Is there a use case for XRP Ledger IOUs in an ILP World

[brianwalden]

If ILP becomes highly implemented will there be a reason to issue IOUs on the XRP Ledger? I can only think of niche use cases; the control and privacy of running your own ledger your own way and then implementing the ILP would seem to win out in all the common uses I can think of.

[JoelKatz]

I agree. But maybe people will think of things I haven't thought of. There are definitely people who think the idea that the ledger enforces the rules of transfer and exchange rather than relying on the issuer to run them makes interesting use cases possible.

[Guest dfault123]

Maybe some sort of worst case scenarios, where you're unable to reliably run your own servers in a country lead by unreliable/hostile government?

[xh3b4sd]

Good that we talk about this. I feel like I have to close the gap in my head. I understood one use case for XRP is to act like a path finder where it tries to bridge two currencies. Lets consider a payment example where two parties are ILP compatible and one wants to issue a payment to the other. How is defined that this goes through RCL via XRP? Is this a specification in the issuing/receiving payment? Does the example require both parties to hold XRP so it can be bridged or is it possible for the parties to just issue IOUs? I would now say when it is going through RCL they have to hold XRP. When they are just doing it via ILP without RCL being involved, they just pay IOUs. Is this somehow correct? 

[tulo]

1 hour ago, brianwalden said:

If ILP becomes highly implemented will there be a reason to issue IOUs on the XRP Ledger? I can only think of niche use cases; the control and privacy of running your own ledger your own way and then implementing the ILP would seem to win out in all the common uses I can think of.

I can see two advantages of IOUs in RCL vs ILP:

• Time of transaction. Since ILP transactions are dependent on the slowest ledger, having for example BTC involved means long time for payment, while if BTC IOU are involved it's only a matter of 3.5 seconds.
• Decentralized nature: a connector is more incentivized to rely on a decentralized ledger than a private ledger, because you need to trust the private ledger, while for the decentralized, as Joel said the rules of transfer are enforced by the ledger itself. (i.e. a bank can decide to not release the funds to the connector even if the ILP transaction succeeded. That's impossible in RCL).

PS: and if we could have IOUs ILP-enabled, probably that would be the missing piece for a real decentralized exchange.

Edited June 8, 2017 by tulo

[Dennis]

2 minutes ago, tulo said:

(i.e. a bank can decide to not release the funds to the connector even if the ILP transaction succeeded. That's impossible in RCL).

I agree with everything except this example.  ILP enforces the transaction requirements and would not release the escrow funds if one side failed to fulfill their obligation.  I get what you're saying, but I think this would not be considered as ILP transaction success.

[tulo]

Just now, Dennis said:

I agree with everything except this example.  ILP enforces the transaction requirements and would not release the escrow funds if one side failed to fulfill their obligation.  I get what you're saying, but I think this would not be considered as ILP transaction success.

No, because a private ledger can implement the protocol as they want. It's a private ledger, so the owner of the ledger can decide whatever he wants

A bank can always freeze your funds, size them, kick you out of the ledger access...but of course it is not in their interest but you need to trust them. 

[tulo]

I trow in a third point:

• Fees...nowadays if you want a "reasonably" time for a BTC transaction you need to pay between 0.2$ and 1$. Imagine micro payments with those fees...it simply cannot work. With IOUs the fee now is ridiculously low. 

[brianwalden]

@tulo I think you're technically right that there's nothing in ILP that prevents a ledger from approving a transaction but not transferring the funds between its account holders. But in any real world application, no one's going to have a connected account without a legal agreement that goes with it and the situation you described would be a clear violation of the agreement. The ledger might freeze your account, but they wouldn't mess with ILP transactions. ILP makes any ledger a gateway to the internet of value, it has the same types of risks as an XRP Ledger gateway.

[brianwalden]

@tulo I think you're right about transaction times and fees. But that can be solved by having a BTC "bank" where connectors can deposit their bitcoins and then use the bank's ledger to connect all their other accounts to. It's essentially a Ripple gateway without Ripple.

Also interesting thought, cryptocurrencies derive their value from the utility of their ledger. If a ledger is slower and more expensive than other ledgers in an interconnected world, what will happen to its coin's value?

