Re: Seeking clarification about the use of the HTTP binding

Hmm.  So an interesting question is whether the HTTP binding ever sends a 
202.  I find the table in 7.5.1.2 somewhat ambiguous.   It contains the 
following:

Status Code
Reason phrase
Significance/Action
NextState
2xx
Successful
 
 
200
OK
The response message follows in HTTP response entity body. Start making an 
abstraction of the response message available in http://www.w3.org/2003/05/soap/mep/InboundMessage .
"Sending+Receiving" or "Receiving"
3xx
Redirection
The requested resource has moved and the HTTP request SHOULD be retried 
using the URI carried in the associated Location header field as the new 
value for the http://www.w3.org/2003/05/soap/mep/ImmediateDestination property.
"Init"
4xx
Client Error
 
 
400
Bad Request
Indicates a problem with the received HTTP request message.
"Sending+Receiving", "Receiving" or "Fail"
401
Unauthorized
Indicates that the HTTP request requires authorization.
If the simple authentication feature is unavailable or the operation of 
simple authentication ultimately fails, then the message exchange is 
regarded as having completed unsuccessfully.
"Requesting" or "Fail"
405
Method not allowed
Indicates that the peer HTTP server does not support the requested HTTP 
method at the given request URI. The message exchange is regarded as 
having completed unsuccessfully.
"Fail"
415
Unsupported Media Type
Indicates that the peer HTTP server does not support the Content-type used 
to encode the request message. The message exchange is regarded as having 
completed unsuccessfully.
"Fail"
5xx
Server Error
 
 
500
Internal Server Error
Indicates a server problem or a problem with the received request
"Sending+Receiving", "Receiving" or "Fail"
I think the 2xx is intended as a separator, and the only response sent by the 
binding is 200.  I agree that we could have called for 202, in which case 
we should have clarified how the state machines at both ends continue to 
implement the Request/Response MEP from that point.
If my earlier note is correct, an explicit SOAP header could be created 
calling for the 202 as you suggest, the spec for which would fill in  the 
missing pieces as to how the binding state machines would proceed from 
there.   Is 202 really an appropriate response if you are going to give an 
application response to the post using some other out-of-band means?  I 
suppose so. 
I still think all this is a bit unsatisfying, primarily because you're 
trying to preserve the appearance of use of the SOAP HTTP binding when in 
fact there's no interop with unenhanced implementations.  That being the 
case, I think it's more appropriate to admit that you're doing a new and 
to some degree non-interoperable binding.  I accept your reading that 202 
makes use of HTTP at least semi-appropriate.  Thanks.

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








Mark Baker <distobj@acm.org>
11/12/2004 02:32 PM

 
        To:     noah_mendelsohn@us.ibm.com
        cc:     Hugo Haas <hugo@w3.org>, www-ws-desc@w3.org, xml-dist-app@w3.org
        Subject:        Re: Seeking clarification about the use of the HTTP binding


On Fri, Nov 12, 2004 at 01:42:04PM -0500, noah_mendelsohn@us.ibm.com 
wrote:
> Finally, I think there's a case to be made that this is a significant 
> misuse of HTTP itself, which has a pretty strong notion of 
> request/response on a POST (you couldn't send the header on a GET.)

I don't think it's quite that strong, Noah.  I believe that the 202
"Accepted" response provides the desired "out" here, by being
"intentionally non-committal".  I've used it myself, on occasion, for
exactly the scenario described.

Mark.
-- 
Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca