Ripple or Bitcoin

[timeofmind]

If MtGox, BTC-e, CaVirtEx, etc. all started opening up all their data with one another, and allowing buy orders on one exchange to be fulfilled across this whole network of exchanges, would that not serve the function that ripple is now trying to fulfill? In that scenario, anyone could go to any of these exchanges to change their money to any currency supported by any of these exchanges... would we not then gain the same functionality as Ripple, but without IOUs or Ripples?

There would still be "IOUs" because gateways have to trust each other not to abuse the API. Typically this is done by each gateway holding a balance at each other gateway but you can see how that would quickly become prohibitive from both a cost and an administrative perspective.

Ya. I follow. So basically, Ripple provides a framework through which all these exchanges can achieve this.

[mmeijeri]

Sounds like a meaningless argument over semantics.  You can call the "gateways" whatever you like.

No, the word has a specific meaning within the Ripple system.

The function they serve is to exchange currency in someone's hand for IOUs, ripples, bitcoins, etc.

A gateway exchanges between external currencies and IOUs. An exchange exchanges between all sorts of currencies. Bitstamp is both a gateway and an exchange, but its own exchange only trades between USD and BTC. Once you've sent them money and transferred the corresponding IOUs to Ripple you can trade inside the Ripple system, and exchange between JPY and EUR if that's what you want. If you want to use Bitstamp's exchange functionality, you'd log into your Bitstamp account on the Bitstamp site. If you want to trade in Ripple, you'd use a Ripple client instead. It's crystal clear which of the two you are using. If the Bitstamp website goes down during a DDoS attack, then you can still trade your Bitstamp-issued IOUs on the Ripple network without any ill effects. Only if you want to buy more Bitstamp IOUs or if you want to redeem them would you need Bitstamp itself.

[misterbigg]

Ya. I follow. So basically, Ripple provides a framework through which all these exchanges can achieve this.

Right. Ripple provides a decentralized, peer to peer mechanism for transferring balances in a cryptographically secure fashion. This doesn't make gateway IOUs "safe" - you are still exposed to loss if they default on the redemption. It's a hybrid of centralization and decentralization:

Redemption of gateway balances: centralized, possibility of default
Transfer of gateway balances between third parties: decentralized, cryptographically secure

Then there's XRP, which is fully decentralized, cryptographically secure, and deflationary (only 100 billion can exist). With usual assumptions about OpenCoin going open source and all the XRP getting distributed soon.

[JoelKatz]

Of course a gateway is the same as an exchange... What on earth would you use a gateway for other than to exchange your money for another currency? OK, sure, you just want to send dollars to someone, so you go to "gateway"... the gateway exchanges the dollars to some electronic currency... the recipient collects in dollars at some gateway close to them.... the money is being "exchanged". Every gateway is definitely an "exchange".

I suppose you can view it as an exchange. One thing is traded for something else, certainly. There are two senses in which it's not like an exchange, at least not in the sense that Bitcoin exchanges are exchanges.

First, the two things traded are denominated in the same currency. A dollar from a gateway in Ripple is worth less than a physical dollar because there's counterparty risk. But a dollar from a gateway in Ripple is worth more than a physical dollar because it provides you access to the liquidity in the ripple network and can be transferred rapidly. For healthy gateways, there's an implicit agreement to treat these two factors as cancelling out for mutual convenience. The expectation is that this will remain true, so there's negligible exchange rate risk.

Second, the gateway acts much more like a bank than an exchange. The gateway is just holding fiat currency until someone claims it, just like banks normally do. The gateway doesn't have to match parties to counterparties or hold anything other than fiat.

I think these two factors make what a gateway does so different from what we think of as an "exchange" that it's not helpful to describe it in those terms.

[timeofmind]

Sounds like a meaningless argument over semantics.  You can call the "gateways" whatever you like.

No, the word has a specific meaning within the Ripple system.

LOL. Yep. Just like every word in the english dictionary? Sure, it has a specific meaning, but that does not clarify the symbolic associations that can be applied to that meaning.

