[cap-talk] What we have here is a failure to communicate

[Jed Donnelley]

At 11:51 AM 12/18/2008, David-Sarah Hopwood wrote:
>Karp, Alan H wrote:
> > If you're wondering why we can't make ourselves understood to
> > security practitioners, consider the following definition from
> > page 55 of the December 2008 issue of CACM.
> >
> > "Authentication: Security measure designed to establish the
> >  validity of a transmission, message, or originator or a means
> >  of verifying an individual's authority to receive specific
> >  categories of information"
>
>:-)

I'll second that :-), or perhaps a :-(  The above really is an
awful example.

At 11:51 AM 12/18/2008, David-Sarah Hopwood wrote:
>Raoul Duke wrote:
> > I guess, statistically, they are in good company; most folks conflate
> > the two all too often?
> >http://en.wikipedia.org/wiki/Authentication#Authentication_vs._authorization
>
>Gack, when even a sentence starting:
>"However, more precise usage describes authentication as ..."
>gets it wrong, you know you're in trouble.
>
>(The most important mistake was using "person" instead of "subject"; I've
>now fixed that.)

Thanks for that fix.

I wonder if it might help any to discuss how we should use these
terms, "authentication" and "authorization" in the context of the
object-capability model.  Neither of these terms are currently
included in the Wikipedia page on the object-capability model:

http://en.wikipedia.org/wiki/Object-capability_model

Might not this in itself lead to some confusion and perhaps less
acceptance of the OCap model?

To me the OCap model does "conflate" authentication and authorization.
That is, the possession (as evidenced by communication) of a capability
demonstrates that the communicating subject is authentic (in that it
apparently possessed the communicated capability) and is authorized to
access (in the sense of "access control") whatever the capability
grants (permits) access to.

Perhaps others have other potentially better ways of relating the OCap
model to authentication and authorization?

I wonder if it would be worthwhile to add a discussion of this
relationship to the Wikipedia page on the OCap model?  The danger
of course is that it would simply lead to more confusion. Thoughts?

--Jed  http://www.webstart.com/jed-signature.html