The Bitcoin Fork Saga has taken yet another strange turn as some Chinese Bitcoin miners – hit hard by Bitcoin losing one third of its value – are now switching sides to the Bitcoin-Core team to activate the SegWit bugfix-cum-capacity upgrade in time for the UASF (User Activated Soft Fork) day on August 1. But it may be too little too late as others still want to hard-fork Bitcoin regardless and create a new coin.
Chinese miners have been blocking SegWit – a.k.a. Segregated Witness, a bug fix, capacity increase and a roadmap to even more capacity increases – since it was rolled out in November 2016. Some oppose SegWit because they say SegWit coins are insecure and can be spent by anyone, and opens up its users to patent lawsuits. Others say it is incompatible with ASICBoost (a technique that gives large miners 20% extra income), although Litecoin has been running SegWit for months now without any mishap.
The thing about SegWit is that there are currently four paths Bitcoin can follow to adopt it. Three of the BIPs (Bitcoin Improvement Proposals) are soft forks (meaning that older software – miners, wallets and bots – can continue to run without being affected, but cannot take advantage of new features).
BIP-91 and BIP-141 are the traditional paths for a Bitcoin network upgrade and require miners to signal readiness for SegWit. BIP-148 (the UASF) is an attempt to run SegWit in parallel on the same blockchain. All of these use the same 1MB block size as the current Bitcoin blockchain.
Then there’s Segwit2x, which is, well, a fudge. It is a soft fork with SegWit on August 1, but with a hard fork to larger blocks in three months time that will be incompatible with any existing, non-upgraded software. If the soft fork works and SegWit is activated, the need for big blocks as a means to increase capacity will be much less.
However, this is where things get interesting.
Wiping out the enemy’s blocks
BitMain co-founder Wu Jihan has announced he will use a portion of his mining power to protect the Bitcoin blockchain by wiping out BIP-148 SegWit blocks. This can be done by privately mining a separate blockchain – the moment a BIP-148 block appears on the blockchain, he would expose his privately mined version, and if it has more combined difficulty behind it, it will nullify the other block.
This poses a risk, as one does not have to be using SegWit, but if the transaction is mined by a SegWit miner, it could be wiped out as if it never happened. Transactions based on old coins can simply be re-submitted, but a transaction that references a newly mined coin, even in part, would be totally wiped out. At the very least, this would mean that pre- and post-UASF coins should not be mixed and pure pre-UASF coins should be more valuable. At most, it would mean that a Bitcoin network transaction cannot be trusted until a clear winner emerges.
Wu’s Antpool controls around 22% of Bitcoin network mining capacity, meaning he has the power to carry out this promise. It also means he can (and did) block BIP-141 and can easily block BIP-91.
Bitcoin exchanges closed for business on August 1
As for why some Chinese miners are having a sudden change of heart, Bitcoin exchange Bity announced last week it was halting transactions on August 1 until things become clearer and it appears safe to proceed. Since then, nearly every exchange has also announced it will similarly halt withdrawals and deposits and – in many cases – trading of Bitcoin on August 1.
When mainstream media caught news of the UASF situation (and did not quite understand the difference between a soft fork and a hard fork, for the most part) and the exchanges announced the planned shutdown, the price of Bitcoin plummeted from around $2,500 to just $1,850 over the weekend, wiping $10 billion off its market cap.
Perhaps it was this very real realization that blocking SegWit would be bad for their profit that, on Monday, most of the Chinese mining pools suddenly switched their allegiance and are now signaling BIP-91 to try and lock in SegWit activation before August 1 to avoid the civil war.
Not two but now three forks on the table
Of course, not everyone is doing it. Wu Jihan’s pools have carried out his promise and are now running Bitcoin1, the Segwit2x varian of Bitcoin-Core (the reference Bitcoin software).
Then there is BitcoinABC (ASICBoost Compatible) that totally eschews SegWit and will fork Bitcoin for big blocks. (BitcoinABC itself suffered a fatal error and it was funny to watch the Github discussion teach the difference between 0 and NULL in making comparisons in C++ vs C.)
ViaBTC, with 5.1% of the network hash rate, is pushing BitcoinABC as a launch of a new token. In a way, one must commend them for launching a new token by forking Bitcoin instead of through a presale ICO. It nicely solves the fair initial distribution problem as anyone with Bitcoin prior to the fork will automatically have the same amount of its BCC tokens.
One wonders though if that is the ultimate goal for BitMain and the Segwit2x gang. Segwit2x was never about capacity, but taking control of the future of Bitcoin. But even if they fail, they may be in control of a very widely adopted alt-coin.
If BIP-141 activates within any 2-3 day window, then there will be no fork (leaving aside for one moment the BitcoinABC fork). If that does not happen in time for UASF, then the network will be in for a rough ride, with Wu attacking SegWit blocks and UASF miners promising to wipe out non-148 blocks if they become the majority. Even if Segwit2x wins, that only means there will be a hard fork three months down the line, meaning three months of uncertainty for Bitcoin users.
Things are moving very quickly, and by the end of the week who knows what new wondrous new plan someone will come up with? Until then, enjoy the roller coaster ride.