Re: [Asrg] Spam, defined, and permissions

[Daniel Feenberg <feenberg at nber.org>]

On Wed, 29 Dec 2004, James Lick wrote:

> I'm not sure I understand something completely...
> 
> In order for a paid postage system to work, ISPs would need to 
> block/control all mail transactions on their network.  The logical way 
> to do this is to block port 25 and monitor and rate limit transactions 
> through the authorized servers.
> 

Actually, they might just as well charge for or rate limit usage of their
mail servers, but leave port 25 open for IP addresses of their customers.
So customers with static addresses could send email themselves, and gain a
reputation as legitimate or not with the DNSBLs. But their own server
would not be of use to spammers, and in that way they could keep the good
reputation of their own server, even if other IP addresses in their domain
got blacklisted.

As for customers with dynamic addresses spamming, presumably victims will
eventually learn the address ranges involved and start refusing mail from
them. I am not sure why zombies have remained such a problem, since they
never produce any legitimate mail, they are ideal candidates for a DNSBL.
More complicated mechanisms are required only when a given MTA is the
source of both spam and ham.

BTW, I would expect that the typical owner of a zombie, told that his
email quota was used up, would not fix the computer, but would switch to
Yahoo or Hotmail. 

Daniel Feenberg





_______________________________________________
Asrg mailing list
Asrg at ietf.org
https://www1.ietf.org/mailman/listinfo/asrg