Which Bitcoin Should You Run?

Do you want to help scale Bitcoin to billions of users?

➜ Run BCH or BSV. Enjoy your 32MB blocks of nothing — or 4GB of Craig's legal docs.

With Lightning and programmable money, we can scale to billions without sacrificing decentralization.

Do you like privacy?

➜ Run PanoptiCoin. Every transaction observed and pre-approved by your compliance officer. Freedom is slavery.

I'll use Coinjoins to prepare Lightning channels (potentially in an Ark channel factory) and accept payments via Bolt12 offers.

Should privacy be built-in rather than a separate coin?

➜ Run Monero (XMR). Do you accept Monero? Nothing says private like announcing your tiny anonymity set to every merchant.

Cashu and Fedimint give me private ecash backed by real sats.

Is miner decentralization important to you?

Honest miners should coordinate to censor naughty transactions. I'm sure the Censorship Committee will never be co-opted, and financially-incentivized miners will never just include them anyway. ➜ Run BIP 110: Censorship by committee. What could go wrong?

The goal isn't to transfer power from large miners to small miners, but to ensure NO miner has any power. In particular, no miner should have censorship power.

Are Layer 2 solutions the right approach to scaling?

➜ Run Bitcoin Unlimited. All txs on-chain and visible to surveillance forever. Chainalysis thanks you for your service.

Cheap blockspace + L2s is the best of both worlds. On-chain access secures L2s, and L2 adoption frees room for more scaling.

Should all monetary txs be relayed regardless of purpose?

➜ SatoshiDice Derangement Syndrome. Silently shipped blacklists in Gentoo, i.e. sneaky auto-updates. Dev apologised for this, but he might not apologize for the next ones.

SatoshiDice was monetary. You don't get to judge what a payment is for.

Do you understand that not all mempool filters are equal?

➜ Wishful Thinking Node. Your node rejected it. But the miner with a direct submission pipeline didn't. Congrats, you played yourself.

The question is whether the transaction still gets mined out-of-band (OOB).

Do you understand that a filter has FAILED if the transaction still gets mined out-of-band?

➜ Run Knots. I'll rely on LibreRelay and CoreV30 to secure decentralization, freeing me to run my virtue signalling node.

That's the worst case for decentralization. OOB mining concentrates fees with large miners, achieves nothing, and makes mining less competitive.

When demand is high and it incentivizes OOB mining, what should you do?

➜ Run BIP 110. Bans OP_SUCCESS, breaks OP_IF, activates at 55%, and admits in its own text that spam is best fought with policy, not consensus.

Run LibreRelay. Reluctantly allowing txs we dislike protects decentralization.

Do you understand that the success of the dust limit filter does not prove all filters work?

➜ Survivorship Bias Node. The dust limit works because nobody cares to bypass it. High-demand txs? Ask the OOB miners.

Dust works because incentive is low. But that doesn't mean that every filter works, see the sub-1sat-per-vB summer of 2025.

Do you understand that steganography makes it impossible to ban spam without banning money?

➜ Run Ordisrespector. Filter either fails (due to relaxed nodes) or succeeds (leading to OOB and mining centralisation). Heads you lose, tails you lose.

One man's spam is another man's data. You can't ban content without banning money.

Given that spam is inevitable, where should it go?

➜ Run BIP 110. Congrats, you forced spam into fake pubkeys. UTXO bloat forever.

OP_RETURN and Witness data — keep the junk prunable and out of the UTXO set.