Consensus, the key to decentralisation POW,POS,PO?...

[Mastercore]

I want to open a discussion about consensus algorithms.
I am super interested in all the exotic ideas being tested in the crypto space but I am so far a little unimpressed by the variety of consensus algorithms being tested. POW (bitcoin and all its clones and cousins), and then POS DPOS, etc... (EOS, Icon, etc...) seem to be the only real options being tested as far as I can tell.

From my point of view, the consensus algorithm should be a set of easy to follow (easy to understand so anyone can create a node with compliant rules) rules that set fare and scalable way for a network of nodes to find consensus, i.e. agree on the winner or winners of a lottery, the prize being who wins the right to extend the blockchain by one more block, and win the block award.

The goal of the consensus algorithm is to include as many participants (Ideally every wallet is a node) into this game. The game to determine, in a truly random and unpredictable way, who will be the next block winner(s). The more that play, the better the contest is, the more secure the network.

It would be ideal (in my view) if the consensus algorithm concentrated maximum CPU power to verifying transactions and minimum to establishing which node(s) will participate in creating each block.
Also we are assuming that the block creation and the distribution of tokens is linked. I understand the economic reason for this, but apart from that, there is no technical hard link, so we may experiment with variations.

My wishlist for a consensus algorithm

• is lightweight enough to allow every wallet to be a node (even a feature phone with shaky internet connection). I think it may be a more flexible approach if we think that more than one node is the block winner. Nodes cross verify each other. A bit like kids in a classroom checking each other's homework.
• 2 (or more) speed blockchain. Integrate something like lightning into the main consensus algorithm. After each main chain block, n number of sub chains split off. These side chains are reconciled at the time of the next main block.
• idiot proof. Being a node should be as simple as installing and running the a piece of SW. As I said ideally the node is in the background of a normal wallet.
• vote rights: nodes have voting rights built in to them. When a node is chosen to be block producers, they can also vote (not token holders. Not POS).

Some features I can imagine...

• I see a system where each block is produced by 10’s or 100’s of nodes working in concert
• The nodes all share in the block reward.
• When nodes are selected to be block producers
• In the block duration, chosen block producers can vote. Voting can span multiple block durations.

P.s.
I have been interested in the crypto space since 2013 when the bitcoin 'light went off in my head'. Since then I have not been able to look away.
The consensus issue has been one of particular interest to me.
What I have hinted at here, I have elaborated on in a sort of white paper.(80% done) My end goal is to release this white paper, maybe together with inputs from others, and to hopefully have it further the entire space to a better place.
I sheepishly cringe at discovering ideas like mine have been discussed to minute detail and rejected already, but because I cannot find anything like this, I reach out tentatively to test the water with mine.
I have chosen bitcoin talk to dip my toe in the water and see if my ideas have any merit.

I look forward to getting constructive feedback. If you have something to contribute, I would love to start a conversation. If you have criticism, I want to hear too, but please back it up with fact and science.

[1714022538]

1714022538

[1714022538]

1714022538

[1714022538]

1714022538

[1714022538]

1714022538

[Mastercore]

Hi ETFbitcoin, I wills start at the end...

Past thread about consensus algorithm has proven it's impossible

Can you please link to this thread, and sorry it I´ll ruffle your feathers here, but the word "impossible", in this context is used by those with a lack of imagination and/or a desire to limit other's imaginations.

AFAIK PoCW and some PoS variation already achieve this. Obviously the bigger their hashrate or coins, they have have bigger reward or have higher chance to make a block and claim it's reward.

My feeling is that PoS and its variants are flawed approaches due to their reliance on staking, meaning the more coins you have the more you will earn by staking, and this, I fear, will lead to less and not more decentralisation.
I have the strong feeling we must explore options outside the PoS and PoW duopoly.

Can you please link me to more info on PoCW. I am curious and have not been able to find anything online about it yet. Which coin is using it? Any links would be appreciated.

NO, transaction verification should uses little computational power to make it scalable.

My bad, this was obviously going to mislead now I re read it. What I am getting at is that it is not sustainable that most of a node´s CPU is invested in simply determining who will mine the next block. This should be the other way around. Transaction verification should be the main CPU role of the node and this should also be optimised to be as streamlined as possible. i.e. max throughput.

Only can be achieved with Sharding where each nodes don't broadcast or verify all transaction. But there's plenty of disadvantage, have some degree of centralization and IMO it can't be considered as full nodes.

