Re: CGA Use with HBA in Shim6 IETF Meeting July 10, 2006

[Shinta Sugimoto <shinta@sfc.wide.ad.jp>]

Hi Marcelo,

Sorry for the late response, but by answering this email, I would
like to address my opinion on the security mechanism of the SHIM6
protocol.  please find my comments below.

On Wed, 19 Jul 2006 22:39:18 +0300
marcelo bagnulo braun <marcelo@it.uc3m.es> wrote:

> 
> El 19/07/2006, a las 22:28, Bound, Jim escribió:
> 
> > You and I are disagreeing on where the market is going to go 
> > technically.  I don't think we will compromise or that we can 
> > converge.  So trying to save a lot of email here.
> >
> > What we need the working group to discuss now is how much and when do 
> > we worry about the market doing our engineering/scientist work to 
> > identify a protocol.
> 
> exactly
> 
> i think we (you and me) send enough emails about this point, the point 
> is whether the wg thinks that basing the shim6 security in widespread 
> availability of certificates issued by a global PKI is acceptable or 
> not
> 
> comments form the wg?

The choice of IPsec is not realistic, I think. Because, SHIM6 protocol
is expected to run among anonymous nodes.  That is, we cannot assume
any soft of pre-shared secret.  If the peers try to authenticate with
public key, then we must rely on global PKI.  I agree with your statement
that the assumption (wide deployment of PKI) is difficult.

So, I support the current security mechanisms (HBA, CGA, HBA/CGA) specified
in the SHIM6 base protocol which do not require any infrastructure.
I think its more scalable and easy to deploy.


Regards,
Shinta

(rest of the original email snipped)