Topic: [Megathread] The long-known PoW vs. PoS debate - page 3. (Read 3828 times)

Of course, because millisatoshis not divisible by 1000 are enforced on-chain, right?

I can say the same about my idea, just cut this "it’s not even a chain" part, and you can copy-paste it 1:1 into my proposal.

I don't need any new coins. Each validator can take fees, in the same way as LN nodes take fees.

They always can. There is no way to stop them, they can put their coins in and out as they will, it is just a matter of signing some transaction.

The problem of creating new features without any forks, of course. People have a lot of ideas, and create a lot of altcoins to reach them. There is no reason to create any new coins out of thin air, if they could sign their bitcoins instead.

And people accept that by signing their own coins. They accept to gain and lose coins, by signing messages, and there is no way to stop anyone from signing any message they want. That's the best part of this "unstoppability", that there are some ideas, where even Proof of Work enthusiasts cannot destroy easily. That's why if you try to stop my proposal, you may hurt Lightning Network at the same time. Then, you have to choose wisely, to not kill too many networks with one shot.

I still cannot see any difference. Each Lightning participant decide, how to make a closing transaction. The same with sidechains, the act of moving coins on-chain simply means, that everyone agreed to destroy sidechain coins, and take them on the mainchain.

Simple, in LN you cannot transact with all participants directly, in sidechains you can. If you are completely new, with fresh keys, you cannot be introduced into LN, without touching on-chain coins. In sidechains, it is possible.

That’s making it too easy.

Lightning doesn’t create coins out of thin air, it just takes routing fees, and if there’s a direct channel then not even that. There’s no peg to be broken here. It’s just a trust minimised way to transfer Bitcoin off-chain and hasn’t much to do with a consensus mechanism, it’s not even a chain. A huge difference. It also delivers another indicator that speaks against block demand going too low. Because it will always be smart to open and close channels when demand is low, making transaction revenue more stable and distributed more evenly on the main chain. Kinda like an options contract.

Your argument was that a sidechain could reach higher transaction fees(in monetary value) than the mainchain. And thus incentive miners to do less PoW. Now isnt there a problem?

1. Your sidechain will need a fixed exchange rate, i put 1 Bitcoin in, i can get 1 Bitcoin out.
2. You said you want PoS, now you have the option to create new coins as a staking reward and/or just distribute the transactions fees to stakers.

Now there’s a problem, Bitcoin is fixed supply, so you can’t just create  new coins as a staking reward out of thin air on the side chain, without breaking the exchange rate, because then the side chain has more Bitcoin in circulation, than can be redeemed or actually were put in. If the peg breaks it will make the Bitcoin of the side chain actually worth less, and in monetary value the main chain still pays more in fees for miners.

If you don’t create coins to reward staking, then if the transaction fees are higher on the side chain, why wouldn’t people just use the main chain again? Then miners get less fees on the PoS chain again, because people are disincentivized to even use it. So what problem was fixed in the first place?

In the a peg case, it’s always dependent on the reserves of Bitcoin that can be redeemed. Total value of the sidechain = amount of Bitcoin that can be redeemed. No matter how you change the units. The only thing the PoS does in this case is moving the relative wealth to the top holders, meaning they can redeem more Bitcoin over time, while everyone else can less. How isn’t this vastly different from Lightning. This is Fiat 2.0. Fixed exchange rates always have this problem and idk how you will solve this.

Because it’s different to get paid a small amount for a service in an competitive market, than to use an potentially broken consensus mechanism that favors someone for just holding funds and gets centralized over time. Because lightning doesn’t break sane economic principles and strengthens Bitcoin.

There’s definitely potential for sidechains to take load off the main chain, if fees get too high there, i just don’t think they can disincentivize mining and im still not sure how the basic problems of PoS will be solved on there. And then there’s still the question how will they do it better than lightning? The infrastructure lightning is building at the moment is impressive and it complements Bitcoin really well.

Of course I don't have all the answers yet. But I know the basic layer is technically possible to build, and that can be improved later, because a single on-chain transaction could move coins from one sidechain to another, so making improvements will be quite easy.

