What about using l=0 in combination with
bodyhash=sha:w2iu2y32iur2iu3yriuy2r3== ?

This is consistent with current implementations while allowing people to
calculate the bodyhash separately. The only thing that a legacy application
looses is the ability to verify the message body.

This provides a very nice, clean solution to deployment of new C18N
algorithms as well provided that the header c18n is not affected.


> -----Original Message-----
> From: [EMAIL PROTECTED] 
> [mailto:[EMAIL PROTECTED] On Behalf Of Jim Fenton
> Sent: Tuesday, March 28, 2006 12:02 PM
> To: [EMAIL PROTECTED]
> Cc: [email protected]
> Subject: Re: [ietf-dkim] mailing lists and -base
> 
> [EMAIL PROTECTED] wrote:
> > Is signing the body at all an essential requirement? Yes, some 
> > potential risk for a replay attack but otherwise "whoami I 
> sent this" 
> > should be sufficient for some providers,
> >
> >   
> As long as people support the l= tag, they could use l=0 to 
> not sign the body.  This capability has been cited as a 
> reason to get rid of l= because it facilitates such 
> "dangerous" behavior.  IMO, if they want to sign such 
> messages, and recipients want to accept them, let them do that.
> 
> -Jim
> _______________________________________________
> NOTE WELL: This list operates according to 
> http://mipassoc.org/dkim/ietf-list-rules.html
> 
> 

Attachment: smime.p7s
Description: S/MIME cryptographic signature

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

Reply via email to