At 10:22 AM 5/9/99 -0700, Ned Freed wrote:
The long and short of it is that I don't see either extreme position on this
issue as tenable. The notion of "all XML goes under the application/xml media
type and a different mechanism is used to distinguish different instances of
XML usage" is a nonstarter, and so is "every little twiddle to something done
in XML gets a new media type". I also don't think that XML is even close to
mature enough for us to attempt to establish definitive rules for when a media
type is needed and when it isn't.
Hard to argue with that. This discussion has been peeving me seriously
because with each exchange, I realize I understand less than I thought
about how the Internet is supposed to fit together.
I honestly don't have an opinion I'm comfy with, but here are a few
data points that might be useful:
- I'm currently helping work on a language whose mime type will
almost certainly be encoded as "model/x3d" - it will be encoded
in XML
- Internet Explorer 5 has a facility I've found very useful - send
it a well-formed XML document with no stylesheet and it constructs
a nice collapsing-tree view of it in real time. Tremendously
helpful when you're getting some chunk of XML vocabulary for
the first time and want to get a handle on it. This is, in fact,
a useful general-purpose XML processor, something I seem to
recall denying the existence of here not too long ago.
- There are going to be a lot of new formats that happen to be
encoded in XML, for which the fact of the encoding syntax is
an irrelevant side-issue.
It's obvious that there is a real need for application/xml; among
other things, it clarifies i18n immensely. Three days out of each week,
I don't believe that text/xml is useful. I know this is heresy, but
sometimes I think that what you need is not a media type but another
parameter akin to charset, so that you can say for some arbitrary mimetype
content-type: text/foo; charset="ISO-8859-1" syntax="xml-1.0"
which lets you know that there's a really useful fallback (ie your
XML-savvy-browser) in the case that you're not set up for text/foo.
Or maybe the top-level xml/* type has the same utility.
Or maybe the whole notion that a rigid 2-level type hierarchy is a useful
vehicle for interoperability is becoming bogus.
I just don't know, so I'm going to shut up and listen. -Tim