ietf-asrg
[Top] [All Lists]

[Asrg] Re: Stamping

2003-03-21 11:51:26
On Fri, 21 Mar 2003 12:07:43 -0500, Kee Hinckley 
<nazgul(_at_)somewhere(_dot_)com> writes:

At 11:36 AM -0500 3/21/03, C Wegrzyn wrote:
I have to ask this one question: why e-stamps?  It seems a solution
looking for a problem to solve.


That certainly seems to be the attitude of the systems I've
seen. However e-stamps do address an ISP concern (namely, they don't
get reimbursed for incoming email, and they can't charge the user, but
the user doesn't *want* most of the incoming email).  Of course
hashcash type solutions don't address that.  Just e-stamp solutions
where the final MTA actually gets real money.

There are a few problems with this.

  1. If the price were over $.001, it would kill mailing lists. For
  instance, this mailing list would cost IETF a $.10
  /subscriber/day. At, say, 500 subscribers, thats $15,000/year. If
  one puts in an 'exception' mechanism for lists, how does one avoid
  having it be abused?

  2. If the price were under $.001, would it really affect unwanted
  email that much?

  3. How does one enforce it. Many organizations just get a leased
  line. How would they get charged? Would you require that every
  cross-domain router MUST have a content-aware packet inspector to
  redirect all email to the 'billing machine'? This severely breaks
  end-to-end, as showable on #4.

  4. With respect to #3, such a mechanism would seriously impact
  robustness of email. Now, its not enough that the bits get there,
  but that every cross-domain router's content-aware inspector must
  work.

  5. Administrative overhead and disputes. Imagine the administrative
  horror in unravelling the accounting of something like Melissa or
  I-LOVE-YOU.

  6. Abuse. Now, someone can recieve money for getting email. They can
  use cracked machines, worms, proxies, and viruses to transfer money
  to themselves.

In any case, we already have a billing mechanism. If I want to send
email to you, do I not incur half of the technical cost? I incur
bandwidth charges and CPU utilization. I may not incur HD charges,
(but, assuming stuff is removed from the ISP server weekly, those are
pretty small in comparison to the others.)

If those sender charges are insufficient to stop spam, why would
merely doubling them, with a 'fair' fictional sender-pays model,
suddenly stop spam? Two-three years ago, internet costs were twice
what they are now, and it didn't stop abuse.

I think what you want is to charge the sender for the time wasted by
the receiver, but is that relevant to MTA's or ISP's?

This assumes that ISPs have control over the email being sent.  If all

Not true, and not necessarily a good idea; many people are already
noticing breakage because their ISP's are unfortunately breaking
end-to-end.

ISPs did (and all ISPs wore white hats), then we wouldn't be sitting
here.  From the white-hat ISP perspective, outbound spam is a bug in
their security.  e-stamps won't fix that.

Outbound spam is not a bug in security. Outbound spam is like any
other outbound communication. Bits are sent to the internet sourced
from their customers. Though it may be annoying to you, is my ability
to give you a prank phone call a bug in the telephone system?

Scott
_______________________________________________
Asrg mailing list
Asrg(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/asrg