It is not an invalid block to only issue justifications.
Nash believes: Whether or not you get traffic should be a matter of luck. All validators in a shard should be equal.
Greg believes: Validators should be rewarded for bringing in user traffic.
In essence, we want to pay for value. User transactions or finalization.
Do we look at the justification tree? Can we constrain the DAG.
Come up with a set of properties that we do not want in the DAG & make those slashable
Any paths in the justification tree have to come back to the main path.
Implement configurable 'knobs' that allows us to adjust as the attacks come in. Attackers will know what the knobs are. Makes this a very hard problem.
Greg's idea: Validators place bets. The bets are multipliers, and if the multipliers have a shape The gossip about the bets acts as an iterative function.
Transaction fees
Monetary Expansion
Action items
Lucius Meredith: Provide his research for creating an algorithm that will lay down the shape of the DAG, which can be used to form the basis for rewards and punishments.
Doesn't mean that what it outputs will fit with our intuitions for rewards and punishments.