The Rainbow Network: An Off-Chain Decentralized Synthetics Exchange [pdf] | Hacker News Reader

The Rainbow Network: An Off-Chain Decentralized Synthetics Exchange [pdf]

rainbownet.work

48 pointsbpierre11 comments

This "synthetic" balance does closely remind me of "uninsured" balance I proposed for XLN (Lightning, extended with credit lines to solve capacity problem) https://github.com/fairlayer/wiki/blob/master/04_four_balanc...

The difference is synthetic balance is offered to be enforced from the collateral from the escrow. In XLN I proposed to try to enforce it by taking it from the other party's onchain balance.

Both options would work together even better ("synergy" haha)! From the first example:

1. blockchain dispute resolution code tries to take USDC from Alice's onchain balance (since Alice owes 750 to Bob).

2. if Alice still owes something, try to convert any other Alice's onchain balance into USDC using oracles, then return to Bob.

PS. will finish reading the paper sometime soon, so not sure if got it wrong.

Overall, happy to see new solutions coming to payment channels that do not imply full collateralization (via cross-asset synthetic collateralization as in here). Full prepaid collateral is stupid and limiting the growth.

Sounds cool. At the end of the day, someone will probably built it and deploy it and we'll get to see it succeed or fail. Very interesting time to be alive. A lot of good science going on. I think we will look back at this time as a fairly golden age in whatever we eventually call this (decentralized finance, or is that just one angle?). This will be the case regardless of what succeeds or fails. Really an industry you can make advances in.

> decentralized finance, or is that just one angle?

People are grouping under that term, #DeFi.

Upvoted because I'm pretty sure this comment was written by a neural net and I'm hoping for a reply.

A neural network trained on HN comments would probably give a scathing critique of a tangentially related project.

I'm 99% sure the original comment was written by a neural net. I'm also 99% sure your reply was written by a neural net. I'm 100% sure my comment was written by a neural net.

Boring old wet neural nets :-)

> 3.8 Risk of undercollateralization

> In Example C, if the price of ETH falls by 50%, Alice’s balance in the channel is effectively worth nothing. If the price of ETH falls further than that,the channel would become undercollateralized—Bob would no longer be able to withdraw his USD at its current value.

This can and will happen in the real world, and seeing how the "crypto-sphere" is teeming with scammers running pump-and-dumps, and exchanges manipulating prices, it's pretty much guaranteed.

Let's restrict the discussion to the top 10 cryptoassets. All of those would be pretty hard to manipulate in to a price decrease of 50% considering how much money you'd have to spend. So it seems like this concern is really only a function of how volatile or risky the asset is. And as for price manipulation, any exchange doing that is not going to be a top-10 exchange either.

> And as for price manipulation, any exchange doing that is not going to be a top-10 exchange either.

You mean... Like MtGox or btc-e or GDAX?

Okay, so I have no evidence that GDAX is actually running the bots that comprise a tremendous amount of its transactions, but you get my point.

There's some strong research to support Tether being used to manipulate Bitcoin's value, and Tether's site was recently updated to clarify that it is in fact not backed 100% by USD.

Both of these are top 10 crypto assets...

Not to mention MtGox manipulated prices during the 2013 bullrun.

You're trying to paint this as some sort of hypothetical scenario when this has already happened! And will happen again if this "rainbow network" gains any significant traction.

tldr; does this add a breakthrough or is it just a variation of a consensus + payment channels solution?

Reply on news.ycombinator.com