Some things in a cat's world are governed by time rather than permission. The garden gate that stays shut until the builders finish. The room that's off limits until the paint dries. Tuga cannot argue with any of these, and neither can I — the constraint isn't a decision somebody keeps making, it's a condition that hasn't been met yet.

Bitcoin can do this with money. You can create coins that cannot move until a specific moment, enforced by the same consensus rules that enforce the 21 million cap, with no company, lawyer or trusted party involved.

This is the most underused feature in self-custody, and it solves problems that keys alone can't.

1. What a Timelock Actually Is

A timelock is a condition attached to coins that makes them unspendable until a point in the future — either a block height (say, block 1,100,000) or a timestamp.

Two flavours, and the distinction matters:

• Absolute timelocks point at a fixed moment. These coins cannot move before block X.
• Relative timelocks count from when the coins were received. These coins cannot move until 6 months after they landed here.

Nodes enforce it. A transaction that tries to spend early is simply invalid — not rejected by a policy, invalid the way a forged signature is invalid. You cannot appeal to anyone, including yourself, which is precisely the point.

2. What It's Good For

Forcing your own hand. The oldest use: locking savings so that a panic at 3am cannot reach them. You can want to sell as much as you like; the coins won't move until the date you chose while calm. This is a commitment device in the classic sense — you're binding your future self because you know what your future self is like.

Inheritance without a lawyer. The strongest case. A wallet can be set up so that you can spend at any time, but if you don't touch it for, say, a year, a second key — your partner's, your child's, your executor's — becomes able to spend it. Your normal use keeps resetting the clock. Your absence eventually opens the door.

This is a genuine deadman switch, and it solves the problem in my inheritance article far more elegantly than instructions in an envelope: the fallback key has no power while you're active and needs nobody's cooperation once you're not.

Vaults. More advanced: an arrangement where moving coins out starts a delay during which a second key can cancel the transfer. A thief who compromises your main key can start a withdrawal but can't complete it before you notice and claw it back. Real vault designs are still an active area of work, but the pattern — theft becomes visible before it becomes final — is the direction.

3. How You'd Actually Do It

You don't need to write scripts by hand. Wallets like Liana are built specifically around timelocked recovery paths, and Sparrow and Specter can handle miniscript-based policies of this kind. The setup is a policy — "key A always, or keys A+B after 12 months" — rather than code you author.

The practical shape of a good inheritance setup:

• Your everyday key, which can spend at any time.
• A recovery key held by someone you trust, or split between two people, which becomes valid after a long inactivity window.
• A habit of moving the coins to a fresh address once a year, which resets the relative timelock and proves you're still here.

That annual movement is the whole maintenance burden. Put it in the calendar next to the recovery drill you're already doing.

4. The Warnings, Because They're Serious

This is the most "here be dragons" article I've written, so:

• There is no undo. If you lock coins for two years, they are locked for two years. Not for you, not for a developer, not for anyone. A typo in a block height is a real and permanent problem.
• You must back up the policy, not just the seeds. Like multisig, a timelocked wallet needs its descriptor — the full policy description — to be reconstructed. Seeds alone may not be enough. This is the most common way people lose access to setups like this.
• Test with trivial amounts, on a short lock, first. Lock ten thousand sats for an hour. Watch it refuse to spend. Watch it become spendable. Only then do it for real.
• Block heights are approximate time. Blocks average ten minutes but drift. A lock "for a year" in block terms might land weeks either side. Fine for inheritance; not fine if you need a precise date.
• Complexity is a risk, not a feature. Everything in the multisig article applies here doubled. If you don't fully understand the setup, you have built a more elaborate way to lose money.

5. Who Should Actually Bother

Not most people. That's the honest answer.

If you're holding an amount where losing it would be unpleasant but survivable, a well-executed single-sig with a metal backup and a letter is the right tool, and adding timelocks adds risk without adding much.

The threshold is roughly the same as for multisig: an amount large enough that its permanent loss — to your own panic, or to your own death — is the dominant risk. At that point, a mechanism that enforces your intentions without needing anyone's cooperation is worth the learning curve.

The Point

The gate stays shut until the builders finish. Nobody enforces it in the moment, nobody can be persuaded, and Tuga's opinion is not part of the process. The condition simply hasn't been met.

That's a strange and useful thing to be able to do with money. Not a lock somebody holds the key to — a lock made of time, that opens on its own, exactly when you said it would.

A promise you can't break is worth more than one you can. 🐾⚡