Re: [wsrf] Nillable resource properties

[Steve Graham <sggraham@us.ibm.com>]

ok, good enumeration of the cases and
good description.

Is the conclusion that the WS-ResourceProperties
specification as written satisfies our expectations for these cases, or
is there some correction we need to make in the normative text in the spec?

sgg

++++++++
Steve Graham
(919)254-0615 (T/L 444)
STSM, On Demand Architecture
Member, IBM Academy of Technology
<Soli Deo Gloria/>
++++++++

Samuel Meder <meder@mcs.anl.gov> 07/29/2004 01:54 PM Please respond to meder

To:        wsrf-oasis <wsrf@lists.oasis-open.org>         cc:        Jarek Gawor <gawor@mcs.anl.gov>         Subject:        [wsrf] Nillable resource properties

Here is the promised writeup on nillable resource
properties:

Nillable Resource Properties

During the work in the TC the question of what how to deal with nillable
resource properties was raised. The following cases of interest were
identified:

     1. The ResourceProperty was not declared nillable, but is
nil. This
        is in violation of one of the fundamental assumptions
that the
        specification makes, i.e. that the resource
properties document
        is validatable at any given time, and should
thus never occur in
        a compliant implementation.
     2. The ResourceProperty is declared nillable and is nil.
        Irrespective of its minOccurs value a call
to
        getResourceProperty on this property should
return the
        ResourceProperty element decorated with a xsi:nil="true"
        attribute, e.g. <wsrl:TerminationTime xsi:nil="true/>
     3. The ResourceProperty is declared nillable and minOccurs="0"
and
        there is no occurrence in the resource properties
document. In
        this case a call to getResourceProperty on
this property should
        result in an empty response. 


/Sam

-- 
Sam Meder <meder@mcs.anl.gov>
The Globus Alliance - University of Chicago
630-252-1752