gradientskier

Post Nubila Phoebus. Developer.

[gradientskier @gradientskier]

@BtcpayServer Can be exploited if btcpayserver is not exposed to the internet, but LND node is exposed on the lightning port (9735)?

[gradientskier @gradientskier]

@krakenfx Please put the “best bid price” button back!

[gradientskier @gradientskier]

@peter_szilagyi top right without “dot stand” in light gray. Can’t read well and is also different format.

[gradientskier @gradientskier]

@giacomozucco Indeed, but is bolt12 multi-asset? AFAIK it’s not. How could you receive stablecoins (as claimed in the quoted tweet), or any other asset that is not btc, with bolt12?

[Kügi @mkuegi]

its my internal name for the idea of a new OP_RETURN.

inspired by LFS in git. Add an OP code that defines "this is just random data, this is the hash and relevant meta data, if you want you can download the data, but you don't have to to verify the validity of the tx".

And as you suggested make the bytes in OP_LFS count less, so that its cheaper to use that instead of fake UTXO signatures etc.

[gradientskier @gradientskier]

@MrKukks @TTrevethan @bergealex4 @benthecarman @arkade_os @BtcpayServer AFAIU safety and liveness of the Arkade bridge (from VTXO to onchain UTXO) depend on Arkade Signer (or TEE), and covenants could remove the role of Arkade Signer (and TEE) from the whole design of the bridge protocol, reducing interactivity and trust. Am I missing something?

[gradientskier @gradientskier]

@TTrevethan @MrKukks @bergealex4 @benthecarman @arkade_os @BtcpayServer From the doc: The Arkade Signer is an independent entity that manages the cryptographic keys used for cosigning user transactions. Operates within a TEE and generates a single key that all Arkade addresses require for VTXO cosigning, isolated from operator control.

[gradientskier @gradientskier]

@NickSzabo4 @maxtannahill @Bitcoinpoet_JP @phyrooo @lorandimecs @Excellion @_jonasschnelli_ @adam3us There are financial use cases that require storing data on the chain. Bitcoin needs a financially incentivized OP_RETURN, and users need the ability to retain the OP_RETURN data they use, eg based on protocol/signer, and to prune everything else.

[Nick Szabo @NickSzabo4]

If Core added feature(s) for archival node operators to optionally and non-disruptively delete OP_RETURN content, and incentivized content creators to move from witness data to OP_RETURN data, then the increase in the OP_RETURN allowance would reduce legal risks. I am not in favor of Core's approach until they do that.

[JJS @JJS36038159]

Nick, there are THREE different concepts being conflated in this discussion.

Let me distinguish them:

TYPE 1: Standard Pruning (Already Exists)
Deletes entire old blocks after validation. Keeps only recent blocks + UTXO set. Reduces storage from ~650GB to ~10GB.

Limitation: Cannot bootstrap new nodes (missing historical blocks).

Status: Built into Bitcoin Core since 2015

TYPE 2: OP_RETURN-Specific Pruning (What Ghost Describes)
Keeps block structure and transaction history, but removes only OP_RETURN data from stored blocks.

Limitation Ghost correctly identified: Still needs to download OP_RETURN to bootstrap itself, cannot serve OP_RETURN data to new nodes.

Status: Possible but not implemented

TYPE 3: Prunable Lanes (Economic Incentive Approach)
This is NOT a pruning mechanism.

It's an economic incentive layer that guides where data goes in the first place.

Why Bitcoin Core v30 Changed:
Bitcoin Core developers faced a reality: the 80-byte OP_RETURN limit wasn't working.

Large miners were already ignoring the limit (more fee revenue)

Users bypassed the public mempool, dealing directly with miners (centralization)

Data was going into witness/taproot anyway (unprunable formats)

It acknowledged what was already happening and removed a restriction that was creating centralization pressure.

The intent: Align relay policy with miner behavior to reduce centralization, not to encourage spam.

The Core Economic Problem (Why OP_RETURN Isn't Used):

Witness/taproot data receives a 75% discount (segwit witness discount), making it 4x cheaper per byte than OP_RETURN.

Current pricing:

Witness/taproot data: 1/4 weight (75% discount, but unprunable - stored forever)

OP_RETURN data: Full weight (no discount, but prunable)

This is backwards.

The cheaper option (witness) forces every node to store data forever.

The expensive option (OP_RETURN) could be pruned, but nobody uses it because it costs 4x more.

Removing OP_RETURN or limiting inputs doesn't fix the economic incentive. Data will continue flowing to witness/taproot because it's cheaper.

The witness discount is the root cause. Fix that economic signal, and behavior changes.

The Prunable Lanes Approach:
Add a fee discount for OP_RETURN outputs (while maintaining witness discount for its original purpose - encouraging segwit adoption and reducing UTXO bloat).

Result: OP_RETURN becomes economically competitive with witness for data storage.

Current state:
Data goes to witness (cheap, unprunable)
Every node forced to store forever
No relief, no choice

With prunable lanes:
OP_RETURN becomes economically competitive
Data flows there naturally (price signal)
Nodes choose: archival (store all) or pruned (drop OP_RETURN)

Network maintains bootstrapping capacity through nodes that choose archival

Ghost's limitation (can't bootstrap) is real for pruned nodes. But it's not fatal because:

Not every node needs to bootstrap

Network maintains archival nodes through self-interest and altruism (businesses need full history, some operators want to provide service, storage costs are manageable for those who choose)

This model already works with current pruning (we have thousands of full archival nodes despite pruning being available since 2015)

The core issue?

