The Benevolent King

As an American, I was raised to believe that "Democracy" was the pinnacle of good government. Never mind that the United States was set up as a Constitutional Republic, with many safeguards placed specifically to limit the negative effects of Democratic levers. We are encouraged to worship at the feet of Mob Rule. And Monarchy, well, we all know it is strictly evil, right?

Hans-Hermann Hoppe's Democracy: The God That Failed, argues persuasively that Monarchy is actually preferable to Democracy, largely because elected officials do not have as much skin-in-the-game and are incentivized to think and act with high time preference (short-term thinking) as opposed to a king who sees the national resources as his property to protect and nurture. But what if the king is a tyrant? The monarch is supported by a family and royal court which would have the means and motivation to depose a king whose abuses became unbearable. Given enough abuse, regicide will be carried out by a conspiracy of the inner circle or even a sole individual close enough to slip a knife into the king's back. This is historical fact.

Let's now turn to Bitcoin. A lot of vitriol has been thrown at Luke Dashjr over Knots and BIP-110. Many who claim to hate spam and JPEGs on Bitcoin stop short of supporting BIP-110 because of arguments like this one:

"I'm not under any illusions Core is perfect. But, however well charges against them hold up, switching Bitcoin's reference implementation over to a SOLO DEV project, especially when it's led by someone of questionable mental health, would be disastrous!"

Solo dev. Questionable mental health. Disastrous consequences.

The man with "questionable mental health"

I have never met Luke in-person, but I have engaged with him online. He strikes me as both highly intelligent and highly autistic, which is not surprising given he is a genius-level C++ developer. For example, I've gone back and forth with him about his beliefs concerning Catholicism. I'm likely on the spectrum myself, so I understand him probably better than most. Autistic types tend to have trouble seeing things from the perspective of other people, and that is often where his argumentation can be off-putting. For example, rather than just stating that he's part of a small sect of Catholics who believe the majority of the Catholic Church (including the Pope) are not real Catholics, he engages in circuitous arguments that sound a lot like the No True Scotsman fallacy. From my vantage point as a Protestant Christian, he seems to have much more in common with Protestants and Martin Luther (who was a Catholic priest with the goal of reforming the Catholic Church) than not, but he doesn't see it that way and seems highly offended if you make such a suggestion. If you understand that he doesn't mean "Catholic" like most people understand it, then you start to see that his logic is internally consistent. Here is how Luke's own bio opens:

"I am first and foremost a Roman Catholic. This is not to be confused with the heretics and pedos who have taken over Vatican City; Leo XIV & his entire sect are NOT Catholics."

He's not crazy. He's just obtuse. His own bio says it plainly: "MBTI classifies my personality as INTP. (I'm not actually being a jerk, that part is your perception only. Meet up with me in person sometime and you'll understand.)"

A temperament that will not soften a premise to win an argument will also not merge a bad patch to win approval. Judge Luke by his code, his review record, and fifteen-plus years of not being wrong about Bitcoin's failure modes. On that record, Luke is the longest still-active Bitcoin Core developer alive. Dismissing him as "crazy" is intellectually lazy.

Bitcoin is not a democracy

The people warning about a Knots “takeover” want you to imagine a coup.

Bitcoin does not work that way.

You do not vote in Bitcoin. You run software. Every node enforces the rules unilaterally, and the network's rules are simply the overlap of what all those nodes accept. Nobody can loosen or tighten that by winning a popularity contest. Nobody can force you to run their client. Nobody can vote your node into accepting invalid garbage.

That is the whole point.

“Democratizing” Bitcoin development to make it more "inclusive" sounds noble, but it is also an attack surface. Majorities can be manufactured. Sentiment can be bought. Dissenting voices can be censored. Sock puppets are cheap. Corporate money is patient. This hostile takeover of Core has been carefully documented by Hodlonaut and you need to understand it.

Node enforcement is different. You either run the code or you do not. Network consensus is emergent. Be wary of anyone claiming to know what the majority of Bitcoiners believe or want.

The intolerant minority wins

Nassim Taleb wrote the definitive essay on this: The Most Intolerant Wins: The Dictatorship of the Small Minority. Less than half a percent of Americans keep kosher, yet nearly every drink on the shelf is kosher certified. Why? The kosher buyer will not bend, the other 99% do not care, so the market quietly reorganizes around the inflexible few.

Bitcoin already ran this experiment.

In 2017, miners had stonewalled SegWit for months. A minority of users responded with BIP-148, a user-activated soft fork. Beginning August 1, their nodes would reject any block that did not signal SegWit. No permission requested. No majority claimed.

How large was this army? Smaller than people remember. The UASF Node Tracker, built on node data from Bitnodes, shows BIP-148 signaling nodes hovering between 6% and 13% of all reachable nodes from April through late July. A June 12 snapshot reads 11.51%. Contemporary reporting in late July counted 1,095 UASF nodes out of 7,896 reachable, under 14%.

