Risto, this week has been amazing for me:

https://bitcointalk.org/index.php?topic=1049048.msg12140971#msg12140971

That is concerning:

• Instant 0-conf transactions that can't be double-spent with any of the typical attacks on Bitcoin.
• Holy grail of on chain anonymity (superior to Cryptonote).
• No more 51% attack possible.

Those claims are so incredible, I doubt anyone will believe me until they see white papers. And they should not until they do.

Well if you can offer peer review maybe that can be arranged. I will get back with  you when I am prepared.

Btw, I tried to sleep and suddenly I solved the last remaining snag in my consensus design. I suddenly realized I indeed solved it. And yes mine will be fully proved in a white paper.

Just a few days ago, I solved the holy grail of anonymity:

https://bitcointalk.org/index.php?topic=1049048.msg12127776#msg12127776

It is all over for Bitcoin. Seriously. I will go silent. Next you hear from me will be via white papers.

Good luck.

Okay I've digested the white paper. I appreciate john-conner providing more details. That was the honorable and helpful action.

Unfortunately afaics, there are elementary attack vectors that he has not addressed in this white paper:

• No cost to being a peer, thus a Sybil attack on nodes in general. The adversary could insert a unlimited nodes if he wants to in order to dominate voting.
• How to squelch DoS attacks on the vote? A node which always votes against consensus can't be distinguished from one voting honestly since lock conflicts result in an indeterminate vote.
• He employs a Bellagio Algorithm to avoid gaming the selection of which nodes to poll for votes, so this removes the reputation weights of Skycoin's consensus, but it doesn't address the spamming of total node count nor the spamming of the lock conflict.
• Nodes are not paid for voting. If they receive bribes from the double-spender, the algorithm has no defense other than it hopes that 50% of the nodes in the network are not on-the-dole.
• Lock request spamming or DoS attacks. Are transaction fees hardcoded for the entire network?? Preset constants are anti-decentralization.

Also this algorithm requires the entire network to see all the transactions which of course won't scale without centralization, so if he is trying to enable real-time microtransactions (1 second confirmations) which could explode transaction volumes up to the 100,000s or more per second then he has a problem. Refer to the GavinCoin fork debate.

Apologies to harp on this thread, but it would be sneaky if I added this as an edit to a prior post.

Zerotime is a very significant claim. When you make bold claims that have the potential to seriously challenge Bitcoin, you will be held to a higher standard of scholarly proof.

Indeed the necessary proofs can be explained with math, computer science, and networking research.

Einstein said you don't really know something until you can explain it to a n00b.

The only ways I know of to squelch DoS and Sybil attacks are:

• Participant expends resources.
• Participant must have a reputation.
• Participant performance can be measured.

I already stated that I don't know how to do #3 for his proposed lock protocol, because there is no verified correct vote on a lock. Disagreement isn't a verified crime in this protocol.

I already stated that I don't know how to do #1 for his protocol, because transaction fees can't be taken in the lock stalemate case where Sybil attack applies. Others suggested burning resources (I assume the voting nodes) and I responded saying which nodes would participate in voting if they must burn resources? If instead you require every transaction to burn or include a proof-of-work hash or fee, then the innocent spender is penalized by any lock stalemate.

So that leaves us with reputation. But again how do we measure performance in order to assign reputation? And then there are other problems with reputation.

There is a reason that solving the Byzantine Generals Problem remained unsolved until Satoshi invented proof-of-work. The problem is essentially that there is no reference point, because either value is correct and incorrect at the same time (e.g. lock or not lock) because we can't prove the network is reliable.

It is as if John isn't aware of how significant Satoshi's accomplishment was. Or somehow John thinks he has a clever improvement, but hasn't yet proven it to us.

Btw, I think I have a solution of how to achieve zerotime. But I do it an entirely different way and my solution doesn't require all peers to see all transactions, so even if John fixes his, I've still got him beat (assuming my white paper comes together which is not finished yet).