Witness discount makes unprunable storage cheaper than prunable storage.

[Giacomo Zucco (Bear Market Edition) @giacomozucco]

Ok, but I'd add some relevant details to this nice metaphor.

The engineer who originally built the airplane disappeared mid-flight, parachuting off. Then the second most influential engineer went rogue and was kicked out of the plane by the others (and the passengers), together with other 2 very important engineers (Jeff/Mike). Then two of the original engineers, namely the second most prolific ever (Luke) and one of the most influential (Peter), started disagreeing vocally, in front of the passengers, about how to manage the plane windows (the mempool). Then one of the main engineers unintentionally almost crashed the plane (inflation bug), and the others didn't see it coming because of reputation/trust dynamics. Btw, the same engineer launched a DEI campaign to hire black obese bisexual women with acne from the passengers to maintain the plane, because "diversity". Then another engineer went around changing the wording in all the emergency instructions, because he was afraid to be called "racist", and some passengers started having some mild doubts about the whole team. Then most of the other experienced OG engineers disappeared and retired, parachuting off the plane. Then two of the remaining OG engineers started going around the plane dressed like women (not an avionic issue per se, of course, but passengers may have prejudices and feel uncomfortable, so it's worth mentioning), and another one invited his girlfriend with less than 2 years of experience in avionics inside the cabin to become lead engineer (which says nothing about her skills, but again, you could expect some passengers to raise eyebrows). Then everybody on the plane realized that engineers botched up the installation of a screen against intense sunlight (spam). One off the two fighting OG engineers (Luke) freaked out in front of the passengers about "intense sunlight destroying the plane", and proposed to fix the screen, while the other (Peter) was happy it was broken and considered the bug actually a feature, because he had (rational) arguments about the screen not working at equilibrium anyway, and possibly making things even slightly worse. The screen didn't get fixed, but all emergency instructions were changed to reflect the behavior, with most engineers mocking passenger for even pointing out the screw-up. Then the contrarian engineer who proposed the fix was robbed of his wallet, and an FBI agent on board said it may have been one of the other engineers (of course it could have be somebody else, including him: never trust cops). Then the same engineer solicited a hip-round to mitigate the issues of a broken engine (mining centralization), and another engineer publicly called for the rich passengers (Jack) to withdrawal funding from the initiative, due to unrelated past personal beefs. Then, AND ONLY THEN, most of the engineers set out to make it an absolute priority to change the color of the paint around the windows, in a way that the contrarian engineer considered "fatal", and many passengers just disliked. After a huge push-back, they withdrew the controversial proposal, just to rush to implement another one, almost identical in effects but worded in a subtler way. Now, a growing, substantial minority of passengers are looking at the contrarian engineer as "maybe not as crazy as they thought". But it's not about him being very persuasive, or popular, or good at marketing. It's mostly about them being fed up.

Here, fixed it.

[David Seroy 🏔️ @david_seroy]

@LukeDashjr ShieldedCSV seems like a legitimate use-case. Perfect privacy payments.

It requires nullifiers, which are basically compressed blobs to be posted on-chain. Do you not see this is a legitimate use case?

[Bitcoin News @BitcoinNewsCom]

NEW: Tether CEO Paolo Ardoino explains why he prefers to build on top of the Bitcoin Lightning Network rather than other trendy Layer 1’s

[gradientskier @gradientskier]

@MrKukks @bergealex4 @sethforprivacy @benwgold @spark @ArkLabsHQ @secondhq how do you enforce the “covenant” in the round transactions, that is the fact that the round root tx can only be spent by a tree of txs whose leaves are the actual VTXOs? Who pre-signs the round transactions of the tree?

[gradientskier @gradientskier]

@bergealex4 @sethforprivacy @benwgold @spark @ArkLabsHQ @secondhq Would non-recursive covenants a-la-CTV solve the issue of being online for long periods of time? I understand once CTV or the like is a thing, users caring about security of their funds can refresh with an online interaction with the server and then go back offline. Is it true?

[gradientskier @gradientskier]

@bergealex4 @sethforprivacy @benwgold @spark @ArkLabsHQ @secondhq Would non-recursive covenants a-la-CTV solve the issue of being online for long periods of time? I understand once CTV or the like is a thing, users caring about security of their funds can refresh with an online interaction with the server and then go back offline. Is it true?

[gradientskier @gradientskier]

@bergealex4 @benwgold @sethforprivacy @spark @ArkLabsHQ @secondhq You mean “Ark technically can have a better trust model”, right?

[gradientskier @gradientskier]

@bergealex4 @benwgold @sethforprivacy @spark @ArkLabsHQ @secondhq Ark OOR payments are like a small Spark statechain. In theory, Spark is a permanent statechain, while Ark is a statechain until VTXO refreshed, so it gives flexibility to the client. A
concern with Ark is interactivity in rounds, which may easily make any Ark a Spark in practice.

[gradientskier @gradientskier]

@adam3us @GrassFedBitcoin @Zatoichi42 @adam3us Assuming one tolerates data in the blockchain,for liveness of L2s or CSV protocols, but does not want non-prunable outputs. Will unlimited OP_RETURN actually discourage taproot abuse, or the latter is still more economically viable? Why not a discounted prunable output?

[gradientskier @gradientskier]

@callebtc @btcplusplus What’s the main issue with this design? It seems a proposal from 2024, why it has not gained traction? Just a matter of implementation friction, or more deep concerns related to privacy?