I think it should be demonstrated in practice. Imagine talking about Lightning Network without having a working system. It is hard to convince anyone it could work. So, now I am trying to write some code, and build a basic layer of what I already described. Then, I could run that as a some kind of test network, where you can easily get some test coins, just by signing some mainnet coins (that could be also done as a commitment, to make it during sending some regular transaction, so nothing more than tweaking a signature will be needed, and it would cost zero additional on-chain bytes, and one tweaked signature could commit to the hash of the block, to cover the whole test chain at one shot). Another thing is that if it will ever reach some popularity, then it could be possible to add commitments inside the coinbase transaction, just by tweaking some miner's Taproot address, that would also require zero additional bytes, and would allow reaching sidechain finality every 10 minutes, without increasing mainchain size at all. Imagine Bitcoin as a huge clock that would confirm all sidechains every 10 minutes, without any additional bytes, the whole burden of commitments will be taken by each sidechain separately.

1) Merged Mining done wrong: they should always follow the highest source of the Proof of Work, and grant coins, accordingly to that. Having 10% of Bitcoin's hashrate should allow reaching 10% of NameCoin's hashrate, there is no need to create a separate difficulty, each chain should follow the heaviest Proof of Work.
2) No coins should be created out of thin air, they should be based on Bitcoin, may be pegged in 1:1, may be signed and cloned in this way, there were many possible solutions to that.
3) "The spent fee goes to the miners in the next block fee" - burning 0.01 NMC to register a domain is pointless, it can be done safely without it, and handled in the form of fees. Also, by throwing coins out of circulation, there is a hard limit of available domains, which is not needed.
4) Names can be attached to Bitcoin signatures or addresses, and revealed later, when needed. There is no need to explicitly include hashes.
5) No "modernised arrangement with the Merkle Tree on top" was created, no commitment-like style of revealing names was added. It should be based on sorted commitments, that should be timelocked, and then revealed in a given timeframe, to get it recorded, and committed into NameCoin, which should commit into any Bitcoin transaction every sometimes (preferably once per 10 minutes, if sidechains would get some traction, and if they would be supported by miners, and not only by users, as it is possible today).
6) "it's cryptographically possible to make a risk free trade", and we have 2022, and it is still on some TODO lists, but not yet implemented. Also, NameCoin developers didn't even try to reach that goal.
7) "A longer interval than 10 minutes would be appropriate for BitDNS" - and they picked of course 10 minutes for no reason. I thought about increasing block time on some networks, into two weeks, maybe even into one halving period. There are some nice properties, when the block time is increased (for example, Proof of Work reduction is then possible, without breaking any consensus rules, that are adjusted to faster block times, so the difficulty is then forced to drop). On some networks, it can be also decreased, like P2Pool did with their 30 second blocks, this method can be also used to do some Proof of Work fragmentation and decentralize mining, or give some people a choice between staking and mining, so they can find a balance between that, for example they could go into 90% mining and 10% staking, it is a good start, and could go further later, when people will be more convinced, that it works.
8­) "In that case, domain objects (domaincoins?) could represent the right to own a domain for a year" - guess what, of course it is not "one year".
9) "You might consider a certain amount of work to generate a domain, instead of a fixed total circulation" - was it considered to utilize something like "vanity address" to have a "vanity name", where people will mine a name that will be unique, will have some matching characters, and will be cryptographically secure at the same time? Do people have a choice between paying for some characters and mining other parts? Of course not. It is not possible to mine "bitcoin" as a vanity string, and pay for "eater, don't send" to own that name. No, you have to claim the whole "bitcoin eater, don't send" part, you cannot mine names, no matter that vanity addresses are better, and they even not require a blockchain! What about mining vanity names and supporting Bitcoin Proof of Work at the same time? They didn't even start thinking about that!
10) This list is longer, but I think it will be too much offtopic. Maybe this reply should be moved to some NameCoin-related topic, feel free to do that if needed, but also note that in all of those points, you can put "Proof of Stake" instead of "Proof of Work", and it will still make sense, so you can also discuss it here.

Edit: "But doesn't the Lightning Network need a peg? If it mirrors Bitcoin 1:1000 then the transaction reward can't be higher than on Bitcoin, and you can't create new coins without breaking the peg, or where will the new Bitcoin be coming from?"

