This thread is dedicated to all the back-and-forth that’s happened around these two consensus mechanisms. As a fan of Proof of Work, I want to share my thoughts on why I believe Proof of Stake falls short, especially in terms of centralization and the ability to reach consensus. Those are crucial for a type of currency that aims to be innovative.
But before diving deeper, it’s vital we clarify what we mean by centralization. Simply put:
Centralization is often a fuzzy concept and is frequently misinterpreted, whether intentionally or not, to serve someone's agenda. For instance, someone might argue that Bitcoin is less centralized than an altcoin because its hash rate is more evenly distributed. But then again, you could counter that it’s more centralized due to a more uneven wealth distribution (think no ICOs and all). Then there’s the other side that points out how a handful of mining pools control most of the hash rate and claim that makes it centralized; they might also argue that centralization can’t be pinned down quantitatively, etc.
So we ought to lay out what we mean by centralization. My definition would go like this:
When something is more centralized, it means that a small group can exert a greater influence over it. You find a direct correlation there. That small group could be developers or just regular users.
Proof of Work
How it works in theory: (Think lottery tickets here)
Anyone can create lottery tickets. Scratching one off needs computational power. You can’t tell if a ticket is a winner until you scratch it. Each ticket has its own rarity. If someone shows they scratched a ticket with a one-out-of...
Yes, everyone can do that, but not everyone can generate "shares". This problem can be partially solved by Merged Mining, but it is still not yet ready in production, because if/when it will be, we will no longer have those huge pools: https://www.truthcoin.info/blog/security-budget-ii-mm/#f-the-trilemma
With PoW, a miner needs to make an investment in mining equipment for to mine the cryptocurrency. If that equipment cannot be used to mine that cryptocurrency, there is no other way for the equipment to create value for its owner. If the miners as a whole were to act badly, the mining algorithm could be changed, and the mining equipment would become worthless.
Similarly, with PoS, a miner must "lock up" their coin in order to receive mining rewards. This is similar to a miner buying a miner, except without the mining manufacturer. If a PoS "miner" were to produce invalid blocks, their "locked up" stake would be lost. It would be up to the other PoS "miners" to determine if another miner is producing invalid blocks. If the miners as a whole were to act badly, the value of the coin would become worthless.
In both PoS and PoW, if the miners as a whole act badly, they will lose their investment. However, with PoW, the underlying coin will survive. This is an important distinction. If a large PoS miner were to successfully act badly, they could potentially enter into derraritive contracts to offset the coin they have "locked up" that is mining. This removes the requirement for a PoS miner to actually invest in their "mining" operation. OTOH, a PoW miner cannot easily hedge their investment if they intend to act badly while mining.
I can't believe some people are still thinking that Proof of Stake is the future, just because it spends less energy than Proof of Work, and it fits the green narrative propaganda.
Just look what happened with fake news about ethereum switching over to eth 2.0 that is probably never going to happen, except maybe in metaverse space
Intel just entered Bitcoin mining with new power saving dedicated chips and it's not impossible to think that in future we are going to have even more improvements in this field.
Real question is if enough people care about decentralization and proof of work, or they just want quick gains and fake promises.
I think this is what is behind the recent push to get bitcoin to convert to PoS. There is not a technical rational behind wanting to switch. There is a desire to implement the green new deal.
It should be clarified that there is a difference between "wealth distribution" and not having an "ICO, premine, etc.".
Having an ICO means initially there were coins that were created out of thin air by the creator and sold to people so that the creator can make money for something that doesn't yet exist. That is a big centralization sign.
On the other hand, wealth distribution means how the circulating supply is distributed among adopters. I'd argue that it has no effect on decentralization. Simply because you can't affect the network with whatever amount of bitcoin you have (bitcoin is PoW after all). Which brings us to the fundamental flaw of PoS.
Why "economy"? It is limiting the definition. I'd say the network as a whole not just the economy. For example the "economy" was centralized back in 2013 when there was one major bitcoin exchange (MtGox) controlling more than 85% of the total trading volume and it was manipulating the market. Yet bitcoin was still decentralized.
P.S. Here is a realistic risk of PoS: a lot of people have proven to see cryptocurrency as a trading tool and they will deposit their coins on a centralized exchange. All of a sudden we have one entity with a huge amount of coins aka stake that can control the PoS too.
One could enter this debate, but their is little reason.
If the person supporting PoS is winning the debate, then the mods lock the topic.
If the person supporting PoW is winning the debate, then the mods allow it and others point to it as proof.
The Truth is these random arguments no longer have any weight in the real world.
In the real world. The top 12 coins on coinmarketcap.
#1 on CMK: Bitcoin is the only PoW coin
#2 on CMK: Ethereum is switching to PoS.
#3, #4, #5: Are Tokens
#6 on CMK: Ripple is neither.
#7 on CMK :Cardano is already PoS.
