Re: XMIME

[Dave Crocker <dcrocker@xxxxxxxxxxxxxxx>]

FOlks,,

Thursday, February 6, 2003, 1:50:35 AM, you wrote:
GK> As it happens, it embodies a quite different approach to identifying header
GK> fields:  the approach in draft-klyne-message-xml aims to embody the header
GK> field names as XML element names, namespace-qualified.  The other draft
GK> uses a fixed set of element names and includes header field names as
GK> attribute values.

To carry Graham's reference to people's "expectations", a bit further:

One essentially treats header names as something to encapsulate in XML.
This might be good for quick "tunneling" of 822 data.

The other integrates header names fully into XML processing.  This is
good for long-term.  It allows email headers to benefit from the rich
repertoire of XML tools.

Think of it this way:

Is the goal simple to interface to the RFC822/RFC821 world, or is the
goal to emulate it in XML, permitting a "native" XML environment for
email headers?

Another way of saying the latter:  What would the XML look like if we
were creating email headers from scratch, starting with an XML syntax?

The result would have the same semantics as RFC822/RFC821 headers, but
the syntax would be entirely native XML.

d/
-- 
 Dave <mailto:dcrocker@xxxxxxxxxxxxxxx>
 Brandenburg InternetWorking <http://www.brandenburg.com>
 t +1.408.246.8253; f +1.408.850.1850