Re: [xacml] Make "*-on-and-only" an AttributeDescription option

[Polar Humenn <polar@syr.edu>]

On Thu, 17 Jun 2004 Anne.Anderson@Sun.COM wrote:

> Sorry, yes, I meant AttributeDesignator.  -Anne

Then, I am opposed to this modification. Why do you want it?
To cut down on XML characters one has to type? %^D

The type of the <*AttributeDesignator> construct would then be overloaded
to represent a Bag or a singleton depending on this attribute. That would
have typing issues with the <*Match> constructs as they take
<*AttributeDesignator>s that must have type Bag.

Granted, that the type can be figured out by the looking at this
attribute, but in effect it is really a different construct all together,
i.e. one with the "*-one-and-only" function applied.

This modification will not add any functionality. It is merely a
convienience, and furthermore with typing issues.

If we were to have it, it should be named differently and appropriately,
especially since it will not be able to be applied within a <*Match>
construct.

Cheers,
-Polar


> >What is an AttributeDescriptor?
> >
> >Do you mean AttributeDesignator or have I missed something?
> >
> >-Polar
> >
> >On Thu, 17 Jun 2004, Anne Anderson wrote:
> >
> >> I would like to propose that we add an optional "OneAndOnly" XML
> >> attribute to the <AttributeDescriptor> element.  If this
> >> attribute is "true", it causes the <AttributeDescriptor> to
> >> evaluate to a single element of its DataType, rather than a bag
> >> of such elements.  If this XML attribute is used, and the
> >> <AttributeDescriptor> matches no element or more than one
> >> element, then the evaluation result will be "Indeterminate".  The
> >> default value of the OneAndOnly Attribute is "false".
> >>
> >> The intent is that
> >>
> >>   <SomeAttributeDescriptor AttributeId="x" DataType="y" OneAndOnly="true"/>
> >>
> >> be exactly equivalent to
> >>
> >>   <Apply FunctionId="y-one-and-only">
> >>      <SomeAttributeDescriptor AttributeId="x" DataType="y""/>
> >>   </Apply>
> >>
> >> Rationale:
> >>
> >>   All our standard functions except for the specific bag
> >>   functions expect a single value, so the case for expecting and
> >>   wanting a single attribute is extremely common.  It is awkward,
> >>   verbose, and probably confusing to wrap all instances of
> >>   <AttributeDescriptor> with an <Apply> of a "*-one-and-only"
> >>   function.
> >>
> >> Anne
> >> --
> >> Anne H. Anderson             Email: Anne.Anderson@Sun.COM
> >> Sun Microsystems Laboratories
> >> 1 Network Drive,UBUR02-311     Tel: 781/442-0928
> >> Burlington, MA 01803-0902 USA  Fax: 781/442-1692
> >>
> >>
> >> To unsubscribe from this mailing list (and be removed from the roster of the
> >OASIS TC), go to
> >http://www.oasis-open.org/apps/org/workgroup/xacml/members/leave_workgroup.php.
> >>
> >
> >To unsubscribe from this mailing list (and be removed from the roster of the
> >OASIS TC), go to
> >http://www.oasis-open.org/apps/org/workgroup/xacml/members/leave_workgroup.php.
> >
>
>
>
> To unsubscribe from this mailing list (and be removed from the roster of the OASIS TC), go to http://www.oasis-open.org/apps/org/workgroup/xacml/members/leave_workgroup.php.
>