Re: [xacml] MInutes for TC Call 13 January 2011 - UPDATED

[Erik Rissanen <erik@axiomatics.com>]

See one more correction inline:

On 2011-01-17 16:30, Bill Parducci wrote:
> { corrected date of call, noted proper status of Doron Grinstein }
>
> I. Roll Call
>   Voting Members
>    Hal Lockhart (Chair)
>    Bill Parducci (Co-Chair, minutes)
>    Erik Rissanen
>    David Staggs
>    Gareth Richards
>    Rich Levinson
>    Sridhar Muppidi
> | Doron Grinstein
>
>   Members
>    Gregory Neven
>    Franz-Stefan Preiss
>    Duane DeCouteau
>    Remon Sinnema
>
>   Quorum met: (72% per Kavi)
>
> II. Administrivia
>   Approve Minutes:
> | 13 December 2010 TC Meeting:
>    APPROVED unanimously
>
>   IDTrust Call for Papers
>    http://middleware.internet2.edu/idtrust/2011/
>
>   Test Cases, V2, V3
>    Rich raised the question of where the v2 schemas are for the test
>    cases in the subversion repository. Discussion moved to the list.
>
>   PCAST
>    David reviewed recent work by the President's Council of Advisors on
>    Science and Technology (PCAST). A document was posted to the TC list.
>
> III. Issues
>
>   BTG
>    Postponed since David was not on call.
>
>   Attribute Assertions in XACML request
>    Gregory discussed the concept of introducing an new DataType,
>    AttributeID and FunctionId. This is described in a note posted to the
>    TC list (immediately prior to the call). Predicates are not limited
>    to simple comparisons, rather any/all XACML functions could be used
>    within the predicate.
>
>    Erik noted that this approach tightly binds knowledge of the Request
>    with the Policy composition, so what are the benefits over attributes.
>    Gregory offered that this approach would require no modification of
>    the Context Handler. Further discussion has been directed to the list.

Gregory's answer was that with the attribute approach it is necessary to 
define an attribute identifier for each conditional one needs to use in 
the policy, while this is not necessary with the predicate approach, 
which, I agree, is a great benefit.

>   PIP directive
>    Hal briefly reviewed the philosophy of knowing all necessary
>    information in advance by being able to gather information various
>    sources as necessary. Rich noted that he believes that the
>    specification is underspecified in this area as callbacks are
>    inferred only in non-normative text and that this should be
>    normatively defined. Hal discussed work on AMF that is being done in
>    this area.
>
> meeting adjourned.
>
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail.  Follow this link to all your TCs in OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
>