So 64 notary nodes share majority of mining rewards between them, and non notaries can mine kmd but at a disadvantage.
So consensus is PoW like bitcoin, but the notaries work together and share block reward equally like a cartel.
And notary cartel membership is voted on by coin holders, preventing cartel from misbehaving.
A very interesting combination, not fully decentralized, but much more advantages from structure than other crypto coins.
Trust in the notary cartel will depend on election process, very interested to see more details when available.
51 13 12 16 -1 -1 -1 38 -1 34 28 -1 -1 -1 50 49 -1 -1 1 -1 5 -1 -1 2 53 -1 -1 14 -1 17 -1 -1 42 20 10 56 -1 -1 55 -1 -1 4 15 60 -1 9 8 23 54 41 0 59 44 19 35 -1 -1 -1 37 47 29 -1 36 -1 -1 <- prev minerids from ht.187826 notary.1 gpucount.28 43.08%
Whether a block goes to a notary node or not is random. In exchange for the ability to mine at easy diff, the notary nodes provide high end servers for network services, like basilisk mode.
Notary nodes dont have the power to create transactions, nor do they ever control anybody's funds. they are elected based on a Proof of Stake election
So while the notary nodes are in a semi-trusted position, this trust is used for getting consensus on what checkpoint to write to the bitcoin blockchain, availability of network services and other things that otherwise would be unavailable or unreliable
The network is at the same time a PoW network and a cartel mining (as you call it) which are coexisting and seamlessly shifting between each other. The added randomness of which group will mine the next block makes it that much harder to conduct attacks as no matter how much hash power you bring, the notaries mining costs dont go up.
From the other side, if any group of notaries start misbehaving, then they are subject to not receiving subsidies or outright replacement and since any notary mined block must be signed by the notary, it is not possible to hide behind a brand new address
Thank you for this detail jl777! I am interested in governance aspects of crypto, what you have designed is very ingenious, a good blend of decentralization but with elements of structure that allow stability and increased reliability. The other governance model of DAO looks promising too, but so far untested to run a sophisticated platform like komodo. I can't see any weakness in the notary setup for possible attacks, only problem might be human element, large number of capable candidates competing at elections, but that is a good problem to have.