Edited June 9, 2017 by brianwalden

[tulo]

26 minutes ago, brianwalden said:

@tulo I think you're right about transaction times and fees. But that can be solved by having a BTC "bank" where connectors can deposit their bitcoins and then use the bank's ledger to connect all their other accounts to. It's essentially a Ripple gateway without Ripple.

Also interesting thought, cryptocurrencies derive their value from the utility of their ledger. If a ledger is slower and more expensive than other ledgers in an interconnected world, what will happen to its coin's value?

Yes, but why trust a private bank and build something new if everything is ready in RCL? If IOUs could become decentralized, with an automatic deposit/withdrawal from all other crypto network then it would also become "counterparty risk free".

[brianwalden]

So here's my thoughts on when IOUs on the XRP Ledger might be used. First, I don't think not trusting the ledger to implement the rules of transfer correctly is a real reason. If you trust a ledger enough that you're willing to deposit your money and expect to be able to withdraw it, then you implicitly trust them to handle exchanges correctly.

I think that in the future there will be software that even individuals can run to create an ILP server to create tokens for the neighborhood yard sale or whatever. So setting up a bare bones ILP ledger will be as simple as issuing IOUs from a ripple wallet. One reason for going with the XRP ledger is if you either don't want to host a server yourself or don't want to pay to host it. I think this reasoning really only applies to individuals, even a small business wouldn't want to trade privacy for saving a few bucks per month.

Another reason I can think of is the decentralization. If it's critical that your ledger is always up, it may worth it to just use the XRP Ledger instead of setting up servers with different hosting companies all over the world.

The last type of thing I can think of is if something is a record for the public good. If you merely wanted all the account holders to see everyone's balances, you would just make them visible. I can't think of an example, but I'm sure there are some types of transactions that you want recorded for posterity's sake.

[tulo]

2 minutes ago, brianwalden said:

So here's my thoughts on when IOUs on the XRP Ledger might be used. First, I don't think not trusting the ledger to implement the rules of transfer correctly is a real reason. If you trust a ledger enough that you're willing to deposit your money and expect to be able to withdraw it, then you implicitly trust them to handle exchanges correctly.

I deposit money in private ledgers because it's the only available way right now. I use and deposit money into Paypal, but do I trust them? NO! I have money in my bank, but do I completely trust them? No! I have money on Poloniex, but do I trust them? Absolutely no!

[brianwalden]

5 minutes ago, tulo said:

I deposit money in private ledgers because it's the only available way right now. I use and deposit money into Paypal, but do I trust them? NO! I have money in my bank, but do I completely trust them? No! I have money on Poloniex, but do I trust them? Absolutely no!

You're mixing up the everyday meaning of trust with the DLT jargon meaning of trust. Regardless of how you feel about your bank, PayPal, and Poloniex, you trust them enough to give them your money - and that's the definition of DLT trust.

20 minutes ago, tulo said:

Yes, but why trust a private bank and build something new if everything is ready in RCL? If IOUs could become decentralized, with an automatic deposit/withdrawal from all other crypto network then it would also become "counterparty risk free".

If you trust the automatic program to handle deposits and withdrawals, you would trust the same automatic program to implement ILP. In fact, this method has less counterparty risk. You're only trusting the program and don't need any trust in the XRP ledger.

[jmar42]

@brianwalden The biggest benefit IMO is the use of the XRP native for cross currency for the IOU. The IOU could be based in only 1 currency such as USD.  Person A in Europe may want to buy the IOU but has Euros and the IOU (GatewayUSD) only accepts USD.  Person A can open an account w/ the Gateway and fund their account in XRP then convert it to GateWayUSD to purchase the IOU.  This is very similar to the ICO's popping up right now.  A person buys the ICO w/ ETH and the ICO is based on USD or whatever currency they want.  The startup then receives the ETH and converts it to USD or whatever other currency to operate their business or go party.

If there wasn't a need for the XRP then the ledger would look like a HyperLedger w/ no native. To me, the biggest benefit of the Ripple is the use of the XRP w/ various IOU's. But, really depends on what you're doing from a regulatory perspective and speed need of the ledger such as HFT.