Re: [POLL] Possible scaling compromise: BIP 141 + BIP 102 (Segwit + 2MB)
Bitcoin already has a sigops per block limit to mitigate the risk of this attack (despite the FUD that introducing such a limit is all that's needed to solve the problem, which is pretty dumb considering that's already what happens 
)
who the damned F*ck should be allowed to build a single tx that uses 20% of a block!!)who the damned F*ck should be allowed to build a single tx has 16,000 sigops!!
Bitcoin is permissionless. The proper answer to "who the damned F*ck should be allowed to build..." is anyone who wishes to.
But another aspect of permissionless is the fact that no miner is compelled to include that transaction into a block. In fact, there is a natural disincentive to any miner including such a transaction into a block. Blocks that take a long time to validate are likely to be orphaned by other blocks which validate quickly.
but this is where the rules need to be defined. so that if a malicious pool did add it. the LOWER tx sigops would be knowingly rejected by consensus so pools wont bother.
by keeping it at 16,000 malicious pools 'could' accept it and know its a valid block and only have to worry about 'timing' as the disincentive, rather than RULES.
however if you prefer not to have RULES and just rely on belief or faith of pools.. then that is a weakness.
however if belief and faith were strong for such an occurrence not to happen. than quadratics has never been a problem and never be a problem because bitcoin is protected by belief that it will get orphaned due to 'time'.
meaning segwit doesnt need 100% of users to move their funds to segwit keys (weeks after activation) just to possibly get a segwit fix to fix it(never gonna happen anyway), because belief alone in 'time' has protected the network thus far and will continue to