Re: Primer draft with GET Additions

> Paul Prescod wrote:
>  >
>  > Anyhow, the text you disagree with is merely a restatement of the
>  > relevant TAG finding in a SOAP context!
> 
> If the TAG finding you are referring to is 
> http://www.w3.org/2001/tag/doc/ws-uri-05042002.html, then it 
> specifically refers to the 4K limit (actually, it specifies 4000), and 
> it contains a link to 

You did not mention that the context of the mention is "Myths, Bugs, and
Ephemeral Limitations". And it says "the limits evolve as the legitimate
uses of application developers evolve." In other words, developers are
encouraged to push the limits and thus encourage vendors to fix their
products.

> http://www.w3.org/2001/tag/doc/ws-uri-05042002.html as "representing the 
> initial investigations into requirements and proposed solution".  That 
> latter document specifically limits its scope to Simple RPC style SOAP 
> requests, which are without headers (it expresses this requirement as 
> BODY only) and with no structured parameters.

The key phrase is *initial investigations*. I don't know what the
current state of the normative text of GET support SOAP is, but I'm
pretty sure ws-uri-05042002 is not it. My understanding is the Primer is
closer to what will be in SOAP than the "GET Binding" draft. For
instance, the Primer says that GET and POST are part of a single
binding, which makes more sense than having two different bindings. 
 
>...
> Why should the TAG and/or the XMLP working group consider imposing a 
> higher bar on the future of SOAP than has ever been contemplated on HTML?

I don't know what you are trying to say here...

Perhaps we can just agree that you should use HTTP GET when you can use
HTTP GET and HTTP POST otherwise. That is the gist of the TAG finding.
You can use HTTP GET when the operation is safe, ideally
side-effect-free and the volume of data in the request is acceptable to
the software you are using (at the very least 4K!).

 Paul Prescod

Received on Monday, 17 June 2002 14:35:46 UTC