Ripple Q & A @ Joel Katz and Ripple inc.

[🏰 TradeFortress 🏰]

Correct me if I am wrong this stuff is admittedly over my head. I really dont care about the details of why it works just that it works. I leave it up to tech nerds to figure out the detials

But dont I have to trust miners?

And with asics coming out that are tera-hash units I dont care what anyone says having that much computing power does re-centralize a decentralized network anyone can see that cpu mining would be less centralized then asic mining and from my view point it makes it much more probable that the network could be taken over and parts of the blockchain rewritten?

So my question is dont I have to trust third party institutions like KNC to not pervert the network?      

No. You are not trusting miners, you are trusting the proof of work and that no hostile party will control more than 51% of the hash power.

Anyone can mine. I'm sitting by my two Avalon ASICs as we speak . You don't need to be trusted to mine, because nodes trust your cryptography and proof of work - not trust based on who you are.

This is trust on a completely different level - it is like trusting your code is free of bugs, instead of trusting Bitstamp won't defraud you.

[1714304053]

1714304053

[DeathAndTaxes]

But dont I have to trust miners?

...

So my question is dont I have to trust third party institutions like KNC to not pervert the network?       

Miners can't change the rules.  Miner can't validate invalid tx.  Miners can't simply decide to print 100 billion BTC on a whim.  The Bitcoin trust model involves more than just miners.  All full nodes independently validate transactions.  If a miner (or cartel of miners) try to break the rules their blocks/txs will simply be discarded by the rest of the network.   Miners simply resolve "disputes" in the network by selecting one tx to be canonical.   If I send 1 BTC to you and the same BTC to TradeFortress only one of those can be valid or I printed money from nothing.   Without miners your node and TradeFortresses node will never reach consensus (as it involves one of you losing 1 BTC).   Miners force that consensus because one of those double spent tx will be included in the next block and the network will then consider the other "version" as invalid.

I recommend reading the Bitcoin wiki as it outlines what miners can and can't do.  

As for specifically trusting one ASIC provide.  Well that is the elegance of the Bitcoin reward system.  There is an economic incentive to "do the right thing" and that economic incentive as it relates to ASICs has spawned a true free market of competition.  KNC doesn't have 51% of the network, nor do they have any reason to try.  They can become incredibly wealthy by simply "doing the right thing" and providing superior hardware at a price the market accepts.

[sublime5447]

Correct me if I am wrong this stuff is admittedly over my head. I really dont care about the details of why it works just that it works. I leave it up to tech nerds to figure out the detials

But dont I have to trust miners?

And with asics coming out that are tera-hash units I dont care what anyone says having that much computing power does re-centralize a decentralized network anyone can see that cpu mining would be less centralized then asic mining and from my view point it makes it much more probable that the network could be taken over and parts of the blockchain rewritten?

So my question is dont I have to trust third party institutions like KNC to not pervert the network?      

No. You are not trusting miners, you are trusting the cryptography and that no hostile party will control more than 51% of the hash power.

Anyone can mine. I'm sitting by my two Avalon ASICs as we speak . You don't need to be trusted to mine, because nodes trust your cryptography and proof of work - not trust based on who you are.

This is trust on a completely different level - it is like trusting your code is free of bugs, instead of trusting Bitstamp won't defraud you.

That is exactly my point if KNC or some other asic manufacture where to decide that it would be advantageous to control the network could they not plug in their tera-hash units and have 51 percent of the network and do as they pleased? 

[DeathAndTaxes]

That is exactly my point if KNC or some other asic manufacture where to decide that it would be advantageous to control the network could they not plug in their tera-hash units and have 51 percent of the network and do as they pleased?  

51 percent of the network doesn't allow you to "do as you please".  That is a misnomer.  

With 51% of the network they could attack Bitcoin one of two ways
a) reverse tx they CREATED (double spend)
b) prevent tx from being confirmed (denial attack)

both of which would likely do significant damage to the value of Bitcoin and thus the value of their multimillion dollar enterprise which builds .... chips that are designed solely for high speed mining.

