[ietf-dkim] Base issue: multiple linked signatures

[Hallam-Baker, Phillip]

> [mailto:ietf-dkim-bounces at mipassoc.org] On Behalf Of Charles Lindsey

> I just want to point out that similar arguments also apply to 
> the "MUST sign the From header" case.

I disagree, I can easily write a test vector for that MUST so it is emprically verifiable. Hence it can be made a MUST.

In general we can place MUSTs on senders but its much harder to do this on receivers. A Must on a receiver means 'if you do this and it breaks in future we will hang you out to dry if we choose unless you manage to make the practice so widespread that we cannot ignore it'.