Goes live in …
Bitcoin Splice
A real Bitcoin fork where your own computer can mine again. Difficulty starts at 1 and re-tunes itself every single block, so the chain settles around whoever actually turns up: a laptop, a gaming PC, a Raspberry Pi.
Measuring your machine…
Your browser is slow at this. A real mining program on the same computer is fifty to a hundred times faster. The comparison holds either way.
Why this exists
Mining left home, and it was never coming back on its own.
Once upon a time you mined Bitcoin on the computer you already owned. That stopped being possible a long time ago. Today a handful of pools decide what goes in most blocks, the machines live in warehouses, and the honest advice to anyone who asks "can I mine?" is no, and don't bother trying.
Nothing in the software changed to make that happen. The maths is the same maths. What changed is one number, the difficulty, which grew until only an industry could reach it.
Then, this month, a group of miners split off from Bitcoin to enforce BIP-110. They took Bitcoin's difficulty with them but almost none of its mining power. In three and a half days that chain produced four blocks. Bitcoin produced 561 in the same stretch. The difficulty was simply set to a number nobody on that chain could afford.
Bitcoin Splice picks up that stalled chain and fixes the difficulty first. It starts at 1 and adjusts on every single block, where Bitcoin adjusts once a fortnight. Too many miners and it gets harder. Too few and it gets easier. Nobody sets it, nobody votes on it, and there is no schedule to be stuck waiting for.
Several other things had to change alongside it. Splice has its own network greeting and its own ports, so a Splice node and a Bitcoin node cannot reach each other even by accident, and its signatures are deliberately incompatible with Bitcoin's so nothing signed here can ever be replayed to move your real coins.
The effect is that a chain can start from nothing and find its own level in a few hours. That is exactly the thing Bitcoin cannot do, and exactly why a fork of Bitcoin normally dies on the first day.
It was only ever one number.
What happens
Point a computer at it and this is roughly how the day goes.
A block in minutes
At difficulty 1 a block takes about 4.3 billion tries. It sounds enormous. A normal desktop gets through it in well under a minute, so your first block will probably arrive before your coffee does.
It tightens up
Every block you find makes the next one harder, automatically. If more people join, it tightens faster. If everyone leaves, it loosens again. You don't configure any of this and neither does anyone else.
Ten-minute blocks
Within a few hours the chain settles at roughly one block every ten minutes, the same rhythm as Bitcoin, at whatever difficulty matches the people actually mining. Then it just keeps going.
Getting in
Three steps, and one of them is a download.
One Docker command. It's a normal Bitcoin node with a different difficulty rule bolted on.
The chain continues from a real Bitcoin block, so your node needs Bitcoin's history up to that point, and only a Splice node can serve it. Today that means one machine, so this is the slow bit. Everything after it is fast.
Three are known to work, unmodified: BFGMiner, cpuminer-opt and an ASIC. Give one your address and let it run.
The full instructions are short and every command is copy-pasteable.
Get started →Who runs this
One person. That's on purpose.
Bitcoin Splice is written and maintained by one developer, kosinskihenry. There's no foundation, no company, no working group and no weekly call.
Think about what actually killed the chain we forked. A committee did. Months of process built to manufacture agreement produced 2.53% support and a chain that stopped moving. That process had plenty of developers. It had nobody with the authority to decide.
One person can obviously be wrong, and probably is about something. But one person can't be lobbied into a compromise nobody wanted or stalled for eighteen months, and when they are wrong you can see it, name it, and copy the code and do it better. Fewer developers, fewer places for the argument to hide.
- Developer
- kosinskihenry
- Foundation
- none
- Company
- none
- Investors
- none
- Meetings
- none
- Coins premined
- 0
- If you disagree
- fork it
Questions