BigTrees

You know what can stop Denial of Service, or Distributed DOS attacks? Requiring a tiny fee to be paid that has a state transition contract defined, which only moves one direction: forward. The data packets themselves are then tied to this state output, gated by configurable acceptance time, that anyone can verify quickly, globally. Senders are then required to place proofs linking a valid ledger-borne state in the data packets themselves for them to be accepted by intermediate forwarding systems. No invalid packets get through. Evidence trail is established back to funder for every packet. Complete privacy enhancing integration through deterministically anchored addressing source/destination via IPv6 specification. It won't stop all attacks, one could take over wallets and such to generate bad traffic with someone else's funding, but importantly this provides bounds. The funding will eventually run out. It's a lot harder than just operating a botnet. Chainalysis can yield more clues than were available before. Additional distributed, immutable logging can play a support role in the investigation process. Downstream recipients have more to filter on to differentiate source attribution, and this can lead to better infrastructure defense posture. It forms a natural, built-in rate limiting function. All data must be valid, intact, hash-conforming, and verifiable through Merkle-tree traversal back to widely distributed information. I'm working on this and similar ideas now. I've already proven UDP and TCP/IP transmission end-to-end. I now have SPV verification working in under 1 millisecond. That's cheap enough to be run by every router, every system. The future is looking amazing.

From the lab: I've been upgrading my on-site testing capabilities. I'm measuring Teranode multicast delivery throughput this morning. The results are encouraging. 2 million TPS on old 8-core AMD machines, without really breaking a sweat. More hardware, and more testing on the way