ietf-dkim
[Top] [All Lists]

RE: [ietf-dkim] user level ssp

2006-09-06 14:23:45

[mailto:ietf-dkim-bounces(_at_)mipassoc(_dot_)org] On Behalf Of J.D. Falk

On 2006-09-06 10:45, Hallam-Baker, Phillip wrote:

The main value I see in user level policy is easing phased 
deployment. 
If you are a bank with 100,000 employees with email and you want to 
deploy DKIM you probably want some form of hook that lets 
you do it in 
stages.

So they'll have 100,000 SSP records?

Perhaps there's an easier, more flexible, more scalable 
hook...like "we don't sign all mail."

We don't sign all mail is utterly useless as a policy record. There are only 
two usefull policy positions ALWAYS SIGN and MIGHT SIGN. There is no value in 
distinguishing MIGHT SIGN and NEVER SIGN.

Since you can wildcard the most common case one would need 50,000 SSP records 
at most. It is likely that they would be generated automatically as individual 
mail servers were configured to use DKIM.

BITs are cheap. I see no problem in deploying 100,000 DKIM records in such a 
situation. 


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

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