Email bindings of SOAP (was: Re: [SOAP] SOAP Digest - 2 Oct 2002 to 3 Oct 2002 (#2002-200))

Anne Thomas Manes writes:

>> While the PocketSOAP spec is great, has no official standing.

Agreed.

>> The W3C XML Protocol Working Group is developing an official
>> email binding for SOAP 1.2. See: http://www.w3.org/TR/soap12-email

Actually, speaking as one who worked on that draft, its status is not what
you imply.  As noted in the "Status" section:

"This document was written by members of the Transport Binding Task Force
(TBTF)-- a part of the XML Protocol WG-- on behalf of the XML Protocol WG.
The XML Protocol WG agreed to the publication of this document. The XML
Protocol WG has no plans for further work on this document."

and below that under "motivation":

"The motivation for this document is to illustrate the SOAP 1.2 Protocol
Binding Framework and the creation of an alternative protocol binding
specification to the Default HTTP binding. This second binding is meant to
validate the Protocol Binding Framework for completeness and usability.
Please note that this document is a non-normative description of an Email
Binding."

The reason the WG did this work is that we found ourselves in the position
of having inventented what purports to be a general framework for letting
you build new bindings to transports [1, but only having shown its use for
one binding to one underlying protocol (HTTP) [2].  Our remit in doing the
"email" binding was merely to go far enough to demonstrate the generality
of the framework.  We leave it to others to do deployable binding(s) of
SOAP to email.

Also, as an aside:  I should note that there was some unresolved
controversy as to whether the most effective such binding would be to SMTP
specifically, or to more of a meta-protocol based on RFC 2822.  Without
starting that whole flame over again here (it is indeed unresolved), I
note that much of the email in the world spends only some of its time
moving over SMTP.  For example, this note will move through our corporate
network using proprietary protocols until it reaches a gateway that will
send it via SMTP, and I'm sure the reverse will be true as it gets closer
to some recipients on this list.  What makes this work is that we all
agree on the RFC 2822 equivalent fields like From, To, Reply-to, etc.
Those are used consistently through the process, or else their equivalents
exist at each step of the relay.  A mapping to SMTP only covers the case
where the mail is actually going SMTP.  The debate centered on whether
there was indeed a meta-protocol implied by RFC 2822, which primarily
documents the fields and sort of implies their correct use in email
exchanges -- SMTP is the spec that gives you (one way) that exchange
actually looks on the wire, thus making the protocol concrete.

Of course, I don't speak formally for the Protocols WG, but I'm confident
that this explanation is essentially correct.  I hope this is helpful.

[1] http://www.w3.org/TR/2002/WD-soap12-part1-20020626/#transpbindframew
[2] http://www.w3.org/TR/2002/WD-soap12-part2-20020626/#soapinhttp

------------------------------------------------------------------
Noah Mendelsohn                              Voice: 1-617-693-4036
IBM Corporation                                Fax: 1-617-693-8676
One Rogers Street
Cambridge, MA 02142
------------------------------------------------------------------

You can read messages from the SOAP archive, unsubscribe from SOAP, or subscribe to other
DevelopMentor lists at http://discuss.develop.com.