See? If you have some questions about sidechains, you can try putting "Lightning Network" in the same place, just as a thought experiment. Many times, you will see that it is quite similar, and many times you will see it is also similar to Proof of Stake, if you do the same trick, and if you notice, that there is no mining inside LN. I still wonder, why some people accept LN, not accept sidechains, and not accept staking. Those three concepts are quite similar, if you think more about it. And of course the answer is that fees will go to other nodes, as it is in LN, so "the fee", calculated as the sum of all outputs, minus the sum of all inputs, will cover only bare, minimal, on-chain requirements, the rest will flow inside channels, and will be signed in that way or another, by LN-validators, by sidechain-validators, or by some other validators.

Another part of the answer is that there is no "the peg", but "a peg", so if some mainchain is making some changes, then each sidechain can decide, where it should be pegged, because all coins up to the point of the split are identical. Also, even if the mainchain is splitted into BCH, BSV, or anything else, the sidechain can decide to not split, and to merge coins, instead of doing replay protection. Many rules can be altered, everything is always in users' hands.

But doesn’t the sidechain need a peg? If it mirrors Bitcoin 1:1 then the transaction reward can’t be higher than on Bitcoin, and you can’t create new coins without breaking the peg, or where will the new Bitcoin be coming from? And if you start to break the peg, you’re creating an inflationary spiral that makes the sidechain more and more worthless, and not even fully redeemable, because this many coins don’t exist on Bitcoin.

Creating new coins forever also diminishes the monetary qualities of a good(worse store value), and it won’t be able to compete with a chain that had 100 years of adoption already and has higher monetary qualities, this is the problem i see.

It’s just too much based on the assumption that Bitcoin needs to give out free money forever to survive, but i think the incentives speak against this.

From what im seeing is that the fee market can actually work out. That there’s actually potential to bring the variable cost of mining(electricity) closer to 0, because the more the block reward diminishes, the more miners are forced to use otherwise unused energy(most of the worlds energy). If the variable cost of miners is close to 0, then there will always be a point where the transaction fees are enough to break even, after a certain time.

This is a real problem in the energy sector, there’s countries like Germany that pay other countries to use their energy overproduction, there’s potential here for example to actually make a profit instead of paying others to use your electricity.

But this is just one example, i think the block subsidy will actually help build the energy sector to get more and more efficient, and then have a buyer for unused energy forever. To me this is the most genius subsidy to build a renewable energy sector ever made, and energy producers won’t be solely reliant on mining to be profitable, because it will always be more profitable to sell the energy to the grid, except when the grid doesn’t need it, and then it’s more profitable to use the energy on mining than to waste it. A perfect complement to the peak energy demand problem.

This sounds like you don't have all the answers yet. I'm aware that Proof-of-Stake is possible to implement in a sidechain, but the devil is hidden in the details. Anyway, if you've thought of a Proof-of-Stake-like idea that works better, I hope you make it real. I'm really interested to see how it'll work. Otherwise, you can try to debunk the disadvantages of Proof-of-Stake we've all remarked.

Which were?

Just by producing blocks every 40 minutes. If any difficulty reduction will be needed, then soft-fork is more than enough to cover it. But I think moving coins off-chain will be enough to decrease on-chain fees, and to decrease on-chain rewards, which will encourage miners to turn off their machines.

By making signatures. Instead of signing a regular transaction, you sign something else. For example, in signet, if you want to sign a block header, you make two transactions, there is "to_spend" and "to_sign" transaction, you sign this "to_sign" part, and place your signature in the coinbase transaction.

I thought about 1:1 peg (of course 1:1000 or other values are possible, but then it would mean just moving the comma, so for example using millisatoshis, instead of satoshis, as in LN).

As every sidechain, it has to reach "a peg", not "the peg". As long as all coins are covered by the chain they are pegged into, it works.

It is in progress, I just decided to mention that it is ongoing, but I could patiently and silently work on it without informing anyone, and then suddenly reveal everything at once. So far, I shared parts, where there is very little chance that someone will destroy that easily (also, Core developers already discussed some parts of it, for example "temporary forks", so they are aware of what is possible, they may not be aware only about that "no fork required" part, but I think they should be, because this was also posted (not by me) to the mailing list). But there are other parts, that needs to be hardened, they are too fragile to be shared with Proof of Work enthusiasts, because then they will abuse the system by making chains, and pretending that "this idea was tested, and it failed", which is not the case. In the same way NameCoin was presented as a BitDNS implementation, while it is clearly not the case, as it lacks some features (described by Satoshi).