On the other hand the network model (either directly by mining or indirectly by hardware sales) gives them the ability to make significant profits at less risk by just "doing the right thing".

[🏰 TradeFortress 🏰]

That is exactly my point if KNC or some other asic manufacture where to decide that it would be advantageous to control the network could they not plug in their tera-hash units and have 51 percent of the network and do as they pleased?  

51 percent of the network doesn't allow you to "do as you please".  That is a misnomer.  

With 51% of the network they could attack Bitcoin one of two ways
a) reverse tx they CREATED (double spend)
b) prevent tx from being confirmed (denial attack)

both of which would likely do significant damage to the value of Bitcoin and thus the value of their multimillion dollar enterprise which builds .... chips that are designed solely for high speed mining.

On the other hand the network model (either directly by mining or indirectly by hardware sales) gives them the ability to make significant profits at less risk by just "doing the right thing".

Think about this: Why would KNC miner want to double spend or destroy the Bitcoin network when they are earning a steady stream of BTC through coinbase and transaction fees?

[sublime5447]

But dont I have to trust miners?

...

So my question is dont I have to trust third party institutions like KNC to not pervert the network?       

Miners can't change the rules.  Miner can't validate invalid tx.  Miners can't simply decide to print 100 billion BTC on a whim.  The Bitcoin trust model involves more than just miners.  All full nodes independently validate transactions.  If a miner (or cartel of miners) try to break the rules their blocks/txs will simply be discarded by the rest of the network.   Miners simply resolve "disputes" in the network by selecting one tx to be canonical.   If I send 1 BTC to you and the same BTC to TradeFortress only one of those can be valid or I printed money from nothing.   Without miners your node and TradeFortresses node will never reach consensus (as it involves one of you losing 1 BTC).   Miners force that consensus because one of those double spent tx will be included in the next block and the network will then consider the other "version" as invalid.

I recommend reading the Bitcoin wiki as it outlines what miners can and can't do.  

As for specifically trusting one ASIC provide.  Well that is the elegance of the Bitcoin reward system.  There is an economic incentive to "do the right thing" and that economic incentive as it relates to ASICs has spawned a true free market of competition.  KNC doesn't have 51% of the network, nor do they have any reason to try.  They can become incredibly wealthy by simply "doing the right thing" and providing superior hardware at a price the market accepts.

Thanks for the clarification my question would then be what if a government wanted to control the network to shut it down couldn't they go in with guns drawn and confiscate the equipment plug it in  and then create so much havoc that it collapsed the network? 

I know this is getting to the point of off topic on my own thread, but I guess the larger point to me is that all transactions require some level of trust.

Even transfers for gold or cash   

[🏰 TradeFortress 🏰]

Thanks for the clarification my question would then be what if a government wanted to control the network to shut it down couldn't they go in with guns drawn and confiscate the equipment plug it in  and then create so much havoc that it collapsed the network?  

I know this is getting to the point of off topic on my own thread, but I guess the larger point to me is that all transactions require some level of trust.

Even transfers for gold or cash  

They require different kinds of trust. Trusting cryptography versus trusting entities is like trusting physics to trusting deities.

And you're detailing a 51% attack - it's possible, but that's about a hostile government takeover instead of trusting entities or centralization vs decentralization.

[DeathAndTaxes]

Thanks for the clarification my question would then be what if a government wanted to control the network to shut it down couldn't they go in with guns drawn and confiscate the equipment plug it in  and then create so much havoc that it collapsed the network?  

Which would be easier try to find and confiscate and mine with Petahashes worth of equipment or hand a siezure warrant demanding OpenCoin turning over the private keys for the validator nodes?  For added bonus take the owners/operators away in handcuff and throw them in jail without trial under a content of court charge until they hand over the private keys.

