P2P Exchange for bitcoin

[gollum]

3 Exchanges:
a) One and only one exchange
b) Many independent virtual exchanges running in the system

I can think of one positive aspect of having one and only one exchange, namely the quality and consistency of the data feed (used to plot bars and perform technical analysis).

When there is one exchange, for a particular instrument e.g. CME for S&P 500 this exchange distributes its data to companies like Trading Technology and IQ. Every end user buying data from these distribution companies can be 99% sure that the data plotted on his monitor are the same as data plotted on monitors of other traders - this is good.

Now try the same in forex ... tens of brokers for EUR USD, each broker reports different prices and different volumes (and different spreads). Gosh - it is a nightmare.

Bitcoin is a currency, not a stock, and our goal is decentralization, not centralization. Therefore the decentralized OTC-markets of FX should be our role model.
In the FX-world there are many OTC-exchanges and yes you will not see exactly the same price of EurUsd over all exchanges all the time, but a lot of algos are doing arbitrage 24-7 so most of the price differences will be wiped out before you even notice it.

[mobile4ever]

Bitcoin is a currency, not a stock, and our goal is decentralization, not centralization. Therefore the decentralized OTC-markets of FX should be our role model.

I agree. Decentralization in a form that takes it out from under the control of a small group of people.

[TierNolan]

So, what about the following framework.

Transactions would be kept to a few limited types.  Only one type of trade would be allowed per chain (say BTC vs dollar).  Brokers would issue both fiat and BTC IOUs.

Broker introduction

This transaction includes broker id and broker public key.  It could also include an SSL cert or something to verify identity.

Brokers would need to reissue these every so often.

Broker trust update

This allows brokers to change how much they trust another broker.  It is the maximum debt.  It must be signed by the broker.

Broker balance update

This allows the brokers to change the balance.  It must be signed by the broker who clears debt (or has their debt increased).

Trader introduction

This allows a trader id to be created.  A balance is held for each trader.

IOU credit

This assigns an IOU to a trade.

IOU redeem

This cancels an outstanding IOU.  This must be signed by the owner when they collect fiat.

IOU convert

This moves an IOU from one broker to another.

If a broker says they trust another broker for $10k, and the current balance is -$2k, then up to $8k of IOUs can be converted before the link is saturated.

It must be signed by the owner of the IOU.

Offer

This says that the owner will sell fiat for that price in BTC or the market price if higher.

When this link is added if there are any matching bids, then the fiat is sold at the highest bid price.

Otherwise, the offer is added to the order book for the stated lifetime.

Bid

This says that the owner will buy fiat at that price in BTC or the market price if lower.

When this link is added if there are any matching offer, then the fiat is bought at the lowest offer price.

Otherwise, the offer is added to the order book for the stated lifetime.

--------------------------------------------------

Binding to the bitcoin network is an open question.

You could add a rule that a broker can designate a bitcoin address in his introduction.  If a transaction is more than 6 blocks deep and sent to that address, you can automatically create an IOU.

For redeems of bitcoins, the rule might be that the redeem link is valid if a tx with <some random number the owner picked> is added to the main chain and has been confirmed by 10 blocks.

Keeping the network fast is also an open question.

[gollum]

So, what about the following framework.

Transactions would be kept to a few limited types.  Only one type of trade would be allowed per chain (say BTC vs dollar). Brokers would issue both fiat and BTC IOUs.

Why just one asset pair / chain?
If we are going to use the colored coin model I dont think there is any limit to how many type of contracts that can be created in the same chain.

If we trade FX we want to have this theoretical FX-crosses in the same exchange:
[ EUR USD CNY RUB JPY SEK GBP BTC LTC XAU XAG ] x [ EUR USD CNY RUB JPY SEK GBP BTC LTC XAU XAG ]

[TierNolan]

Why just one asset pair / chain?

I was trying to keep that chain clean and hopefully faster.  There are both BTC and fiat IOUs on the chain.  As I said, you can link them back to the main chain if required.

