Heard Bad Things About Ripple? Bring Your Questions Here

[Apollo]

Welcome, to all the new members, investors and crypto enthusiasts taking a dive into Ripple for the first time. The Ripple Consensus Ledger (RCL) is a wonderful and complicated piece of tech, but there are a lot of misconceptions and mistruths floating around various forums. Like any good technology, Ripple made choices with real tradeoffs when designing RCL. But the cost and benefits of these decisions are not well understood in most of Cryptoland. 

Ripple also made controversial decisions with regards to their business strategy, and delay of full decentralization. I will be the first to tell you that not all of Ripple's decisions worked out. But they have always been well thought out. One thing you notice following Ripple is there are usually very good reasons for the decisions they make. So, if you have questions or heard something bad and can't figure it out, post it here. Hopefully, a few of us oldtimers can help make a little sense of Ripple. 

[stuffie]

I have a question: is 100b XRP enough when Ripple is going mainstream and when you can buy candy and houses with it in every country in the world?

You can't transfer $0.01 if the price of 1 XRP exceeds $10,000. Or am I wrong?

Edited April 5, 2017 by stuffie

[Apollo]

1 hour ago, stuffie said:

I have a question: is 100b XRP enough when Ripple is going mainstream and when you can buy candy and houses with it in every country in the world?

You can't transfer $0.01 if the price of 1 XRP exceeds $10,000. Or am I wrong?

 

lol, I have actually had this conversation, not as a joke.

Many economists believe more or less you need a little inflation. However, XRP will likely not have the same function of legal tender, even if it goes mainstream.  For example, here is no price stickiness if you're not denominating good and services in XRP. So it might not matter in the same way, even if you do accept the argued need for inflation with normal money.  

[Guest]

What exactly is the difference between the ILP and RCL? which is Ripple pushing to banks and where does XRP fit into both? 

[rootvegetable]

23 minutes ago, Apollo said:

lol, I have actually had this conversation, not as a joke.

Many economists believe more or less you need a little inflation. However, XRP will likely not have the same function of legal tender, even if it goes mainstream.  For example, here is no price stickiness if you're not denominating good and services in XRP. So it might not matter in the same way, even if you do accept the argued need for inflation with normal money.  

The set of economists who believe you need a little inflation and the set of economists who are wrong resemble a solar eclipse. I've included a visual to help clarify the point.

http://pop.h-cdn.co/assets/17/14/980x490/landscape-1491314824-solar-eclips-1999-5.gif

[Dsimmo]

2 hours ago, aaksut said:

What exactly is the difference between the ILP and RCL? which is Ripple pushing to banks and where does XRP fit into both? 

ILP stands for Interledger Protocol. It is an open source, currency and ledger agnostic, protocol that allows ledgers (e.g., the Bitcoin network and the Ethereum network) to be cryptographically connected through the use of market makers (called connectors) in a secure way https://interledger.org/. In more detail, it allows person A (who only wants to hold BTC) to seamlessly and securely transmit BTC from their account and have person B receive a corresponding amount of ETH. It is akin to the internet protocol (IP) in that it allows inter-networking of disparate ledgers, just like how IP allows inter-networking of disparate communication networks.
 

RCL stands for the Ripple Consensus Ledger. It is an open source distributed ledger technology, allowing arbitrary tokens (representing ounces of gold, USD, BTC, etc) to be transmitted from one ledger entry to another. For this, it uses a standard (non proof-of-XXX based) Byzantine fault tolerant consensus mechanism (I believe it uses the PAXOS algorithm) to achieve distributed consensus among the RCL nodes. During the consensus process, nodes that run the RCL software (anyone can run it) simply vote on which transactions are valid. The voting is public so that nefarious voters will be quickly outed by other RCL users. This allows ledger consensus to continue forward in the presence of nefarious behaviour. RCL also has a native currency, called XRP. 100 Billion were created at the inception of the RCL; no more can be created. Because RCL uses a voting mechanism (rather than proof-of-XXX), forming consensus is incredibly fast (1000 Tx per second and ledger close times sub 5 seconds) and cheap (fractions of a penny).

