Bitcoin-like implementation of Ripple

[Benson Samuel]

It has Vitalik.
There has been a lot of discussion and work.
I am currently awaiting a few comments from Jorge Timon who is an authority on Ripple & Bitcoin.
Will post after they come about.

As of now, there are 4+ ways to integrate Bitcoin and Ripple.

[1713511973]

1713511973

[1713511973]

1713511973

[freequant]

Ripple is an interesting concept and I discussed it with Ryan some years ago.

It tries to solve a different set of problems to BitCoin. The biggest issue I have with it, is that the benefits over the existing monetary system are not that big and it's a very complex way to handle debt. In particular the way I become transitively liable for the debts of my friends is something that's difficult to swallow for many people.

The automatic detection and destruction of circular debt was always more interesting to me, but Ripple in general is such a radical change to how money works that it's going to be much harder to get traction than BitCoin - which itself is seen as radical and risky by many folks, even though conceptually it's just another currency.

The Bitcoin economy has always had a major problem with trust.
The fact transactions aren't reversible, the heavy inclination for anonymity in the community, and the track record of scams, make lending a very risky activity.
In turn, this makes it difficult for businesses to raise funds, which prevents the growth of the economy.

A system like Ripple could be of great help to facilitate lending.
The chain of trust acts like a path to recall the funds, making the transactions somehow reversible.
Anonymity is preserved in a way that reminds of TOR. Each intermediate knows who is his immediate debtor and creditor, but doesn't know who are the end points, or if the immediate creditor or debtor are endpoints themselves or just forwarding the debt along.
Scams are not practical because defaulting on one's debt is going to wrong a close acquaintance and no one else.

Beside, there is another major quality of a system like Ripple: debt can be distributed. Instead of owing a huge amount to a single creditor, one can instead contract tens of small debts for the same amount on tens of paths that involve tens of direct friends. This reduces greatly the risk borne by each creditor, and gives flexibility to arrange for selective default in case of difficulty (family or closet friends will likely be more flexible and understanding). Debt can be distributed even further. You may accept to lend a large amount to a friend, and distribute the debt on 5 more friends.
That can even be combined with a system of interests that would be proportional to the number of levels involved. If you have many friends, and you borrow a small amount to all of them, you will get away with very low or no interets. If you borrow to friends of your friends, you will have to pay interests to your friend's friend as well etc. The huge advantage is that, while the interest rate grows linearly with the distance, the sum that can be raised grows exponentially.

[Sultan]

Has this gotten anywhere? I just found it now and thought that there truly is a need for an open-source, decentralised implementation of the Ripple protocol.

[CIYAM]

I would suggest looking into what is going on with Open Transactions and its recent integration testing with BitMessage.

If this is packaged well then I think it could well be the "next big thing" (and it doesn't require XRP).

[Sultan]

I would suggest looking into what is going on with Open Transactions and its recent integration testing with BitMessage.

If this is packaged well then I think it could well be the "next big thing" (and it doesn't require XRP).

What is Open Transactions, sorry?

[CIYAM]

What is Open Transactions, sorry?

To quote from the github project:

Open-Transactions democratizes financial and monetary actions. You can use it for issuing currencies/stock, paying dividends, creating asset accounts, sending/receiving digital cash, writing/depositing cheques, cashier's cheques, creating basket currencies, trading on markets, scripting custom agreements, recurring payments, escrow, etc. Open-Transactions uses strong crypto. The balances are unchangeable (even by a malicious server.) The receipt history is destructible. The transactions are unforgeable. The cash is untraceable. The cheques are non-repudiable. Etc.

https://github.com/FellowTraveler/Open-Transactions

[jtimon]

Oopps, I missed some action in this thread.

@blablahblah

LETS and other local currency systems work perfectly well, there's more than 4000 local currencies in operation. It's interesting to note that many complementary currencies advocates often say exactly the same you said about LETS, but about Bitcoin: "Bitcoin deflects attention away from social currencies and other real alternatives."
Goldbugs, greenbackers, "local currency hippies", gesellians, bitcoiners, trade clubs, etc are all on the same boat, we're all monetary reformists that recognize that that out monetary system is the root of most evils. Even if we have different opinions on what's exactly is wrong with the current system.
We should try to understand each other better and fight together against the banking cartel, usd, eur, etc instead of fighting each other.
We bitcoiners are just the last faction of a much bigger movement.
All local currencies, time banks and "barter" clubs combined have far more users than Bitcoin.
It would be huge for Bitcoin to integrate with those trade networks.

@CIYAM Open
OT is useless for many use cases because you can't even trade two assets in two different servers atomically.
Ripple.com has a lot to do with this proposal, replacing proof of work with their new consensus mechanism. I don't know if that will work, but it's worth trying.
It is currently free software although the network is not p2p yet.
I think it has several design flaws though, and it doesn't look like their developers want to fix them.

@Everyone interested
For me Freimarkets is the continuation of this proposal.
I also wanted to integrate this proposal with Ryan's Two phase commit ripple distributed protocol and "private chains" are the way we decided to do that.
Well, not exactly with Ryan's design, but we also use the two phase commit and support more scalable and private off-chain transactions.

Some advantages over Ripple(tm):

- Support for off-chain assets/transactions.
- Support for interest bearing assets.
- Open orders (and trust lines, which in this case are emulated using regular orders) are off-chain.
- More scalable KYC support (and more flexible for some local currencies related special needs).
- Compatible with Bitcoin's design, code and scripting language (it's just an extension).
- Unspent outputs instead of accounts (no sequencing, better privacy).
- Although miners are likely to prefer the hostcoin, fees can be paid in any asset type.
- Support for unique non-fungible assets (better for some smart property use cases).
- Auctions, options...