[SamS]

Of course a gateway is the same as an exchange... What on earth would you use a gateway for other than to exchange your money for another currency? OK, sure, you just want to send dollars to someone, so you go to "gateway"... the gateway exchanges the dollars to some electronic currency... the recipient collects in dollars at some gateway close to them.... the money is being "exchanged". Every gateway is definitely an "exchange".

I suppose you can view it as an exchange. One thing is traded for something else, certainly. There are two senses in which it's not like an exchange, at least not in the sense that Bitcoin exchanges are exchanges.

First, the two things traded are denominated in the same currency. A dollar from a gateway in Ripple is worth less than a physical dollar because there's counterparty risk. But a dollar from a gateway in Ripple is worth more than a physical dollar because it provides you access to the liquidity in the ripple network and can be transferred rapidly. For healthy gateways, there's an implicit agreement to treat these two factors as cancelling out for mutual convenience. The expectation is that this will remain true, so there's negligible exchange rate risk.

Second, the gateway acts much more like a bank than an exchange. The gateway is just holding fiat currency until someone claims it, just like banks normally do. The gateway doesn't have to match parties to counterparties or hold anything other than fiat.

I think these two factors make what a gateway does so different from what we think of as an "exchange" that it's not helpful to describe it in those terms.

So a gateway is a bank and a exchange is an matching engine/broker for FX transactions -- combine the two or not. Risk still appears to be risk in this system. I don't see the big deal so far -- which I'm willing to admit may be due to what's been called "a lack of vision on my part."  Simply put, what's the value add here over the existing processing system(s). Is it really just the single communications protocol? Is it wrong to say that Ripple is an initiative to "disrupt" an existing cartel via use of a single communications protocol?

[timeofmind]

Of course a gateway is the same as an exchange... What on earth would you use a gateway for other than to exchange your money for another currency? OK, sure, you just want to send dollars to someone, so you go to "gateway"... the gateway exchanges the dollars to some electronic currency... the recipient collects in dollars at some gateway close to them.... the money is being "exchanged". Every gateway is definitely an "exchange".

I suppose you can view it as an exchange. One thing is traded for something else, certainly. There are two senses in which it's not like an exchange, at least not in the sense that Bitcoin exchanges are exchanges.

First, the two things traded are denominated in the same currency. A dollar from a gateway in Ripple is worth less than a physical dollar because there's counterparty risk. But a dollar from a gateway in Ripple is worth more than a physical dollar because it provides you access to the liquidity in the ripple network and can be transferred rapidly. For healthy gateways, there's an implicit agreement to treat these two factors as cancelling out for mutual convenience. The expectation is that this will remain true, so there's negligible exchange rate risk.

Second, the gateway acts much more like a bank than an exchange. The gateway is just holding fiat currency until someone claims it, just like banks normally do. The gateway doesn't have to match parties to counterparties or hold anything other than fiat.

I think these two factors make what a gateway does so different from what we think of as an "exchange" that it's not helpful to describe it in those terms.

Basically, people have to honor these issued IOU's to make anything work. If "gateway" X is shutdown, all his IOUs become worthless. So if my exchange..er...gateway, starts honoring the IOUs of another "gateway", I'm risking the day that gateway goes under.

What I'm really interested in is understanding the individual motivation for both exchanges and gateways to want to use this, rather than just operating outside the network. I definitely understand the value of having this network for its price-setting capabilities for good-of-all. I'm just not so sure this will fly as a way to achieve it. I'm not sure this network would be a reliable enough way to achieve it.

[JoelKatz]

Simply put, what's the value add here over the existing processing system(s).

The existing systems are absurdly inefficient. Each payment system is tied to its own proprietary liquidity providers. Connections between systems are ad hoc, expensive, and slow.

Is it really just the single communications protocol? Is it wrong to say that Ripple is an initiative to "disrupt" an existing cartel via use of a single communications protocol?

I think that's one way of looking at it. Think about things like email and text messages. What makes them so ubiquitous and useful is the fact that they provide a unified namespace and work across providers seamlessly and without high costs, long delays, randomly different policies, or the like. Payment systems today are where email was before it was federated by SMTP.

[timeofmind]

Concern #1. Are individuals likely to honor each other's IOUs?

Concern #2. Are gateways likely to honor each other's IOUs?

