> [mailto:[EMAIL PROTECTED] On Behalf Of Frank Ellermann

> Charles Lindsey wrote:
> 
> >> The folks supporting to list used algorithms in the SSP apparently 
> >> think that receivers could care about this nuance.  And the folks 
> >> opposing that idea note that spammers would try to abuse this info.
> 
> > Eh? This info is provided to counter a possible exploit. Nobody has 
> > yet suggested that this extra info will open the way to yet further 
> > exploits.
> 
> I'm too lazy to dig through the last 250 or so messages to 
> find the source (probably posted by John or Dave), but IIRC 
> the idea was this:
> 
> A signer publishes to support a new algorithm "rot13".  If 
> spammers happen to know that certain receivers don't support 
> "rot13", they can forge (invalid) "rot13" signatures in 
> phishes to these receivers.

John introduced ROT13 but unless I am severely mistaken he was arguing that 
this attack was unimportant.

My argument is that the attack is a very important one and that policy MUST 
meet it.

Since I only see the need for two DKIM policy tags, DKIM meaning I always sign 
and DKIM-TEST meaning I am testing my policy to see if it works, I don't see 
why adding what is necessary here should lead to endless complaints about 
over-complexity.

All we need to add here is DKIM=<selector-sufix>

The algorithm for verifying compliance with policy is only slightly more 
complex.

_______________________________________________
NOTE WELL: This list operates according to 
http://mipassoc.org/dkim/ietf-list-rules.html

Reply via email to