RE: Another try at HTTP binding

This sounds like an excellent idea to me - it also provides a template for documenting additional bindings in the future.

Paul


-----Original Message-----
From: www-ws-desc-request@w3.org [mailto:www-ws-desc-request@w3.org]On
Behalf Of Roberto Chinnici
Sent: 22 January 2004 02:16
To: Philippe Le Hegaret
Cc: Jean-Jacques Moreau; Web Services Description
Subject: Re: Another try at HTTP binding



I'm just thinking aloud here, but wouldn't it be better to keep the
two bindings in two separate documents? I'm not convinced that there's
going to be that much overlap in readership, plus that approach
would allow us to publish them indipendently, thus avoiding delaying
one binding because of issues with the other.

Thanks,
Roberto


Philippe Le Hegaret wrote:
> On Thu, 2004-01-08 at 11:38, Jean-Jacques Moreau wrote:
> 
>>Philippe, is there any chance you could provide a diff over the 
>>existing spec text (or any other mechanism that you might deem 
>>appropriate), to make it easier to evaluate your proposed changes?
> 
> 
> Jean-Jacques, here is the document:
> http://www.w3.org/2004/01/21-httpbinding.html
> 
> Note: to avoid messing up the Part 3, I'm working on a standalone
> version of HTTP binding document. The intent is to fold it into our
> current documents after appropriate reviews and changes.
> 
> I rewrote the proposal to make it closer to the approach taken by XForms
> 1.0. As Mark noted, XForms 1.0 does not allow you to put data into the
> path of the URI. I left it in the document however, marked with an issue
> on it (as discussed during last week teleconference).
> 
> Changes from earlier proposals:
> - HTTP POST does no longer let you put data into the URI.
> - List of simple types are no longer as pairs of name/value, but instead
> using a single pair (that simplify the definition of the URI style).
> - New attribute to select the character separator in URIs.
> - New mechanism to select the serialization format (and the HTTP method
> at the same time).
> - Introduced the URI and Multipart operation styles.
> 
> I still didn't manage to come up with a proposal for HTTP headers,
> especially since we don't @headers attribute anymore.
> 
> Issues are inline in the document.
> 
> Philippe

Received on Thursday, 22 January 2004 02:10:25 UTC