https://blog.bitmain.com/en/uahf-contingency-plan-uasf-bip148/Bitmain is announcing that if BIP148 succeeds, then they are essentially going to mine a new SHA-256 altcoin instead of Bitcoin. If you don't do anything, it'll be as if mining power dropped somewhat, but you'll otherwise be completely unaffected.
I feel like some people might read this statement as trying to force a decision between Bitmain's altcoin and BIP148, but you should be roughly equally unaffected whether or not you enforce BIP148. It might help BIP148's chances a little due to removing Bitmain from consideration and therefore increasing the chances of activating the traditional BIP141 SegWit deployment, though I continue to be pessimistic about BIP148's chances; I consider continuation of the status-quo and a somewhat later BIP149 UASF as the most likely outcome.
I kinda doubt that they're actually going to stick by what they say here, but if they do, it only seems like good news. If you'd like to use a currency controlled by a single company that puts its ability to use its patented Asicboost tech above all other considerations, feel free! I'll stick with Bitcoin, thank you very much.
Actually, after a dangerous *soft fork* split with a minority hash rate, there's in fact nothing else to do for the initial majority legacy chain than to protect itself by a hard fork. I said this a few weeks ago already.
A soft fork imposes itself onto everybody if it is majority. As such, a soft fork that starts out as (large) majority before being activated is not a problem from a chain splitting point of view, because the minority of miners that would like to continue the legacy chain are DIRECTLY orphaned. This was the principle behind soft forks and things like 95% majorities. The user will not see any difference: the chain seems to continue, and has uniformly switched to the new protocol.
A bilateral hard fork is a "clean split". Yes, the chain splits, but once you have your two new coins, they live their independent lives happily. You can trade them on exchanges, and by doing so, you vote with your money ; hash rate will follow. See ETC/ETH.
However, BIP148 is a major cluster fuck: it wants to instore a chain split with a soft fork, where the soft fork is a minority split. The whole idea of BIP148 is that bitcoin splits, that two bitcoins are listed on exchanges, and that, so is the hope, users will buy 148-bitcoin, and dump their legacy bitcoins, attracting more and more hash rate to 148-bitcoin, until it gains majority. After a while, the 148-coin chain takes over the legacy chain, and orphans them. But that can happen after a month or so, and all people having transacted on the legacy chain (in buying/selling coins on exchanges) suddenly see their chain (and their holdings on it) disappear !
This is a clusterfuck that should be avoided: once the two coins are appearing on exchanges, one should consider them independent, to protect the holdings of people on each chain (as well as for exchanges to be able to respect their customers and let them withdraw coins of the types they bought IOU of). This is why in such a case, the legacy chain has NO CHOICE but to hard fork, to protect itself from re-organization, and to keep the split permanent, to turn the dangerous soft fork split into a bilateral hard fork. Note that this could be done with a very minor modification, that doesn't even alter anything to the original chain (but honestly, if one does a HF, one could take the opportunity to increase the block limit too). It would even be wise, in this HF, to alter slightly the signature scheme, to avoid the night mare of replay attacks, which would make it almost impossible to do what 148-coin wants people to do: to sell their legacy bitcoin, and to buy 148-bitcoin. Because with replay attacks (especially with a soft fork !) they will have a hard time NOT selling at the same time, their 148 coins.
In other words: bip 148 is such a mess if ever it gets activated, that the only way to clean it out, is to have the legacy chain HF and make the split clean.
