Re: Dialback authentication

In the case of c2s (to borrow the xmpp nomenclature), we'd need to use
something like client certs - dialback won't work. In that case, we'd want
my proposed split, since otherwise we'll have two protocols.
On Sep 7, 2012 7:01 AM, "Michiel de Jong" <michiel@unhosted.org> wrote:

> Hi Blaine,
>
> On Wed, Sep 5, 2012 at 5:45 PM, Blaine Cook <romeda@gmail.com> wrote:
> > "hosts negotiating a secure / shared channel.
>
> i would not make the assumption that the client is also a host. What
> if the client is (honouring its name) client-side? Then the split you
> propose would not be as useful, i think?
>
> or maybe i misunderstood what you're proposing, in that case sorry!
>
> Ciao,
> Michiel
>