If you are a corrupt government which would you find easier to control, manipulate, or subvert?

If you run the Ripple client you aren't part of the trust network anymore than you having a banking account makes you part of the interbank network.  Your client will accept whatever validators say are correct and the client implicitly trusts OpenCoin controlled valiators.  A single seizure warrant would give a government complete control of the Ripple network tonight with no massive hashing farm to run or the rest of a global network to compete against.   

I know this is getting to the point of off topic on my own thread, but I guess the larger point to me is that all transactions require some level of trust.

All trust is relative.  All risk is relative.  I could die tonight falling down the stairs but that doesn't mean climbing stairs is equally risky to playing with high voltage line in the rain.

[sublime5447]

I get your point

So does trusting the ripple network require the same trust as trusting the ripple currency?

edit- you where one step ahead of me. I see the answer is yes 


The one is for Joel if he decides to jump back in but anyone can interject obviously.  

Why if you can transact in the ripple network for any currency would the ripple currency itself ever become valuable and doesn't the business model require XRP to appreciate to become profitable?

[sublime5447]

"And you're detailing a 51% attack - it's possible, but that's about a hostile government takeover instead of trusting entities or centralization vs decentralization"

I cant quote to save my life 





All governments are hostile. That is the nature of government and no one has more incentive to stop it than them.

[r3wt]

Correct me if I am wrong this stuff is admittedly over my head. I really dont care about the details of why it works just that it works. I leave it up to tech nerds to figure out the detials

But dont I have to trust miners?

And with asics coming out that are tera-hash units I dont care what anyone says having that much computing power does re-centralize a decentralized network anyone can see that cpu mining would be less centralized then asic mining and from my view point it makes it much more probable that the network could be taken over and parts of the blockchain rewritten?

So my question is dont I have to trust third party institutions like KNC to not pervert the network?      

with the network currently setting at a petahash and change, your "end of the de-centralized bitcoin network" scenario seems higly unlikely. sure there will be large portions of network hash rate controlled by various large scale mining ops, but isn't that how its always been? back in may i knew a guy raking in 200+ btc a month with a 1500 ghash gpu farm. now what can he make with that? a fraction. my point being that things change so quickly, its highly unlikely that any normal human being with less than a couple million dollars could acquire enough hardware in a short enough period of time to effectively "centralize" bitcoin. i think you need to quit reading fairy tells and stick to the facts. I won't comment on ripple because frankly i have no interest in it, but i wish David and Co. all the success in the world with their endeavors.

r3wt

[sublime5447]

Correct me if I am wrong this stuff is admittedly over my head. I really dont care about the details of why it works just that it works. I leave it up to tech nerds to figure out the detials

But dont I have to trust miners?

And with asics coming out that are tera-hash units I dont care what anyone says having that much computing power does re-centralize a decentralized network anyone can see that cpu mining would be less centralized then asic mining and from my view point it makes it much more probable that the network could be taken over and parts of the blockchain rewritten?

So my question is dont I have to trust third party institutions like KNC to not pervert the network?      

with the network currently setting at a petahash and change, your "end of the de-centralized bitcoin network" scenario seems higly unlikely. sure there will be large portions of network hash rate controlled by various large scale mining ops, but isn't that how its always been? back in may i knew a guy raking in 200+ btc a month with a 1500 ghash gpu farm. now what can he make with that? a fraction. my point being that things change so quickly, its highly unlikely that any normal human being with less than a couple million dollars could acquire enough hardware in a short enough period of time to effectively "centralize" bitcoin. i think you need to quit reading fairy tells and stick to the facts. I won't comment on ripple because frankly i have no interest in it, but i wish David and Co. all the success in the world with their endeavors.

r3wt

That was my feelings when I first heard about it too, but I am starting to come around a bit. The Ripple currency it's self is not of much interest to me but the network has wild possibilities.

