ietf-dkim
[Top] [All Lists]

Re: [ietf-dkim] DKIM does not claim content is correct

2009-01-28 21:50:06

On Jan 28, 2009, at 5:56 PM, Al Iverson wrote:

Forgive me in advance for a really basic, quite possibly dumb,
question. What does DKIM do that receivers find valuable? What does an
ISP or large corporate receiving entity hope to get out of DKIM
adoption?

It provides an (optional) persistent identity for senders, one that
doesn't need any updating or cross-referencing between identity
and IP address. History for those identities can be maintained
mechanically, with fairly minimal human overhead.

There's no reason that a sender/author ever has to change that identity,
so the problem of moving IP addresses doesn't arise, meaning
that the receiver doesn't have their reputation data invalidated
for no good reason on a daily basis by authors moving IP address
(or ESP) and the receiver doesn't have to deal with the manual
overhead of either supporting senders/authors who change
their source IP address or supporting customers when mail that
was being delivered fine yesterday is discarded today.

That'll allow receivers to maintain a view of the reputation of  
legitimate
senders much more reliably and at lower cost, meaning that their
users will get more of the mail they want at lower cost to the ISP,
which is good for the bottom line.

It provides all of the advantages, in a much cleaner and more
specific way, as the current use of SPF by AOL to map whitelisted
domains onto IP addresses mechanically. And it also allows those
advantages for domains that aren't big or loud enough to sign up
for whitelisting, as their reputation can be monitored by receivers
with no need for any communication between author and receiver
other than the DKIM signed messages themselves.

There are other advantages to a small subset of large receivers,
around feedback loops and suchlike, but they're gravy.

All of these advantages to the receiver could probably be provided
by something a tenth the complexity of DKIM, but that's what
seems to happen to protocols in this segment of the industry
when you make the mistake of paying too much attention to
everyones input.

Cheers,
   Steve

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

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