Re: 6. Proposal - Sender Pays - Article (was Re: [Asrg] ziff-daviseditorial on sender-pays)

["Eric S. Johansson" <esj@harvee.org>]

I apologize for disappearing on this topic for well but I have been in Web 
framework hell.  I will probably only pay attention for a short while before I 
go back to Sisyphian labors.
---eric


Yakov Shafranovich summarized:

> To summarize the article, the following problems apply to sender-pays:
more specifically, money based sender-pays.  None of these issues exist with 
proof of work stamp based sender-pays.


> 1. The Internet is global - "So implementing Sender Pays would require 
> an international treaty".
Postage in this case becomes an end to end protocol and not a cost of entry to a 
controlled access monopoly delivery service.  All it would require it is private 
agreements for various issuers to settle accounts.  If anything, there would be 
a strong need for government intervention to prevent monopolization of postage 
issuance as has almost happened in the certificate market.

> 2. "Micropayments remain a problem."
quite true but since this is a closed domain not an issue..  Generating a coin 
is no problem.  It is purchased from and cleared from the same source.  also can 
be dealt with by settlement agreements between two issuers.

> 3. "Payment systems would have to be set up so that innocent parties 
> wouldn't be stuck covering the cost of other persons' e-mails. Probably 
> everyone would have to pay some amount in advance to cover the mail we 
> might send."
think bearer instruments.  Mail will not be accepted without a valid coin.  No 
need for book entry.

> 4. "by adding a cost to a message, Sender Pays would certainly impact 
> the free e-mail business. Such services would have to decide whether to 
> pass the costs along or pick up the tab."
not necessarily.  using a bearer instrument, the service is not an originator of 
the postage but merely passes it along.   One could also have an account of 
postage coins with the e-mail service so one could send e-mail while they are on 
the road and no longer have access to their stamp pad.

> 5. "Here's another interesting problem: allegations have been made that 
> there are Internet worms and other attacks that create an open proxy for 
> spammers on an infected system" - "While this particular form of 
> hijacking is built on old-fashioned SMTP, it would still suggest that 
> Sender Pays will require that we first solve the open proxy problem—yet 
> another difficult issue."
proxy relaying will have no impact on sender pays systems.  They still have to 
have a stamp of some form in order to be delivered.  The real problem is a worm 
that gets into your system and takes over your e-mail client and sends lots of 
mail to mailboxes in Nigeria.  If you compromise 300,000 hosts (like they did 
last time), it only takes $3.50 per machine to give you $1 million.  ka-ching!
Now Chairman Bill really can get a dime every time one of his machines are 
compromised...

things he didn't consider:

cost of infrastructure.  And who pays.

invariably the cost of infrastructure is very high because it must be designed 
for peak loads.  Additionally, the double spending database must record every 
single issued coin and its state for a given epoch.  This is not a small database.

The issuer pays for their infrastructure out of fees tacked onto the issuance of 
stamps.  One could make this model more sender pays if the fees were tacked onto 
the redemption of stamps.  But in the end, somebody pays and pays a lot.

latency of double spending database

spammers will double spend.  It's important to keep the latency of the double 
spending database to a minimum which means stamp redemption as soon as the mail 
is received and subsecond update times from stamp redemption to entry in double 
spending database.  That way the window of opportunity for a double spent stamp 
becomes very small.  The acceptable size of this window is a function of how 
many double spend checks per second one can make of the database.

how to cope with inaccessibility of double spending database.

since then that is not perfectly reliable, invariably the double spend database 
will be unavailable at times.  The decision is then whether to hold mail or pass 
mail.  Both are problematic in different ways.  Additionally, if there are 
multiple network caches of the double spent database providing service to a 
greater portion of the net.  If they lose connectivity to the double spent 
database, they should go silent and not give potentially false answers.

issuer noncooperation

some folks might see it as a competitive advantage to have a private currency. 
They may perceive it as a "lock in" and if they can just get a monopoly, they 
will be set for life.

What's more likely is that there will be a large number of issuers some of which 
will settle with others and some of which won't.  As a result, people probably 
have dozens of little accounts where postage accumulates.  This will trigger a 
market in software to purchase postage from all these different accounts if 
there is a sufficient positive balance in any of them thereby solving the 
settlement problem with external tools.

many issuers will go bankrupt.

if anything, this will trigger the most discussion of government regulation and 
a significant loss in confidence in a sender pays by money world.  Underfunded 
issuers will regularly go bankrupt causing individual losses on the order of 10 
or $15 per person.  if this happens once or twice to a person.  It won't be a 
big deal and will push them towards big bank backed issuers.  We will see a 
similar cycling of fees up and down as we have seen in the personal banking arena.

But the threat of people losing money will be the biggest kick in the pants to 
government regulators to try and "protect people's interests".  It probably 
won't go far if the banking industry steps in and office to provide a stable and 
safe environment.  If anything, you'll see regulation at that point to reduce 
competition and freeze out the independents.

the loss of confidence is a more serious problem especially during the 
transitional phase if people get stamps issued by a defunct bank that has been 
taken over by a spammer.

international regulations

or more to the point, transit fees to governments or private parties imposed by 
international regulations.  folks may look at all that money flowing around the 
net and decide that they want to piece of the action for no other reason than 
the traffic is transiting the network.

Money is so much fun...

---eric


>
[1] this is a SWAG number.  It's a gut feel based on how long you will let a 
double spending event persist and allow Spam to traverse filters.




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