Concern #3. Is there enough incentive for users to use this system over other services out there that already allow people to exchange fiat for crypto and vise-versa?

Concern #4. Is there enough incentive for companies to want to use this system, rather than just doing this on their own? If you are a company wanting to start a gateway, are your IOUs going to be valuable until you get other gateways honoring them? Are people going to be able to take your IOUs and use them to get other currencies?

[timeofmind]

If bitcoin becomes ubiquitous enough before Ripple ever catches on, then what would be the incentive to use Ripple to exchange dollars for bitcoins, when you can just do this directly with your brother or your neighbor without ever having to deal with credit relationships?

[timeofmind]

Another question that interests me... Lets say that either Ripple or Bitcoin become the international currency of the world. Which one would be more reliable? Which one would be more impervious to attack? What would the topology of the Ripple network ultimately look like, compared to Bitcoin in such a world? Which one would be more resilient?

[timeofmind]

Another question that interests me... Lets say that either Ripple or Bitcoin become the international currency of the world. Which one would be more reliable? Which one would be more impervious to attack? What would the topology of the Ripple network ultimately look like, compared to Bitcoin in such a world? Which one would be more resilient?

Of course, in such a scenario, you might find the gateway/exchange features of Ripple rather redundant.

[SamS]

Simply put, what's the value add here over the existing processing system(s).

The existing systems are absurdly inefficient. Each payment system is tied to its own proprietary liquidity providers. Connections between systems are ad hoc, expensive, and slow.

Is it really just the single communications protocol? Is it wrong to say that Ripple is an initiative to "disrupt" an existing cartel via use of a single communications protocol?

I think that's one way of looking at it. Think about things like email and text messages. What makes them so ubiquitous and useful is the fact that they provide a unified namespace and work across providers seamlessly and without high costs, long delays, randomly different policies, or the like. Payment systems today are where email was before it was federated by SMTP.

Thank you. The basic thesis and business proposition seems clearer to me now. I don't have huge problems with the payment systems as they exist now -- other than the fee structure and exchange rates of course. They seem to function "reasonably" efficiently -- certainly not instantaneously but within tolerable limits. Seems like a different conversation, and maybe a different product, that deals with alternative, secure, non-inflationary stores of value, which may or may not be anonymous -- which is where I have seen a number of criticisms of Ripple focused on various threads. It might be nice to be all things to all people, but establishing a single communications protocol is a large enough goal in its own right to be worthy of respect.

[JoelKatz]

Concern #1. Are individuals likely to honor each other's IOUs?

Not any time soon.

Concern #2. Are gateways likely to honor each other's IOUs?

I think only in very specific use cases, such as a gateway that has a huge cash in / cash out imbalance. Imagine a cash over the counter gateway or a merchant only gateway.

Concern #3. Is there enough incentive for users to use this system over other services out there that already allow people to exchange fiat for crypto and vise-versa?

If Ripple catches on as a payment network, then the liquidity will make it a good platform for exchanging fiat for crypto. The big problem with exchanging fiat for crypto is the mismatch between hard and soft money. If Ripple makes hard fiat doable, I think that's enough to make it a competitive platform for exchanging fiat for crypto.

Concern #4. Is there enough incentive for companies to want to use this system, rather than just doing this on their own? If you are a company wanting to start a gateway, are your IOUs going to be valuable until you get other gateways honoring them? Are people going to be able to take your IOUs and use them to get other currencies?

You can provide liquidity between gateways yourself if market makers don't do it for you.

[timeofmind]

This is a quote from the "The Holy Grail! I wish I could kiss the author of Bitmessage on his face." thread (https://bitcointalk.org/index.php?topic=212490.220).

Yes, I've thought about this stuff before. Reputation systems are good for lots of little transactions, but there is a subtle but important problem with this approach.

First, decentralization always breaks if the cornerstone of the system is trust in individuals/nodes/people. That's exactly how the banking system came about. People with more and more money, and more and more power, get more of the money and power because they're the only ones you trust to keep your money. People's trust concentrates in those "nodes". This is what we have to avoid in our P2P solution.

We need to make it so the trust is in the protocol itself. This requires public verifiability of transactions.