#8 on CMK: Solana is a version of PoS
#9 on CMK: Tera Luna is a version of PoS
#10 on CMK: Avax is a version of PoS
#11 on CMK: Token
#12 on CMK: Polkadot is a version of PoS
Of the top 12 coins
1 is PoW , 4 tokens, 1 custom XRP
The other 6 are PoS.
So no matter anyone's opinion here for or against, the marketplace has already chosen.
And their is an undeniable lean toward PoS or anything other than PoW.
Does Bitcoin have to move to PoS, absolutely not.
To be honest, bitcoin low number of coins makes it unsuitable to be a proof of stake coin.
PoW does have issues and something will have to be done at some point.
The current Bans and the future bans, because of the limits of the modern power grid to support the ever increasing energy requirements of the PoW miners.
Their will be no green energy solution fix to the above, as China one of the largest hydro energy producers was the 1st to have banned PoW mining.
Europe could ban PoW mining by 2025.
Most other coins will switch to PoS, as ethereum is doing.
So they will avoid the bans.
Your only real concern , if you are dead set on wanting Bitcoin to be PoW and still be used.
Is what will you do when PoW mining is banned world wide?
Brainstorming answers for PoW real problem is what you should be focused on,
because the PoW verses PoS debate is dead.
Ethereum, Cardano, Algorand and all of the other PoS designs will continue regardless of what happens with PoW or bitcoin.
The final question coming is will the world government let your PoW survive, and that is what you need to worry about if you want PoW mining to continue.
Technically, we can move to fractional satoshis when needed. Also, in LN you have millisatoshis, so it is possible to increase coin precision without changing the main layer.
That's why mining decentralization is important. When a lot of computers will be mining single satoshis or millisatoshis, then they can collectively mine new blocks. In case of huge computing power drop, it will be needed to somehow "push chain forward", so trusting "the strongest shares" will be needed. If it happens gradually, it will be similar to gradual computing power increase, but just in reversed direction. That way or another, it will be a failure of Proof of Work, and then trusting "shares" is needed to do anything without a hard fork and hacking the difficulty.
As long as Open Hardware and Open Software exists, I think we are safe. The government can block huge miners, but they cannot block millions of CPU's. If they will, PoW mining will be the least important thing of our problems, because then all wallets can be considered unsafe.
So you are proposing that PoW algo be changed to something that can only run on CPUs as a way to spread out mining and the energy usage.
Doubtful the ASIC miners would agree.
The days of the PC are numbered for the average person, as normal people switch to smart phone or tablets,
the increased heat and faster destruction of said devices , would prevent the majority from mining , plus you have a new strain on the wireless cell towers to content with.
And illegal botnets would have a major advantage in PoW mining.
If a business still used PCs, they would still concentrate the devices to one area to save housing costs and you are right back where you started.
Plus any new CPU PoW algo can have a ASIC designed for it, so you still don't get off the asic treadmill.
I'm as pro PoW and anti PoS as anyone, but your case is not strengthened by the sloppy reasoning at the end:
The same thing happens in proof of stake: To acquire voting power you buy stake from existing stake-holders, who end up having less power.
Financial gains has little to do with it. If you buy ASICs from another miner, then you also increase their gains.
What I don't understand is that if a proof-of-stake "consensus" algorithm alone is not sufficient to reach consensus, especially when it comes to those nodes that have just joined the network or been offline for a long time, then how can we call it a consensus algorithm in the first place? And if it is not possible to reach a consensus with a pure PoS, then the whole debate becomes meaningless: it simply cannot compete with PoW due to its inherent drawbacks that cannot be fixed except through relying on some less secure punishment mechanisms or timestamp servers based on, again, PoW algorithms.
The trouble is that current stakeholders aren't incentivized at all to sell their stakes for they would make more money and would have more power by just doing nothing. No rational player will give up their power and ever-increasing amount of money for the sake of short-term yields. You aren't incentivized to make others more powerful by selling them tokens either.
Clearly, the market has chosen bitcoin with its PoW since it holds first position, above all shitcoins with PoS you enumerated.
I see PoW vs PoS also in a philosophical way: it's meritocracy vs oligarchy.
For PoW one has to work hard and good to obtain rewards and trust.
For PoS the rich ones rule; you have no other choice than just trust them because they've bought that coin.
I know that PoS lovers will come and tell that ASICs can also be bought on big numbers by the rich. Still not the same. PoW is the right direction.
The only thing I'd like to see more is that large ASIC farms start building their own sources of regenerable energy (although I know that the PoW vs Environment drama is fake)
This.
I am not aware of any serious argument that suggests PoS will result in mining to become centralized, over time or otherwise.
Bitcoin PoW mining today is not controlled by any single entity, but the number of major miners has decreased from what would have been predicted even a few years ago. There have probably been times in which a small group of unrelated entities could theoretically band together to execute a 51% attack on the network, but the economic incentives around PoW seem to have held up and we have not seen that.
