W3C home > Mailing lists > Public > public-ws-desc-comments@w3.org > November 2002

RE: SOAP 1.2 Specification and confusion around Subcode

From: <Ed.Morassut@bmo.com>
Date: Mon, 28 Oct 2002 09:56:02 -0500 (EST)
To: "Martin Gudgin" <mgudgin@microsoft.com>
Cc: public-ws-desc-comments@w3.org
Message-ID: <OF8E6759E1.1B1DF48F-ON85256C60.004FBB26@notes.bmo.com>





Thanks for the reply but we're still confused.  You mentioned "You could (
in WSDL via XML Schema ) restrict the type of soap:Fault and specify what
enumerated values are allowed".  I have seen examples (from Microsoft and
others) that show an instance with the following syntax:

     <soap:faultcode>Server.my:DivisionByZero</soap:faultcode>

What we don't understand is how the connection to WSDL is made?  For
example we've seen the corresponding WSDL written as follows:

     <fault message="tns:divisionByZero" name='DivisionByZero'/>

This would suggest that the fault name becomes the subcode?  But we don't
understand how the tooling can make such determination.  Do you folks have
an example of how this is to be described in WSDL and what the accompanying
instance would look like?

Lastly, don't understand how in WSDL you could restrict the type of
soap:Fault and specify what enumerated values are allowed?  The following
1.2 spec reads:  "The value of this element is an application defined
subcategory of the value of the Value child element information item of the
Subcode element information item's parent element information item (see
5.4.6 SOAP Fault Codes)".   This is a hard read and both me and my
colleagues are having difficulty comprehending.  To us the fault code would
need to be described as a complex type if it is to accomodate a subcode
(especially if it is to be restricted from an application spoecific
namespace) but that's not what we're seeing in the limited examples out
there.  Support for faults is very weak in the industry to date but its
fundamental to programmatic flow (especially when it comes to orchestration
going forward) and to us its still fundamentally brioken.  Again, do you
folks have examples that clearly describes what you are intending with
subcode and how it is to be described in WSDL.  Thanks in advance for your
reply.





                                                                                                                                       
                      "Martin Gudgin"                                                                                                  
                      <mgudgin@microso         To:      <Ed.Morassut@bmo.com>                                                          
                      ft.com>                  cc:      <public-ws-desc-comments@w3.org>                                               
                                               Subject: RE: SOAP 1.2 Specification and confusion around Subcode                        
                      10/28/2002 05:42                                                                                                 
                      AM                                                                                                               
                                                                                                                                       
                                                                                                                                       




Ed,

You asked two questions in your e-mail below.

1. Can the Value element ( when appearing as a child of SubCode ) be a
'faultCodeEnum' from an application specific namespace?

2. How do you specify in WSDL what the valid values are?

The answer to the first question is Yes, given that the type of the
Value element is xs:QName, the values can be from any namespace. You
could ( in WSDL via XML Schema ) restrict the type of soap:Fault and
specify what enumerated values are allowed.

The answer to the second question is really in the purview of the Web
Services Description Working Group, whose comments list I have cc'ed on
this e-mail.

Regards

Martin Gudgin
For the XML Protocol Working Group

> -----Original Message-----
> From: Ed.Morassut@bmo.com [mailto:Ed.Morassut@bmo.com]
> Sent: 02 October 2002 18:18
> To: xmlp-comments@w3.org
> Subject: SOAP 1.2 Specification and confusion around Subcode
>
>
>
>
>
>
> I read through the specifications and I have a question around Subcode
>
>
>
> 5.4.1.3 SOAP Value element (with Subcode parent)
>
>
> The Value element information item has:
>
> *     A [local name] of Value .
> *     A [namespace name] of http://www.w3.org/2002/06/soap-envelope .
>
> The type of the Value element information item is QName in
> the " http://www.w3.org/2001/XMLSchema" namespace. The value
> of this element is an application defined subcategory of the
> value of the Value child element information item of the
> Subcode element information item's parent element information
> item (see 5.4.6 SOAP Fault Codes).
>
>
> NOTE. What is not clear here is whether you can make this
> element information item a  faultCodeEnum from an application
> specific namespace. A further problem is that Subcode appears
> to be disconnected from WSDL implying a developer has no way
> of understanding the domain values of this Enum because it is
> not documented as part of the interface.
>
>
> Is this someting you folks have already thought through?  The
> debate in our shop is that we should avoid Subcode and put
> such meaning into the fault detail (especially being that the
> current tooling doesn't yet support Subcode).  If we end up
> putting such meaning into the detail we'll end up coupled to
> non-standard detailed parsers in order to isolate an
> equivalent of Subcode.  Am I missing something here?  Can we
> make the Enum from an application specific namespace and how
> do we connect to WSDL?
>
>
>
>
>
>
Received on Monday, 4 November 2002 16:24:05 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:20:30 GMT