Jump to content


  • Content Count

  • Joined


About jn_r

  • Rank

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Exactly I agree with you it is not the old autobridging from the DEX, but, if it looks like a duck and it quacks like a duck, maybe might as well call it a duck.
  2. Yes. I think the exchange itself in principle only delivers the rails. For e.g. EUR/USD market it probably uses 2 external market makers. The ODL input/output is used for the connector functionality in ILP. The connector can also extend to a second exchange and then the XRPLedger speed comes in to play. Remember that the risk in an ILP transaction is with the connector, so in this case Bitstamp is the connector. If the ODL transaction to Bitso would fail because e.g. Bitso or XRP/MXN market maker messes up, then Bitstamp simply does not receive the money from bankA (because it was in escrow). In that sense ILP does create atomic transaction from the bank point of view, it either succeeds (money transferred to BankB) or fails (no money sent). Not sure yet about the Payment Channels, will have to read in to that Also, Payment channels are currently not much used, see https://xrpscan.com/metrics tab 'types'
  3. I forgot about that. Up till now we have only seen xVia (moneygram) + xRapid. The xCurrent implementations however (banks and such) transact with a form of ILP. So if an xCurrent implementation routes a transaction via an ODL enabled exchange, that would imply that the on-ramp is possible via ILP and the off-ramp also..Maybe @hamasugu is correct with the Joel Katz explanation. like this (ODL only on bitstamp exchange): EUR bankA -> ILP(via Bitstamp) -> USD bankB or (ODL via bitstamp/bitso exchanges) EUR bankA -> ILP(via Bitstamp/Bitso) -> MXN bankC
  4. Well, this is something different and requires some knowledge about ILP. (ILP is also not used in ODL afaik..) With an ILP transaction between 2 parties you need a connector that facilitates the ILP transaction. It seemed at the time that exchanges would be very well fit to take that roll. So in this example someone sends from the Bitcoin ledger some bitcoin, via ILP (via a hypothetical Poloniex connector) to an Ether adres at the Ethereum ledger. And in doing so, Poloniex uses its XRP orderbooks from BTC/XRP and XRP/ETH. Joel Katz uses the term 'XRP auto-bridging', but I doubt it has ever been implemented as such on an exchange. It could definitely be done, but then, why would any exchange use XRP as intermediary. Better use BTC, just as fast (your're not on the bitcoin ledger, remember?) and much more liquid on the orderbooks.
  5. Autobridging is a beautiful thing. It allows for combining two orders in to one atomic order. Because it is atomic (only possible on one ledger) it takes away an important risk for arbitrage, namely either both orders succeed or both don't. So you never end up with an asset that you don't want (XRP in this case). ODL can't do the order in an atomic way. ODL is dependent on normally two exchange ledgers, the XRP Ledger and two market makers. That is not atomic. But, it is guaranteed to work for a predetermined price because 2 market makers guarantee a certain price. This only works with market makers. ODL could be atomic if both on-ramp and off-ramp are on one exchange, so, in this case EUR/USD could be like that. But I think market makers are still part of the flow and therefor the trade is not atomic. Just the on-ramp and off-ramp on one exchange and no need to transact the XRP because it is already on the target exchange.
  6. Without a doubt is ODL inspired by the DEX, autobridging and XRPLedger, the similarities are clearly there. But look at the dates, autobridging existed already in 2014 (Yes, brilliant even then) and works only on the DEX - not on Bitstamp or Bitso. To my knowledge there is nothing on https://xrpl.org that directly describes ODL. edit: except for sending XRP from account A to account B
  7. Maybe the old rippleAdminConsole is an option for you? It is no longer officially maintained, but I still use an older version and recently r0bertz fixed and created a new version with the newest ripple-lib. The wallet can be used offline and stores the keys in a specific format - a wallet file which is encrypted with a password: This other wallet from ripplefox apparently can also handle the same wallet-format
  8. David Schwartz puts it like this:
  9. I think the software is not built for that, so I think 'no'. But I am no insider in Ripple ODL software, so really can't tell for sure... On a non-technical note, it would require market makers on every possible orderbook for Bitstamp, one for USD->MXN, one for USD->PHP, one for EUR->MXN, etc.. If you use one base currency, then on Bitstamp you only need one type of market maker, XRP->USD. And since Bitstamp is also home-base for euro-zone, it requires also market makers for XRP->EUR
  10. Technically it is possible, but that is not how the ODL rails are built. XRP is always involved as one of the pairs. 'XRP the base' as they used to say
