Thread

your node stores whatever ends up in a block. If you don’t want CSAM on your disk then run a pruned node or don’t run bitcoin at all. even if its not in a flat contiguous format (op_return) you will still be able to recover images with a line of bash when it gets in your utxo set, like the bitcoin whitepaper. This you can’t prune and can be done without op_return. my take: you can’t stop it, adversarial actors will do it to attack bitcoin, stop giving them ideas. good luck!

You seem to know a lot about recovering CSAM from the UTXO set. I'm not interested in doing that myself personally. I just want to know if it is technically possible to exclusively prune op_return data

Good luck

I’m actually not aware of any script that does that, its not something that any sane bitcoiner would share. I only have done it with the whitepaper. op_returns aren’t stored in the utxo set by design. So if you have a pruned node it won’t keep it around. but you might still have it in the utxoset somewhere else if there is some not stored in op_return. an attacker likely would store them in the utxoset just to fuck with bitcoiners, not op_return

Check this video. Its very good analisys and touches on this point too. OP_RETURN can be pruned but its still on the full non pruned nodes which is bad. Also the other places where data can be stored can be fixed too. nostr:nevent1qqswx4g0mvh29pq2jszqaflekklvtsee3d3zwwxat4ag5k0e78ftgzspz4mhxue69uhhyetvv9ujuerpd46hxtnfduhszxrhwden5te0wfjkccte9enx7atww3skjm3wvekj7wgjyyx

No, it's a silly meme. Pruning OP_RETURN means dumping all archival data along with it. This is one of the most annoying misrepresentations about this. OP_RETURNs just don't exist in the UTXO set so people present them as harmless.

This is fine art. nostr:nevent1qqsfpwdekvqdp5p0gnz30kg9970g63rglqgyf4feamfa2sjxqzvyl3gpzpmhxue69uhkummnw3ezumt0d5hsz9thwden5te0wfjkccte9ejxzmt4wvhxjme0dldjxr

Mechanic you should do a video on this. Many people think prunable means they can remove from their node. The pruning is only for utxo validation but once in block it stays there forever. This means if you run full node you will have it. Already everyday so much junk is put in the OP_RETURN (even with <80bytes). I run a spam check (I have a simple ML classification model ) It detects | 🧬 Malware indicators : `MZ`, binary blobs, shellcode, etc. | 🧱 Overly large payloads : Likely trying to store malware | 🧾 Structured address spam: Recognizable crypto address formats | 🧟 Botnet / C2 : Random hex data, possible signal packets | 🧨 Garbage entropy :Nonsense strings, high entropy content

Our adversaries aren't idiots. We can either officially sanction 100KB file storage or we can legitimately present it as something we push back against. No technical difference so devs are blind to the world of difference in that.

why would an attacker care about an unenforceable "sanction"

So hypothetically if literally everybody starts pruning their nodes, does that make it impossible for anyone to spin up a new node? nostr:nprofile1qqsr9cvzwc652r4m83d86ykplrnm9dg5gwdvzzn8ameanlvut35wy3gpz3mhxw309aex2mrp0yhx5c34x5hxxmmd9uqsuamnwvaz7tmwdaejumr0dshszythwden5te0dehhxarj9ekxzmny9u0ljp2l nostr:nprofile1qqs8fl79rnpsz5x00xmvkvtd8g2u7ve2k2dr3lkfadyy4v24r4k3s4spz9mhxue69uhkummnw3ezuamfdejj7qg4waehxw309aex2mrp0yhxgctdw4eju6t09uqsuamnwvaz7tmwdaejumr0dshsu0ur8h

this would never happen because pruned nodes are a bit more inconvenient, since you need to redownload everything on rescans/index rebuilds. I still run a full archival node for convenience and so I don't have to redownload everything. you would only want to do this is you don't want to hold onto all the crap thats in the chain. you can still get your balance just fine without storing everything. there will always be copies of the full chain somewhere

Mechanics point proved