It's been long time since i saw PoS discussion on this "Development & Technical Discussion". I hope there won't be any chaos.
And it already happened several times. For example, Binance, Houbi and Poliniex helped Tron foundation to perform hostile takeover on Steem network[1]. Binance only apologize[2] while Houbi make an excuse[3] without any real action.
[1] https://cryptoslate.com/big-exchanges-conduct-a-hostile-takeover-of-steem-blockchain-following-tron-acquisition/
[2] https://www.cryptoglobe.com/latest/2020/03/binance-apologies-for-its-role-in-justin-sun-s-steem-takeover/
[3] https://coingape.com/huobi-exchange-responds-steem-network-takeover-by-justin-sun/
Couldn't you replace the exchanges in question with major PoW miners? The exchanges in question hold customer money, and customers could trivially move their crypto to their own wallet, or another exchange.
There are still pools today, but mining has become increasingly done by entities that own large amounts of mining hardware.
Maybe some algorithm that doesn't utilize cryptographic algorithms (I.e. energy) or tokens is what is necessary to create a low-energy PoW alternative that doesn't put your coins at financial risk (locking tokens I.e. POS).
For example - Proof of Time (I made this one up).
Most operating systems support nanosleep(2) for pausing execution for a certain amount of nanoseconds.
So instead of solving a problem where you must find successively smaller hash inputs, we turn it into a problem of sleeping for a random amount of nanoseconds (1 second = 1 billion nanoseconds) up to a certain target.
The idea would be that the network keeps a global 256-bit counter that tells how long all of the nodes have slept, and that all of the nodes will compete with each other to sleep, each amount of time slept increments the global counter, and the node that whose sleep time causes the counter to hit the target will get the block reward. Then the counter time is reset to zero and the target is increased or decreased based on a difficulty adjustment algorithm similar to the ones in Proof of Work.
The sleep time will be hardcoded, to e.g 100ms.
A "stale message" multiplier will also be hardcoded, e.g 20.
This means that nodes have 2 seconds to sleep for 100ms and relay the messages over the network to other peers before they start rejecting the message.
So, what happens after a node sleeps (it actually doesn't have to sleep, it just need to wait for the sleep time doing something else but the only logical activity that wakes programs up after particular intervals is sleeping), then it sends this message to all of its network peers, with the following content:
{
current_UTC_datetime: string
public key: string
hash: ECDSA signature of current_UTC_datetime against private key
}
A message with datetime in the future is rejected.
Messages with datetime more than (sleep_time * multiplier) nanoseconds in the past are also rejected.
Otherwise, the sleep time is added to the global counter, stored in a ledger I.e. an array of entries that look like:
{
current_counter: 256-bit int
message: {above message embedded here}
}
If the node finds that their public key is associated with an entry that has hit the target, then it generates a block reward for itself (as the consensus is coded in the program) .
Then each node only does one ECDSA operation per "sleep time" interval instead of several trillion SHA256d iterations, while keeping a decentralized blockchain.
This schemes relies on a majority of nodes being honest in relaying the correct time, just like PoW. But it neither uses vast amount of energy (ASICS are made redundant for this), nor locks up finances.
Im the stupid one in the forum, but from learning, listening, and reading what the smart people say and write in interviews or technical papers, their description of Proof of Stake drew this picture in my mind.
Another image in my mind is a cat chasing its own tail.
Not at all. I thought about solo mining a fraction of the block reward without trusting centralized mining pools. ASIC miners could agree to that, because they would get more satoshis than their competitors mining on their CPU's. The algorithm should be left as it is, as long as it is not broken. The only problem is splitting the coinbase transaction to move from a "winner-takes-all" to "miners-take-shares" system.
It was decentralized, but less than it is today. If we're going to agree with my centralization definition, then it's true; few people who owned an exchange had a fortune of 850,000+ BTC. The manipulation was easier and more intensive than it is today. Charts speak themselves.
Economy isn't a strictly connected term with money. It's related with administration, in general.
Don't you choose a pool to stake?
This place is known for its freedom of speech. If you don't want to try, then don't blame mods. The mods don't lock topics subjectively. Only if it's against the rules.
Which proves what exactly? The market had also chosen to put a dog-looking coin in the third position. Is that what you want? To simply reach a certain position among other currencies? Is that your project's goal?
And now you're comparing China with Europe...
Okay, so... May I assume you didn't read my writeup, right? I've detected flaws to your mechanism and you keep avoiding them. You're the one who asked for serious conversation, troller.
You do have the option to not buy them from another miner, though.
You can't even begin to compare a "major miner" with an exchange that holds a lot of users' funds. A miner has spent a lot of money and put a lot of effort to acquire that large hashrate while the exchange has done virtually nothing. Not to mention that the miner has a lot at stake while the exchange doesn't.
No.