On Jun 23, 2018, at 10:08 PM, Alyssa Hertig <alyssa@coindesk.com> wrote:

Sure thing! Few questions for you:
- Why are you attacking bitcoin cash?

It is a stress test to validate the integrity of the Bitcoin Cash network on behalf of it’s investors. In the past Bitcoin has been attacked for good and for bad but it’s integrity remains intact which provides investor confidence. The #1 basic rule to all decentralized computer networks is that if it’s not attacked enough to resolve any problems then it will end badly. Skype is the perfect example of a decentralized computer network that was reverse engineered, attacked and subsequently 3 indivduals collapsed all 100,000 backup supernodes leading to a full centralization of the protocol to Microsofts datacenter.

- What kind of attacks are you executing?

Everything from low-level TCP/IP stack attacks to high level bitcoin cash protocol attacks. The combination guarantees several things: 1. The weakest nodes running on VPS’s will crash or run out of bandwidth and become non-responsive. 2. All other nodes will begin to stall out from multiple back-to-back pre-mined 32 Megabyte blocks. Our LevelDB stress testing shows that the Bitcoin Cash UTXO database on a complex 32 Megabyte block can require up to 200 Gigabytes of RAM to fully process and if this RAM is not available the UTXO database will become corrupt and if LevelDB doesn’t release the memory the OS will become unresponsive. Our LevelDB stress tests have been active for over 7 months since we migrated our Bitcoin implementation’s codebase's UTXO set to it’s own database; that said LevelDB wasn’t designed to handle such complex Big Data being read, written and deleted in-parallel.

- Why do you think you'll be able to fork the blockchain?

There are only a handful of mining pools and not enough nodes to enforce network rules; isolating majority of these nodes allows us to utilize our own nodes to withhold blocks and/or headers, reject blocks and/or headers, purposefully fail to relay block’s and/or headers and so on. Recently the Bitcoin Cash network has hard forked to accept 32 Megabyte blocks. With a combination of sybil attacks and our farm producing 32 Megabyte blocks (in-advance) we can inject enough blocks to induce latency and churn into the network so that miners will fall behind on consensus and begin to build their own chains since we will have isolated all of the miners nodes to our own nodes with different rules regarding blocks sizes they are willing to accept. At this point anyone can double-spend at any Bitcoin Cash Zero Confirmation merchant, even from a light client and with zero-effort; the funds will simply show back up once connected to another isolated node.

We believe we have already perfected our approach and our studies are conclusive; LevelDB cannot sustain our complex blocks and Bitcoin Cash is 85% centralized to only a few data-centers and miners. If Bitcoin Cash were 15% centralized then our efforts would fail except for the LevelDB weaknesses; our confidence as always is high. When we attacked the Bitcoin LN we failed across the board except for the massive nodes that crashed however they were quickly fixed and running again. This was a win-win for Bitcoin LN; only time can tell how Bitcoin Cash will withstand the assault.