Hacker News Reader

Comment by bacheaul | Hacker News Reader

What I've never understood about XRP and Ripple is that originally XRP was meant to only be used for paying for transactions and the ledger was meant to track IOUs between people denominated in some other "real" currency. They had videos floating around explaining the concept. In fact, they were handing out XRP for free to bootstrap the network of users [0] (that post says 1000 XRP, originally it was 30,000 XRP when the post was first created).

XRPs themselves were basically worthless, but it was required to interact with the ledger (you pay maybe 0.000001 XRP per transaction). The idea was that you pass a small amount on to friends and others interested in using it to create their wallet and start tracking IOUs, and you needed to have at least 20 XRP in your wallet.

Somewhere along the way all hell broke loose, Bitcoin got popular, crypto currencies in general got popular and XRP itself suddenly became worth something, the original concept got lost and everything seemed to go off the rails... That's when I lost interest.

[0] https://bitcointalk.org/index.php?topic=145506.0

View on HN

No, that was the idea of Ripple, invented by Ryan Fugger long before bitcoin even existed, and requires no new coin or token for its operation. OpenCoin, a company founded by Jed McCaleb, bought the trademarks other IP for Ripple and rebranded their bitcoin competitor altcoin as XRP. They never had any interest in the original concept, and quickly threw it away when it was no longer fashionable.

The original Ripple site was available for a while under https://classic.ripplepay.com/. I just noticed that Ryan is now operating it under the new domain https://rumplepay.com/.

I don’t remember if I had ever talked to Ryan, but I remember talking to people about this in forums and in person at Stanford and Cal in 2007. By 2005, the financial system was showing signs it was headed toward a point of no return because there was no revenue basis to justify the debt leverage, and the only way out was either extreme dollar scarcity, or previously unthinkable monetary base injections. Either way, there was some concern that we’d need a system of barter for those soon to be left behind by the banking system, and some hope that a massively decentralized barter network would be a more just and stable and productive economy regardless. The main theoretical discussion was whether a system could exist without a central currency. Most ripple believers thought that central currency was basically antithetical, although one based on unskilled labor hours might be acceptable as fallback. I believe this transformed into a resurgence of interest in HashCash, which had only been interesting for spam mitigation, but now looked like a possibility on Ripple. There were some weaknesses in the protocol, however, and as soon as the sci.crypt community got pulled into it, they dove into the double-spending problem. The old methods of Chaum’s blinded signatures didn’t work for decentralized exchange because you’d have to trace every coin back to its original minter. So the question was whether to use the network of confirmations to create an acceptable level of trust, or to maintain a central ledger, which looked a lot like a central bank. Satoshi’s concept was to use hashcash to both secure the ledger and issue currency. This was brilliant but antithetical to the no-currency concept, and re-ignited the debate, with Szabo and Dai reviving their own currency ideas and heading toward implementation with Finney. The ripple crowd were mostly interested in new ideas in economics, and as nobody could find a way of securing a decentralized exchange network without a currency, most of the interest faltered. Sure; you could decentralize the currency, but it doesn’t succeed in prioritizing reward to those performing useful work. At some point Mazieres entered with a more rigorous defense of trust generated by the network, but it still relied on a central token of value to process each exchange. Bitcoin emerged from the impending doom of financial collapse as a less hopeful but more realistic option, and interest in cryptocurrency in general fell off dramatically as the Bitcoin network went through its dead period of only one consistent miner, Satoshi himself. Ripple as XRP emerged much later in the post-Doge altcoin boom with huge investments, big claims about XRP in deals to replace Forex, hiring a lot of people, and were the only job option for the people that would be their critics. To the original ripple supporters, XRP looked no different than a central bank with, except that it was even more autocratic, unaccountable, and unequally-distributed. In no way did XRP fix the basic problem of allocating more rewards to useful work.

I could go on... but... just my recollection...

Thanks for that - very interesting. I think it is possible to create a pure credit/IOU monetary system and I’m working on it at the moment. Though reality is that I spend most of the time reading economics journals... The key is to track reputation and enable the minimum necessary amount of transparency. Eg instead of a single currency like BTC or ETH, everyone becomes an issuer and then the problem then becomes how to efficiently make a market in these currencies by swapping IOUs - essentially what banks do today but in a decentralised way. Furthermore, I think it can be done without a blockchain... Simple tech will do. For one who currently works in the blockchain space - I really like the idea of not using blockchain :)

If you are interested, it’s worth reading the articles “what is money” and “the credit theory of money” by Alfred Mitchell-Innes. He maintained that “credit and credit alone is money” and I mostly agree with him. Another good paper to read is “money is memory”. Oh and Knut Wisksell “interest and prices” is also very good. A lot of literature has been written on pure credit economies but it is difficult to find as it wasn’t ever a popular school of thought. I’m hoping for a revival though!

