[Cryptography] Digital currencies

[Natanael]

Den 24 juni 2016 01:32 skrev "Ray Dillinger" <bear at sonic.net>:
> If that were the full basis of comparison, it might very well
> be hard to pick which system is more efficient.
>
> Unfortunately, as far as I can see, in any universe where the
> block chain based solution scales, transactions actually on
> the block chain become so expensive that they are limited to
> very large transactions.  Smaller transactions would have to
> be handled "off-chain," which ultimately opens up the system
> to the traditional model of banking - lending at fractional
> reserve, closed ledgers, individual transactions not recorded
> anywhere they can be checked or proven immutable, etc.
>
> In short, you cannot scale the block chain solution without
> reenabling wholesale fraud and deception, and the expense it
> entails.

Are you sure of that? Because Lightning Network (LN) is a system within
Bitcoin built on extending what I'd call valid "draft transactions" in
between users and LN nodes, using timelocked multisignature transactions.

At first you create a payment channel between you and your LN node where
you commit done coins. After X time, the commitment expire and you've got
control of then again. Before time X and after commitment, you and your LN
node act together to iteratively assign coins in both directions by
updating the "draft transactions" as a result of payments. With networked
LN nodes, you'll tell your LN node to pay X to Y, so you assign coins in
your payment channel to your node, they assign coins to the node of the
recipient, and then the recipient gets coins assigned to him.

Upon withdrawal your "draft" holds the final tally and gets published to
the Bitcoin network, giving you your final balance.

There's multiple levels of safeguards in here, so the LN node *can not*
steal coins or pretend to have more coins than it really does, and
everything is logged by the users, and they can not seize funds (only
temporarily block transfers). Censorship is the worst plausible attack in
the system. Even internal rollbacks is hard (somebody else would have to
explain why, but as I mentioned users always hold a valid transaction ready
to publish, with their final balance).
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.metzdowd.com/pipermail/cryptography/attachments/20160624/887febc8/attachment.html>