  11. I think it's a valid use case (e.g. Moneygram converting EUR to USD) using the available rails from ODL. It follows as a consequence from having ODL market makers available on XRP<->EUR and XRP<->USD. And because it involves the same exchange the XRP does not have to be transferred via the XRPLedger. If the exchange rate is cheaper than other markets, then why not? I think this can happen from time to time due to imperfections in the EUR<->XRP and/or USD<->XRP markets. And of course, enough liquidity on both currency pairs, such that the order books are not hollowed out
  12. Yes that's what I meant, a ledger technically implemented with some sort of private database managed by one entity without consensus mechanism. But since it is private we can't know that last part for sure :-) So, a private database.
  13. ODL works with 'normal' exchanges, such as Bitstamp and Bitso. Each has its own 'proprietary' ledger. On Bitstamp there is an orderbook USD<->XRP and on Bitso there is an orderbook XRP<->MXN. An ODL order basically looks like this: Exchange USD on exchange A (Bitstamp) for XRP Send the XRP to exchange B (Bitso) Exchange XRP on exchange B (Bitso) for MXN Those are the simple steps. As it turns out there is more to it and Ripple is not very open about it how it works exactly. Imo it could look something like this: ClientA wants to send MXN to clientB using USD. He checks with Moneygram for a price offer Moneygram checks with (a set of) MarketMakerA (working on Bitstamp USD/XRP) and asks for a price offers USD/XRP Moneygram checks with (a set of) MarketMakerB (working on Bitso XRP/MXN) and asks for price offers XRP/MXN Money gram now offers ClientA the price USD/MXN The ClientA accepts Moneygram offers the USD at Bitstamp USD/XRP exchange for the agreed price from MarketMakerA MarketMakerA then takes the trade (if not already taken) Moneygram sends the XRP to Bitso exchange Moneygram offers the XRP at Bitso XRP/MXN exchange for the agreed price from MarketMakerB MarketMakerB then takes the trade (if not already taken) Moneygram sends the MXN to clientB Something like that. The thing that might raise questions is, how do the market makers determine the price and how do they rebalance their USD, MXN, or XRP. Because in the end MarketMakerA will have a lot of USD and needs XRP, and MarketMakerB will have a lot of XRP but needs MXN. There is some discussion about this elsewhere on the forum, but my take on it is, that the market makers will have to rebalance simply via other liquidity sources. Those other liquidity sources are perhaps less fast than ODL, but this is exactly what marketmakers are for, this is part of the service they deliver. On top of that the market makers could probably do a better job if there were financial instruments for XRP available, such as lending, borrowing, leverage, etc. As it seems, Ripple is currently working on these instruments for XRP. So, quite a story, yet nowhere are IOU's used that are available on the XRPLedger. Might be an interesting thought experiment to imagine one of the exchanges (or both?) as a gateway on the XRPLedger-DEX, but that is not how ODL is currently used. btw, everything you read at https://xrpl.org is not about ODL (except for the sending XRP part)
  14. First of all, it is important to realise that ODL is not making use of the IOU/DEX mechanism, it only makes use of simple XRP transfers. The IOU/DEX functionality is another aspect of the XRPLedger. From a technical point of view more interesting imo, but not used in ODL :-) Anyways, beside that, I kinda agree with you that issuers trusting each other is an interesting concept. But it would expose the gateway A to the systemic risk of gateway B. Let's say that gateway B gets hacked and turns out to be insolvable, then a fully automated bankrun will take place where gateway A will end up will all useless IOU's from gateway B. Gateway A could reduce that risk to the amount of trust as defined in the trust line. I have been playing with that concept to see what it would look like:
  15. The following tweet is food for speculation Reading it as it is, you must conclude that Ripple is not working on Codius. But, if you are into some speculation and reading between the lines, you could interpret it as that Polysign is building on the Codius technique in some form, i.e. take in to account that next to working at Ripple, David Schwartz is also CTO for the new business Polysign. And it is actually not that far fetched, Codius in adjusted form fits the Polysign use-case as a glove. It can hold secrets of any type, it allows for decentralization of keys (multi-sig), it can do smart contracting (for DeFi), it can transact every type of digital asset, and what more. Create a trusted set of highly qualified nodes with HSM for the node operator security and you have a perfect solution. If not Codius itself, then at least some of the knowledge and techniques from Codius are being used.
  • Create New...