That’s precisely what RipplePay was. This was all sorted out a decade ago. The idea works, and is still relevant today.

I am too thinking about Self-Sovereign banking.

That’s good to hear! How do you envision it working?

> Most ripple believers thought that central currency was basically antithetical, although one based on unskilled labor hours might be acceptable as fallback.

This stuff always sounded like it came from people who've never hired. Unskilled labor is not all equal.

> Satoshi’s concept was to use hashcash to both secure the ledger and issue currency. This was brilliant but antithetical to the no-currency concept, and re-ignited the debate

Satoshi's most important advancement was in making a currency and using it as its own rate-limiter by paying, in it, to publish.

Bitcoin's weakness is that computation isn't priced in BTC, so scripts had to be nerfed. Vitalik applied Satoshi's rate-limiting solution to computation and Ethereum scripts can now be allowed to grow arbitrarily complex.

The unskilled labor concept was not often discussed in forums or in most of my interactions at Stanford, but when I went to Cal, I realized that the labor value theory foundations were a lot more important to the early supporters than I had thought. I learned about the origins of the movement from a UW student visiting Cal. Somebody may have given a talk there and it caught on.

Smart contracts seemed to me to be another sticking point that nearly killed the whole idea. Most people were looking for a better system of simple exchange, and smart contracts made that difficult. Now transactions became difficult to understand and fraught with error. Running a node meant trusting a fairly large code base you downloaded from an insecure Internet forum, and then let strangers run code on that. Before Script came into it, anybody could look at a shell script implementation and understand what was going on with all trusted components. For a long time it looked like Bitcoin would never release, and vulnerabilities in the early versions made everybody realize that they weren’t trusting a mathematical concept; they were trusting software written by some guy on the internet. Then there was the debate inside the smart contracts community about the halting problem and how it could be used to attack the network. This is probably practically solved with transaction costs alone, but I think it was a practical software engineering compromise do scripting without looping and backward jump instructions because they made the software much more complex and vulnerable to attack. Same with the block size, maybe even block time. This was all an experiment. By the time Vitalik started pushing for improvements, the Satoshi compromises had become canon. I spoke to him and he seemed to be in agreement with upgrading the protocol through miner consensus, but, by that point, an industry of rather unsophisticated mining operations had emerged, and they were not interested in changing anything. Briefly I lived with one of the biggest miners in the world, and he would ask questions and I would explain to him how transactions work etc. So Vitalik gave up on on-chain scripting and proposed an off-chain scripting scripting layer that would use a separate gas currency just for smart contracts with settlement on the Bitcoin blockchain. This was very important to distinguish itself from the rest of the already-defunct altcoins. But practically, I don’t know how it was supposed to happen. We have this now, but it’s been a long road. Is there any reason that Bitcoin doesn’t have Turing completeness, faster blocks, oracles, etc? Not really; but a fork would have been risk without reward, whereas a new token offers the ability to premine and speculate. One of the biggest learnings was how much speculation would dominate over anything non-financial, and how hard it would be to modify the Bitcoin experiment. Still, the use cases for smart contracts are very few, and Turing completeness itself is basically unnecessary. But to more directly address the point, Bitcoin has transaction fees already, and it would be simple to assume that a similar consensus on gas price would emerge and doesn’t need to be part of the protocol. To go further, the gas price itself doesn’t solve the problem of computational complexity because only the miner that wins the block actually gets the reward, while all others downstream are expected to verify it, but they have no incentive to do so, and frankly, today, most don’t. As an example, almost nobody boots up an eth node with verification from origin, and they would be greatly in the minority if they corrected a bug. Eth nodes boot to a trusted state handed to them. Most of the early eth discussions I was involved with implied doing this in snarks or some other provable computation in the initial discussions, but that never made it into eth. So no; that problem was never solved, although we are getting closer.

> One of the biggest learnings was how much speculation would dominate over anything non-financial

Yup. And still does. There was a huge sea change circa 2013(?) when everything became about currency and speculation. Every idea had to pivot to being its own coin, or including some useless utility token (a bit later) because that was 99%+ of the value to investors.