 

My understanding is that Ripple is heavily pushing both ILP and RCL, and believing that the future of payments lies in:

1) the internetwork of ledgers (i.e., ILP),

2) counterparty-less digital assets (e.g., XRP, BTC).

The first point is because, with ILP, ledgers no longer need to compete on user adoption. This is because all ledgers are connected. The fact that one user refuses to operate on a particular network matters not a penny for whether he can receive payments from users who do operate on that network. Instead, ledgers must compete on quality of service, and it is unlikely that any other distributed ledger can beat the quality of service offered by RCL (fast (1000 Tx per second and ledger close times sub 5 seconds) and cheap (fractions of a penny)).

The second point regards cross boarder payment, and the concept of a world reserve currency. For cross boarder payments to be most effective and not tie up capital, a (world) bridge currency will be required. The most `fair' bridge currency is one that is not tied to any particular country (i.e., America and the USD). Ripple believes that a digital asset (e.g., BTC or XRP) will be the most effective reserve currency. See the video below to understand more about why digital assets are good for this. Because of the immense quality of service offered by RCL, Ripple believes specifically that XRP will play the role as world reserve currency.

 

Edited April 5, 2017 by Dsimmo

[Sukrim]

4 minutes ago, Dsimmo said:

For this, it uses a standard (non proof-of-XXX based) Byzantine fault tolerant consensus mechanism (I believe it uses the PAXOS algorithm) to achieve distributed consensus among the RCL nodes.

No it doesn't - they use their custom consensus algorithm, called "Consensus". This matters because while anyone can run a node that takes part in this process, these nodes do NOT matter to the outcome unless other nodes include them in their local lists of nodes to trust. Currently Ripple Inc. runs 5 of such nodes that only trust each other, so adding your own node is like voicing your opinion about Trump if you are not an US citizen - maybe interesting, but you won't be able to vote on the actual US president unless they grant you citizenship. "Nefarious voters" are hard to detect, since the list of people whose votes you are willing to follow is secret and it is entirely possible to come to a different conclusion than large portions of the network initially. It is  not possible to visualize the underlying trust network between validators (well, other than the current situation: Ripple Inc. nodes only trust each other and nobody else matters - they either also trust these nodes or don't participate at all) which makes it hard to detect secret collusion - one of the main attack vectors against Consensus.

[Dsimmo]

2 minutes ago, Sukrim said:

No it doesn't - they use their custom consensus algorithm, called "Consensus". This matters because while anyone can run a node that takes part in this process, these nodes do NOT matter to the outcome unless other nodes include them in their local lists of nodes to trust. Currently Ripple Inc. runs 5 of such nodes that only trust each other, so adding your own node is like voicing your opinion about Trump if you are not an US citizen - maybe interesting, but you won't be able to vote on the actual US president unless they grant you citizenship. "Nefarious voters" are hard to detect, since the list of people whose votes you are willing to follow is secret and it is entirely possible to come to a different conclusion than large portions of the network initially. It is  not possible to visualize the underlying trust network between validators (well, other than the current situation: Ripple Inc. nodes only trust each other and nobody else matters - they either also trust these nodes or don't participate at all) which makes it hard to detect secret collusion - one of the main attack vectors against Consensus.

Their `Consensus' algorithm uses PAXOS. Trust should be granted tentatively. Also, I don't think you know who Ripple's nodes are trusting. This video explains RCL in much more detail than I was prepared to write

 

[Takio]

Sukrim knows a lot about how ripple/xrp functions.

[Hodor]

@Apollo thanks for posting - I also want to mention that it's important for us to welcome new members when they arrive, and remind them of the links that Karlos created to provide additional info.  Also, thank you to people like @enej and @tomxcs for activating new wallets!   Let's grow this ecosystem to new heights!  

