ietf-dkim
[Top] [All Lists]

Re: [ietf-dkim] Issue #1265: Signing by parent domains

2006-05-30 13:31:18

On May 29, 2006, at 8:53 PM, Steve Atkins wrote:


The only valid reason to require it, I think, is for the benefit of users who use wildcard MXes, to enable them to make up subdomains on-the-fly, and who send mail using from addresses in those subdomains.

The parent domain can still sign messages. The question is whether a signing-domain should be precluded from linking with an email-address of a sub-domain. This mechanism is intended to establish trust for an email-address via the signing-domain. A verifier could create special policies for when an email-address is a sub-domain of the signing-domain, which may result in the same practical limitations as would excluding the use of this linking mechanism. In general, reducing a profusion of right-hand-side names better establishes trust. Precluding the use of wildcard MX records also reduces some concerns related to various attack scenarios.

If this i= provision is included, the informational RFC1591 should be updated. By allowing a link to an email-address be fully independent of any domain delegation, TLD managers will be able to offer DKIM related services where keys for domains can be stored in a sub-domain directly below the TLD. When one assumes the signing-domain establishes trust, keys at these locations might be considered elite property generating substantial revenue. If it become a common practice for TLD managers to offer DKIM key services, security becomes far more critical due to a catastrophic failure mode.

-Doug






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

<Prev in Thread] Current Thread [Next in Thread>