Proof of Late Work (PLW): addressing the problem of trusted intermediaries in layer 2 solutions
TLDR: Miners on the Bitcoin base layer who build a valid block, which is not included in the blockchain, are ipso facto authorized to be a layer 2 intermediary (server, validator, facilitator, etc). This addresses the problem of choosing who the intermediaries will be, not how the system operates. This is not a layer 2 proposal; it is agnostic, subject to compatibility, of which I am ignorant. This could potentially be used with any type of side-, state-, or other L2 chain, or possibly with a channel system.
–
Has there been a proposal to somehow utilize the “wasted” energy in mining? I’m not being critical of the mining algorithm itself. I’m referring to the mining that is done by the non-winning miners, those who are “runners up” in creating a block. A very small amount of bitcoin mining hashrate creates a single block; the vast majority of the mining either doesn’t create a block, or doesn’t create a block as quickly as the winning block. Or it is created at the same time, followed by a “playoff" (so to speak), after which the loser is deemed " stale “. Most miners will not have created a block, will accept their competitor’s block, and restart with a fresh mempool, energy wasted.
What if there could be a use for the work done attempting to create a block, or the blocks which are abandoned on stale forks (essentially wasted energy). Imagine being able to harness some of that Proof of Work to benefit the system.
Could the miners be rewarded in another way, for having done nearly as much work as the winning miner? I’m going to call this the “Second Place Miner” (“2PM”) because I don’t know of a proper term for it, and I think that is intuitive to understand.
Could the 2PM, for example, continue building their block after being notified that a block has already been made? They could then complete their block (seconds or minutes later), which will be valid but not part of the blockchain. They then use proof of building this block as proof of work (“late” work), giving them authority to be a second-layer agent (hereafter referred to as “Layer Two Server” or “L2S”, but can refer to operator, validator, facilitator, sequencer, etc).
So the 2PM (Second Place Miner) becomes L2S (Layer 2 Server) on an L2 (Layer 2).
This solves the problem of “just anyone” (permissionless) versus “federation member” or individual entity (permissionful) being able to be the L2S. This way, actual Proof of Work is done, raising the cost of attack to the L2 and doing work on the main chain at the same time. Note again that the block created is not added to the blockchain– its transactions are “invalidated” (so to speak) and are of course returned to the mempool.
This method could be blended with a federation, where the number of 2PM federation members is set by the percentage of malicious federation members required to steal funds. For example, if a federation setup requires 33%+1 malicious members to execute an attack, then 67% of the federation must be 2PM.
Since there are “disputes” (multiple versions of the blockchain when two blocks are created at about the same time) in which the network isn’t 100% decided which chain wins (the “playoff” event), it may be necessary to wait for this to resolve. Therefore we use n-blocks-ago blocks. Say current blockheight B, anyone who has made a valid block from the mempool at B-n, which is not part of the valid Bitcoin blockchain (a “late” block), is given L2S status. There could be a dynamic time window, depending on the need in the L2. So maybe any valid block from B-n up to time duration T.
The block winner (the block added to the Bitcoin blockchain) does not receive L2S status on the L2. This is so that an adversary does not have incentive to attack L2 and finance it with block rewards. They can't both win block reward AND gain a position to attack the L2.
One question is the “how" of implementing this without a central authority “registering" or ”authorizing". I'm imagining the miner being able to use their created block containing their public key as well as their private key to give themselves permission, or a kind of proof of identity which can be affirmed by other servers in the L2 and ultimately the users. If they don't have this identity, no one will use them. The timestamp of the original block they mined to qualify could also be used to set a time limit on their authority. This would also allow for bootstrapping, as the initial L2Ss would “expire” as 2PMs earn L2S status.
In an ideal world, incentives on the L2 would be matched to Bitcoin ideals, probably with dynamic fees (user-selected fees are probably not appropriate, to optimize speed and UX). If fees go up, more miners are incentivized to attempt to become an L2S, and the PLW system could allow more L2Ss by expanding the time window in which PLW blocks are accepted.
Incentive changes:
-> more people would mine, because there is more profit (NO increase in mining reward! Just the possibility of fees on the L2)
-> miners may not immediately abandon their attempts at a block when another miner finds a block, thus lowering the number of miners working on the newest version of the mempool.
-> adversary attempting to mine BTC “just” to get the L2S status would need nearly as much hashing power as the block winner. Attempting to build up an attack on the L2 REQUIRES doing the work of the main chain, strengthening the Bitcoin network (or being very costly).
Drawbacks:
- the “winners” who become L2S on the L2 are most likely to be those who have enough resources to build blocks on the main chain. Wealth begets wealth, which leads to centralization. The claim that “anyone" can be a validator is only superficially true; nonetheless, it is arguably more egalitarian and decentralized than a federation or centralized authority.
- depending on implementation, a market potentially emerges to buy and sell L2S status. An adversary to the L2 could buy their way in without mining. However, this would reward miners, and still increase the cost of attacking the L2.
- potential for L2S status to be tokenized and susceptible to shitcoinomics.
- it is not clear that miners are correctly incentivized under all conditions. Expansion of the L2 may reduce on-chain transactions, lowering fees for miners (but if it expands adoption, then fees go back up).
Alternate name:
PLEBS : Proof of Losers Expanding Bitcoin Scaling
I'm just a monkey with a keyboard; surely this contains glaring problems in incentives and mechanism, and makes assumptions to JuSt UsE cRyPtOgRaPhY. It was probably brought up and dumped for OBVIOUS reasons ten years ago.
So go ahead and let me have it.