[cap-talk] [e-lang] Second ABAC Google talk is now up

[Jed at Webstart]

At 03:49 PM 7/14/2006, Brett Cannon wrote:
>...
>
>I think perhaps less talk about how object-capabilities handles 
>specific issues that have been presented but more on making it 
>blatently obvious how object-capabilities differ from a 
>permissions-based system inherently would have helped.  I think it 
>was helpful to point out where the misconceptions have propagated 
>from, but more base information on object-capabilities and comparing 
>to permission-based systems to help the uninitiated understand would 
>have been helpful.
>
>-Brett

I agree with the above.  I hope Tyler is able to remedy this 
situation in the third ABAC talk.

Related to this is:

At 07:25 PM 7/16/2006, Kevin Reid wrote:
>On Jul 14, 2006, at 18:41, Mark Miller wrote:
> > ...
> > I think this talk could have been much better, and hope to improve it,
> > so feedback appreciated. Thanks.
>
>Regarding the history review at the beginning, I had the feeling
>that, if I had not already known of the things (ideas, papers,
>implementations, etc.) you spoke of, it would have been completely
>uninformative and uninteresting to me, as it was discussion of the
>relation between those things with little explanation of what those
>things are (what the papers said, what the implementations succeeded
>at).

I read the above as a similar suggestion.  Namely the technology needs
to be put onto a solid foundation as to what it provides (zero based
access control, subjects able to communicate permission tokens
in messages - capabilities, and how they can be used) before it can
be contrasted or put into historical perspective (probably not needed
at this point I think), etc.

I believe this is particularly important at the network level where Tyler will
speak next (as I understand it) and where it would seem Google (OK,
me too!) has the most direct interest and requirements.

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