Why don't the exchanges get together and fork Ripple and use it.

[CLains]

Exchanges can use BitShares DEX to issue their own IOUs and move their orderbooks onto the blockchain. They basically would become gateways using the decentralized exchange.

Interesting.
In that case they will all have the same orderbook? So the price some asset will be the same on all exchanges in the blockchain?

BitShares lets you trade 2 types of assets: user issued assets and market pegged assets. Each exchange could issue a user issued asset as an IOU for their assets (IOUs are basically what you have when you hold crypto on a centralized exchange) and trade against other exchange's IOUs or market pegged stablecoins like bitUSD, bitCNY, and bitGOLD. On BitShares you can trade any asset against any other asset... but not all of the markets are liquid.

So you might see coinbaseUSD, bitstampUSD, btceDOGE, etc. trading on bitshares. The cool thing about this is that each participant still controls the private keys to their balances, and they can also be sure that the orderbooks aren't being manipulated behind the scenes.

More reading:

http://bytemaster.bitshares.org/update/2014/12/18/Benefits-of-Being-a-BitShares-Gateway/
http://bytemaster.bitshares.org/article/2015/01/05/The-Future-of-Crypto-Currency-Exchanges/
http://bytemaster.bitshares.org/article/2015/01/29/How-BitShares-Prevents-Front-Running/

This. BitShares enables KYC/AML compliance for User-Issued Assets (UIAs -> IOUs), and when Exchanges becomes Gateways, liquidity on the decentralized exchange for bitAssets like bitUSD, bitCNY and bitGOLD will go through the roof as a counterparty-risk-free safe haven that grants users who want fiat, derivatives, stocks, etc. all the benefits of crypto.

[1715612283]

1715612283

[1715612283]

1715612283

[1715612283]

1715612283

[tokeweed]

Please be on topic.  We aren't talking about Bitshares.

[tokeweed]

Quoting Sukrim's last post below to avoid the topic getting sidetracked...

1) How would they pay for network fees (= who gets the forkXRPs)?
2) How would they ensure a healthy validator network structure?

These are the 2 main remaining questions for mainline Ripple too...

1) First of all in Ripple nobody get's the fees. XRPs are being destroyed not re-destributed. Second if the exchanges run their own "internal" network then they don't need to worry about fees because they controll all of the forkXRPs and can split them up evenly.

2) Each exchange runs 1 validator which trusts the other exchanges validators. done.

1) I know they are destroyed, you need to own them before you can destroy them though. Customers of these exchanges will also want forkXRP to operate their own accounts, who should pay them? What if 5 exchanges start this project, get 20 billion forkXRP each and then a 6th one wants to join?

2) How do you ensure that this is really the case and some exchanges are not just claiming to have everybody on their UNL but are secretly forming a network that will kick your validator off?

change the consensus mechanism to POS, because it's currently trust based and only 20% collusion attack proof

Ripple's Consensus is 79.9...% collusion proof according to their technical whitepaper. if you have different claims, please write a more formal mathematical explanation.

Proof of Stake only works if the thing that is "at stake" is well distributed...

[monsterer]

change the consensus mechanism to POS, because it's currently trust based and only 20% collusion attack proof

Ripple's Consensus is 79.9...% collusion proof according to their technical whitepaper. if you have different claims, please write a more formal mathematical explanation.

Proof of Stake only works if the thing that is "at stake" is well distributed...

https://ripple.com/files/ripple_consensus_whitepaper.pdf

Lastly we will show that the Ripple Protocol can
achieve these goals in the face of (n − 1)/5 failures which is not the strongest result in the literature...

...FaB Paxos [5] will tolerate (n − 1)/5
Byzantine failures in a network of n nodes, amounting
to a tolerance of up to 20% of nodes in the network
colluding maliciously

edit: I'm not saying this is bad per se, it is appropriate for ripple labs to be happy with this, because they work with the banking industry which is highly trust based already, but for altcoins, which are largely trustless, we need a different way.

[robrigo]

Please be on topic.  We aren't talking about Bitshares.

Not trying to jack your thread, but you asked for our thoughts. What can Ripple network do that bitshares is incapable of? Why must it be a fork of Ripple, if you can achieve the same end result of creating a gateway network between exchanges in a system that is more decentralized? Genuinely curious.

[monsterer]

Not trying to jack your thread, but you asked for our thoughts. What can Ripple network do that bitshares is incapable of? Why must it be a fork of Ripple, if you can achieve the same end result of creating a gateway network between exchanges in a system that is more decentralized? Genuinely curious.

Ripple is designed to be a auto-bridging network between liquidity providers. It will find you the best deal for your given order requirements, crossing several different liquidity providers in one atomic transaction, if that is the path of the best deal.

Bitshares has IOUs, and it has gateways, but it doesn't have the glue which joins them together in a seamless manor.

[Sukrim]

Ah, we were talking about the same thing in different ways:

Yes, if more than 20% of your UNL colludes against you, you'll see that there is something fishy going on and your node will stop validating. Only if 80+% of these nodes collude can they force an actual "bad" transaction upon you though sompared to 50+% of hashing power in PoW or 50+% of current staking equity in PoS.