> Is there any reason that Bitcoin doesn’t have Turing completeness, faster blocks, oracles, etc? Not really; but ...

It feels like the first implementation couldn't have gotten it right if it aimed for too much more than Satoshi did. But technically, yes, BTC is a plausible base for this as much as anything else.

> debate inside the smart contracts community about the halting problem and how it could be used to attack the network. This is probably practically solved with transaction costs alone, but I think it was a practical software engineering compromise do scripting without looping and backward jump instructions because they made the software much more complex and vulnerable to attack.

Unbounded loops are an issue, but functional constructs (foreach) have 99% of the power and little of the risk. Ideally we'd have Satoshi take three months now and make v2, which would be BTC in spirit but with the lessons of the last ten years.

> To go further, the gas price itself doesn’t solve the problem of computational complexity because only the miner that wins the block actually gets the reward, while all others downstream are expected to verify it, but they have no incentive to do so, and frankly, today, most don’t.

This seems like where we need a mechanism that allows a script to be audit-executed, and if its outputs don't match expectations a deposit can be claimed by the auditor and the script will be marked as unreliable. If the precondition of the code is that it returns the correct public key, make sure that key works. If it doesn't you can claim the QA/Security prize - which can be set at whatever amount the script author wants depending on their belief in their code.

> Most of the early eth discussions I was involved with implied doing this in snarks or some other provable computation in the initial discussions, but that never made it into eth. So no; that problem was never solved, although we are getting closer.

Is there a specific blocker for this or just that the complexity is 1000x what is currently supported?

At that time the ZeroCoin protocol was known, but an example was only shown for simple transactions, and anything else would have been a lot more complex, and nobody who claimed to understand the snark construction at that level was willing to do it. Gavin was just cranking it out as an op code state machine because he knew he could do that himself. Vitalik would engage with all the big crazy ideas in the room, bringing as much as possible into the idea that was ethereum, and Joe was kind of like the interface between selling the dream and delivery on what they knew they could do. I realized much later that ethereum needed all three personalities in separation to succeed with the crypto community, because they were three different types of promises, not all fully congruent, but the big unifying vision was in the eye of the beholder, a belief that it would all bend eventually toward their personal vision: a utopian egalitarianism for some, and vast plutocratic wealth for others, a gpu mining business for some and environmentally friendly staking for others, etc etc.

Thanks, that recollection is awesome! I'd love to hear the rest!

I remember their original pitch sounding like it made a lot of sense. Fast, cheap, verifiable trading layer for IOUs that can then be settled externally. A lot like how many stock exchanges work. Ended up selling any XRP I had two years ago when things went nuts. Now it looks like people are re-building that whole concept elsewhere and Ripple is becoming irrelevant.

The ripple name was bought to exploit the positive impressions people had of the idea and was slapped onto something entirely different and unrelated.

https://bitcointalk.org/index.php?topic=144471.msg1547843#ms...

Bitcoin's lightning is in many ways the spiritual successor to the original Ripple idea-- a pathfinding network between pairwise relationships, but instead of IOUs, the parties use smart contracts to twiddle funds around in bitcoin payment channels without having to take the updates to the public network.

This fixes the big security weakness of Ripple while preserving the positive scalability properties. Unfortunately it also requires more complexity and limits flexibility somewhat.

I could be wrong, but I always understood the primary purpose of XRP as a mechanism to increase liquidity. Since Ripple is all about cross-border payments, you're potentially dealing with hundreds of currencies. So if you were a bank that needed to do global business, you'd be required to hold a subset of each traded currency in nostro/ vostro accounts. XRP was then meant to be the "standardized" currency that member-banks in the Ripple ecosystem could transaction with.

Since you as a bank would only need your own fiat currency + XRP, you could materially cut down on additional costs from holding all the other accounts. When it comes time to convert currencies, because Ripple (XRP) transactions settle in seconds, the period you're exposed to market volatility is virtually zero, so you can move in and out of a position in XRP in exchange for whatever other currency you need almost immediately.

I don't see how adding a currency makes this easier. Even if I don't want the currency you have, I want XRP less.

