[ietf-dkim] Charter bashing...

[Dave Crocker]

Folks,

> Frankly, I think this is a huge step backwards. You're changing the 
> charter
> from discussing the goals of the service we're trying to define to 
> discussing
> the details of the mechanisms we use to build the service. IMO this is 
> going
> down a path that is likely to cause far more problems than it solves, 
> as it
> invites confusion with efforts to define very different services using 
> similar
> mechanisms.
> ...
> The existing charter was careful to distinguish between service and
> mechanism. Let's please try and keep that distinction.


This is a point that Ned has been stressing and I believe he is entirely 
correct. 

The benefit of having our discussions consider mechanics as "merely" 
secondary, so that we maintain a focus on goals/purpose, strikes me as 
massive.

> DKIM is _not_ an alternate signature service, and that's precisely the 
> point.
> DKIM only uses signatures as a means to an end, and the end is not to 
> provide a
> nonrepudiatable signature covering the message. Rather, it is to 
> provide a
> means whereby someone can assert responsibility for a message. This is 
> a type
> of authorization service, not a signature service. We are forced to 
> use digital
> signatures as a mechanism because the service has to deal with forgery 
> and
> replay attacks, but that's an (unfortunate) implementation detail.


The main reason I am posting this response is in the hope that folks 
will (re-)read the text of Ned's that I have quoted. 

I believe a very great deal of confusion will be avoided if we can all 
embrace this one, main concept that he has so nicely distinguished.

d/