forked from cheng/wallet
196 lines
8.1 KiB
HTML
196 lines
8.1 KiB
HTML
<!DOCTYPE html>
|
||
<html lang="en"><head>
|
||
|
||
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
|
||
<style>
|
||
body {
|
||
max-width: 30em;
|
||
margin-left: 2em;
|
||
}
|
||
p.center {
|
||
text-align:center;
|
||
}
|
||
</style><title>Ripple</title></head><body>
|
||
|
||
<p><a href="./index.html"> To Home page</a> </p>
|
||
|
||
<h1>Ripple</h1>
|
||
|
||
<p>This rambling explanation fails to explain what
|
||
level of anonymity is provided, or how it is provided.
|
||
We want aggregative anonymity, cryptographically
|
||
secured, with money flowing on chains of trust.</p>
|
||
|
||
<p>Fails to explain need for integration with
|
||
reputation management, fraud problem, ordering, person
|
||
to person transactions</p>
|
||
|
||
<h2>How internet money should work</h2><p>
|
||
|
||
The idea is that an organization will issue some kind of money
|
||
equivalent – promises to deliver US dollars, promises to deliver gold,
|
||
shares, promises to deliver some service. Let us call the organization
|
||
issuing the money Big Money Issue Corp, and the money
|
||
Big Money Issue Corp OUs. </p><p>
|
||
|
||
You will be able to buy Big Money Issue Corp OUs from
|
||
Bob, and use them to pay Carol for something. You will
|
||
be able to do this over the internet by instant
|
||
messaging – dropping some money into the instant
|
||
message. Let us suppose you pay Carol to deliver some
|
||
cigarettes. Carol will know you paid her, and what you
|
||
paid her for, and you will be able to prove that Carol
|
||
agreed to deliver cigarettes for a certain payment and
|
||
that you paid to deliver some cigarettes, prove it to
|
||
the whole world, which ordinarily you will not want to
|
||
do, but you might want to do if the deal goes bad.
|
||
</p><p>
|
||
|
||
Now the most obvious way to make this work is that the
|
||
people operating the financial system know everything
|
||
about everyone, and all transactions are on their
|
||
computers, readable by the government, who can thereby
|
||
tax everyone. </p><p>
|
||
|
||
We don’t want a technology where everyone deals
|
||
directly with Big Money Issue Corp, because then Big
|
||
Money Corp will be under pressure to refrain from
|
||
dealing with this or that person or group, and then it
|
||
will have to know its customers, and for a big
|
||
organization to know its customers is likely to be
|
||
inefficient for the big organization and burdensome for
|
||
the customers. </p><p>
|
||
|
||
We want payments to be provable, and linked to deals,
|
||
so that if something goes wrong, the customer can do
|
||
something about it – but provable and deal linked
|
||
payments run contrary to anonymous payments. </p><p>
|
||
|
||
Cypherpunks had a vision of how anonymous funds
|
||
transfer would work on the internet, which vision ran
|
||
into some disastrous obstacles. The cypherpunk vision
|
||
was built around Chaumian anonymity, but this
|
||
technology does not in itself provide many capabilities
|
||
that are necessary for internet money, and makes
|
||
implementing those capabilities more difficult. In
|
||
the Chaumian vision, everyone dealt directly with Big
|
||
Money Corp, but Money Corp could not link Bob’s payment
|
||
to Carol receiving money. Unfortunately, this meant
|
||
no recourse for Bob if he paid and Carol stiffed him,
|
||
plus the fact that everyone dealt directly with a
|
||
single center provided a single point of control
|
||
</p><p>
|
||
|
||
<a href="%3Chttp://ripple.sf.net/">Ripple</a> is a
|
||
brilliant insight, great idea, but suffers from the
|
||
critical mass problem that afflicts any proposal to
|
||
provide competition to the existing government run
|
||
money system. </p><p>
|
||
|
||
In addition to the critical mass problem, governments
|
||
tend to treat competition as illegal, once the
|
||
competition gets large enough to to make a dent -
|
||
observe what is happening to e-gold. E-gold is
|
||
attempting to comply with overbearing and intrusive
|
||
regulation to the great detriment of its business and
|
||
its customers, but the government has no intention that
|
||
it be possible for e-gold to comply. Of course a
|
||
decentralized system is a lot more resistant to this
|
||
than a centralized system such as e-gold. E-gold looks
|
||
like being the Napster of internet money – one that
|
||
showed the way, though it was ultimately suppressed.