Adding a currency that everyone can agree upon as the "intermediary" prevents you from needing to hold huge sums of random currencies. For example - you're a bank in the UK so your fiat currency is GBP. If you want to manage transactions within just Europe, you would need to have Euros, Croatian Kunas, Danish Krones, Swiss francs, Polish zloty's, Russian rubles, Turkish liras, etc. Therefore you as a bank are subject to the ongoing valuations of these currencies, not to mention you've got hundred of thousands of (insert currency here) tied up in these accounts. As a bank, you have to correct for this inefficiency so you produce conversation rates that are not consumer friendly, but necessary to recoup the money you've lost setting aside all those accounts.

With XRP, you replace all those currencies with some amount of XRP. Because all member-banks use XRP when you need to transaction, say GBP to EUR (simple example), you would take your GBP, convert it to XRP, then use that XRP to buy EUR from another bank, that bank then can hold the XRP or chose to immediately convert it back to EUR through other banks. Like I mentioned previously, that whole transaction takes a few seconds, so you're effectively not subject to market volatility that could impact the transaction itself.

> you would take your GBP, convert it to XRP, then use that XRP to buy EUR from another bank

But I could more easily just use my GBP to directly buy Euros, with one less exchange rate and one less transaction in the process.

> that whole transaction takes a few seconds

For a bank, that's slow. But even if it wasn't, using an intermediate currency simply takes twice as long for each step.

> But I could more easily just use my GBP to directly buy Euros, with one less exchange rate and one less transaction in the process.

In this scenario, the transacting bank would need to have (or want) to have GBP for them to sell Euros. While I agree this is a poor example, think of the more obscure conversions. Swedish kronas to South African rands, etc.

Anyway, I won't pretend XRP is a perfect solution. Just seems to me that it solves the edge cases of liquidity while doing so more cost effectively, which in turn benefits the consumer.

Why wouldn't the bank just use dollars or Euros as the intermediate currency? Have a pile of dollars, a trading pair for everything against the dollar, and you're all set.

Currency roughly speaking doesn't move across borders, correspondent banks in each country have a ledger, you send money by decrementing the ledger on the send side and incrementing it on the receiver side. Someone else does the opposite. The exchange rate is set such that the ledger balance tends around 0.

No new currency needed.

I have no experience here, so I understand this in theory, but in practice, banks would somewhat indefinitely have some digital sum of a range of currencies in which they may not wish to have exposure to. How do you solve for that as a bank?

I suspect they directly trade the major currency pairs, and trade via USD for minor pairs.

I remember thinking it sounded awful. Automated repackaging and reselling of user-valued loans? The original idea was like the subprime mortgage racket, only with less opportunity for the participants to understand the risks.

So what destroyed or displaced the efforts of the original concept - was it the profit incentive associated with Bitcoin's structure, essentially maligning incentives?

Interesting. Do you have any links to some of the new IOU market maker projects? Cheers

Raiden is doing that for the ethereum network, if you count the fact that each node can be a mediator for multi-hop transfers. https://raiden.network

Raiden has fallen short on delivering a workable system and has had funding issues... most people people in the ethereum space are now banking on rollups filling these use cases (a competing approach for low-cost transactions)

Though certainly I'd be pleasantly surprised if Raiden still pulls through on their plans at some point.

Raiden has many issues, but funding is not one of them. Part of the hiring pitch from Brainbot was that the ICO got them funds to finance their operation for a decade at least.

Speaking as someone who worked on Raiden's core team last year and spent a reasonable part of this year working on a project received a grant from the Raiden Trust [0], the last release is workable, but still requires a lot of elbow grease and knowledge of the protocol to get it running. End users will certainly have a better shot when the light client gets parity to the reference implementation and gets a mainnet release.

In any case, I just wanted to point out that Raiden is a project that is related to what OP asked about and that is still alive and kicking.

[0]: https://hub20.io, a self-hosted payment gateway that integrates with Raiden and abstracts all its complexity from the users and node operators to enable fast and near-instant transactions.

So happy to see actual discussion of cryptocurrency concepts and not just visceral hate for all things crypto on hacker news. New line it looks like you guys have really turned a corner I'm proud of you hacker news.

it was doomed from the start, it was never a real crypto currency.

