Bitcoin Forum

Alternate cryptocurrencies => Altcoin Discussion => Topic started by: shaitand on March 18, 2013, 08:23:15 PM



Title: Bitcoin Superset?
Post by: shaitand on March 18, 2013, 08:23:15 PM

I was sitting and looking at Ripple today and dislike it for a number of reasons. Thinking about how I'd solve that problem I had some thoughts and wanted to solicit feedback and inquire if anybody is looking at doing something like I came up with (am coming up with).

My thought is essentially another decentralized network but a P2P trade network. This would have a simple API that would among other possibilities allow merchants to accept a Bitcoin payment from a customer and immediately generate a sell order for a desired alternative currency be it USD, EUR, Gold, whatever.

The network would need a few critical pieces:

* Alternative currency options to choose from.
* Escrow and Escrow time limits related to the above
* Buyers and Sellers (obviously)
* Shared Order and Escrow Ledger
* ???
* Profit

The alternative currency options wouldn't be strictly currencies. They would represent accepted payment methods, they would have properties like transaction size caps, required escrow terms, and necessary receiver data points like addresses. These could be created and voted on by user adoption with the least popular x% being cut to keep the volume reasonable. For example, Paypal might be accepted. Paypal could actually exist as multiple options. Some might be willing to accept Paypal with a short escrow term of 1 week but with a low transaction limit like $20. Others might want no limit (or a high one) but require a 90 day escrow. Cash might be another option with one for each fiat currency and varied escrow terms. A little more thought might allow for a scheme that would allow this to take on an even greater role where these aren't just payment options but are in fact viewed as trade goods that might happen to sometimes be a Payment option like paypal. This view would mean someone could sell toasters via this system or a commodity like gold.

Buyers and Sellers is obvious. Every transaction would involve Bitcoin from at least one party and another party wanting to trade something for it. I'll refer to the non-bitcoin side as the buyer which is probably right for Paypal but wrong for toasters) and the bitcoin side as the seller. When a seller puts in an order his accepted currency options are attached to it and the coin associated is put in escrow for up to the escrow term on the relevant currency option. When matched with a buyer the buyer submits payment and the seller accepts on receipt or sets to automatically accept at the end of the escrow term.

Escrow. This is the tricky part. I'm thinking this might be handled with 2 of 3 multiple signing addresses. If everything goes correctly the buyer and seller sign and the escrow funds are released. The seller will always have submitted funds to escrow so if there is a problem it will be the seller claiming he didn't get paid or the escrow expiring. Either results in the network generating transactions to pay the escrow funds and the seller client and network both signing transactions to pay the escrow funds to unrelated third parties. These transactions should be signed at the earliest point possible and only submitted to the Bitcoin network if things go south. The idea here is that both parties share the risk. If the seller falsely claims he didn't his compensation he doesn't get his money back so at best he breaks even. If the buyer sends nothing or an empty envelope, etc, the buyer gets nothing in reward. The escrow funds could be utilized to reward the escrow processors who are sharing the burden of determining if escrow records are still needed, are expired, processing orders, etc. Order cancellation would be best effort with the result determined by network consensus. There may be options beyond that for the two parties if they come to agreement or via their payment option but that is outside the scope of the exchange.

A Shared Order and Escrow Ledger is needed to keep everything coordinated between clients. This will never be an ultra low latency trading platform with 1 min chart followers trying to follow the up and down ticks and using chart analysis. Thank god for that. But there will a potential advantage to positioning your connection to this network in a fast well connected location like a Datacenter with multiple peerings. This advantage isn't cheating in the way an ultra low latency trading platform is because everyone else would benefit from these well connected and fast nodes being placed in these positions, validating and propagating transactions to the shared ledger.

Most of the buyer/seller interaction could be automated in the client. Especially for a retail type situation where automatic conversion of bitcoin transactions into something like paypal transactions might be desirable along with automatic confirmation of payment receipt.



This all serves to give distinct advantages. A P2P instead of centralized exchange that is not subject to and needs no additional identity verification or regulation. The ability to make a personal choice regarding how anonymous you want to be versus how convenient you want currency conversion to be. It also gives you the ability to trade transaction speed for risk. Merchants have the ability to absorb a great deal of risk because they factor it into their costs and pass it back to consumers. Within reason they are willing to do this if you give them an easy road to follow. Merchants and consumers would benefit greatly by Bitcoin becoming the de facto transaction method.

Thoughts, questions, concerns? By all means, tell me why this is a shitbrained idea.




Title: Re: Bitcoin Superset?
Post by: Sunny King on March 19, 2013, 12:18:01 AM
It's on my todo list for a while.
https://bitcointalk.org/index.php?topic=113615

I would cross out the escrow part for a first release.

You missed the important first step:
Obtain the capital and take the risk to start such a project.

Ripple has been in development for I guess at least a year? No it hasn't turned profitable yet.

What is it that you disliked about ripple?


Title: Re: Bitcoin Superset?
Post by: shaitand on March 19, 2013, 01:52:34 AM
I was thinking of a decentralized peer to peer project not a for profit. The whole thing doesn't work without the escrow.

For starters as far as I can tell ripple has a fixed amount of it's currency and it is permanently consumed through usage. So the project has a fixed point of death. The parent company is retaining a large amount of the currency.

In all the talk of chains of trust on the website there isn't a single world that mentions something actually changing hands. If I trust pointA and you trust pointB and they both trust point C... how exactly does that magically get dollars you have in your bank account but point A doesn't to me? How are you even supposed to get it to pointB and they to point A or is point A expected to be a large central reserve of funds that is supposed to trust B even though they really don't trust you, B or me? Trusting someone doesn't equate to trusting who they trust.

1. Trust someone who trust someone you don't trust
2. ???
3. Profit

Even if it were clear. I don't trust people. Or rather what I trust about people is their unreliability. I trust probabilities. Bitcoin doesn't merely work because people believe in it, people believe in Bitcoin because pretty much every aspect has a solid numerical basis.  People statistically being more likely to hold up a deal they solicited when there is an upside they wanted and no upside to failing to follow through. That just makes far more sense to me than a magical cloud of trustberries.