Topic: P2P EXCHANGE (Read 5301 times)

I've been thinking about this off and on for a couple years, and the only way I can see a zero-trust P2P fiat exchange, is if the issuing authority (or maybe a highly trusted institution) issues their own P2P crypto currency.  All they would need to do, is clone/develop their own PoS fiat system, issue themselves a trillion dollars, then release the network to the public. They can then issue their currency to banks and other institutions for further dissemination to the public at large.

This still does not solve the problem of a zero-trust P2P currency exchange.  A previous poster suggested that this would require a hard fork in one of the existing currencies.  This is true, but a proof-of-concept can be done without a hard fork.  Just create a new currency with the features required to solve the problem.

This new exchange and trade currency will have the ability to process a transaction based on transactions in 3rd party block chains.  A payment script that would be way too experimental to include in an established blockchain before proving itself.

Here's what would happen if Bob wanted to buy BTC for SNC (some new coin):

Bob announces that he wishes to buy 1 BTC for 100 SNC in the next hour.

Alice sees the announcement and sends a reply to accept.

Bob sees the acceptance and sends 100 SNC to Alice on the condition that 1 BTC be sent within the next hour and receives 5 confirmations.

Project launch?  Need some developers and/or investors to make this a reality.  Any takers?

The SNC network receives Bob's transaction and adds it to the public ledger.

Alice sees the transaction in the SNC blockchain, then after a sufficient number of confirmations, she sends 1 BTC to Bob.

The SNC network sees the BTC transaction, which triggers the first transaction to start confirming.  Once the confirmation count is reached, the SNC network completes the transaction, sending 100 SNC to Alice.

If the hour is up, and 1 BTC was not sent, the transaction will be completed/reversed, giving the 100 SNC back to Bob.

The proof of concept will only allow for 1-way transactions that must originate by sending SNC.  To exchange BTC for LTC, you must first buy SNC with one, then sell SNC for the other.

Incentive for mining this new currency would include block rewards with a built-in halving rate, and market driven transaction fees as a percentage and/or a fixed minimum.  IOW, no major change from how Bitcoin works.

Don't forget Nxt The first sole PoS crypto currency and decentralized application network not based on bitcoin or any other altcoin coin. 

Nxt will be implementing decentralized services such as Asset, Currency exchange, Colored coins, Marketplace, Auctions,
Transaction mixing, Encrypted Messaging, Chat DNS, Alias name scheme and more.

I keep hearing hints that "something is in the works" too.. but nobody will say what or when. All these rumours... I'm still waiting for somebody to release a beta already!

If you have any info.. an ETA would be very welcome.

I am sure there can be better solutions than ssl logs..something interesting is already in the works I heard

I'm building a decentralized, anonymous, and p2p exchange called Elqnt.

Take a look at the Whitepaper -- feedback much appreciated!  

I'm currently in talks to secure some funding to continue development into next year -- I'd already built the framework for the exchange nodes, but we're working on refactoring them to decentralize the signaling process as well, which is taking some time.  

If anyone is interested on jumping in on development, let me know, we're looking for node and chrome devs primarily -- also android or other mobile platforms.  

Our platform won't 100% comply w/ the stipulations lined out by OP, but gets damn close to a fluid process for fiat <-> BTC exchange in a decentralized manner.

Stephen

Hmmm... I would definitely not be the first to give access to my online banking, even with all the assurances, single session, etc. I would rather trust an escrow agent, at least in that case my risk would be limited...

The best idea so far is:

SSL logs as proof of money transfer for p2p exchanges

To be fair to them, they have been very valuable in advancing liquidity and driving the growth of the network in the past years... but yes, they are way too vulnerable, we need the next stage in evolution soon...

I think this should be included in the official development roadmap as it is basically vital for anyone holding the currency, as you can see centralized exchanges have only brought problems so far (Mt Gox, BTC China) and they could continue to do so in the future.
 

