[cap-talk] A can of worms, or, "why do we care so much about identities?"

[Jed Donnelley]

At 06:28 PM 12/11/2006, David Hopwood wrote:
>Karp, Alan H wrote:
> > Jed wrote:
> >
> >>Still, I'm struck most by how the situation in computer systems seems
> >>much simpler to me.  Just thinking in terms of public/private key
> >>pairs, they can be created, the public key shared, trust developed,
> >>etc. regardless of the nuances of self behind the identity.
> >
> > It's important to remember that private keys can be shared just as
> > easily as public keys.  We may have problems if we act as if a private
> > key identifies a person.
>
>In some (most?) situations it is sufficient to identify a persona; in others
>it is required to identify a person.

The situations where it's necessary to attribute an action to a person,
the need presumably derives from a requirement to tie the action to a
person.  Any such requirement would of course be difficult or impossible
to meet in a situation where only electronic communication was available.

>(The essay I referred to says as much, but so verbosely that it is difficult
>to make out the point.)

I agree that the "essay" you refer to was way too verbose - even for my
standards ;-).  It focused on a few fairly simple examples and pounded
them to death.

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