Im rather sure this won't work bc of multiple reasons:
All relays you connect to must implement this detection (or/and filtering)
NIP5 -as far as my understanding goes- isn't failsafe or intended for this puprose (they are identifiers, not verifiers. And appear spoofable?)
The suggested bot detection in the script is not appropriate for detection of the behavior as they mangle the OP, needing to both scan the OP and all replies to be somewhat succesful. Meaning determining the spam behavior isn't so trivial and can easily be circumvented by spammer once they notice it stops working and will change the mangling algo.
Maybe determining based on post create date + reply create date is more efficient, as they appear to reply almost instantly once a post launches. But again, this is easily circumvented.
Unlike traditional anti spam measures - which are specific to centralised systems mostly - you can't block a user from creating an account. Nor have a centralized store of blocked IPs or other metadata. Or detecting bots the cloudflare proof of work way (browser checks) also isn't possible.
So. That's where my understanding stops. Nor do I have ideas.
#nostr #asknostr might check my claims for correctness tho
/ 0x80085
npub1n8…mfqkr
2025-06-03 15:52:32
in reply to nevent1q…q30s
Author Public Key
npub1n8qwa5x82cst04y0lfzemz3g0tpvpapy837zvrzrkmmswd9t5r3sjmfqkrPublished at
2025-06-03 15:52:32Event JSON
{
"id": "d5ec4671e4e02815c6380aa9d425e70a32b08662dca6516062106f5d736c8347",
"pubkey": "99c0eed0c75620b7d48ffa459d8a287ac2c0f4243c7c260c43b6f70734aba0e3",
"created_at": 1748965952,
"kind": 1,
"tags": [
[
"e",
"63b1a2e8866889ce1a2233768051955a5f4fbabf183babd035eeac90c3e2d431",
"",
"root"
],
[
"p",
"8671a6e5188f73e7746ccf0dd169bf4a6785438b72c121f83bfffd04f88194d1"
],
[
"t",
"nostr"
],
[
"t",
"asknostr"
]
],
"content": "Im rather sure this won't work bc of multiple reasons:\n\nAll relays you connect to must implement this detection (or/and filtering)\n\nNIP5 -as far as my understanding goes- isn't failsafe or intended for this puprose (they are identifiers, not verifiers. And appear spoofable?)\n\nThe suggested bot detection in the script is not appropriate for detection of the behavior as they mangle the OP, needing to both scan the OP and all replies to be somewhat succesful. Meaning determining the spam behavior isn't so trivial and can easily be circumvented by spammer once they notice it stops working and will change the mangling algo.\n\nMaybe determining based on post create date + reply create date is more efficient, as they appear to reply almost instantly once a post launches. But again, this is easily circumvented.\n\nUnlike traditional anti spam measures - which are specific to centralised systems mostly - you can't block a user from creating an account. Nor have a centralized store of blocked IPs or other metadata. Or detecting bots the cloudflare proof of work way (browser checks) also isn't possible. \n\nSo. That's where my understanding stops. Nor do I have ideas.\n\n#nostr #asknostr might check my claims for correctness tho\n\n",
"sig": "f81595ccd9e8da01285e34560dd01d19268583b46ce921ef33fd327a155ed84089351eaacffa321ace18ce3f5358de11a5e850052e8490906a58595186705d6e"
}