Re: Clean up of state tables

At this point, shouldn't we just get back to the WG and ask for 
guidance?

Jean-Jacques.

Marc Hadley wrote:
> On Monday, Sep 23, 2002, at 15:53 US/Eastern, Martin Gudgin wrote:
> 
>> Isn't the only place that 'RequestingSOAPNode' and 'RespondingSOAPNode'
>> only appear at the very top of 7.5.
>>
>> Why not just change the bullets to read
>>
>> For binding instances conforming to this specification:
>>
>>     A SOAP node instantiated at an HTTP client may assume the role
>> (i.e. the property reqres:Role ) of
>> "http://www.w3.org/2002/06/soap/mep/request-response/ RequestingSOAPNode"
>> or  "http://www.w3.org/2002/06/soap/mep/soap-response/RequestingSOAPNode"
>>
>>     A SOAP node instantiated at an HTTP server may assume the role
>> (i.e. the property reqres:Role ) of
>> "http://www.w3.org/2002/06/soap/mep/request-response/ RespondingSOAPNode"
>> or  "http://www.w3.org/2002/06/soap/mep/soap-response/RespondingSOAPNode"
>>
>>
>> Would that work?
>>
> The following two paragraphs might also need to be changed:
> 
> <current>The remainder of this section describes the MEP state machine  
> and its relation to the HTTP protocol. In the state tables below, the  
> states are defined as values of the property reqres:State (see 6.2 SOAP  
> Request-Response Message Exchange Pattern), and are of type xs:anyURI .  
> For brevity, relative URIs are used, the base URI being the value of  
> reqres:Role .</current>
> 
> We might just be able to add a reference to 6.3, but then the baseURI  
> would be different depending on whether you were executing  
> soap-response or request-response state machines. How would the  
> participants agree which state machine they were executing, would it  
> matter if they disagreed (probably not in this case, indicating that we  
> don't need two separate state machines).
> 
> <current>Failure reasons that are specified in the tables represent  
> values of the property context:FailureReason and their values are  
> relative URIs whose base URI is the value of  
> context:ExchangePatternName . If an implementation enters the "Fail"  
> state, the context:FailureReason property will contain the value  
> specified for the particular transition.</current>
> 
> I think there's also duplication of failure states between the two  
> state machines so I'm not sure we need both.
> 
> Marc.
> 
> 
>> Gudge
>>
>>
>>
>>> -----Original Message-----
>>> From: Marc Hadley [mailto:marc.hadley@sun.com]
>>> Sent: 23 September 2002 12:46
>>> To: Martin Gudgin
>>> Cc: W3C Public Archive; Jean-Jacques Moreau; Nilo Mitra; Noah
>>> Mendelson; Henrik Frystyk Nielsen
>>> Subject: Re: Clean up of state tables
>>>
>>>
>>> On Monday, Sep 23, 2002, at 15:19 US/Eastern, Martin Gudgin wrote:
>>>
>>>>>
>>>>> I think we have two options:
>>>>>
>>>>> (i) rethink the base URI for the states such that they are
>>>>
>>> shared by
>>>
>>>>> both request-response and soap-response - or -
>>>>> (ii) Split section 7.5 into two, one for each state machine.
>>>>>
>>>>> I'd prefer (i) but LC issue 305 might push our choice to (ii).
>>>>
>>>>
>>>> It seems to me that (ii) is probably easier and quicker for us as
>>>> editors to implement.
>>>>
>>> I disagree, (i) is *much* easier editorially, just change a base URI
>>> here and there. (ii) is a lot of work, will make the document
>>> significantly longer (lots of duplication required) and I
>>> hate editing
>>> those state transition tables !
>>>
>>> Marc.
>>>
>>> -- 
>>> Marc Hadley <marc.hadley@sun.com>
>>> XML Technology Center, Sun Microsystems.
>>>
>>>
>>
>>
> -- 
> Marc Hadley <marc.hadley@sun.com>
> XML Technology Center, Sun Microsystems.
>

Received on Tuesday, 24 September 2002 07:28:13 UTC