Todd has been giving me the creeps for a while now - the whole Inscriptions mess is very fishy, and although its an opaque sense of unease, i'm leaning toward the impression actors like Todd, and 'security budget' Stzorc, if not actively colluding with mining entities, maybe at least leveraging the inscriptions problem to pursue their interests.
Miner collusion with hostile forkers in the past is what's probably triggering the unease in me now, but i generally listen to my instincts when they nag me the way they have been recently...
spoiler: as a node operator, i wont accept an inflation fork
TLDR; looks like the ordinal mints are coordinated to stuff blocks with high fee transactions, and it could easily be profitable to do so.
Author suggests fee burning and advocated for Todd's inflation fork:
this solution could effectively complement arguments made by Peter Todd and others regarding the future of block subsidy, which in short go along the lines that block subsidy halvings should be stopped at some point with a hard fork, leaving a perpetual fixed subsidy per block. By itself this would arguably make Bitcoin into an inflationary currency, which many people object to, but if you combine it with partial fee burning, it could very well become deflationary instead depending on how the fee market develops, while still providing a guaranteed minimum reward per block. This would effectively alleviate the danger of a deficient fee market compromising the security of the blockchain due to low miner rewards at some point in the future, while only adding an "about" to the statement "there will only ever be 21 million coins".
What about doing something like: half the fees will distributed as block rewards later. Something like half the fees in block B will be split in K portions (eg., 10) and paid out to the miners of blocks B+9996,B+9997,...B+10000,...,B+10005?
:-) Has a scheme like this been discussed anywhere?
I mean, I am pretty much in the camp of: Bitcoins most important attributes are: (1) Its immutability, and (2) scarcity, (3) decentralization, and (4) censorship-resistance.
So, I would be very very very cautious about changing anything on the incentive structures. However, there was some research that when block-subsidy is very low, there could be re-ordering attacks of high-fee transactions etc. Moving some of the fees to some later blocks would keep block-subsidy high without touching the sacred 21 million limit.
I am just curious if someone has thought about this longer than me :)
I agree that we need to be very very cautious about changing anything with regard to the incentive structure.
That being said, it does feel like current incentives are not as aligned as they could be. I have not yet read the specific post that you linked on bitcoin-dev, but here is one of my favorite (soft-fork) ideas which might better align incentives:
example soft-fork w/ burning: require all transactions over a certain size (including witness) to include an op_return output which burns a certain number of sats/byte
Burning sats is not a very popular idea. If need be, we could probably replace burning with a suitably long timelocked anyone_can_spend output, something like 10 or 15 years would be my preference. Then it is almost like burning, but the sats still are available to be (re)mined in the future.
This mechanism still has some vulnerabilities (for example, people would probably just chop their larger transactions into smaller transactions to get under the limit, so more research is certainly necessary), but the idea of burning sats is at least interesting to me because it fairly (proportionally and instantly) rewards all remaining hodlers -- and the node operators we care most about and want to protect are the hodlers. These are the people who are in bitcoin for the long run.
At least then when your node is verifying all these large transactions, every time you get to the op_return output, you get a little bit richer and, by definition, the creator of that offending transaction gets a little bit poorer.
anyone_can_spendoutput, something like 10 or 15 years would be my preference. Then it is almost like burning, but the sats still are available to be (re)mined in the future.