On 27/Aug/11 16:55, John Levine wrote:
>>> any way to tell the difference.  If it mattered, I'd suggest adding
>>> a tag in the record saying which sort of name the record applies
>>> to, with multiple records if need be.
>>
>> +1, how about Discovery-Origin?
> 
> Seems a bit prolix when the other tags are r= and ru= but whatever

Oops, sorry John, I don't know why I was thinking of feedback fields.

>>> Sec 9, security, the obvious problem is that a malicious sender can
>>> publish "rp=o [email protected]" and indirectly mailbomb people.
>>
>> Should that be equivalent to a redirection?
> 
> No, it might be real and it might not be.  The point is that for this
> not to be a DDoS vector, there needs to be some way to validate the
> address before sending it reports.

Perhaps, storing that in the DNS provides for easier verification
than, say, an HTTP query.  If example.org outsources report handling
to example.com, the latter could publish a confirmation as

  example.org._report.example.com. CNAME _report.example.com.
  _report.example.com. TXT "rp=o [email protected]"

or

  example.org._report.example.com. TXT "[email protected]"
  ; rp= retained from original record, or default.

or

  example.org._report.example.com. CNAME _report.example.org.

This way, a generator can confirm the external domain and fetch any
relevant detail using a single extra query.  In the latter example,
the service accepts whatever settings its customer may define, but it
can always revert to one of the former cases; that is, the real
handler is in control of the target email address.

Just an idea
_______________________________________________
marf mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/marf

Reply via email to