[ietf-dkim] 1193 considered harmful

[Douglas Otis]

On Mar 23, 2006, at 10:23 AM, Arvel Hathcock wrote:

> > 3) +   can hash a body once for redistribution; a fairly marginal
> >        feature that might help mass mailers, but Moore's law is just
> >        as likely to help, um, more.
>
> For mailing lists which sign their outbound traffic surely this  
> benefit is more than marginal; especially for large lists?  I don't  
> have as much knowledge here as Mike and I'm no crypto expert (thank  
> God for OpenSSL BTW) but isn't the hashing part what takes the  
> longest (sorry if I'm wrong on that).

Here is an interesting link on performance.

http://www.eskimo.com/~weidai/benchmarks.html

The MB/S rating of the hashing function is therefore dependent upon  
the size of the message.  The RSA overhead is dependent upon the key  
size.

The advantage found by the hash parameter from the verification  
perspective is found when considering the related overhead related to  
weeding through a series of possibly "broken" signatures (largely  
related to more expensive DNS transactions) and various added  
signatures when this becomes more common place.

-Doug