RE: http binding

GET or sometimes POST sounds worth exploring, but i'm not sure it's something i'd require.

This particular service was initially implemented as a CGI. The POST data contained the SOAP request which was passed onto a Mainframe service, the query string parameters contained routing and security parameters. The GET parameters usually remained static across each call - though URL re-writing could have been employed here. 

When the service was later migrated to a J2EE SOAP toolkit the abstraction caused the query string parameters to be unavailable in the bean.

How GET and POST may be safely combined in a HTTP/SOAP service is unclear (to me at least).

Paul

-----Original Message-----
From: Jean-Jacques Moreau [mailto:jean-jacques.moreau@crf.canon.fr]
Sent: 29 October 2003 11:22
To: Downey,PS,Paul,XSJ67A C
Cc: sanjiva@watson.ibm.com; www-ws-desc@w3.org
Subject: Re: http binding


You mean (in WSDL terms), a single operation that exchanges messages 
sometimes with GET, sometimes with POST?

I don't think we have that flexibility today, but maybe this is 
something we should explore.

JJ.

paul.downey@bt.com wrote:

> I recently saw a SOAP/HTTP service which used a query string in the URL to provide routing and security parameters and POST to exchange SOAP documents 
> 
> - should WSDL allow, bar or ignore this combination of GET and POST ?
> 
> Paul
> 
> -----Original Message-----
> From: Sanjiva Weerawarana [mailto:sanjiva@watson.ibm.com]
> Sent: 29 October 2003 11:05
> To: Jean-Jacques Moreau
> Cc: www-ws-desc@w3.org
> Subject: Re: http binding
> 
> 
> 
> I think the removal of <message> offers interesting possibilities
> for a "direct" HTTP POST binding. I'd like to explore that.
> 
> I'd like to define an HTTP GET binding for RPC style operations.
> 
> Finally, we need to sort out the SOAP Response MEP stuff. 
> 
> So maybe there isn't much difference, but we need to get it all 
> done. I don't expect there will be a MIME binding at all, but to
> be honest have not thought about how MTOM bindings may work. 
> 
> Bye,
> 
> Sanjiva.
> 
> ----- Original Message ----- 
> From: "Jean-Jacques Moreau" <jean-jacques.moreau@crf.canon.fr>
> To: "Sanjiva Weerawarana" <sanjiva@watson.ibm.com>
> Cc: <www-ws-desc@w3.org>
> Sent: Wednesday, October 29, 2003 4:21 PM
> Subject: Re: http binding
> 
> 
> 
>>How different would that be from the text that went in last time (apart 
>>from component model issues)?
>>
>>JJ.
>>
>>Sanjiva Weerawarana wrote:
>>
>>
>>>I'd like to see us make some progress on the HTTP bindings too.
>>>I've been thinking about this for a while, even though I haven't
>>>put anything down on paper yet. However, I do have 20 hrs of
>>>sitting in planes to do just that. 
>>>
>>>Can we get some time to discuss it at the F2F? I don't think
>>>there'll be enough to make decisions, but I'd like to get 
>>>people thinking on how we may want to evolve the HTTP stuff.
>>>
>>>Philippe, I know you've had a long standing action item on
>>>this .. I'm not trying to take that over (sorry for appearing 
>>>to do so though, especially without asking), but I would like
>>>to make progress on it. If you have anything that you can send
>>>with your thoughts on it (before Saturday night my time - Sat
>>>AM yours) then I can go thru that too while writing down what
>>>I have in mind.
>>>
>>>Sanjiva.
>>>
> 
>

Received on Wednesday, 29 October 2003 06:43:12 UTC