Re: [wsrm] Proposal to resolve PR26 - Soap Fault with rm-fault

[Tom Rutt <tom@coastin.com>]

I agree with Doug's suggestions, and have queued this mail for 
discussion at the Tuesday meeting.

Tom Rutt

Doug Bunting wrote:

> Tom,
>
> I have a slight (very slight) preference toward a SOAP fault in the 
> case that an RM fault leads to an unexpectedly empty SOAP Body.  
> However, this text should be focused specifically on that case since 
> the producer may not be expecting any consumer payloads.  SOAP Body 
> content (a SOAP fault) would be entirely redundant and itself 
> unexpected unless a consumer payload was expected.
>
> To avoid such an over-generalized statement, I would suggest adding 
> "and a consumer payload was expected" before the comma in both 
> sentences you propose.
>
> An editorial nit: Should these two sentences be talking about 
> "soap:client" and "soap:server" or "SOAP client" and "SOAP server" 
> faults?  Consistency seems necessary here.
>
> thanx,
>     doug
>
> [1] ... whom, I assume, is the target of the SOAP fault.  This is a 
> bit counter-intuitive since the sending RMP hides the SOAP messaging 
> layer from the producer to some extent.
>
>
> On 02-Jul-04 11:27, Tom Rutt wrote:
>
>> Proposal to Resolve Issue PR26 – Soap Fault with RM-Fault
>>
>> The behaviour that an RM-Fault is returned with a soap fault in the 
>> case that a response
>> payload is not available for response reply pattern was part of the 
>> public review draft
>> cd .992. The changes suggested by Sunil would constitute a 
>> substantive change to the public review draft.
>>
>> This behaviour works, and provides fault information separately 
>> tarteted for the rmp and
>> the producer.
>>
>> However the following sentences were inadvertently removed during 
>> editing after
>> CD .992
>> “
>> If the RM-Fault encountered was due to a problem with the request 
>> header element, a SOAP
>> client fault MUST be returned. If the RM Fault encountered was due to 
>> a problem with processing
>> by the receiving RMP (including the inability to return a response 
>> due to Duplicate Elimination), a
>> soap:server fault must be returned.
>> “
>>
>> We agreed to have section 4.5 only talk about rm faults, so the 
>> parenthetical statement should be removed.
>>
>> Proposed Resolution:
>>
>> Add the following paragraph in Line 1070 of 1.04JacquesContrib, after 
>> the first sentence of the bullet:
>> “
>> If the RM-Fault encountered was due to a problem with the request 
>> header element, a SOAP
>> client fault MUST be returned. If the RM Fault encountered was due to 
>> a problem with processing
>> by the receiving RMP, a soap:server fault must be returned.
>> “
>>
>>
>> Tom Rutt
>>
>

-- 
----------------------------------------------------
Tom Rutt	email: tom@coastin.com; trutt@us.fujitsu.com
Tel: +1 732 801 5744          Fax: +1 732 774 5133