I can share unstoppable ideas, that are hard to destroy, and I already shared that, the only thing left is answering some questions, and clarifying, what is possible. But still, I don't want to share more fragile parts, because I don't want to repeat NameCoin's mistakes.

If the block subsidy becomes lower than the reward from fees, then the basic reward will come from fees. Currently, the miners earn a ~0.1 BTC from fees on average. If we assume there'll be greater adoption in the future, it's highly likely for this to rise. In any case, a difficulty of 1 is, by no chance, justified.

The median fee will definitely decrease if most users transact off-chain, but there will always be on-chain fees.

How can you increase the block time from 10 minutes to 40 without a hard fork?

That was not my point. If no consensus is required, then, naturally, there's no agreement required either. All I meant was to make us comprehend your idea. The way you describe it is vague and foggy. (And that's why you probably haven't got the necessary attention)

The basic idea is very basic and vague, as I said. All you've said is that it's possible to create a separate chain wherein users can mint altcoin units by providing a valid bitcoin transaction. How is Proof-of-Stake implemented exactly? If the altcoin is inflated, how can you return to the main chain that is not Proof-of-Stake based? If there's a direct correlation between bitcoin and the altcoin, how isn't it completely dependent on Proof-of-Work?

Make a thread, explain your mechanism in a clear and detailing way. I'm sure there will be interest if you've thought of something that's significantly utilized somehow.

Why? The coin supply is finite, sooner or later we will reach the case, where the basic block reward will be zero. And then, users could fully decide, if they want to move their coins in a Proof of Work network, or if they want to get their transactions confirmed by Proof of Stake validators. Then, it may be pointless to mine zero satoshis on-chain by using Proof of Work. On-chain fees may be lower than off-chain Proof of Stake fees. What then?

Lightning is similar to staking, but it lacks some features, like direct transfers between all Lightning participants. You still need to touch on-chain, you cannot send coins directly in Lightning, and finalize it later on-chain, you have to finalize that on-chain immediately, no matter what, because each channel should be in the right state to move any coins anywhere.

Well, sidechains at least have some independence, and only finalize things on-chain. But Lightning is fully pegged to the main chain, and that's why all problems from the main chain are always moved downwards to LN (for example, if on-chain fees are high, then there are problems with creating and closing channels).

If the basic block reward is zero, and on-chain fees are very low, then this low difficulty is justified. It just reflects on-chain rewards. And they will be lower if the whole system will move into LN, staking, or other layers.

If you increase the block time from 10 minutes to 40 minutes, then without any fork, the difficulty will drop to the minimal value, and will stay there.

No agreement is needed, if there is no hard-fork, and no soft-fork needed. It is just about users willing to move coins as they want, and sign transactions as they want. They don't even need to move on-chain coins, they can create coins by signing something, then they will get their coins deposited on a separate chain. Later, they can move coins on-chain, then they will be destroyed on the sidechain. In this way, it is possible to create two-way-peg, where nobody can stop it. Then, it is possible to turn chains on and off, it depends only on users, they will have freedom they never had before. And the whole work is about making the right scripts, so that on-chain coins will move if (and only if) all participants will agree on that, and that will be needed only if someone will want to leave the sidechain.

The basic idea is already shared, the original author thinks about utilizing it in the Proof of Work context, but I think it could be used as a Proof of Stake too. Many things are ongoing, because, as you can easily guess, Bitcoin community is mainly against staking, so to make it real, there is a need to allow the minority to use their coins as they want, and to gain and lose them, by testing their ideas in practice. If I would share my ideas in the current stage, they would be destroyed by the Proof of Work community. They should be hardened, and more work is needed to make some unstoppable proposal, but the main idea of unstoppable sidechains (without any fork) is very promising.

If the Proof-of-Stake concept is to get rewarded with coins for proving you own coins, then you can already achieve this in a modern-banking way; lend your coins with an interest, or invest in Lightning's capacity. Unless your no-fork idea includes sidechains?

"Good enough", how? If you have such difficulty it means there's very little work done. And if this internal-Proof-of-Stake system is dependent on Proof-of-Work, there's very little security.