If we are going to use the colored coin model I dont think there is any limit to how many type of contracts that can be created in the same chain.

That system wouldn't use coloured coins.

[master.yoda]

There is already a project that is moving in this direction:
https://ripple.com

Ripple is owned by a company. Im talking about a free model of ripple where everyone can "mine".

Ripple is merely a medium to get bitcoin from point a to point b instantaneously and with extremely low fees. No need to wait for confirmations any more and no need to mine.

[mobile4ever]

There is already a project that is moving in this direction:
https://ripple.com

Ripple is owned by a company. Im talking about a free model of ripple where everyone can "mine".

Ripple is merely a medium to get bitcoin from point a to point b instantaneously and with extremely low fees. No need to wait for confirmations any more and no need to mine.

Ripple is under the control of how many people?

[lunarboy]

A couple more threads that can be added to your list. The concept of a decentralised exchange for bond control has been discussed here.

https://bitcointalk.org/index.php?topic=123271

Within the NMC chain it would be possible to create a truly decentralized exchange:

https://bitcointalk.org/index.php?topic=152869.0

Atomic coin swapping and coloured coins
https://bitcointalk.org/index.php?topic=112007.0

The namecoin protocol seems highly under utilised and ripe for the pickings as its merged mining offers a huge head start with security

 

[mobile4ever]

Just look at Bitinstant. Butterflylabs. All these companies doing exactly your method and look at the results? I don't want anymore amateurware but that is just me.

Enough words about this. I look forward to your perfect P2P exchange that runs without flaws from day one, has all the features people could want from day one, and won't have any scaling issues. In the meantime I'm going to look if I can concretely support some project. Sayonera.

Here is one:

https://bitcoinstarter.com/projects/54

[gollum]

A couple more threads that can be added to your list. The concept of a decentralised exchange for bond control has been discussed here.

https://bitcointalk.org/index.php?topic=123271

Within the NMC chain it would be possible to create a truly decentralized exchange:

https://bitcointalk.org/index.php?topic=152869.0

Atomic coin swapping and coloured coins
https://bitcointalk.org/index.php?topic=112007.0

The namecoin protocol seems highly under utilised and ripe for the pickings as its merged mining offers a huge head start with security

 

Yes we should look at NameCoin as well, could be an alternative to Bitcoin coloring.
My opinion is we should keep the p2p exchange outside the bitcoin blockchain exchange.
Since NameCoin is not primarily used as a currency it feels better to use it as a database.
If Namecoin dont meet the requirements, we should start a new alt-coin that will not have any value but to store/exchange data securely.

[Luckybit]

Just look at Bitinstant. Butterflylabs. All these companies doing exactly your method and look at the results? I don't want anymore amateurware but that is just me.

Enough words about this. I look forward to your perfect P2P exchange that runs without flaws from day one, has all the features people could want from day one, and won't have any scaling issues. In the meantime I'm going to look if I can concretely support some project. Sayonera.

Here is one:

https://bitcoinstarter.com/projects/54

The idea is good, it seems like it could work however based on how it's set up don't expect to receive many donations.
They need to provide some incentives. No one is going to donate without any incentives beyond having their Bitcoin address listed. They can do better than that.

Nothing stops the people developing this from charging transaction fees and paying back the people who donate. We need to contact whoever is behind that project and tell them to set it up so that by making a pledge or donation not only will their Bitcoin addresses be listed in the credits, but actual Bitcoins will be sent to those addresses associated with donating in the form of transaction fees (if you are behind the protect then find a way to set it up like that and I might be able to find some people willing to donate). These fees should last for a limited amount of time so that the donating individuals can make a reasonable profit of perhaps double their money and then make it free for all. This is a better model than making it completely free and taking money from one group of people to provide to another. The right way is to provide money AND reward people who protect Bitcoin similar to how miners get rewarded for protecting Bitcoin.

