> > I guess in a distributed scheme, you could look at every
> > lambda as an anonymous capability?
>
> Assuming that the lambda expression somehow expressed access
> rights as well as functionality, yes.
Maybe I still haven't t it, but if a capability is a pair of
{name,action to perform}
(following in the footsteps of the earlier RPC conversation), then the presence of lack thereof is the really important thing. In other words, if an object doesn't have the capability, they can't invoke it.
In a distributed scheme, the ability to access objects is strictly controlled by the lexical scope, and the one means for passing on capabilities the environment.
Last night, I was thinking a bit more, and in the distributed scheme, objects were never truly anonymous... there was always some form of id used...