Re: LC 47 proposed resolution

Hi, 

I'm traveling so I don't have access to the minutes of previous telcons
so I don't know if the issue was already closed with this resolution,
but if not, let me raise an objection.

For machine-evaluating the result in HTTP, it is only the numeric code
that is relevant. Any RFC-2616-compliant implementation will have to
view "200 OK" as equivalent to "200 bite me", and the same applies to
any other codes, so the reason phrase is really only for the human who
looks at the wire to be able faster to understand the message result.

Of course, "200 bite me" is not what any serious server would respond
with, but I believe that adding whttp:reasonPhrase is completely
unnecessary in WSDL. Any information of this kind belongs to
Documentation.

Best regards,

Jacek

On Fri, 2005-04-01 at 16:09 -0800, David Orchard wrote:
> For LC issue 47 http://www.w3.org/2002/ws/desc/4/lc-issues/#LC47
> 
>  
> 
> I propose that a reasonPhrase attribute of type string is added to the
> whttp namespace and a corresponding http error reason phrase property.
> 
>  
> 
> a xs:string representing a error Reason-Phrase as defined by [IETF RFC
> 2616], to the Binding Fault component. The value of this property
> identifies the Reason-Phrase that the service will use in case the
> fault is returned. If empty, no claim is made by the service.
> 
>  
> 
> There should be a new section 5.8 inserted that roughly copies the
> current 5.7 with the change of code to reasonPhrase, etc.  The syntax
> summary in 5.2 is also updated
> 
>  
> 
> Cheers,
> 
> Dave
> 
>  
> 
>  
> 
>

Received on Friday, 15 April 2005 08:50:47 UTC