, Paul Langer wrote:
Rick Jelliffe wrote
(http://www.imc.org/ietf-xml-mime/mail-archive/msg00064.html):
In particular, I think we are missing a key distinction that the MIME
content-type is not so much the "type" of a resource, but the type of a
particular *publication* of a resource.
I disagree with this definiton of the semantics of "MIME content-type".
XML is portable data; the particular use of this data should not be
specified by the "publisher". The user should be able to detect the
data type and then select an application that he/she likes. How can a
"publisher" know, what applications are available at the user's site?
My point is not that a "publication" selects an application, but that it
is metadata which is expected to select a class of handler: we expect a
MIME content-type of application/xml will select a different class of
handler to text/plain. Otherwise, all XML would be text/plain and we
wouldn't need this mail list.
If you want an example, the DTDs on my site at http://www.ascc.net/xml/
are typically published both as .dtd (text/xml) and .txt (text/plain).
This is because I don't know which class of handler the user has: in the
case of IE5, the DTD will cause a WF error, and it the user cannot View
Source (if they know how) to view the whole DTD, because the file
transmission is aborted.
Providing this kind of multiple content-type doesnt require any
additional MIME parameters, I can do it at on the file system using links.
Rick Jelliffe
Academia Sinica Computing Center
Tapei, Taiwan