Roughly one node in eight, at its peak.

And it still won. Faced with a credible threat to the sale value of their blocks, miners capitulated before the deadline arrived. BIP-91 locked in on July 20, 2017 and began enforcement around July 23, making August 1 a non-event. SegWit itself locked in weeks later and went live on August 24.

The minority never had to fire the gun.

I remember that August well, though for the wrong reasons. I was in Zurich pitching a crypto index fund to an accelerator, a finance guy fresh off four years brokering precious metals with Schiff Gold, certain I was building the new Nasdaq. Wrong about the broader "crypto" market but right about keeping Bitcoin blocks small, I sold my BCH. I'll save that story for another article.

The point stands either way: a proactive, inflexible minority with skin in the game beats an indifferent majority.

The same play, again

BIP-110 is the same play.

It is a user-activated soft fork that temporarily limits arbitrary data in transactions: caps on small outputs, the 83-byte OP_RETURN ceiling restored, limits on data pushes and witness elements. I laid out why this matters in The Retardation of Bitcoin: nearly half of Bitcoin's block space now carries non-financial payloads, and the cost lands on every node operator on earth.

Miner signaling is expected to begin August 7th. If history repeats, we who are running Knots and RDTS don't need a majority. We need what the UASF had in 2017: a minority that will not bend.

The solo dev

Now to the indictment itself.

First, "solo dev project" is factually backwards. Knots is a derivative of Bitcoin Core. It continuously inherits the reviewed work of Core's hundreds of contributors and adds a curated patch set on top. The labor is plural. The final review is singular.

Knots does not have a solo dev. It has a solo gatekeeper.

That is a different argument.

Second, we have already run the solo-gatekeeper experiment for three decades (credit to Mechanic in a recent Krueger Spaces). It is called Linux. Linus Torvalds has been the benevolent dictator of the most important open-source project on earth since 1991. The kernel he gatekeeps runs much of the internet, every Android phone, and, with some irony, many of the machines Bitcoin nodes run on.

Single-gatekeeper review is not an exotic risk. It is the governance model of the most successful software project in history.

Third, and most important: the king is disciplined by exit.

Here is Luke himself, on Knots gaining node share:

"And Knots can lose it just as fast if we do something truly insane. What we need next as soon as we get through BIP110's activation, is real alternatives so users have options."

The man accused of seeking a throne is publicly asking for more competitors.

Skin in the game

Hoppe's pro-Monarchy argument maps onto this fight perfectly. An owner-king treats his realm as capital, something to preserve and pass on, so he governs with a long time horizon. A democratic caretaker holds temporary use of something he does not own, so his incentive is to extract value while his term lasts.

Ownership aligns; stewardship-for-hire extracts.

Apply that to Bitcoin's software. Jeff Booth recounted recently that Gloria Zhao, until recently a lead Core maintainer, did not own bitcoin. A maintainer whose US dollar paycheck comes from grants and sponsors, and whose net worth does not move when Bitcoin's monetary properties degrade, is at best a caretaker. A group of developers can be perfectly competent and still be beholden to whoever signs their checks.

Luke is the owner type. He holds bitcoin. He is, by his own bio, married and the father of 11 children, which is skin in the game in the most literal sense: a large stake in the world his money has to survive into.

And note carefully how he is paid. Luke is CTO of Ocean, which took a $6.2 million seed round led by Jack Dorsey. He is paid to work on Ocean. He is not paid to work on Knots.

Investment in an independent company is not a salary with performance reviews. Dorsey cannot fire Luke from Knots, because nobody can. It is his project, which is exactly Hoppe's point.

What a king cannot do

If Luke is "king" of Knots, let's be precise about the limits of his crown.

Luke cannot force a single node to run Knots. He cannot stop anyone from forking his code; Knots itself is proof that forking works. And he cannot survive doing, in his own words, "something truly insane."

The only absolute power in Bitcoin belongs to node runners. That is the entire point of BIP-110.

As the market dumps Core and moves to Knots, why should we fear that it couldn't move again to another implementation if that was required? It is trivially easy to fork Core or Knots. Earning the faith of the market is the hard part. Rational actors can and should move implementations when they believe their economic interests are no longer being defended by the maintainers of the software they run.

If Luke turns out to be a tyrant, then you can depose him by simply switching to another client, of which there are several available or in the works: The Bitcoin Commons, Production Ready, btcd, libbitcoin.

Run a node. Choose your implementation deliberately. The signaling window opens in weeks.

For now, and until I have a reason to change my allegiance, I'm with Luke and Knots. And if he ever loses his way, I will move on, and you can say it with me: "the king is dead, long live the king."