It’s been a little over a month since The DAO, the most visible and well-funded application on the Ethereum network, crashed and burned at the hands of an anonymous hacker. Until its spectacular demise this June, The DAO was a centerpiece of the Ethereum ecosystem. It was intended to operate as a decentralized, automated investment fund, which at one point had over US $150 million worth of the cryptocurrency Ether at its disposal. Had it succeeded, it would have been a shining early example of Ethereum’s potential as a smart contract platform, adding credibility to the thoroughly-hyped idea that blockchains are poised to replace our legal, political, and social frameworks with code.
Those dreams were smashed on 17 June when a hacker exploited a flaw in The DAO and drained it of some 3.6 million Ethers. Due to the nature of The DAO’s underlying structure however, these funds are stuck and cannot be used in a transaction until 28 July, meaning the hacker has no way of cashing out for another nine days.
Provided with this handy deadline, the Ethereum community has spent the last month passionately debating whether it is possible, and more importantly, whether it is advisable to make investors whole again. In many ways, it is a lose, lose situation. If no action is taken, then there is a good chance that legal action will ensue—initiated either by investors in The DAO or by federal regulators, who are already sniffing at the hem of the blockchain.
No one invested in this technology is eager to see Ethereum developers (many of whom were involved to some degree with The DAO) in legal disputes, as this would only hamper further development. On the other hand, properly cleaning up this mess requires returning Ether to investors, which in turn requires a controversial update to the Ethereum protocol.
Over the weekend, those in favor of reimbursing investors seem to have won the debate. On Wednesday developers will attempt what’s called a hard fork of the blockchain—a risky, backwards incompatible software upgrade that is usually only imposed upon a network when critical fixes are necessary, and which is usually scheduled months in advance.
If successful, the move will refund the original investors while effectively dissolving The DAO. If unsuccessful then we will end up with two competing versions of Ethereum, an old one and a new one, identical to one another in every way but their treatment of The DAO.
Which of these scenarios plays out depends entirely on whether everyone chooses the same path. The matrix of persuasion and influence among Ethereum developers, users, miners, and exchange operators is rather opaque. Developers cannot force any participants in the Ethereum network to install the new software. All they can do is publish the code and hope for the best, which is what finally happened over the weekend.
Ethereum has already survived one hard fork. It happened in mid-March, and it was a breeze. In that case the update had been anticipated and it included software changes that were clearly included, from the outset, in the longterm vision of Ethereum.
Two things set this new hard fork apart: First, to the best of my knowledge, it is the first hard fork—not only in Ethereum, but across all blockchains—that has been designed to make reparations to investors of a failed enterprise. And this is at the heart of the controversy. To a large swath of the cryptocurrency faithful, the blockchain is a political no-go zone. Bitcoins and Ether move between accounts at the behest of code. Likewise, the rules that govern the parameters of the Ethereum economy are etched into the code and they apply to everyone equally. As a consequence, no entity, be it political or corporate, can censor transactions or take control of accounts without first convincing people in the network to adopt new code.
The hard fork makes investors whole by adding a new set of rules to the Ethereum code. The rules only apply to the accounts associated with The DAO. The software moves all the funds that were initially invested in The DAO to a new address. From there, Ether will be doled out to people based on how much they initially invested. If the fork is adopted, it will be the first time a protocol change has been implemented on a mainstream cryptocurrency with the explicit goal of confiscating coins.
It is the opinion of many early Bitcoin adopters that this kind of influence should be exceedingly difficult, if not impossible, to achieve. But do people in the Ethereum crowd feel the same way? This has been the crucial question in deciding whether to push out a hard fork version of the software.
After being drawn into ugly debates on Twitter and Reddit, developers finally unveiled some rudimentary polling tools last week to help formally gauge sentiment about the fork. (There are no perfect ways to poll all the various human cogs in the Ethereum machine.) Most of the polling tools now out there allow people who own Ether to either vote for or against a fork while putting the weight of their Ether holdings behind the vote. In other words, if you have more Ether, then your voice counts for more.
Carbonvote, the most well-known of these tools, has shown a clear preference for a hard fork since people began voting. This weekend, developers pointed to Carbonvote as ample indication of a community-wide consensus and used it to justify the release of new code for the hard fork. A hard forked version now exists for all the major Ethereum clients, although some provide a way to opt out.
The other reason this hard fork is different from the one in March is because it’s happening fast and dirty. Although the hacker is unable to use the stolen Ether in any transactions before 28 July, other niggling maneuvers are available to them before that date, and so it makes sense to attempt the hard fork as soon as possible. Ethereum client developers announced a bug bounty that will start today and extend multiple weeks after the hard fork is initiated. But the circumstances leave very little time for testing and auditing.
By the end of Wednesday, it should be rather clear whether the hard fork is a success. And there is much more than $40 million dollars at stake. What happens on Wednesday will determine whether Ethereum will continue to exist as a single coherent technology and it will clarify the extent to which people participating in the Ethereum network are willing to intervene socially on a system that was designed to care as little as possible about social frameworks.