On Jun 11, 2012, at 8:57 AM, Tony Finch wrote:

> Janne Snabb <[email protected]> wrote:
>> 
>> IMHO it seems pretty pointless to implement DANE for SMTP without the
>> ability to verify both sides of the conversation.
> 
> The point of server authentication for SMTP is to make sure you are
> sending mail where you intend to.
> 
> The point of client authentication is ... what?

...to make the server sure it is speaking to the client it thinks it is.

> What the server cares about is the authenticity of the message data, which
> has relatively little to do with where it came from and even less to do
> with the client host name.

That's one view; another is that the two parties both want to be sure of the 
others' identity.

Jakob and I are dealing with client auth in our S/MIME proposal; it could be 
applied to client certificates as well.
_______________________________________________
dane mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dane

Reply via email to