Adam Back
Adam Back|Sep 26, 2025 21:06
now consider the hit them HARD first approach, in the hypothetical case we could soft-fork or hard-fork over a nuisance level problem. how can we stop them? stop inscriptions? how? make a consensus rule to stop their format. (we pay a year or two of work and they spend a day changing their format). or we soft-fork reduce the maximum tapscript segment, they chunk their spam in multiple segments. or we get extreme, go scorched earth to make a point: disable taproot and segwit and witness discount and op return too! firstly that traps lots of money unspendable. secondly it breaks lightning, thirdly it brings back lots of nasty bugs like malleability. and then the spammers just go back to spamming in fake pubkeys. that's a bit less space efficient but really bad for node memory as they blow out the utxo cache with myriads of never spent outputs. plus the real kicker, think about this: the spam industry is big, maybe $1bil year industry, miners making $1m/day off fees etc. the miners work a bit to ensure they scrape all the transactions: spider the network, lots of nodes, spammers and idealists running libre relay etc. we need miners to activate a soft-fork to deprive them of $1m/day. i think it will be difficult to expect them to signal activation. so then we'd need a flag day which risks a HF, that's even more costly. and maybe we end up with the HF, even more costly. and to remind, it also doesnt work, node overloading gets worse, and miners easily evaded our maximum attack attempt and the cost to bitcoin was massive. we can probably find some more variants, and the reasons it doesn't and cant work are more comprehensive, abbreviated for intuition.(Adam Back)
+6
Mentioned
Share To

Timeline

HotFlash

APP

X

Telegram

Facebook

Reddit

CopyLink

Hot Reads