RE: [xacml] Minutes of 25 March Policy Model teleconference

From
Tim Moses <>
Date
2002-03-26T20:38:12+00:00
ID
Thread
RE: [xacml] Minutes of 25 March Policy Model teleconference
Title: RE: [xacml] Minutes of 25 March Policy Model teleconference

Polar - In earlier days, we had spoken of having a section containing a specification for the PDP.  I think the closest to this is the normative Section 7, which will contain the specification of the combining rules.  Perhaps, this is where the statement you refer to should eventually reside.  All the best.  Tim.

-----------------------------------------

Tim Moses

Tel: 613.270.3183

-----Original Message-----

From: Polar Humenn [mailto:]

Sent: Tuesday, March 26, 2002 1:58 PM

To: Tim Moses

Cc: Carlisle Adams; XACML TC

Subject: RE: [xacml] Minutes of 25 March Policy Model teleconference

Personally I would like to see a version 12 with all the resolutions in

it, so that we can start working on issues with the document, and not the

model itself.

This leads me to one question. I have taken on an action item about

creating text about the PDP's evaluation of a single base policy that we

could all be happy with. This situation came about after I complained

about it, of course. :)

I would rather instead of submitting text for an issues document, which is

largely just maintained for historical purposes, submit editing

instructions for the XACML document itself.

However, as I look for the appropriate place for such text, I find that

the only sections that it applies are NON-NORMATIVE, which serves only as

an explanation but does not stipulate requirements. This IS FINE WITH ME!

:)  However, I was under the impression that we were writing conformance

points for the operation of a PDP and PEP. I believe that is where we get

into trouble. Which is the point of contention with me.

It seems that the normative part of the specification is the language and

its semantics. Which really says nothing about the PDP or the PEP, other

than a PDP shall evaluate a policy and along with some inputs according to

the well specified semantics of the language.

The NON-NORMATIVE part, i.e. Section 4 (Models) specifically, 4.1. Data

Flow Model, is primarily there for explanation and context. The data flow

model is not required implemented as separate components and the whole

thing can be viewed as a PEP from the get-go.

Am I correct on this interpretation?

So, I am at a loss about what to write and where. All I want to write is

specifically, given a XACML policy and inputs of Subject, Resource,

Environment, and a SAML Authorization Decision request, a PDP shall

evaluate such request against the policies and inputs according to the

semantics of the Normative part of the spec, i.e. Rules, Policies, and

Policy Combinations.

Cheers,

-Polar

----------------------------------------------------------------

To subscribe or unsubscribe from this elist use the subscription

manager: <http://lists.oasis-open.org/ob/adm.pl>