I agree, but I think the term "full node" as taken from the bitcoin network is the most robust but the least efficient way to describe what a node is or should be.
I think there is a very large latitude available in the fullness of a node while retaining decentralised security and speed. Why could it not be that hundreds of nodes participate in each block creation and share/overlap the work, and in overlapping they also cross verify each other.

Isn't install and run application (with default configuration) already idiot proof?

When my wife can install something and run it without my help, then we have started to reached the realms of idiot proof (my wife is far from an idiot, but when it comes to computers I am sure she would not mind being the hapless benchmark here)

CMIIW, but AFAIK few cryptocurrency such as EOS works in similar way.

Again, what I am trying to discuss here is a non PoS consensus algorithm.

Thanks for your time and comments ETFbitcoin


 

[Zin-Zang]

AFAIK PoCW and some PoS variation already achieve this. Obviously the bigger their hashrate or coins, they have have bigger reward or have higher chance to make a block and claim it's reward.

My feeling is that PoS and its variants are flawed approaches due to their reliance on staking, meaning the more coins you have the more you will earn by staking, and this, I fear, will lead to less and not more decentralisation.
I have the strong feeling we must explore options outside the PoS and PoW duopoly.

FYI:
Staking is the consensus, the earning more coins is a program setting.
Definable by your code, whether anything is earned at all.
Staking reward can always be set to zero, so no coins are earned thru staking.
You could always make staking for a % of time a requirement to send or receive coins.
IE:  Wallet has to attempt to stake for 1 day or 1 week to be able to send /receive coins for 1 hour.

However you ask for other consensus methods
https://hackernoon.com/a-hitchhikers-guide-to-consensus-algorithms-d81aae3eb0e3
Proof-of-Authority (PoA) 
Proof-of-Weight (PoWeight)
Byzantine Fault Tolerance (BFT) — Siege the blockchain!
Directed Acyclic Graphs (DAGs) — aka the Blockchain Killers!

https://www.newgenapps.com/blog/8-blockchain-consensus-mechanisms-and-benefits
Proof of Importance:
Proof of Elapsed Time:
Proof of Identity:
Proof of Burn:

Good Luck.

[Mastercore]

Thanks Zin-Zang and ETFbitcoin.
I need to digest this information and see where it leads me.
Not for the sake of having the last word here (because I hope it is far from the last), but there is still nothing that negates the idea I have been working on. A consensus algorithm that does a better job of decentralisation, speed, security that what is out there at the moment.
I know that I have not really let onto what I am thinking,... what has gotten into my head, and will not let me go for now,...

but I want to have the idea at least prodded enough from outside before I release it for full scrutiny. Thanks for the prodding guys

So thanks for helping out for now.
Now it's time to read more....

p.s.

Isn't install and run application (with default configuration) already idiot proof?

When my wife can install something and run it without my help, then we have started to reached the realms of idiot proof (my wife is far from an idiot, but when it comes to computers I am sure she would not mind being the hapless benchmark here)

Then the problem lies within person with bad computer literacy (aka computer illiterate), not how hard to install/run nodes.

IMHO This is a dangerous way to think. We are trying to build systems for everyone, not for the sliver of humanity that feels right at home sitting in front of computer most of the day toying with code, having our micro epiphanies and hanging out in chat rooms. The SW we produce should target users for whom programing a VCR or an oven timer is not immediately successful. That is my idea of idiot proof.
I think we should never consider a piece of SW is now reached its easiest to use state. This is a constant target to work toward.

[Mastercore]

In the current paradigm, I fully agree. Way to much of an knowledge gap in the user base.
But if we really subscribe to the idea that open public decentraslised blockchains are the best future for giving back the power of money back to the citizen and away from profit driven entities, then we need to really aim to close the gap between node and user.

I know, a lofty goal.
But I think the same could have been said about Satoshi and his idea in 2008

Just as a side note, came across this article in a print.
(https://photos.app.[Suspicious link removed]/No7HqRLVH8stiXh97)
To succeed, this sw must be idiot proof. (Or there has to be a profit motive for users to overstep any difficulty)

May only be an oblique reference, I admit, but the general idea is there.
Fintech exists in the most acutely uncomfortable corner of the software spectrum.

It is difficult to understand (usually because finance is difficult for the average person to understand)
It is also boring making it yet more difficult to understand.
It is dealing with something we all hold precious, our money. Nobody wants to screw about with SW they barely understand when their money is on the line. This makes people wary and fearful of innovation in the finance sector.

If we fail to win the decentralisation battle, the financial warlords of the world will surreptitiously absorb the abstract notion of blockchain and ruin it to their own ends.

That is why I set the high goal of every wallet is a node on the network benchmark.
Sorry for aiming too high.