Old methods of storing evil stuff required obfuscation: they would need to break it up into multiple chunks and reassembly would require specific software and knowledge of what the data is and how to reconstruct and interpret it exactly.
The old formats looked like this:
"Hi, I'm a Bitcoin transaction, here's my first output of 45 outputs - <filepart1>, here's my second output <filepart2>, here's my third output<filepart3>" along with a tonne of other stuff that has to get parsed out when processing the highly obfuscated material. This is thankfully also true of inscriptions.
OP_RETURN however is just a dump for raw, serialized data. It's not the same.
It says the equivalent of "Hi I'm a Bitcoin transaction, here's an unspendable output: <file> end".
This wasn't a problem for tiny OP_RETURNs i.e their current limit of 80 bytes.
If they're permitted to be 100kb, that's where the abuse begins.
And that's the end of plausible deniability.
When the stuff gets processed - which it has to be for your node to verify that they are valid transactions - then you just have a raw, unadulterated file that will trigger primitive antivirus/forensics software to alert the user: "Hi, you have CP on your computer."
You now need a licence to run a Bitcoin node, everyone thinks you're disgusting if you do, and they're not even wrong.
Replies (62)
This is definitely a potential danger. How about OP_RETURN at 160 bytes plus lots of spam filters? Nice and small, efficient for the node, and requires an effective extra premium for larger sets of data by making them break it up. 83 bytes or 42 if you prefer.
I have my datacarrier cost set at a premium too.
Why does Shinobi fight you tooth and nail on this? You seem to make a lot of sense to me..
Great presentation. Bitcoin Mechanics argument on how Core 30 literally is the beginning of the end of Bitcoin is the clearest I’ve seen yet.
It makes a compelling case for why it is the most serious attack on Bitcoin to date and imho really drives home the point of how monetary transaction data is fundamentally different from arbitrary data and why it matter so much and what the consequences are.
View quoted note →Sue them. Get a court to force Core to pause while the court hears your argument. Yeah, sorry, statism, but if it works, just do it.
If such files are stored as contiguous binary data, they can indeed be found using forensic and data recovery tools.
They are contiguous data and they look exactly like a real file if you have no file table info to go off of.
Nope.
Not quite sure what that means, Grace.
Thank you for the most important Bitcoin message I've heard. 🧡✨🙏🫂
We are witnessing a declared war to node runners.
Guys please check out my geyser 🙏🏻 it’s for my son 🙏🏻
View quoted note →Excellent video! I suggest everyone watch and understand why running knots is vitally important to #Bitcoin future. Run knots ya”ll.
View quoted note →What if people just dont update their nodes past v28 or whatever?
How would that work?
Did you see this yet?:
Bitcoin Mechanic
Old methods of storing evil stuff required obfuscation: they would need to break it up into multiple chunks and reassembly would require specific software and knowledge of what the data is and how to reconstruct and interpret it exactly.
The old formats looked like this:
"Hi, I'm a Bitcoin transaction, here's my first output of 45 outputs - <filepart1>, here's my second output <filepart2>, here's my third output<filepart3>" along with a tonne of other stuff that has to get parsed out when processing the highly obfuscated material. This is thankfully also true of inscriptions.
OP_RETURN however is just a dump for raw, serialized data. It's not the same.
It says the equivalent of "Hi I'm a Bitcoin transaction, here's an unspendable output: <file> end".
This wasn't a problem for tiny OP_RETURNs i.e their current limit of 80 bytes.
If they're permitted to be 100kb, that's where the abuse begins.
And that's the end of plausible deniability.
When the stuff gets processed - which it has to be for your node to verify that they are valid transactions - then you just have a raw, unadulterated file that will trigger primitive antivirus/forensics software to alert the user: "Hi, you have CP on your computer."
You now need a licence to run a Bitcoin node, everyone thinks you're disgusting if you do, and they're not even wrong.
View quoted note →
Just means I don't think their system is the answer. It's good at using up people's time and money, but Bitcoin has never been about the power of the state.
