CPU-mined Tor-native Bitcoin 0.1.0 clone launches with RandomX proof-of-work

The core change replaces Bitcoin’s SHA-256 proof of work with RandomX, a memory-hard algorithm designed for general-purpose CPUs. This allows laptops to compete with servers and removes the economic incentive for ASICs. Block time is set to about two minutes with a 50 BTF reward that halves every 210,000 blocks, preserving the original 21 million supply limit. Addressing shifts from IP-based connections to a system where each node is identified by its public key and reached through an encrypted rendezvous, enabling operation behind CGNAT and preventing peers from learning a node’s location.

Running a node requires the latest Bitflash software because consensus rules have diverged from the original Bitcoin chain; older releases will follow a separate fork. Mining can be started locally with the -nogui -gen flag or by operating a pool that is reachable only over the .btf rendezvous interface. Wallets are restored from a twelve-word seed phrase, and the default storage is a single SQLite file, though legacy wallet.dat files can be converted in one click. Backing up the seed and the wallet file is essential, as keys generated before the seed installation are not recoverable from the phrase alone.

Adoption costs include maintaining an up-to-date client to avoid chain splits, allocating CPU cycles for RandomX mining, and ensuring rendezvous relays (btfpeers.json, nostr relays, or dedicated rendezvous servers) remain available. The system relies on a distributed discovery stack: first trying peers that responded previously, then nostr relays, then a rendezvous relay, and finally peer-exchange descriptors. If any of these layers fail or are censored, nodes may struggle to find each other, increasing latency or causing temporary isolation. Electricity consumption remains a factor, though RandomX’s memory hardness reduces the advantage of specialized hardware.

The design stops working if the Tor-like encrypted rendezvous infrastructure is blocked or if the nostr relays used for descriptor publication become unavailable or malicious. Since the network depends on a sufficient number of nodes to keep the rendezvous relays healthy, low adoption could lead to longer discovery times or higher variance in block times. Additionally, because the consensus rules differ from Bitcoin, any software that does not implement the exact validation logic will fork away, meaning that mixed-version networks cannot interoperate. Regulatory restrictions on Tor or on cryptocurrency mining could also limit usability in certain jurisdictions.