openethereum / block-reward Goto Github PK

I'm trying to understand the block reward contract for use in a private PoA network. What I basically want is to give a validator 1 Ether for each block they produce. Is it correct that I just have to change this line:
https://github.com/parity-contracts/block-reward/blob/3bcc898dbf7e9ca7491a797c22d2f1b65903bd46/contracts/ExampleBlockReward.sol#L48
to 1000000000000000000 Wei (1 Ether)?

edit: Yes, it does.

The constructor within ExampleBlockReward.sol, where it takes in the "systemaddress" rather than directly using the commented out code

constructor(address _systemAddress) public { /* systemAddress = 0xffffFFFfFFffffffffffffffFfFFFfffFFFfFFfE; */ systemAddress = _systemAddress; }

This will fail on a production environment with the equivalent of:

            "blockRewardContractAddress": "0x000000000000000000000000000000073410f33D",
            "blockRewardContractTransition": "0x0",

(with the bytecode being set within the "accounts" subsection & associating the bytecode with the 0x....000073410f33D" address)

as the parameter that is passed through will be incorrect, and indeed even msg.sender in the constructor will return "0x000000000000000000000000000000073410f33D", which is not the system address that will be required to be checked in subsequent reward() calls (which does indeed come from 0xffffFFFfFFffffffffffffffFfFFFfffFFFfFFfE).

My suggestion is to specifically rename this file to a "RopstenExampleBlockReward" or the equivalent, with the "ExampleBlockReward" having the constructor parameter removed & the system address association uncommented.

I should note that this is more than likely an issue that wasn't spotted on Ropsten due to genesis.json configuration - in my tests the Ropsten ExampleBlockReward.sol (with those constructor parameters) will not complain if the bytecode is placed in a blockRewardContractCode vs the blockRewardContractAddress parameter (probably due to the call type difference).