I don't know why this paper needs to introduce so much already common terminology.<p>"Corrupt" is a re-org
"stable transactions" are confirmed transactions<p>"Function maxvalid(C, C). Returns the longest chain from C ∪ {C} that does not fork from C
more than k blocks."<p>This is the recipe for partitioning consensus. All you need to do is broadcast to anyone a lower-work chain and they will be on their own fork.<p>This is easy to do during a new sync in which the node has been offline for more than k blocks, or with a little more work you could stake grind and partition the network at the tip.<p>The former problem is covered in section 3.2.<p>Generally the nothing at stake problem will result in the problem of requiring some trusted source to not lie, regardless of the implementation, see <a href="https://download.wpsoftware.net/bitcoin/pos/" rel="nofollow">https://download.wpsoftware.net/bitcoin/pos/</a>
Some earlier discussion of the protocol itself: <a href="https://www.reddit.com/r/ethereum/comments/52qfwl/provably_secure_proof_of_stake_algorithm/" rel="nofollow">https://www.reddit.com/r/ethereum/comments/52qfwl/provably_s...</a>
Interesting! Great to see people working on a more environmentally friendly blockchain.<p>What are the weak points of this cryptocurrency compared to Bitcoin? The white paper is very technical. Is it required that majority of stakeholders stay online and constantly take part in the network for it to function correctly?
Is anyone actually mining this cryptocurrency? It seems a bit odd to say this is a proof of concept, without proving the concept by actually running the code. It might be happening somewhere, the topic just seemed avoided by the README.