- From: Henrik Frystyk Nielsen <henrikn@microsoft.com>
- Date: Mon, 11 Jun 2001 13:35:06 -0700
- To: "christopher ferris" <chris.ferris@east.sun.com>, "Larry Masinter" <LMM@acm.org>
- Cc: "Simon Fell" <soap@zaks.demon.co.uk>, <xml-dist-app@w3.org>, <xmlp-comments@w3.org>
(Sorry for the intermittent email - I am on vacation at the moment) I think the problem about multi-hop message paths and the SOAP/HTTP binding is a little bit of a red herring - the HTTP binding itself speaks nothing about multiple SOAP hops - only HTTP intermediaries which of course are HTTP constructs and not SOAP constructs. In order to talk about multi-hop SOAP routes one needs a SOAP routing mechanism like for example the SOAP-RP proposal http://www.gotdotnet.com/team/xml_wsspecs/soap-rp/default.html which (not quite) coincidentally talks about a similar "action" construct :) Henrik >I echo Larry's concerns regarding this revised proposal. >It does little to improve the situation and still does not >address how SOAPAction is communicated across different >transport protocols. If a SOAP message starts out being >communicated over the Frobnaz transport protocol, which does >NOT have a SOAPAction header (or even a place to put one) and >the message is being sent via a Frobnaz->HTTP gateway, where >does the gateway get the appropriate SOAPAction to put in the >HTTP headers when it forwards the message to the ultimate destination?
Received on Monday, 11 June 2001 16:58:09 UTC