Hi There,
I would like to reach an attention to consider moving to PoS.
Arguments that Bitcoin is not moving to PoS (like here: Proof of stake in future?) are nonsens. Bitcoins is a symbol and it will stay like this.
PoW is not sustainable:
Fragmenting the hashing power leading to increased amount of successful attacks. Now the incentive to attack aeternity is low because the adoption is low. But once there will be enough dapps, it may be more profitable.
Energy consumption (I’m not talking about validators, but about miners).
Not scaleable.
Lack of finality is an Achilles heel for developers. It creates many traps and bugs.
Ethereum is moving to PoS for a very good reason: sustainability and scaleability.
There has been massive amount of reasarch done to address the crutial problems of PoS:
liveliness
cartels (especially in dPoS versions)
profitability (economical incentives)
Let’s bring back a discussion about PoS.
For the beginning I would like to propose a great reaserch done by Tezos team and their liquid PoS. Last month I was talking with Arthur and I think they come with a very interesting approach to PoS. Short explanation about it is here: liquid proof of stake, reddit.
Yes, I heard about Aion PoS. It doesn’t solve the problems listed above: it’s still uses PoW and it still has a problem with finality. I think current aeternity approach (PoW + Bitcoin-NG) is better than Aion PoS: thanks to Bitcoin-NG we have much better performance.
I haven’t given their paper a full read yet but looking at the references, which don’t seem to include any of the prior work on hybrid leader election, makes me at least a bit sceptical.
we should also look at VDF (Verifiable Delay Functions) and VRF (Verifiable Random Function), which are really close to mimic PoW features.
I wonder if this is really a problem (PoW), if in the future permissionless public blockchain may be only a settlement or anti-fraud platform. All real activity will happen either in private side chains or state-channels-like structures. Those side systems can have any mechanism that is cheap (PoS, PoA, etc.) and handle security as they wish. Also it will naturally partition the system.
I think the simpler PoW the better as it lets create ASICs for cheap and also use various devices to mine.
(contrary to very complex anti-ASIC PoWs that eventually are implemented in advanced expensive circuits not available from multiple vendors)
To conclude, I am infavour of simple custom PoW for public permissionless chain, until we get proven randomness from public resources.
Yes, VRF is very interesting approach. Algroand is pioneering it for Blockchian PoS.
@michalzee (Hi!) PoW requires lot of hash power to protect the network. There is some research that even 33% hash power control already exposes too much risk. Moreover selfish mining is real by pools that command less than 1/4 of the resources (explanation, paper: Majority is not Enough)
The risk is real - look at other smaller blockchains, they have been successfully attacked, eg: Ethereum Classic attack.
With carefully designed PoS we have both $$$ representation of the economical value protecting the system AND a control mechanism to have n-factor protection. Eg: cosmos blockchain is halting when an attack is detected. Ethereum just go on and allows system corruption.
Having both $$$ expression and control mechanism are the important motivation to use / research PoS in blockchains like Ethereum and Tezos (which have very strong team and research).
Exactly - the adoption of PoW is there but it’s perceived as a legacy mechanism.
With current state of the art PoW doesn’t have any benefits (security wise and sustainability wise). Even contrary - it depends on mining pool adoption and once the mining pool will decide to shift their computation power somewhere else, all of the sudden the network is vulnerable (NOTE: PoW requires orders more computation power to secure the network then other appraches).
Recently I stumble up the Avalanche consensus algorithm. It’s perceived as one of the biggest research in consensus algorithms. Additionally it’s backed by solid academia: Professor Emin Gün Sirer of Cornell University, one of the cryptosphere’s most esteemed critical thinkers, and the co-founders of AVA Labs, Kevin Sekniqi, CS Ph.D. at Cornell and Ted Yin, CS Ph.D. at Cornell.
Avalanche introduces extreme decentralization which enables very high immutability in the ecosystem by allowing thousands to millions of network participants by being lightweight, easy to implement/understand and has no specialized hardware requirement.
Quick finality and low latency : It takes around 1–2 seconds to achieve finality across the globe. This is how long it takes to get your payment processed and verified.
Higher throughput : 1000–10,000 transactions per second. 6500 TPS benchmarked at NYC Blockchain Week over 1000 nodes hosted on AWS.
Robust : The network doesn’t need to agree on participants identity to achieve undeniable consensus.
Quiescent protocol : A green protocol that requires no energy or specific hardware resources in order to be secure.
Highly Scalable: The protocol is lightweight and therefore admits scalability and low latency.
Egalitarian ecosystem: The Avalanche protocol gives rise to an egalitarian ecosystem, i.e., all nodes in the network are born the same (there are no leaders and none have special rights). Since there are no miners, there is no centralization of the hash power through “pools”.
Byzantine tolerance: A significant percentage of Byzantine participants can be tolerated with no impact on safety. For example, under certain configurations of Avalanche, up to 50% of the nodes can be Byzantine, i.e., nodes that attempt to trick the network and keep the entire network imbalanced. However, they will be unable to do this in a way that causes two nodes to decide on two different colors at the same time.