Currently I'm busy finishing Freicoin Foundation's website, maaku is busy with the "ultimate UTXO compression" implementation and we're both busy building a crypto-currency business together.
In addition we haven't received much funding for this yet and we would also appreciate more feedback from core bitcoin developers that have also thought about smart property and off-chain transactions.

Anyone interested in seeing this happen please review the design and/or fund the development.
We cannot promise you to get rich like mastercoin or bitshares, or any return in your donation at all, but we think this would open the door to many entrepreneurship opportunities.

[IncreaseMyT]

Can someone explain in laymen why I would need a ripple if I already have a Bitcoin?

thnx

[kwilliams]

Ripple advantages:
 * Instant Transactions
 * No hashing / mining
 * No block chain
 * Ability to have balances in multiple currencies

[grau]

Ripple advantages:
 * Instant Transactions
 * No hashing / mining
 * No block chain
 * Ability to have balances in multiple currencies

Bitcoin makes no compromises enabling trustless trade in a fungible unit of account.

Ripple lets you use any units but does not protect you of colluding servers.

Ripple might work just as secure as an SPV node of Bitcoin and it is likely more user friendly.

I guess Bitcoin remains the choice for ultimate settlement, while Ripple adds features we also need.

[jtimon]

Ripple advantages:
 * Instant Transactions
 * No hashing / mining
 * No block chain
 * Ability to have balances in multiple currencies

These advantages are specific to Ripple labs (previously opencoin) implementation concept, but I think he's asking more generally.
It doesn't have a block chain but it has a ledger chain, not that different really.
The transactions aren't instant, but their confirmation takes seconds instead of blocks.
The lack of proof of work as an advantage is arguable. Bitcoin is 100% trust-less because of that and Ripple has some trust involved in the construction of the UNL. We don't even know for sure if this will work when the network is p2p instead of under Ripple lab's control. Now that it is free software, people can create their own networks to experiment on forks, sibil attacks and other security aspects of this new proof of consensus.
We will see, but I'm glad they're experimenting with new global consensus mechanisms different from proof of work.

Can someone explain in laymen why I would need a ripple if I already have a Bitcoin?

In general, Ripple-like systems like freimarkets allow the trade and transfer of more things than just a currency. Typically the balances represent IOUs, but they can represent many other things like shares, bonds...
The balances can be denominated in any currency.
The more interesting thing of the Ripple concept is the ability to perform transitive atomic transactions. That is, you can exchange the IOUs transitively and pay the recipient in a different currency from the one you spent.
An example:

Merchant D only accept localD_USD (a local currency in his community) as payment. Customer A holds BTC and wants to pay to merchant D.
In a single transaction, he sells his BTC for issuerB_EUR, which then sells for issuerC_USD and then for localD_USD and then sends them to merchant D.

Maybe you don't want to pay anyone and just want to buy localD_USD for BTC at the lowest price you can find and this path results to be the cheapest.
You could make another example where the payer pays USD and the merchant receives BTC, of course.

This feature is ideal for a p2p exchange.
And Bitcoin could have p2p exchange and transitive transactions too (among many other things) if we modify the protocol as proposed in Freimarkets.

[jtimon]

Bitcoin makes no compromises enabling trustless trade in a fungible unit of account.
...
I guess Bitcoin remains the choice for ultimate settlement, while Ripple adds features we also need.

As said there's no technical reason impeding Bitcoin from having those multi-currency and transitive transactions features.
Providing more privacy than Ripple, since Bitcoin uses unspent transactions and Ripple uses everlasting accounts (you cannot destroy them right now).

I think you're right that proof of work remains the choice for ultimate settlement but Ripple also has limits to sclability and we ultimately need off-chain transactions in one way or another. Whether those off-chain transactions will happen on Freimarkets-like private chains or in Open Transactions centralized but trust-less servers we don't know yet.
I think that the ability to atomically trade in-chain and off-chain assets is critical, but I could be wrong.

[grau]

I think of Bitcoin as ultimate digital settlement, like gold over IP.

Bitcoin transactions are however rather expensive (including block reward) and slow to be final.

For higher abstractions or fast micropayments OT and Ripple seem better suited, but proof are outstanding. I actually prefer OT over Ripple for its capture of contracts and focus on auditability of the server actions.

[jtimon]

I think of Bitcoin as ultimate digital settlement, like gold over IP.

Bitcoin transactions are however rather expensive (including block reward) and slow to be final.

For higher abstractions or fast micropayments OT and Ripple seem better suited, but proof are outstanding. I actually prefer OT over Ripple for its capture of contracts and focus on auditability of the server actions.

Ripple transactions are (or will be) expensive too because all nodes in the network have to proccess them, they're still in-chain transactions (just another chain without proof of work).
Will Ripple scale to 1M tx/s ? I highly doubt it.
This thread was proposing to have multiple assets and transitive (ripple-like) transactions on Bitcoin's chain (or an altcoin), but that doesn't help with scalability, just adds a lot of new interesting use cases.
That's why integrating with something off-chain like Ryan's 2PC was critical and that's why we have "private chains" in Freimarkets to account for off-chain assets.
OT is more interesting than Ripple for scalability because its transactions are off-chain but its main limitation is the lack of transitive transactions between different OT servers.
I think that's a huge deal. Freimarkets has contracts (with scripts) and could be extended further but OT will never have ripple-like transitive transactions between different servers.