I have become disillusioned with bitcoin because it is so highly manipulated, is difficult to get, there are many attempts to regulate it, it is rampant with fraud, it is treated as commodity and not a currency, it has 0 potential for wide spread adoption (no mass appeal), it is not a threat to the banks or governments and I cant see how it ever will be, there is almost no p2p transactions with it, and the biggest problem I see is that it is incapable of bridging the gap between the traditional banking system and itself.

Ripple has the potential to fix a lot of those problems especially, If they are successful doing three things, 1. providing positive identification, 2. importing trust from other areas, and 3. providing insured transactions.

I need to think about it more though

I wonder what would happen if a node became to big to fail if it could create systemic failure that could create a a scenario where people in the network could lose their money due to no fault of their own?
I also wonder about its capacity to facilitate ponzi schemes?
 

     

[sublime5447]

Currency is to be spent

Money is to be saved

Right now bitcoin is a hybrid that doesnt do either well.

[sublime5447]

I also wonder about the adaptability of the trust network. For instance if a member that I trust decided that they wanted to use the ripple network for crowd funding for a highly speculative venture. If a public company did that it would more often then not be forced out into the open and I would have the opportunity to sell off my shares, but if someone did that in the ripple network it would be difficult to detect. I wonder if there could be an alert mechanism?
That could say hey this guy that you trust is engaging in risky behavior.

I am thinking out load here, just trying to run through likely scenarios     

[bitcoinBull]

Are you seriously saying that when OpenCoin unilaterally changed account reserve requirements, which is akin to Federal Reserve tampering with interest rates?

{
    "Account" : "rrrrrrrrrrrrrrrrrrrrrhoLvTp",
    "Fee" : "0",
    "Sequence" : 0,
    "SigningPubKey" : "",
    "TransactionType" : "SetFee",
    "hash" : "C6A40F56127436DCD830B1B35FF939FD05B5747D30D6542572B7A835239817AF",
    "ledger_index" : 562177,
}

Decentralized - OpenCoin unilaterally changed account reserve requirements.

The change has to be accepted by the majority of validators. You already know that, unless you didn't read the thread where you learned about that SetFee transaction.

Bitcoin will work if I have completely different nodes with someone else as long as we're on the same blockchain. It's not the same with Ripple.

No it won't, you will not be on the same blockchain if you are connected to completely different nodes. That's why the bitcoin client connects to a bootstrap IRC channel, so that everyone can connect to the same set of ip addresses. The UNL basically serves the same purpose, except where bitcoin requires hash power as a defense against an ip address Sybil attack, ripple uses an explicit list of non-colluding nodes. The function of the UNL is analogous to a miner choosing to mine at three different pools because she trusts that all three will not collude against her.

Take the barrier of entry. The Bitcoin network has a very low barrier to entry, being that if you run Bitcoin-Qt, the recommended client of that time, you are now part of the network. The current recommended client on Bitcoin.org, MultiBit still makes you part of the network. With Ripple, the recommended client, ripple.com/client does not make you part of the network, and does not allow you to choose your own validators. Ripple inherently favors centralization.

Running the bitcoin client only means you can broadcast transactions, that's somewhat shallow participation in the network (you aren't mining blocks). The ripple client has a setting in Advanced -> options to set your validator node / rippled server, so its just not true that it doesn't allow you to choose your own validators (the setting is an explicit field right there in the UI).

I also believe that a trust free solution is inherently more decentralized than a solution that requires trust, even if granting trust is decentralized (which isn't with Ripple at it's current stage).

That's fine, but bitcoin alone is only a trust-free solution if you never do any trading on exchanges. You are the one comparing apples to oranges, because the ripple protocol decentralizes multi-currency exchange. The bitcoin protocol decentralizes bitcoin transactions.

Miners can't change the rules.  Miner can't validate invalid tx.  Miners can't simply decide to print 100 billion BTC on a whim.

Ripple did not print BTC, they printed XRP. Every alt-chain miner decides, on a whim, to print their own pet alt-coins (litecoins, feathercoins, ppcoins, etc). May the best alt-coin win.

