Use UpgradableProxy? Some thoughts on settings, overengineering and its consequences

[kleros]

Use UpgradableProxy? Some thoughts on settings, overengineering and its consequences

[greenlucid]

CHALLENGE_WINDOW ended up being important for the UX. Just like Uniswap (by itself, or wrappers) passes safeguard arguments to revert in case something goes wrong, here you need to pass the editionTimestamp in case the item gets accidentally (or, purposely) edited on the lapse of time between choosing the challenge, and the tx getting included.

If you are giving the challenger leeway to choose past editions, then they could choose to challenge old editions that don't matter anymore, just to obtain rewards. So you need to limit it somehow, like with this CHALLENGE_WINDOW.

Another topic I reflected about, why 4h for commit timeout? Isn't that excessive? 1h should be enough time to reveal under regular circumstances, and it would allow for either BURN_RATE or MIN_CHALLENGER_STAKE_RATIO to decrease significantly.

In that case, you could go with these settings and still get sniper to burn 10% stake for every item and day they perform sniping attacks:

uint16 public constant MIN_CHALLENGER_STAKE_RATIO = 2_083; // 20.83%
uint16 public constant BURN_RATE = 200; // 2%
uint32 public constant MAX_TIME_FOR_REVEAL = 1 hours;