How's block time relevant?

Here's a friendly suggestion: Let's talk about it. Don't implement theory if you haven't discussed it with others. There might be a "hole" you haven't thought of. You've only created a thread; make a paper, let us comprehend it.

I have a question about this. Let’s say miners really mine less and the difficulty drops, wouldn’t it weaken their competitive edge?

Now new miners can come in easily, forcing them to mine more again. So how does the incentive really play out in a competitive environment? Miners can’t afford to mine less, if they have running costs, new people could come in and just outmine them. „Kicking“ the ones that decided to mine less out. Idk if a PoS can outincentivize PoW like this.

Aren’t you basically handing over the biggest Bitcoin whales, the control over your network for free?

And in case just newly mined coins can be staked, then there would be no incentive for non miners to use this network. And in case miners mine less on Bitcoin to use this network more, they would be „kicked“ out by new miners, because they became uncompetitive, so they can’t be staking in this network anymore, if they don’t find new blocks.

I think this concept doesn’t play out, or im missing something.

Note that people can be rewarded in this Proof of Stake network directly, so miners could find it more profitable to mine less on Proof of Work, and validate more on Proof of Stake. It is a matter of incentive, dropping the difficulty from the current level to the bare minimum 2^32 hashes is technically possible.

Why not? Using the difficulty equal to one will be "good enough". Going even lower is also possible, because it is possible to increase block time, for example from 10 minutes to 1 hour or longer.

I do. And I have plans with staking, be ready for them.

Isn't this Merged mining Proof of Stake?

It is definitely backed by bitcoin, and since bitcoin relies on PoW, it's indirectly backed by PoW. Not something "environmental evangelists" like LegendaryK are going to support of.

So... Why don't you buy bitcoin?

Here's another shower thought. What if a developer builds, and bootstraps a new cryptocurrency network based on POS, and if you want to join the network as a validator, you can buy the tokens, BUT, only through Bitcoin. Does that make the new POS cryptocurrency network backed by Bitcoin's POW? Because, Bitcoin needs to be mined, to buy the POS token.

The truth is that Proof of Work is a True/Proper Consensus Mechanism when it comes to Trustless Cryptocurrency which requires Proof that a Job has been done before dishing out rewards. If you view things this way you would likely understand that it's only a based meaning of the more complicated Bitcoin PoW.
Proof of Stake seems to require Money as Proof that a Job has been done. This is a serious deviation from the real purpose of using proof to show that work has been done to achieve consensus in a Trustless environment.

Nope, I am still alive and well, still trying to put my idea into practice: https://bitcointalksearch.org/topic/merged-mining-proof-of-stake-5390027

Considered that point but what about to debate next to it thats the only truth that 80% of community with Long term Speculation stand with POW and I think newones with new investment are more interested in POS thats the reason LUNA, SoL, ADA got so many users and their Destination we saw clear to the Heaven. But POW is still here competitive in market not only competitive with title of King 👑.

I hope the reason they left the discussion is because they were educated and understood why PoW > PoS! That's what debates are usually for.

Then there goes the debate, POW will always be > POS. Has anyone noticed the POS advocates/promoters of the topic were nowhere to be seen when the point was being made after this post? https://bitcointalksearch.org/topic/m.60314887

They are usually very aggressive in convincing the newbies that "POW = wasteful" without truly understanding Bitcoin's underlying nature which makes POW superior. But we could let it go as chosen trade-offs.

It wasn't that young, but the principle applies to everything that even includes bitcoin because if you check, a lot of bitcoins are stored on exchanges by traders and newbies. The situation is even worse for altcoins because we don't have as many altcoin exchanges as we have bitcoin. There is essentially only Binance with majority of the volume. It is even worse for shitcoins like Ethereum because they have massive premines (72 million in this case) that is controlled by one entity.

It makes no difference; it's a general, inherent problem of PoS. It's just as possible on the 'latest and greatest generation X' PoS blockchain, as it is on a 'young / nascent' or a 'developed / mature' PoS blockchain.
The very principle of PoS that the voting power is determined by the amount of coins you hold is the issue at hand.

It's one option; or they might buy it off-chain (buying keys / lending the coins / keys). You can't notice that on-chain and it won't cause price fluctuations.