W3C home > Mailing lists > Public > xml-dist-app@w3.org > December 2001

RE: issue 168 proposal: xsi:type of external references in Encoding

From: <noah_mendelsohn@us.ibm.com>
Date: Wed, 12 Dec 2001 14:20:01 -0500
To: dorchard@bea.com
Cc: andrewl@microsoft.com, jacek@systinet.com, xml-dist-app@w3.org
Message-ID: <OF367E7D3A.D27E4D4E-ON85256B20.006A6DFA@lotus.com>
David Orchard writes:

>> I don't want to deal with package scoped references in SOAP,

Neither do I.  I want to deal with them in the specifications for 
transport bindings, and features implemented by transport bindings. That's 
where you know whether you're using S+A (Mime), XML Packaging, or 
something else.  All I'm advocating is that we say something like:  "the 
specification for a binding SHOULD indicate the means used to reference 
attachments, if any, carried with a message (as distinct from other Web 
resources.)"

>> I'd rather deal with them in the packaging work.

If XML packaging comes up with something good, that's fine, and I hope 
that bindings will use it where appropriate.  I think it's much too 
limiting to assume that all bindings will use any one such mechanism. Even 
for the envelope, we allow arbitrary representations of the Infoset on the 
wire.  Many of the attachments will not even be XML (binary XRays seems to 
be the cannonical counterexample).  I'm not sure that the XML Packaging 
effort would undertake to deal with such non-XML resource, much less come 
up with a solution that we could mandate as the solution for use in all 
cases.  We should let the implementors of particular transport bindings 
decide how to carry attachments.

------------------------------------------------------------------------
Noah Mendelsohn                                    Voice: 1-617-693-4036
Lotus Development Corp.                            Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------------
Received on Wednesday, 12 December 2001 17:17:45 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 5 February 2014 22:28:13 UTC