It's good that the contract is customisable, but if that ends up meaning that we only transact with trustworthy nodes, we get centralization.

I'm curious what people think of this viewpoint in the context of Ripple...

[timeofmind]

You can provide liquidity between gateways yourself if market makers don't do it for you.

Wouldn't this require people to honor each other's IOUs? This would require a credit-link between individuals would it not?

[JoelKatz]

Thank you. The basic thesis and business proposition seems clearer to me now. I don't have huge problems with the payment systems as they exist now -- other than the fee structure and exchange rates of course.

2% for credit cards. 7% for remittances.

They seem to function "reasonably" efficiently -- certainly not instantaneously but within tolerable limits.

I think people have grown to accept as "just the cost of doing business" what they would consider intolerable if they were forced to look at it objectively.

I had a conversation about this with a man who owns a chain of grocery stores. And I asked him to imagine a hypothetical where 1 in 50 of his customers steals their groceries -- just walks out the door with a cart full of groceries without paying. Of course, he said his business couldn't survive with such a high level of theft, his margins were too tight and his costs too high, and that he would make fixing that his number one priority. Meanwhile, he takes credit cards which costs him just over 2% -- slightly more than 1 in 50.

It might be nice to be all things to all people, but establishing a single communications protocol is a large enough goal in its own right to be worthy of respect.

Thank you.

[misterbigg]

The reason TradeFortess is so much against ripple is because right now he is the market leader in BTC lending, he is afraid

LOL...another misunderstanding of Ripple. Even though Ripple has the concept of "IOUs", the debts they represent are NOT the same as loans! They are substitutes for money already on deposit at a gateway.

[JoelKatz]

You can provide liquidity between gateways yourself if market makers don't do it for you.

Wouldn't this require people to honor each other's IOUs? This would require a credit-link between individuals would it not?

It wouldn't require individuals to do anything. The gateway operator would either hold or accept balances from one or more other gateways, depending on which half of the imbalance he had to fix.

In my example of a gateway that's primarily aimed at merchants, the gateway operator could simply set up an account that offered to accept, say, Bitstamp USD balances in exchange for balances from his gateway. This would allow his customers' customers to buy from his customers using Bitstamp balances. Bitstamp would then become part of his cash in pathway -- he'd redeem those balances with Bitstamp directly.

If you imagine a "cash over the counter" gateway, the gateway operator could use Bitstamp as his cash out pathway. He would hold a balance at Bitstamp and offer to exchange his own balances for Bitstamp balances. This would ensure his customers could easily pay people who want to accept Bitstamp balances. He would take some of the money he took over the counter, send it to Bitstamp, and thereby acquire the balances he would trade for his own to keep his balances liquid.

Eventually, market makers would probably do this for him. But he may find his business model unusable in the absence of sufficient liquidity and he may at least need to do it himself to bootstrap.

[timeofmind]

You can provide liquidity between gateways yourself if market makers don't do it for you.

Wouldn't this require people to honor each other's IOUs? This would require a credit-link between individuals would it not?

It wouldn't require individuals to do anything. The gateway operator would either hold or accept balances from one or more other gateways, depending on which half of the imbalance he had to fix.

In my example of a gateway that's primarily aimed at merchants, the gateway operator could simply set up an account that offered to accept, say, Bitstamp USD balances in exchange for balances from his gateway. This would allow his customers' customers to buy from his customers using Bitstamp balances. Bitstamp would then become part of his cash in pathway -- he'd redeem those balances with Bitstamp directly.

If you imagine a "cash over the counter" gateway, the gateway operator could use Bitstamp as his cash out pathway. He would hold a balance at Bitstamp and offer to exchange his own balances for Bitstamp balances. This would ensure his customers could easily pay people who want to accept Bitstamp balances. He would take some of the money he took over the counter, send it to Bitstamp, and thereby acquire the balances he would trade for his own to keep his balances liquid.

Eventually, market makers would probably do this for him. But he may find his business model unusable in the absence of sufficient liquidity and he may at least need to do it himself to bootstrap.

Yes. So it requires a gateway to honor the IOUs of another gateway, thank you. Got it.