Re: [higgins-dev] RP Policy discussed during the higgins call

[Dieter M Sommer <DSO@xxxxxxxxxxxxxx>]

Hi,

our current discussions are on an abstract level, yet independent of a
concrete syntactical representation. Thus, we are still open to go for a
new language or using extensions of an existing one, provided that this
fulfills all our requirements, such as the general evidence specification
used for third-party endorsed attributes and the requirements Paul was
referring to.

An extension of an existing language would have to be clean in the sense
that extensions are only done for well-defined extension points.
Furthermore, the extensions of existing evaluation engines would have to be
doable in clean way.

Looking forward to discuss those issues such that we can come up with the
best possible solution for Higgins.


Cheers, Dieter.

higgins-dev-bounces@xxxxxxxxxxx wrote on 10/26/2006 06:52:53 PM:

> http://wiki.eclipse.org/index.php/Relying_Party_Security_Policy
>
> I joined the call late, and just caught the very last part of this
> discussion so I apologize if this issue was covered.   I was curious
> about why we are proposing a new assertion language for the Relying
> Party when there are several language definitions already out there,
> as well as execution / evaluation engines for those languages.    I
> would hate to have many policy languages evaluated and translated
> across the Higgins infrastructure.
>
> One example of using XACML as WS-Policy constraints was described here
> http://research.sun.com/projects/xacml/ws-policy-constraints-current.pdf
> I have successfully used the http://sunxacml.sourceforge.net/
> evaluation engine to evaluate predicates outside of authorization policy.
>
>
> Duane_______________________________________________
> higgins-dev mailing list
> higgins-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/higgins-dev