Re: FW: LC Comments: Web Method Feature

Mark Baker wrote:
> On Mon, Jul 08, 2002 at 11:13:27AM +0100, Marc Hadley wrote:
> 
>>Just to try to bring this back around to Stuart's (and my) original 
>>problem with the current formulation. I don't think either of us 
>>disagrees (Stuart, please jump in here if you do) that GET is an 
>>application semantic, but at the moment the GETness of the operation is 
>>duplicated in two places: the MEP in use and the value of the web method 
>>feature (which has an implicit MEP). As it stands the MEP and value of 
>>the web method feature can disagree and the spec is silent on what 
>>happens if they do.
>>
>>My suggestion of refactoring the existing formulation to use MEPs and a 
>>'safe' feature in place of the web method feature was intended to remove 
>>the duplication from the two whilst preserving the ability of the HTTP 
>>binding to accurately model and make use of the web architecture.
> 
> 
> Right, you said this earlier in this thread.  My response was that there
> is no duplicate information; a MEP describes how messages are exchanged,
> independant from the meaning of those messages.
> 
I couldn't find this response in any of your replies, but I'll take your 
word for it. Your reply is puzzling, are you are claiming that POST and 
GET don't describe how entity bodies are exchanged - doesn't the 
description of POST say that it can carry an entity body in the request 
and response whereas GET can only carry an entity body in the response ?

Marc.

-- 
Marc Hadley <marc.hadley@sun.com>
XML Technology Centre, Sun Microsystems.

Received on Monday, 8 July 2002 09:29:32 UTC