RE: Introducing the Service Oriented Architectural style, and it's constraints and properties.

For the record, the TAG had an action item to look into defining a Query
method, which would examine the feasibility of a GET method with a body.
[1] But as of the last f2f, that was removed.  There isn't an explicit
action or decision recorded (not sure why) but that's what happened.  As I
recall the discussion, the rationale was that there's not a hope that HTTP
is realistically going to change, so why bother.  Further, another thought
was that it doesn't seem likely that a cache would be created that would use
the body to determine whether to retrieve the cached representation - the
cache would need to look into the body to figure out if it had a match on
the query parameters.  As in, HTTP GET already hits arguably the best ever
80/20 point in the history of software, so why would folks upgrade their
servers and clients?

Maybe the WS-Arch group should have an issue around "Need for GET semantics
with information in the body", and simply ask for a formal resolution of
this.   

Cheers,
Dave

[1] http://www.w3.org/2001/tag/open-summary.html#whenToUseGet-7

> -----Original Message-----
> From: www-ws-arch-request@w3.org [mailto:www-ws-arch-request@w3.org]On
> Behalf Of Ugo Corda
> Sent: Monday, February 17, 2003 1:15 PM
> To: David Orchard; www-ws-arch@w3.org
> Subject: RE: Introducing the Service Oriented Architectural style, and
> it's constraints and properties.
> 
> 
> 
> > You'd have to somehow mark the POST as being idempotent, so 
> that the cache 
> > would know that it didn't have to get a response from the 
> resource.  
> 
> This looks like a good idea to me. It would mark POST 
> requests that should actually be just GETs but, because of 
> reasons like presence of headers, had to be turned into POSTs.
> 
> A similar proposal was launched by Hugo Haas a couple of 
> months ago at [1]. His proposal referred to WSDL, but 
> evidently that info should also be carried over the wire for 
> the caching infrastructure to be advised about the nature of 
> the operation.
> 
> Ugo
> 
> [1] http://lists.w3.org/Archives/Public/www-ws-arch/2002Nov/0157.html
> 
>