Please read my email more carefully; the replay threat would be moot because there would be no alternative chain to replay the TX on,
In order to *get to that point*, you need >51%.
Not only that, but, if you started out with <51%, then you need >>51% in order to *catch up* and replace the large number of blocks added to the legacy chain in the mean time.
So, since >51% is _required_ for BIP148 to succeed (and likely >>51%)... you might as well do as SegWit did originally, or lower the threshold to 80% or something (as BIP91 does).
Without replay protection at the outset, BIP148, as far as I can tell, isn't a threat to miners.
--
Please do not email me anything that you are not comfortable also sharing with the NSA.
Please read my email more carefully; the replay threat would be moot because there would be no alternative chain to replay the TX on, as the non-148 chain would have been reorganized into oblivion.
-------- Original Message --------
Subject: Re: [bitcoin-dev] Replay attacks make BIP148 and BIP149 untennable
Local Time: June 7, 2017 3:26 AM
UTC Time: June 7, 2017 12:26 AM
I don't know what you mean by "render the replay threat moot."
If you don't have replay protection, replay is always a threat. A very serious one.
--
Please do not email me anything that you are not comfortable also sharing with the NSA.
Hmm, that's not the difference I was talking about. I was referring to the fact that using "post-chainsplit coinbases from the non-148 chain" to unilaterally (ie. can be done without action on the 148-chain) taint coins is more secure in extreme-adverserial cases such as secret-mining reorg attacks (as unfeasibly expensive they may be); the only large-scale (>100 block) reorganization the non-148 chain faces should be a resolution of the chainsplit and therefore render the replay threat moot.