Atomic swap tokens across chains

[benjaminbollen]

Research the existing technologies, this has successfully been covered; then see what and how we integrate it into mosaic. This may need to become its own repository, but for now tracking progress here.

Candidates:

• Interledger
• Uniswap - single chain
• Kyber
• Raiden
• AirSwap (?)
• 0x (?)
  or others.

[0xsarvesh]

Kyber swap only supports single chain ERC20 swaps.
https://github.com/KyberNetwork/KyberSwap
https://medium.com/kyberswap/meet-the-new-kyberswap-3d36b6685e1

[hammadtq]

Kyber is doing interesting work with Wrapped Bitcoin (WBTC) and ETH-EOS Waterloo. They have also started doing the same with ETH-ZIL bridge. Majority of this work involves a custodian that locks the tokens on one chain and releases the equivalent amount on the other layer-1 chain. Nothing fancy but this the only approach I have seen that is actually practically implemented and released for the general public to use.

Other than that, Summa has done interesting work with BTC-ETH atomic swaps (MITBitcoinExpo 2019 talk) and Arwen went from another approach of using existing centralized exchanges Market Makers to solve liquidity problem while performing a swap (MITBitcoinExpo 2019 talk).

Another interesting pitch is of Clearmatics - these guys I bumped into at ETHDenver. You can read about their approach here. I was doing a ETH-ZIL cross-chain project at the hackathon, couldn't comprehend the gas costs associated with parsing a whole ZIL block on ETH main chain to find a certain event.

Polkadot and Cosmos are still focused on building and launching their mainnets. Cosmos has only started doing IBC (Inter Blockchain Communication) specs after the mainnet launch. The first meetup video can be seen here. To get included into the IBC bi-weekly meetups, email Jack Zampolin. They are writing these specs by collaborating with agoric, however, I believe these specs are related to Comsos's own BFT blockchains and not external blockchains that use PoW or other consensus protocols.

Cosmos, however, had an earlier implementation for a ETH bridge. It's called peggy, I believe the project was largely abandoned because of the gas costs associated with parsing a block on-chain.