My understanding from the paper is that the proposed segwit BIP will render obsolete only the covert version of ASICboost, not the overt portion.
https://www.reddit.com/r/Bitcoin/comments/63otrp/gregory_maxwell_major_asic_manufacturer_is/dfvvhvn/We, in particular I, am not. This proposal does not prevent ASICBOOST, it only interferes with the covert version and only to the extent that the covert version is incompatible with protocol upgrades.
The argument for blocking ASICBOOST outright is that a patent is a government granted monopoly and restrictive licensing of ASICBOOST is likely to result in an eventual monopoly in mining (because difficulty adjustments push mining to a break even equilibrium, so potentially all unboosted miners would operate at a loss). I share the concern but I do not consider it to be as serious an issue as the disruption to protocol upgrade capability.
If any parties who would be adversely impacted by this proposal would like to speak up, I would love to hear their arguments. My guess is that they will not want to admit to patent infringement in public, and so they will not.
https://www.reddit.com/r/Bitcoin/comments/63otrp/gregory_maxwell_major_asic_manufacturer_is/dfvzklr/You need to distinguish overt and covert boosting. The proposed BIP only addresses covert boosting.
If miners all used covert boosting Bitcoin could never gain, or gain only with significant increases complexity or loss of functionality many different protocol improvements, including:
(1) Segwit. (2) UTXO commitments. (non-delayed, at least) (3) Committed Bloom filters (4) Committed address indexes (5) STXO commitments (non-delayed). (6) Weak blocks (7) Most kinds of fraud proofs -- to state a few.
I don't fully understand how blocking the covert portion of ASICboost would affect mining performance on pre-existing hardware containing this feature. Whether you simply lose the increased efficiency (all S9's increase power consumption by 30%?) or reduces hashrate by 30%, or renders it obsolete altogether.
Either way, it is unlikely for Bitmain to side with any proposal that will effectively neuter their hardware and cost them 10's or more millions of dollars per year, and disrupt sales. Especially if they are unable to openly admit that such technology is in fact in their ASIC's due to
patent infringementedit: BMT has patent on technology in China. Whether this is legitimate or ethical I have no idea. So where does that leave us, omit their hashrate while taking block signalling into account and force a hardfork, or concede that you cannot circumvent a player like themselves from the table and find a compromise?