|
||
</p><p>
|
||
|
||
Therefore, in my opinion, the best prospect for a
|
||
decentralized money system such as ripple is in some
|
||
activity that is already somewhat illegal, for example a
|
||
file sharing system, in which the participants grant
|
||
each other upload credits. </p><p>
|
||
|
||
Ripple, as it stands at present, does not really address
|
||
the routing problem, (in the sense that the proposed
|
||
routing algorithm sucks), and indeed does not need to,
|
||
for there are very few participants. For a file
|
||
sharing system, would need to be more DHT like.
|
||
</p><p>
|
||
|
||
To the extent that a payment system is anonymous, tends
|
||
to be used by ordinary criminals and ponzis. This gives
|
||
government justification and excuse to shut it down.
|
||
</p>
|
||
|
||
<h2>Requirements of an internet payment system:</h2><p>
|
||
|
||
Payments should be identifiable by the participants, but
|
||
not identifiable by outsiders, and should automatically
|
||
generate and maintain the appropriate records on the
|
||
participants machines. If you pay a net presence for X,
|
||
you should be able to prove to outsiders that you paid
|
||
that particular identifiable net presence for that
|
||
particular X, but outsiders should not be able to prove
|
||
it except with your assistance. This requirement
|
||
presupposes a mechanism for identifying net presences -
|
||
Zooko based yurls, and/or authorized certificates.
|
||
Since we obviously want to transfer money between people
|
||
who do not necessarily have authorized certificates,
|
||
have to support yurls, which means we have to first have
|
||
a yurl system, which at present we do not. </p><p>
|
||
|
||
To resist shutdown, and keep costs low, a payment system
|
||
needs to inherently implement "ripple", making every man
|
||
a bank. Ripple, however tends to reduce the profits of
|
||
the currency operator, and is some work to implement, so
|
||
is generally not done. </p><p>
|
||
|
||
The problem with Chaum’s anonymous cash, and Brands'
|
||
anonymous cash, as generally envisaged, is that it
|
||
relies on a single central server, with which everyone
|
||
has a relationship – a star shaped network. Internet
|
||
networks really have to be power law networks – neither
|
||
star, nor entirely equal peers. A single central server
|
||
is a central point of failure, a point that governments
|
||
can attack. </p><p>
|
||
|
||
"Ripple" means that if the Bank of Whatsit implements
|
||
ripple, then you don’t need an account with the Bank of
|
||
Whatsit. You can do fine with an account with someone
|
||
who has an account with the bank of Whatsit. Hey
|
||
presto, a power law network. </p><p>
|
||
|
||
So if you want to buy a domain name registration from
|
||
Wonton, a registrar that has an account with the Bank of
|
||
Whatsit, and you have an account with Joe, who has an
|
||
account with the bank of Whatsit, then Joe’s machine
|
||
tells Bank of Whatsit to pay Wonton. The Bank of
|
||
Whatsit only knows that Joe payed Whatsit. Joe knows
|
||
that you paid Wonton, but does not know what you paid
|
||
Wonton for. Wonton knows you paid them through the Bank
|
||
of Whatsit, does not know or care who Joe is, knows what
|
||
you paid them for, but cannot prove what you paid them
|
||
for. You know, and can prove, that you paid Wonton, and
|
||
can prove what you paid them for. </p><p>
|
||
|
||
Now what I have described is all account based – no
|
||
Brand style or Chaum style anonymous tokens. But if the
|
||
Bank of Whatsit implements ripple, Joe, or anyone else,
|
||
*can* implement Chaum style or Brand style anonymous
|
||
tokens. Further, you can pay Wonton using Joe’s tokens
|
||
even if Wonton has no relationship with Joe, and has
|
||
never heard of these new fangled tokens, and the Bank of
|
||
Whatsit has never heard of them either, for through
|
||
ripple you can turn the token payment into an ordinary
|
||
account payment from the Bank of Whatsit. </p><p>
|
||
|
||
Thus if we had payment guys implementing ripple, they
|
||
could avoid the heat that is likely to arise from
|
||
implementing Chaum style tokens, indeed be entirely
|
||
unaware of either those tokens or that heat, combining
|
||
the anonymity that comes from Chaum style tokens with
|
||
the anonymity that comes from the faceless crowd of a
|
||
power law network. </p>
|
||
<p style="background-color : #ccffcc; font-size:80%">These documents are
|
||
licensed under the <a rel="license" href="http://creativecommons.org/licenses/by-sa/3.0/">Creative
|
||
Commons Attribution-Share Alike 3.0 License</a></p>
|
||
|
||
</body></html>
|