W3C home > Mailing lists > Public > public-ws-desc-comments@w3.org > March 2006

Re: CR Comment on WSDL Version 2, part 2: Adjuncts

From: Mark Nottingham <mnot@mnot.net>
Date: Mon, 27 Mar 2006 13:57:23 -0800
Message-Id: <FE37F5D0-4369-4764-8403-FC7E2CC2C43C@mnot.net>
Cc: <www-tag@w3.org>, <public-ws-desc-comments@w3.org>
To: "Rogers, Tony" <Tony.Rogers@ca.com>

Hi Tony,

Thanks. The WG has addressed many of the details of the issue (and  
reasonable people can disagree on their resolution), but not its  
overall substance.

I'm not lie-down-in-the-road against the HTTP binding -- even if I  
don't think it's a great solution for the problem space -- provided  
that there were actually interoperating implementations of it, and  
people wanted to use it for itself, not just with the SOAP binding on  
top.

However, I don't see that this is the case. So, I'm not satisfied by  
the proposed resolution. Note that I don't expect the WG to spend  
more time on this; closing the issue and marking it for discussion  
during the transition meeting would be adequate.

Kind regards,



On 2006/03/26, at 5:23 PM, Rogers, Tony wrote:

> Dear Mark
>
> thank you for your Last Call comment on WSDL V2.0 part 2: Adjuncts  
> [1]. This comment was assigned issue number CR011, and addressed at  
> the most recent Face to Face meeting of the WS-Description Working  
> Group. It was discussed at some length, both as a whole, and as  
> individual points. In response to the individual points we reached  
> the following conclusions:
> We agreed to remove section 6.4 (HTTP version) from the document  
> (this has been done)
> We did not agree with the assertion that 6.6 constrains HTTP header  
> field values - we believe that simple types (particularly string)  
> suffice, particularly as this is not a general HTTP header  
> description language
> 6.6.6 does not put all HTTP headers into a WSDL-specific namespace.  
> We are identifying components, not headers, and using the WSDL  
> restriction of one component per HTTP header to allow us to  
> identify the component with the header name - we are not  
> identifying the header
> Although the interposing of an intermediary may well make transfer- 
> encoding information (6.9) worthless, we believe that it is a  
> useful optimisation for a large percentage of cases, so we decided  
> to retain it
> We decided to take no action on the question of cookies (6.10) -  
> this topic has been discussed before, and although opinions are  
> divided on it, we are not advocating cookies, merely allowing the  
> documenting of their use. It is not really appropriate for the WSDL  
> standard to take a general position on cookies
> We decided that no action was required to addressing WebDAV,  
> because it can be handled by extensibility, by a separate binding,  
> or via a different language (such as WADL). Note that the HTTP  
> binding supports the description of arbitrary HTTP methods beyond  
> GET, POST, PUT, and DELETE
> It would be helpful if you could respond to the Working Group,  
> letting us know if you accept these resolutions of the points you  
> raised.
>
> [1]: http://lists.w3.org/Archives/Public/public-ws-desc-comments/ 
> 2006Feb/0002.html
>
>
> Tony Rogers
> CA
> Senior Architect, Development
> tel +61 3 9727 8916
> fax +61 3 9727 3491
> tony.rogers@ca.com
> co-chair, W3C WS-Description Working Group
> co-chair, OASIS UDDI Spec Technical Committee


--
Mark Nottingham     http://www.mnot.net/
Received on Monday, 27 March 2006 21:57:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:20:32 GMT