Ratel is an experimental project for the purpose of research work.
The current focus is to extend an existing smart contract programming
language such that it can support the writing of MPC (multi-party computation)
programs. Roughly speaking, ratel's current goal is simply to be a kind of
hybrid language that would allow a programmer to write the smart contract
part and MPC part in one language, one file.
When compiling a ratel program, the output would be in two main parts:
- smart contract ABI and bytecode
- MPC program ABI and bytecode
The smart contract ABI and bytecode can be used to deploy the contract onto the targeted blockchain and to interact with it.
The MPC program ABI and bytecode is meant to be passed to an MPC application that would execute the MPC program.
For the near term, the MPC program ast or source code can be parsed by an MPC application which will then execute the corresponding program.
Eventually, assuming that an MPC framework would have a virtual machine, then the abi & bytecode would be passed to that virtual machine.
In terms of implementation, the current goal is to extend the Ethereum's
vyper smart contract language so that it can support MPC program function
definitions.
It's not clear how exactly this will be done but one approach that is
currently considered involves modifying vyper's compilation phase such
that vyper's compiler produces the usual expected vyper output, and in
addition the mpc's ABI and bytecode or for now just the source code or ast.
The total output can be then passed to a HoneyBadgerMPC application that will
consume the output such that the smart contract ABI and bytecode will be used
to interact with the contract on Ethereum meanwhile the MPC program ABI and
bytecode (or source code / ast for now) will be used to execute the MPC program.