[cap-talk] Why is EQ so dang fascinating?
toby.murray at comlab.ox.ac.uk
Sat Oct 27 06:20:17 EDT 2007
On Fri, 2007-10-26 at 16:44 -0700, Jed Donnelley wrote:
> On 10/20/2007 12:31 PM, Dean Tribble wrote:
> > The real question: why is EQ so dang fascinating?...
> Isn't this because EQ is the one primitive that seems
> to need to exist besides "invoke"?
This is exactly the question I've been trying to answer when raising EQ
on this list.
The version of the object-cap model as described in Robust Composition
(Mark Miller's PhD thesis) specifically avoids dealing with EQ. In
trying to come up with a more formal definition of the object-cap model,
one needs to address the issue of EQ -- should it exist? If so, in what
The best way to decide these questions, I believe, is to know what
cannot be expressed without EQ. If useful patterns cannot be expressed
without EQ, then this gives strong evidence that EQ should be included
in a formal object-cap model. Otherwise, we may choose to leave it
For the record, I'm still undecided here. I now see exactly why the
issue of EQ was purposely avoided in Robust Composition.
More information about the cap-talk