IOweYou Credit Network

by Aniket Kate @Purdue

centralized (Amazon, Uber, etc.) --> decentralized business model

crypto-currencies may or may not survive, but the concept of distributed ledger/blockchain remains


protocol: application level, middleware/service level, infrastructure/base level

thing-thing trade: problems rise from lack of communication medium
stone money: oral history, no  physical movement

Questions:
How well do we understand their consensus process?
Proof of Work vs Proof of Stake

Bitcoin network has scalability problem because of all the communication required.
Credit Networks solves this problem.

Essense of network: confidence on your friends

Problems of credit network:
Path selection (how do we find and select paths)
Liquidity of the network (restrict to certain nodes and paths, what's the probability of transition success?)
Game prevention --> loss due to misbehaving identities is bounded and (sometimes) localized ---> assumes introducing nodes is much easier than drawing trust from well-behaved nodes

Examples:
1. Bazaar (NSDI'11) --> seems to look on simulation of eBay data

2. Ripple Credit Network (realized)
    allows for currency exchange (node performs exchange, you need to find a path with such nodes)
   

Comparison from Bitcoin network:
transfer: bitcoin directly from two wallets, credit network via a path with enough credit
liquidity: good vs. restricted by path availability
scalability: imited (<100 bps="" high="" nbsp="" p="" scalability="" vs.="">
Can augment the credit network with social trust

Privacy might be a problem in Ripple: if I can link one transaction to you, I can find all your transactions.
How to define privacy?
transaction value privacy and transaction receiver transaction