[robrigo]

Ripple is designed to be a auto-bridging network between liquidity providers. It will find you the best deal for your given order requirements, crossing several different liquidity providers in one atomic transaction, if that is the path of the best deal.

Bitshares has IOUs, and it has gateways, but it doesn't have the glue which joins them together in a seamless manor.

Ah good point, thanks monsterer. Maybe that will be possible in the future.

[monsterer]

Ah, we were talking about the same thing in different ways:

Yes, if more than 20% of your UNL colludes against you, you'll see that there is something fishy going on and your node will stop validating. Only if 80+% of these nodes collude can they force an actual "bad" transaction upon you though sompared to 50+% of hashing power in PoW or 50+% of current staking equity in PoS.

I don't think that is correct. That 80% number you're referring to is the amount of agreement that nodes in the UNL must have before a given transaction is approved, not the amount of collusion the network is resistant to, which is a different metric.

since a transaction is only approved if 80% of the UNL of a server agrees with it, as long as 80% of the UNL is honest, no fraudulent transactions will be approved. Thus for a UNL of n nodes in the network, the consensus protocol will maintain correctness so long as: f ≤ (n−1)/5 (1) where f is the number Byzantine failures

[Spoetnik]

why ?

what would be the difference between any other premined / IPO coin ?

ripple is garbage I don't get it..

[Sukrim]

Ah, we were talking about the same thing in different ways:

Yes, if more than 20% of your UNL colludes against you, you'll see that there is something fishy going on and your node will stop validating. Only if 80+% of these nodes collude can they force an actual "bad" transaction upon you though sompared to 50+% of hashing power in PoW or 50+% of current staking equity in PoS.

I don't think that is correct. That 80% number you're referring to is the amount of agreement that nodes in the UNL must have before a given transaction is approved, not the amount of collusion the network is resistant to, which is a different metric.

since a transaction is only approved if 80% of the UNL of a server agrees with it, as long as 80% of the UNL is honest, no fraudulent transactions will be approved. Thus for a UNL of n nodes in the network, the consensus protocol will maintain correctness so long as: f ≤ (n−1)/5 (1) where f is the number Byzantine failures

20% to block, 80% to get something conflicting in. Everything in between will vote "no" on both transactions since consensus won't be reached.

So yes, if there are 20%+ colluding, you have a problem. If 80%+ are colluding, you have a REAL problem.

why ?

what would be the difference between any other premined / IPO coin ?

Ripple is mainly a distributed exchange, not a "coin".

[monsterer]

20% to block, 80% to get something conflicting in. Everything in between will vote "no" on both transactions since consensus won't be reached.

So yes, if there are 20%+ colluding, you have a problem. If 80%+ are colluding, you have a REAL problem.

No consensus is essentially a DOS - 20% is too low a threshold for a trustless environment like the altcoin world, IMO.

[Sukrim]

DoS with cryptographic proof that you are being attacked though. I agree however, these numbers might need to be tweaked depending on requirements. Back on topic I still don't see how a fork would give exchanges anything else than a larger amount of (worthless) native tokens but a LOT of additional infrastructure to maintain.

For the costs of a few full rippled nodes alone (as a comparison: Ripple has about the same amount of tx/s as Bitcoin right now with no arbitrary blocksize limitations, requires more than 1 TB of SSD space if you want full history and won't even compile on something with less than 8 GB RAM) you can buy a LOT of XRP. Add to this people to maintain the infrastructure, handle issues...

[monsterer]

For the costs of a few full rippled nodes alone (as a comparison: Ripple has about the same amount of tx/s as Bitcoin right now with no arbitrary blocksize limitations, requires more than 1 TB of SSD space if you want full history and won't even compile on something with less than 8 GB RAM) you can buy a LOT of XRP. Add to this people to maintain the infrastructure, handle issues...

You can't use ripple for this, ripple labs has too much control.

[Sukrim]

For the costs of a few full rippled nodes alone (as a comparison: Ripple has about the same amount of tx/s as Bitcoin right now with no arbitrary blocksize limitations, requires more than 1 TB of SSD space if you want full history and won't even compile on something with less than 8 GB RAM) you can buy a LOT of XRP. Add to this people to maintain the infrastructure, handle issues...

You can't use ripple for this, ripple labs has too much control.

...which they haven't yet used, even in cases where it would have been much cheaper and faster to do so (e.g. Jed McCaleb's threat to sell off his founder's share or the recent sale of 1 million USD worth of XRP by him or one of his relatives) or where people clearly were defrauded on Ripple. I get the argument though and still hope that their decentralization project at https://rippletest.net/ takes off.

[monsterer]

I get the argument though and still hope that their decentralization project at https://rippletest.net/ takes off.

I hadn't heard about that, thanks, I will read up on it...

...First thing I saw was this:

focus on public trust

which doesn't bode well, but I'll keep reading

[Spoetnik]

what scares me is they may get Banks doors open via Bribery..
Say for example they convince some random bank to use Ripple and in turn they will donate millions of Ripples..
ugggh scary