Tell me how I can print 100 billion XRPs.

Start your own ripple ledger.

[JoelKatz]

You can't get rid of OpenCoin and have a functioning Ripple, even the wiki admits that:

If everyone chooses a completely disparate sets of validators the network will be unlikely to reach consensus that a particular version of the ledger is the one true and accurate ledger.

What happens if everyone chooses a completely different set of rules for what constitutes a valid Bitcoin block? Yes, people have to agree to work together for any system to function.

But no central authority is needed to make this happen. The rules for what constitutes a valid Bitcoin block today are de facto decided by Gavin. But he's not a central authority because nobody has to listen to him. People do because they know that cooperation is needed for the system to work and he is doing a good job. He can't sustain his position unless he is responsive to the community. If he were to make some absurd decision, few would listen to him and that would be the end. Yes, lots of people update their code to whatever release he publishes but only because they trust him and can withdraw that trust at any time.

Any number of organizations can publish lists of validators that they recommend. These lists can be algorithmically combined. If someone publishes a list that is nonsensical, you can just stop following them.

[JoelKatz]

I get your point

So does trusting the ripple network require the same trust as trusting the ripple currency?

edit- you where one step ahead of me. I see the answer is yes

Not really. You're just trusting validators not to collude against you. When you extend a trust line, you are trusting someone to hold money for you or owe you money.

Why if you can transact in the ripple network for any currency would the ripple currency itself ever become valuable and doesn't the business model require XRP to appreciate to become profitable?

The business model requires XRP to appreciate. Nobody would disagree that what we're trying to do is high risk. Success is definitely not guaranteed.

XRP is unique though. It's the only currency that can flow between any two parties with no trust path needed. It's the only currency in Ripple that has no counter-party. While other currencies can have transfer fees imposed on them by their issuers, XRP never has any transfer fee other than the transaction fee. We expect that people will make assets liquid on Ripple by making them liquid to and from XRP, and we expect that people who don't know what asset they'll need will tend to hold XRP.

[DeathAndTaxes]

No it won't, you will not be on the same blockchain if you are connected to completely different nodes. That's why the bitcoin client connects to a bootstrap IRC channel, so that everyone can connect to the same set of ip addresses.

For the record Bitcoin hasn't bootstrapped using IRC for almost two years now.  Last version was 0.5.9 IIRC.

[sublime5447]

Running the bitcoin client only means you can broadcast transactions, that's somewhat shallow participation in the network (you aren't mining blocks). The ripple client has a setting in Advanced -> options to set your validator node / rippled server, so its just not true that it doesn't allow you to choose your own validators (the setting is an explicit field right there in the UI).

Are you saying that people could ignore Ripple inc. If they made a change that people didnt like?

[🏰 TradeFortress 🏰]

You can't get rid of OpenCoin and have a functioning Ripple, even the wiki admits that:

If everyone chooses a completely disparate sets of validators the network will be unlikely to reach consensus that a particular version of the ledger is the one true and accurate ledger.

What happens if everyone chooses a completely different set of rules for what constitutes a valid Bitcoin block? Yes, people have to agree to work together for any system to function.

But no central authority is needed to make this happen. The rules for what constitutes a valid Bitcoin block today are de facto decided by Gavin. But he's not a central authority because nobody has to listen to him. People do because they know that cooperation is needed for the system to work and he is doing a good job. He can't sustain his position unless he is responsive to the community. If he were to make some absurd decision, few would listen to him and that would be the end. Yes, lots of people update their code to whatever release he publishes but only because they trust him and can withdraw that trust at any time.

Any number of organizations can publish lists of validators that they recommend. These lists can be algorithmically combined. If someone publishes a list that is nonsensical, you can just stop following them.

We are not talking about different rules, we are talking about different entities with the same rules. How do you propose on solving the centralization situtation with the Ripple light client?