[Asrg] Re: Pay-for-attention (Was Re: article on spam)

[mathew <[e-mail]>]

On Thursday, May 22, 2003, at 17:13 US/Eastern, Shannon Jacobs wrote:
> If you know of an anti-spam email system that will block any 
> advertising
> UNLESS the advertisers pay MY price for MY time, then please tell me 
> about
> it. I'll sign up and consider my spam problem solved.

Yes, this strikes me as exactly the right model.

People I know get to mail me for free.  Everyone else has to pay a fee 
*I* set if they want to contact me.  I'll call it "postage" in the 
discussion below, for want of a better word.

People sending mail would sign up with a postage provider. There could 
be many of these, just like there are many online banks and payment 
systems. I imagine that people like PayPal would add Internet postage 
as a part of their system, in fact.

The system receiving mail for me would check to see if the person's in 
my address book. If not, it would check for postage--that is, it would 
check my account to see that the appropriate amount had been credited. 
No postage, and either my mail server bounces it, or passes it through 
with a flag so I can choose what to do.

[Maybe valid postage means they get added to my address book, so they 
only have to pay once.]

 From the sender end, the way it would work is that a Postage 
Requirements Server would be resolved, much the same way as an MX is 
resolved, based on my e-mail address. Their client would contact the 
Postage Requirements Server, tell it their e-mail address, and it would 
reply with how much postage I want from them before I'll accept their 
mail. They'd then make a decision whether to use their postage provider 
to send me the money along with the e-mail, or not. Once their system 
confirmed that postage had been sent, they'd send the actual e-mail 
with an appropriate header indicating postage was paid.

What's in it for the end user? Well, apart from going a long way 
towards solving the spam problem, most people could probably fund their 
occasional need for outgoing postage from the proceeds of incoming junk 
mail, and still make a profit.

What's in it for the people deploying it? Money! Postage providers 
(sending and receiving) could charge a penny on every piece of non-zero 
digital postage. These fees would be taken into account when the 
sender's client contacted the Postage Requirements Server.

Transaction charges for the real-world transfer of the money would be 
minimized by rolling up the transactions at the end of each month, just 
like Amazon rolls up my tip jar transactions into one deposit, Paypal 
rolls up my transactions into one bank credit or debit, the Apple Music 
Store rolls up my track purchases into one bill, and so on.

I imagine ICANN or some such organization would keep a limit on the 
number of postage providers, enough to allow competition and keep rates 
low, but not so many that the market would be fragmented and 
transaction costs make it unprofitable.

As long as we had a list of "official" postage providers you wouldn't 
need any fancy cryptography; just SSL connections. Sender client 
contacts postage provider, gives it details of my account and amount to 
send. My mail client contacts my provider, checks amounts received and 
e-mail addresses or message IDs to whitelist. The providers would need 
to know about each other, and agree to interoperate, much like domain 
registries do.

Another nice feature would be zero cost postage for friends. That would 
prevent spammers from being able to break through by sending e-mail 
pretending to be from someone I know, as my system would check for a 
zero-cost stamp and only allow through the right message IDs, yet at 
the same time my friends wouldn't have to pay money to e-mail me.

As far as I can see, something like this can be built using technology 
available today. Although eventually client support would be added, 
initially it could be done using a web interface to your postage 
provider to pay to get you added to my address book, on the "pay once" 
model. The receiving server needs to know enough HTTP and SSL to do the 
equivalent of a PayPal account query, but that's about it.

In fact, I'm rather surprised nobody has built something like this 
already. It seems pretty obvious. Maybe there's some massive flaw I'm 
missing, hence this e-mail. I'm sure if anyone can nit-pick the idea to 
death, it's the people on this list...


mathew

_______________________________________________
Asrg mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/asrg