#uasf

Thunderbirds, U.S. Air Force

BITCOIN IS SPLITTING IN TWO TODAY! As a sensible person who knows better than to care about cryptos, beyond maybe how to make money from them, you're just sitting back with your popcorn. But people have been asking me to explain it, so ...

Like everything in a Bitcoin crisis, the details are both obscure and inane.

THE ACTUAL PROBLEM: Bitcoin is full. The transactions have been clogged since mid-2015. Nobody has been able to agree how to fix this. Everyone involved despises each other.

THE SOLUTION: a Judgement of Solomon where the two women argue how to slice the baby in half.

THE UPSHOT: if you had BTC before in your own hand, you have them now, and you have them on the BCC chain too! If you had BTC before on an exchange, some exchanges credit you with BCC and some don’t, and it’s a huge mess.

And no, I don’t know why everyone chose August 1st.

THE PLAYERS:

Segwit2x - this is backed by a supermajority of the mining power. It puts into place SegWit (Segregated Witness), which is a way of (mostly) solving the Transaction Malleability bug (of which more below). The "2x" means it's set to go to 2 megabyte blocks in a few months. With most of the mining power behind it, I think this is likely what the mainstream Bitcoin will be going forward.

UAHF (User Activated Hard Fork) - something called BitcoinABC - I believe this is now the same as Bitcoin Cash (BCC) - a plan to just straight-up go to 8 megabyte blocks, no SegWit. This is explicitly creating a forked blockchain. So everyone who had BTC before will have BCC now! So the attraction is FREE MONEY. Also, transactions will be faster with much lower fees! If anyone takes it up.

Bitcoin Cash doesn't have huge miner support, so there are questions as to difficulty level of mining on that chain if all the miners leave, potential to attack the chain if you have a lot of mining capacity to hand, long block times until the difficulty adjusts, etc.

This has endless potential for disaster, e.g. replay attacks (where you replay transactions on one chain against the other chain, so you can take coins on the other chain without them having been intentionally sent to you). The same sort of thing happened when Ethereum split into Ethereum and Ethereum Classic, because neither side would change things to make replay attacks not possible. (Ethereum Classic eventually did.) You will know replay attacks are happening when people start yelling about them, going "HOW COULD WE HAVE FORESTALLED THIS" (by not being completely pigheaded, probably). Bitcoin Cash claims replay protection.

UASF (User Activated Soft Fork) - or BIP 148 - where people who want to keep small blocks (1 megabyte) put in SegWit. This is not the miners, but the user nodes. This also adds some transaction capacity, because how the “1 megabyte” is calculated is different. The politics here is that they thought they could swing this without the mining power (nobody else thought this). This got a lot of publicity but is pretty much not a happener.

Bitcoin Unlimited - a plan to just go to 2MB blocks, expanding further as needed. Mostly Roger Ver. I think they're all for Bitcoin Cash now.

HOW DO I KNOW WHAT'S HAPPENING?

btcforkmonitor.info as keeping track of the forks and monitoring them for divergence. It’s getting ABSOLUTELY HAMMERED right now and gives me only a blank screen, but here’s a snapshot from earlier today.

CoinDesk is live-blogging the tyre fire: https://www.coindesk.com/bitcoin-fork-watch-bitcoin-cash-hard-fork-august-1/

BUT WHY ON EARTH

Of course, none of the disputes are technical - they're all the politics of who gets to make money. Everyone involved hates everyone else and characterises their opponents as "evil" and opposed to "Satoshi's vision". They figure if they shout abuse at each other they'll get rich faster or something I dunno.

The Transaction Malleability bug ... well, there's a thing in Bitcoin called the transaction ID. You might think this could be used to, say, uniquely identify a transaction! But welcome to Bitcoin. The transaction ID can be changed in mid-flight by a hostile intermediate node as long as it hasn't made it into the blockchain yet. (A lot of bitcoins were stolen from Mt. Gox this way in 2013.) So SegWit mostly fixes this.

SegWit is controversial because the planned Lightning Network - to work around the transaction clog by bolting another network on the side - requires transaction IDs that can be used as transaction IDs. LN is being advanced by Blockstream, who also employ a lot of Bitcoin core developers, so are widely blamed for the blocks staying at 1MB. Also, LN is still vaporware. So if you see bitter opposition to SegWit, it's because they don't like Blockstream. People who don't like Blockstream treat SegWit as EVIL even though anyone else would call it an obvious protocol fix to have transaction IDs that could be used to identify a transaction. (Though it’s arguably a more complicated fix than is necessary.)

I would characterise all this mess in terms of the Judgement of Solomon - where Solomon says "ok, we'll slice the baby in two" and the women think that's a great idea and then start bitterly fighting over whether to slice the kid horizontally or vertically. It's at the stage where cutting the baby in two sounds like a plausible idea to them.

I can only guess how this is all going to shake out. I think probably the main Bitcoin is going to go to Segwit2x. BCC has tempted people with FREE MONEY and there's a pile of exchanges saying they'll go along with this idea, and give people free BCC to match their BTC deposited. (Coinbase notably has not.) And of course miners will mine on whichever chain is most profitable. Cryptos hardly ever die completely, so I can see BCC lingering quite a while.

The essential point is that all this is actually good news for Bitcoin, because it seems everything is.

Update from CoinDesk’s liveblog: 0 blocks mined, 1% of hash power:

15:06   Bitcoin Cash is Forking But New Blockchain Could Take "Hours or Days"

An unknown number of bitcoin users and miners have split off from the main network and are now attempting to start a new blockchain called Bitcoin Cash.

At press time, it seems the effort may be struggling to attract miners, as the bitcoin blockchain is now five blocks ahead of the Bitcoin Cash blockchain, which has yet to produce a block and create a new cryptocurrency. The development has set off a wave of new commentary.

“It's likely going to take at least several hours,” tweeted BitGo engineer Jameson Lopp.

Bloq CEO Jeff Garzik, who's backing a rival scaling proposal mentioned that since Bitcoin Cash has about 1% of the computing power that the main bitcoin network has, it might even take 24 hours to mine a new block.

The comments add notable context to the current conversation on whether the effort to fork the bitcoin blockchain should be considered a success – or if it will even occur at all.