Indeed. I was involved back in 2014 and never understood why people didn't saw ripple as an open source central bank back end infrastructure. Anyone can start a network running rippled (you, me, the central bank), it is simple, ressource efficient, with our own "internal" XRP which are only used as rate limiters for transactions and then interconnect those networks at will. I don't get why there have been so many governments developing they're own infra for making their national currencies as a cryptocurrency, all they need to do is run a parallel ripple network

> What I've never understood about XRP and Ripple is that originally XRP was meant to only be used for paying for transactions and the ledger was meant to track IOUs between people denominated in some other "real" currency. They had videos floating around explaining the concept. In fact, they were handing out XRP for free to bootstrap the network of users [0] (that post says 1000 XRP, originally it was 30,000 XRP when the post was first created).

So the thing is, isn't this kind of system (very similar to a concept I can only remember as Hawala[0]) explicitly illegal?

> Hawala or hewala (Arabic: حِوالة‎ ḥawāla, meaning transfer or sometimes trust), also known as havaleh in Persian,[1] and xawala or xawilaad[2] in Somali, is a popular and informal value transfer system based not on the movement of cash, or on telegraph or computer network wire transfers between banks, but instead on the performance and honour of a huge network of money brokers (known as hawaladars). While hawaladars are spread throughout the world, they are primarily located in the Middle East, North Africa, the Horn of Africa, and the Indian subcontinent, operating outside of, or parallel to, traditional banking, financial channels, and remittance systems. Hawala follows Islamic traditions but its use is not limited to Muslims.[3]

> Some government officials[which?] assert that hawala can be used to facilitate money laundering, avoid taxation, and move wealth anonymously.[citation needed] As a result, it is illegal in some U.S. states, India, Pakistan,[14] and some other countries.[citation needed]

The whole IOU system doesn't need crypto to work -- but you just can't set up the kind of clearing house/federation that you'd need in the US because it's illegal, I thought.

[0]: https://en.wikipedia.org/wiki/Hawala

Ripple (old-school Ripple, not XRP) doesn't need any clearing house or federation. It's just you, me, and the hops along the way interacting long enough to facilitate the transaction.

My point is that how is that not a hawala system, just facilitated by a distributed ledger instead of phone calls? If it is, why isn't illegal just like those other systems?

The similarity here I thought was the trading of IOUs -- this seems to be very similar to what hawala is/was.

There is no middle man facilitating ripple. (Well, ideally. RipplePay was centralized. But it didn’t need to be.)

Also, hawala is legal in most jurisdictions I believe.

Thanks, I think this might be the big differentiator -- the existence of middlemen is probably the large difference here, but it feels like at some point you're definitely going to need someone to keep track of value across IOUs somehow and at the very least who is trust worthy.

Well to be accurate there are middlemen in most ripple transactions, but not a middleman of the traditional sense. They're not pure facilitators, but rather active participants in the transaction. I'm not a lawyer, but I understand that by some interpretations this makes a difference in the regulation.

Why not use something like Splitwise for tracking IOUs? What use is a blockchain/cryptocurrency for this?

Yup, the original vision of Ripple was payment channels, like what the Lightning Network is trying to be today. Pretty shameful how they abandoned all that to become a shitty version of PayPal.

There are all pyramid schemes. All the alt-coins exist only because BTC got too expensive and they wanted in. A lot of people are going to be left holding the bag.

While I generally agree, there are a few exceptions.

Ethereum exists because Bitcoin isn't capable of running complex scripts.

Monero exists because Bitcoin is too transparent.

And the vast majority of people "investing" in it have no concept of the underlying tech/utility/scalability/whatever, they only care "token go up!" and essentially treat it as a penny stock.

It'll eventually tank, but not until the insane FOMO ends. For example, the completely useless Dogecoin still has a $600M "market cap", and at one point popped over $2B.

Yup, I think the same. I think there’s a lot more “hype runway” left though. It’ll prob hit at least 50k

Absolutely all of them? You may want to look for some subtlety. There are coins with a specific purpose, for example Monero, that have a clear purpose which is separate from Bitcoin.

>All the alt-coins exist only because BTC got too expensive and they wanted in

Ethereum was explicitly created because Bitcoin doesn't support decentralized applications or really anything beyond sending tokens between humans. Vitalik tried to work on the Bitcoin codebase but was shunned and was forced to start a new project.

>There are all pyramid schemes

Do you include ethereum here? if so you are woefully misinformed.

XRP went down, interestingly enough.

I would have hoped for the community to collude and vote/bid its price up to give the SEC a big "ha" and also cover the costs of lawsuits.

After all crypto is one place where markets can be manipulated and the proceeds put to good use.

I'm pretty sure this is precisely what has happened, but I'm not as happy about it.

Reply on news.ycombinator.com