Take the static reward out of the coinstake and put it in the coinbase. The coinstake reward would now be only the coinage based portion. A block would now contain as many valid coinstakes as it can find, spending all of them and giving them their coinage reward. The coinbase static portion would be given to the minter that submits the block, and it can be a function of the number of included orphans. To keep incentives proper, each included orphaned coinstake beyond the original valid coinstake must contribute less than 1 additional unit of static reward, to incentivize creating multiple blocks as opposed to condensing one’s own stakes.
If using coinbase isnt technically possible, it can just be a separate txn with no input. This reveals the only downside I can think of for this approach: it will create a small static-reward utxo on each block which will pretty much by definition be too small to mint. We can do some fancy footwork with the combine function on future mint events to rectify this without proliferating small utxos. I believe the simplicity of this orphan inclusion proposal is worth it.
This proposal is beyond my understanding, but I have a question on the above line. My understanding is that the static reward has the effect of tapering the reward, so that as the number of minters rises, the reward per minter declines slightly, but the reward increases per minter when minters fall in number, thus acting as an incentive.
If the proposal splits the static and coinage-based rewards, would that affect the tapering effect?
The current system is merely a coinage-based percentage plus a portion that is globally fixed to the total coin supply. If there are more minters, the pos difficulty is higher, and so you find less blocks, thereby reducing your portion of the fixed, or static, part of the reward simply by virtue of finding less blocks. This proposal on its face does not need to adjust that amount at all, except by virtue of how we incentivize orphan inclusion. There is no technical requirement for the inflation rate to fundamentally change besides devising this basic incentive, which really can just be any number greater than 0. Incentivizing orphan inclusion introduces some uncertainty, but we could easily decrease the normal static reward to compensate for the orphan inclusion incentive to make approximately the same inflation rate. No matter what, we are talking about a portion of the inflation that only equates to 0.25% and tweaking that, so it would likely be negligible no matter what we choose to do there.