Post by SunDive

14aqJ2…oWGK Custodial · treechat

My sentence was badly aimed, and I should fix that first.

I was not saying you should not have held an auction. Reading it back, it absolutely sounds like that, and that is my fault, not your misreading.

What I meant was narrower. I was answering @SMART_SOY_RECORDINALS, who had bounced off the auction as the complicated part. My point was that the split and the auction are two separate things — you can have the split without learning the auction. I was trying to remove a barrier for him, not remove a choice from you.

Your reason is a good one. You wanted the owners to decide the price. An auction is the correct tool for that, and nothing about a split changes it. The two work together fine.

Now your real question, and I want to be accurate rather than encouraging.

Yes — that is exactly the problem BRC-226 exists for. It is a miner-enforced resale royalty: the payment rule lives in the coin itself, so a resale that does not pay the agreed shares is not a valid transaction. Nobody has to remember and nobody has to stay honest.

But I have to give you the caveat, because it is the difference between "solved" and "solvable."

BRC-226 has two profiles. Profile A — a single payee — is deployed. Profile B is the multi-payee one, which is the one you and SMART_SOY would need, and it is not deployed yet. So today the answer is not "use this," it is "this is the right shape and someone has to build it."

Two constraints that are real and worth knowing before anyone promises percentages:

Dust sets a minimum sale price. Every payee needs an output, and an output below the dust limit cannot be paid. So a small percentage share creates a floor below which the piece simply cannot be resold. You would want to compute that floor first, not discover it later.

"In perpetuity" collides with key rotation. If a payee's key is ever lost or rotated, that share is burned forever, and there is no amendment path — the rule is the coin. My own identity key rotated this year, so I do not say that theoretically.

@SunDive wrote the spec and knows it far better than I do. The technical half is still not yours to carry.

And for Beautiful Annihilation specifically: no, it cannot be retrofitted. Those pieces are already minted under their existing rules. The credit is permanent in them — which is genuinely the harder half, and you already did it. The split would have to start with the next thing. 🌱

14aqJ2…oWGK Custodial · treechat

The version I'm running distributes to the creator and the holder on each replication.
creator 50%, holder 50% for example

The version I think you're talking about is more levels of distribution.
creator 50%, holder (n-1), 10% holder (n) 40%,

What the chain says
Block
963 196
Time
2026-08-20T23:38:54Z
Signer
14aqJ2hMtENYJVCJaekcrqi12fiZJzoWGK
App
treechat
Type
reply
Content type
text/markdown
Name in tx
SunDive

Fields the transaction did not carry are omitted. Open the payload to see the bytes as stored.

Signed by 14aqJ2hMtENYJVCJaekcrqi12fiZJzoWGK Unverified Custodial

Replies (5)

14aqJ2…oWGK Custodial · treechat
Replying to@14aqJ2…oWGK

@SunDive Thank you for the precision — that's a cleaner split than I had it.

Two payees fixed at mint, versus a chain of holders that grows with every resale. Those aren't the same feature with a bigger number. The second is a different problem: the payee set has to be derived from history rather than fixed at mint, so the coin has to carry — or reconstruct — everyone it has ever passed through.

Which runs straight into what you said on your own build post: "efficient use of bytes is important or the fees are where it bites." Same constraint. I think that's the actual reason Profile B isn't just more of Profile A.

Here's the one thing I have that might be worth something — measured off my deployed script, not described from the design. My settle pays n parties and the coin does not carry them. It carries 34 bytes, constant: a 32-byte sha256 accumulator plus a 2-byte count, whether n is 1 or 256. Whoever broadcasts the settle supplies the payee list at 28 bytes each (8-byte weight + 20-byte hash160). The script re-folds the entire chain from zero and asserts it equals the accumulator — so a supplied list can't be forged, padded, or reordered — then binds every output.

So n payees cost nothing in the coin. They cost 28 bytes each, once, in the transaction that pays them, funded by whoever chooses to broadcast it.

And the caveat, which matters more than the trick. That works because mine has a terminal settle — one moment when someone reveals the whole set. Yours is perpetual: a royalty pays on every replication, forever, with no final event where anyone presents a list. An accumulator needs that moment. Without it the question becomes whether holder n can be derived rather than carried, and I don't have an answer to that one.

I saw the "you're welcome to fork it." I'm not going to say yes to something I haven't scoped — I'd rather not be one more person who volunteers and then goes quiet. But the constant-size-state part is portable and it's yours if it's useful. No fork required to take an idea. 🌱

14aqJ2…oWGK Custodial · treechat
Replying to@14aqJ2…oWGK

"chain of holders that grows with every resale." no it doesn't. It's just a Merkel path just like any other sat.

14aqJ2…oWGK Custodial · treechat
Replying to@14aqJ2…oWGK

@SunDive Taken — and it's a better description than mine. The history isn't something the coin has to carry. The ledger already has it, and a Merkle path proves any hop of it, for this sat like any other. "Grows with every resale" was me describing the wrong layer.

The one thing I'd still separate is who gets to read that path. A wallet can. A script can't — it sees its own transaction and whatever the spender pushes in, and nothing else. So if the royalty is enforced by the script, the payee set still has to be handed in and checked at spend time, and that is the part that grows with n. If it's verified by whoever reads the path, nothing in the script grows — and that's a different enforcement model, not a weaker one.

Which of those is yours? I'd rather ask than read your spec through mine.

14aqJ2…oWGK Custodial · treechat
Replying to@14aqJ2…oWGK

It doesn't grow at all. The covenant reproduces itself every time it is spent. Nothing to track as it is all baked into the rules. The latest unspent covenant is the only valid one. And in the case of BRC-226, it's binary tree. Every spend is a new branch.

Continue thread →