[winston]

4 hours ago, rootvegetable said:

The set of economists who believe you need a little inflation and the set of economists who are wrong resemble a solar eclipse. I've included a visual to help clarify the point.

http://pop.h-cdn.co/assets/17/14/980x490/landscape-1491314824-solar-eclips-1999-5.gif

The point is you need to look at why they say inflation is good!

And if you take a look at that you can clearly see they aren't wrong, they have a different kind of perception and therefore nobody is wrong we just have a different kind of percention.

Let me explain you: we need inflation because people will be spending money because of it(next day it will be worth less). However, we don't want to force people to spend xrp anyways. It's a currency that will add value to the world and I believe that it will be more resistant to inflation than every other worldwide accepted currency ever seen before(well I consider usd sort as a worldwide accepted currency).

[Sukrim]

3 hours ago, Dsimmo said:

Their `Consensus' algorithm uses PAXOS. Trust should be granted tentatively. Also, I don't think you know who Ripple's nodes are trusting. This video explains RCL in much more detail than I was prepared to write

No it doesn't, it uses https://ripple.com/files/ripple_consensus_whitepaper.pdf. Their Chief Cryptographer says so too:

I know that Ripple Inc's nodes have very recently trusted only each other and there is little evidence that this has changed. This was even cited as something Ripple Inc did well (keep the ledger centralized under their own control for now, so they can react quickly in case of bugs).

Please don't spread wrong information unless you have very good proof.

 

[JoelKatz]

1 hour ago, winston said:

Let me explain you: we need inflation because people will be spending money because of it(next day it will be worth less). However, we don't want to force people to spend xrp anyways. It's a currency that will add value to the world and I believe that it will be more resistant to inflation than every other worldwide accepted currency ever seen before(well I consider usd sort as a worldwide accepted currency).

I'm not an economist. And of course I'm really not qualified to argue with the expert economists who hold this view. But it sounds like total nonsense to me.

Unless you're a thief (or a politician or preacher, but I repeat myself) to get money, you have to give something for it. If you don't spend that money, what you've done is given something and taken nothing back. That sounds pretty awesome to me. Now, you might take something back later, but if you do, you've given a loan of that thing you'll take back later (but could have taken as soon as you gave something for the money) to everyone. I don't see why you aren't entitled to interest on that loan.

The idea that people will defer purchases to the future when an asset will be worth more also doesn't hold any water. Say we all agreed that gold would be worth $9,000 per ounce next year. Would you have to wait until next year to get that price? Of course not. You would just offer someone the right to sell one ounce of gold one year from now for just a bit less than $9,000. And they would be happy to take it.

In other words, it's impossible to have sustained, predictable deflation because that requires someone to value gold now at $6,000 per ounce while they know for sure that gold will be worth $9,000 per ounce next year. Why would a person do that?

The other reason it's silly is this:

1) Deflation is bad because a person won't buy a car for $16,000 today if they know they can get it for $10,000 next year.

2) Inflation is bad because a person can't buy a car today unless someone is willing to sell it to them, and a person won't sell a car for $10,000 today if they know they can get $16,000 for it next year.

To be against deflation but okay with inflation, you have to find 1 plausible but 2 implausible. How do you justify that?

[winston]

However, getting stuff more expensive will make it less risky.(diversified value of 1 cent compared to 1 euro is more risk in % than 1 cent compared to 10,000 dollar) I guess we just don't agree at this point and that's totally fine by me. It's just my opinion.

[JoelKatz]

3 hours ago, Dsimmo said:

Their `Consensus' algorithm uses PAXOS. Trust should be granted tentatively.

It's not exactly PAXOS or PBFT. The reason we couldn't use them is that they, at least as commonly explained, require complete agreement among all nodes of who the participants are. That works for a private/permissioned ledger system but not particularly well for a public/permissionless one.

As it happens, there actually are conditions under which PBFT and PAXOS will work even if there isn't complete agreement. And you could, in theory, use them to form consensus on a public ledger system. But it wouldn't tolerate as much disagreement as Ripple's consensus algorithm does simply because it isn't made to.

But the Ripple network behaves much like it would if it used PBFT -- both algorithms solve the same kind of problem in the same kind of way.