I'll donate only when it's set up where I have incentive to donate. The best incentive would be simply to add my Bitcoin address to one of the receiving addresses for transaction fees over a certain period of time and to be fair the more I donate the greater the share I should receive, just to be fair. Otherwise if there are any Bitcoin millionaire types out there then someone like that should just pledge 100 Bitcoins or so to this and its done.

[Luckybit]

https://hacks.mozilla.org/2013/04/web-payments-with-payswarm-identity-part-1-of-3/

[mobile4ever]

Just look at Bitinstant. Butterflylabs. All these companies doing exactly your method and look at the results? I don't want anymore amateurware but that is just me.

Enough words about this. I look forward to your perfect P2P exchange that runs without flaws from day one, has all the features people could want from day one, and won't have any scaling issues. In the meantime I'm going to look if I can concretely support some project. Sayonera.

Here is one:

https://bitcoinstarter.com/projects/54

I'll donate only when it's set up where I have incentive to donate.

The incentive is there already. Becoming a part of the system early on means money for those who do. It is based on the system automatically signing up new members in a fair way to those who adopt it.

Early adopters are the winners, just as we are with bitcoin. This system is nothing but free software that sets up a system that makes its users money.

Decentralized to the max.

[btcmind]

Solving the Fiat <> BTC gateway problem would essentially circumvent many banking regulations and be illegal in most countries. If you pool money, you're bank. Essentially any good system would work within the banking network which is not based on pools or central identities. But who watches the identities? A BTC exchange currently serves as a bank, a identification system and a fiat-BTC conversion. To take one example. A bitcoin adress can be easily shared, but running a bank account in a foreign name without declaration would be illegal. It will be more than interesting to see how this plays out. What a strange coincidence that BTC was invented shortly after the crisis of 08.

[mobile4ever]

A BTC exchange currently serves as a bank, a identification system and a fiat-BTC conversion.

Was this a mention of Mt. Gox?

[btcmind]

B24 was closed down with the mention of money laundering and Terrorism. I'm not sure what happened to Bitfloor. This means if you operate an exchange you want to better have the acknowledgement of the government. But then BTC depends on these gateways. BTC could then easily be attacked by shutting down those single points of failures. I'm a bit surprised that people easily say BTC is legal. Most people (99.9999%) haven't understood what it is and certainly allmost all people who have something to say. It would interesting to track this on a global level but the b24 incident is telling. The question is: if you have fiat money in a bank or cash at hand how do you convert that into BTC independent of what anybody tells you what is right and wrong to do with your money. If you're rich you hire a bunch of lawyers who then use the superhighways to offshore centres. Apple Inc has all of it's money in an offshore haven called Ireland.

[mobile4ever]

This means if you operate an exchange you want to better have the acknowledgement of the government. But then BTC depends on these gateways. BTC could then easily be attacked by shutting down those single points of failures. I'm a bit surprised that people easily say BTC is legal.

Is software legal? What if software can change the way people think about money? Machines and software exist to make our lives better. There are no laws against that, that I know of.

Most people (99.9999%) haven't understood what it is and certainly allmost all people who have something to say. It would interesting to track this on a global level but the b24 incident is telling. The question is: if you have fiat money in a bank or cash at hand how do you convert that into BTC independent of what anybody tells you what is right and wrong to do with your money.

Bitcoin will be the money. I have seen the early adopters are sometimes millionaires and bitcoin is still in BETA. We are here in this thread to change the bottleneck that centralized markets have created for our money; bitcoin.

[Bitcoin Oz]

Localbitcoins.com is the closest you can get to a distributed exchange.

[gollum]

Localbitcoins.com is the closest you can get to a distributed exchange.

Yes but you might get robbed if you meet random people at the street to sell/buy bitcoin.

[BTCLuke]

Localbitcoins.com is the closest you can get to a distributed exchange.

What a lazy thing to say.

If an exchange is set up as a software like bitcoin-qt, and all functions are distributed between clients like the trade ledger, escrow, and processing, then we can have a distributed exchange quite easily... It just may only yeild regional results because of pingtime latency, but that's going to be a far, far larger area than localbitcoins can create.