[cap-talk] Demonstration and engagement (was: If not God)

[Jed Donnelley]

At 07:23 PM 4/7/2008, David Wagner wrote:
>Jonathan Shapiro writes:
> >You don't open doors by engagement. You open doors by demonstration.
>
>I'll second that.  I'd say the way to go is to demonstrate first, and
>then try to raise its visibility second (e.g., publish it, sell it, promote
>it, etc.).  Engagement without having built something real to demonstrate
>is dismissed as vaporware or speculation.
>
>Of course, building a system without educating the world about it may
>not achieve your ends either.

I'll second that.

We have CapDesk, PLASH, Waterken, whatever is current in the
KeyKOS, EROS, CapROS, Coyotos thread, Polaris, Tahoe, SAML
assertions, BitFrost, the DARPA browser, and likely many others
I'm missing or otherwise inappropriately categorizing.

Is there something essential in these demonstrations that's missing?
What additional demonstrations do we need?

<OK, I admit that I still want to get what amounts to
a Web capability wiki/explorer available.  Whenever
I want to show something along the lines of what we
had at LLNL for 30 years - I don't have anything to show.
All I have is wideword, which is inadequate in many ways.
That is the one demonstration I want to work on.>

I believe we have plenty of demonstrations.  I also believe the
situation we face with demonstrations of the POLA value of
capability systems is particularly difficult in that the
essential value of POLA is invisible.  Show somebody a
demonstration of CapDesk and it looks essentially like
Windows or MacOS or a Unix windowing system.  What is there
to "demonstrate"?  We're not talking about something like a
Mosaic demonstration (probably the best I ever saw) where you
*see* the essence of the value on the surface.  The value is
hidden.  The essence of the demonstration is that it does
look the same.  It can look the same, act the same, but be
more secure (resistant to attack) inside.

The DARPA browser was a demonstration.  A demonstration for who?
David Wagner and his team were involved.  A few folks at DARPA.
The developers.  Was that demonstration a failure?

Truly, what more demonstrations do we need?

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