@noise 6 November 2020

If the IFP in Bitcoin Cash activates, it suggests that cryptocurrencies are doomed to centralization

There are many arguments against cryptocurrencies, and this is one of them: How can it be decentralized when the developer with the GitHub repo can give themselves any number of coins, arbitrarily and without recourse? The answer has always been “don’t worry, miners will reject it” and “they will just fork off from the rest of the network”. In about a week, this theory will be tested for real in Bitcoin Cash because this is *exactly* what ABC is trying to do by diverting 8% of the blockreward to themselves. (The only difference is they use Phabricator instead of GitHub). What harm can a coder do? Some might say hey, the important thing is *miner decentralization*, so that transactions cannot be double-spent. Whatever code a developer introduce hardly matters, right? But it turns out that developers can do an incredible amount of damage. If left unchecked they can for example: Block necessary upgrades (like Bitcoin’s blocksize increase). Censor transactions (see the aftermath of the DAO hack). Push through poor code that can be abused to harm the network (the DAA). Arbitrarily change the emission schedule (what ABC tried to do with Grasberg). Assign themselves an arbitrary amount of coins (the IFP). Or in other words, with the code they write (or don’t write) they can completely destroy everything that makes a cryptocurrency valuable, and this is why safeguards against the developers’ power are so important. Historically, the reference client’s code is law While I’d like to think that the community would reject Bad Code™ so far the opposite seems to be true. Early on in Ethereum’s history people subscribed to the idea that “code is law”, and that the rules you inscribed into a smart contract made outdated things like human decision unnecessary. But this was thrown out the window after the DAO hack, when a bug in a smart contract was exploited and the Ethereum developers moved to quickly freeze the funds, effectively censoring transactions on the chain. Preventing the DAO theft isn’t morally wrong, but why weren’t similar hacks counteracted in a similar manner? Maybe because they were too small or didn’t affect the Ethereum developers enough for them to care? Freezing of funds is an arbitrary human decision that cryptocurrencies were created to remove. Otherwise we’ll just end up with the same problems that plague PayPal and VISA, where money of innocent people are frozen all the time. Yet in Ethereum, the community followed the reference client.

Another example is how Bitcoin Core managed to block the blocksize increase in Bitcoin, even though an increase to 2 MB had broad support by miners and the community. The reason they managed to do this was that the miners ultimately decided to be passive, and to wait for Core to implement the 2 MB increase, which of course never happened. The third example is how in Bitcoin Cash every change has been dictated by ABC (so far at least). By threatening a fork they managed to push through their preferred changes like the DAA and CTOR, while blocking others like on-chain tokens. This pattern where the reference client *always* dictate the rules is repeated all over in the cryptocurrency space, raising serious decentralization concerns. (The single counterexample I could find is how Monero replaced the original developer team with a new one, which happened early in Monero’s history when it was still very small.) https://monero.stackexchange.com/questions/1011/monero-inception-how-did-bitmonero-become-monero What if the IFP succeeds? It seems unlikely that the IFP would activate in BCH, because all the metrics we can come up with shows a clear preference for the no-IFP / BCHN side: More than 70% of blocks are signaling for BCHN. https://ec.loping.net/ichundes/bchnblocks.png 58.68% of all nodes are signaling for the no-IFP side. https://cash.coin.dance/nodes 86.21% of all upgraded nodes are signaling for the no-IFP side. https://cash.coin.dance/nodes Coinex futures values BCHA/BCH at 0.0599, valuing ABC at around $15. https://www.coinex.com/exchange?currency=BCH&dest=BCHA 98.75% of coins support BCHN against ABC, and only 0.02% support ABC against BCHN. https://votes.cash/ Coinbase will run BCHN and will not give out ABC coins. https://mobile.twitter.com/CoinbaseSupport/status/1324156279489622017 Users on social media have overwhelmingly declared against IFP. https://www.reddit.com/r/btc/comments/f4qgry/lets_have_a_show_of_hands_who_supports_the_ifp/ And if the IFP still succeeds it means that all this is meaningless and that the code of the reference client trumps all. It shows that cryptocurrencies cannot break out of developer centralization, and we might just as well replace proof-of-work with proof-of-GitHub or proof-of-Phabricator. What if the IFP fails? But if the IFP fails this would be the first time in cryptocurrency history that a reference client of a major cryptocurrency has been kicked out in a “hash war”. This is huge because it proves that Bitcoin Cash is resistant to rogue developers and can kick them out if they try to change the coin supply, redirect the blockreward to themselves or something else that could cause serious harm. It shows that cryptocurrencies aren’t necessarily centralized around a developer team, and that the hope of truly decentralized peer-to-peer electronic cash is still alive.