I think this problem is going to be solved, but not in a way that meets the conditions of the OP.

Added.

Happy to pledge 1 BTC

December 18, 2013

FT:  China bans new Bitcoin deposits

http://www.ft.com/intl/cms/s/0/6707013a-67af-11e3-8ada-00144feabdc0.html#axzz2npDcRmGA





This is why we need a P2P exchange.

"but you can't do that online. That's the big challenge."

Having seen how operations like western union work, there are a few possible ways to move fiat money around, but it has to be relatively small amounts, less than ~$1,000.

is there a limit on transaction values for a p2p exchange?

For cryptographic purposes, what assumptions may we make about fiat currency serial numbers? Nonrepetitive? Unpredictable for a given denomination (above guessing chance)?

The 'trustless' requirement for a cryptocoin<->cryptocoin exchange seems to be the show stopper
because it would probably require a hardfork in one of the coins.

What if the exchange does require trust but
-makes it easy to split orders into reasonably small parts
-is really just an user friendly piece of software so everyone can set up an exchange (making it p2p)
-records all trades in the blockchains so the exchanges could earn trust over time
This would still be an improvement over the current state of things.

This looks interesting... Paysty
https://bitcointalksearch.org/topic/alpha-testers-needed-fiat-btc-exchange-paysty-326233

It attempts to resolve disputes over bitcoin seller denying having received funds and refusing to release escrow...  but it is no way close enough to what we are aiming for.. it still requires a trusted third party for the escrow, it still relies on bank wires and other slow payment methods.. transactions are still not automatic..

And even if there were a Bitcoin-MTC and a Litecoin-MTC-clone you still couldn't exchange Btc vs Ltc
with them because they would be totally separated.

Well, if you only need to involve 3 parties (actually maybe 4, as the Bitcoin block chain is also a party in this, although very transparent concerning trust), you can already use the platform I hinted about... it's good enough for me. The parties involved for a BTC trade are: BTC blockchain --> web wallet funds storage (in BTC) --> distributed exchange --> web wallet funds storage (in USD) --> whatever payout method is supported by the web wallet funds provider (e.g. SEPA, PayPal...). These providers might need a license as e-money issuers in most western jurisdictions, but not as money exchangers as all they do is issue a balance on the distributed exchange, not do the actual trading.
Currently the fund storage on both ends as well as the trading engine is supplied by 1 single entity (e.g. MtGox) which is likely what you want to overcome. Most likely there will still be a need to give away funds to a trusted third party and to trust the same or another third party to release funds on a different channel unless there are ways other than cash to truly send fiat money that you own and that is not at a service provider (bank, paypal, dwolla...) electronically without ANY POSSIBLE way to withdraw the transaction after the fact.

The actual other person on the other side of the trade is abstracted away, since I don't know or care about her, only about the offer that is put on the distributed exchange.

2. No third-party trust required (no third party escrow, server, website, etc..)


I agree, it is pretty difficult. However necessary trust is greatly reduced if there is no single provider on which you must rely.. for example localbitcoins lets you list different payment methods (SEPA wires, OKPay, Paypal, Swift wires, Western Union, Neteller, Skrill, etc...) Similarly with Bitcoin-otc which lets you choose payment method.

I would be fine with trusting an external provider as long as the P2P exchange did not add another level of required trust... The greater the number of required intermediaries, the more points of attack/failure... The greater choice of payment mechanisms, the more flexibility and resilience of the system.



https://en.bitcoin.it/wiki/Payment_methods

So if an exchange involves only 3 parties (Buyer-money processor-Seller) it would be a great step up from the current situation (Buyer - money processor - exchange - seller). Of course you can do true P2P in person (buyer - seller) thanks to localbitcoins and other ad listing sites.. but you can't do that online. That's the big challenge.

I think it's difficult, but not impossible.

If you think that it is impossible, feel free to add your pledge to the bounty in the hope that you won't have to pay up