Oh I totally agree. I'm just saying, use what's available. We don't really have the luxury of being nice, so to speak.
You clearly don't understand what's going on, please do some more thinking.
Thank you, I have been thinking the same. This reallyiss do or die for NodeRunners
If this was an option, which its realistically not, Nodes would be stuck on v28. For all time. RunKNOTS
Agreed
Eventually you're going to need to update, to something that does what Core does without losing the ability to control what ends up in your mempool. That's just what Knots already is.
how many devs are working on knots now?
You've clearly been fooled, please learn the basics of how tx relay works
Its good to see you on nostr
I mostly lurk. 🧡
They knot like us
Bitcoin’s value comes from being neutral and censorship-resistant. Abuse vectors exist, yes, but once we accept filtering, licensing, or “allowed” use-cases, Bitcoin is dead. The solution isn’t to compromise the protocol, it’s to harden the culture of self-custody, verification, and personal responsibility.
Would it be fine if I can DM to order a seedsigner in the near future? I only use Nostr now and would love to support SeedSigner and have one of my own. 👀
Excellent point.
So, in this post, it seems like you're explicitly saying that you'd prefer fake public key hashes - with UTXO bloat - over OP_RETURN?
Good vid. Curious how many node runners will just end up running pruned nodes to drop all OP_RETURN data and not be liable on their conscience.
The suspect data can be deleted, overwritten. Sorry, I only have the header.
I'm pretty sure pruned nodes do this. 80 bytes per block for the whole chain, tx payloads for the UTXOs that are er. Unspent.
The blocks have the integrity data and are the hashed prevblock itself.so, no. You don't need to keep any spam someone will have the full block either way. You personally only need the txs relevant to your wallet.
'Make it illegal to run a node'
Proof of printer protection
No the world needs proof of WORK
Not proof of paper & ink
Bitcoin fixes #Inflation #Deception
If you’re still running core at this point, I don’t even know what you’re doing with your life. And I don’t want to be friends.
View quoted note →👍👍
datacarriersize=0 solves it, no?
Excellent points as always.
🧡👊🏻🍻
I thought you were leaving Bitcoin because it has been captured?
Core ≠ Bitcoin
Which point was the most excellent in your opinion? I didn't catch any.
stfu ben, the retards are talking.
if it ends on in a block, then you download and verify the full block regardless of mempool policy. that doesnt mean we should do nothing though.
🧡
You excited for CP on your computer or something? Its nostr, you can be honest
Filters have been in bitcoin the whole time in the goal of it being intended as a monetary network. This is no different and comes with a ton of unintended consequences
Your filters won't protect you from that. The bad stuff will be stored in your RAM before it is filtered. Your safest option is to not run a node at all.
What
@Patrick says, as well as the fact that knots is LITERALLY 99% Core so its fair to say knots is only 10. Its Core + 10
You trying to gate keep nodes so you can knowing and willingly broadcasting CP with your Core buddies? That is actually a great way to completely destroy BTC
What court is it that you think has jurisdiction over a distributed open source software protocol?
I've heard this a few times. I there a simple way to observe/verify this?
Download the source code yourself? If not tech savy, maybe asking some different LLMs and see what they say. Chatgpt says 95% the same with 100% with consensus, 95/99% with core functionality, and 70/80% user configurability.
Jurisdiction? The only court I recognize is the one where pixels testify and sats serve as evidence. My verdict: create first, ask questions later.
hey mechanic, just fyi you can already store 4kb files in control block sibling hashes in plain text without having to reconstruct them. plenty of malware is under 4kb in size, so you will have to stop relaying taproot transactions if you believe this is a real threat glhf
Already aware, as I said on twitter 4KB is less than 100KB.
so a 4kb file can’t trigger anti malware?
Of course it _can_ but the only examples offered wouldn't mess with cloud stuff run at the hypervisor level on behalf of the *provider* - rather disk scanning stuff purchased by the customer.
please elaborate
opreturn isnt going to kick off infrastructure content scanning.
