The Ethereum community is in emergency mode as showstoping findings from the Cornell team sent the soft-fork option out. The only choice now available is whether to hard-fork or not, a decision that must be made in just 16 days.
DDos vulnerabilities were found in Ethereum’s softfork which allows an attacker to exploit the gas limit (comparable to bitcoin’s transaction fees) and, in effect, bring the whole network to a halt. Therefore, the “[s]oft fork is out” says George Hallam, spokesman for Ethereum’s foundation, speaking to CCN.com.
The community now has to move incredibly quickly and make the biggest and most defining decision, perhaps of Ethereum’s entire existence, in just over two weeks. Our investigation indicates that, currently, the majority are in favor of the hardfork, with Ethereum’s subreddit, the main public space for Ethereum’s community, littered with threads calling for a hardfork, but not everyone is in agreement.
One of the few Ethereum developers who responded to us in time for publishing, Alex Van De Sande, who has been against a hardfork since the beginning, re-iterated his position in light of the new information, stating “I don’t like [the hardfork],” before adding in response to our request for elaboration that he had made his position clear in many public statements.
We spoke more in-depth to Emin Gün Sirer, one of the authors of the DDoS Vulnerabilities report who stated:
“I prefer a hard fork. It’ll put a decisive end to the saga. So we can get back to work,” because “[i]t’s just the quickest, cleanest, simplest way of putting this mess behind us. I respect the no-fork point of view as well.”
A hardfork requires an upgrade of the entire ethereum ecosystem, including all businesses, all nodes and of course all miners. We wondered whether such monumental task could be achieved in just over two weeks. Technically, it ought to be achievable, says Sirer, but:
“Whether the community can reach the level of social consensus required to roll out a hard fork is something that we can only discover empirically.”
The Ethereum community acted very quickly on the softfork, reaching consensus within a day or two, followed by miners upgrading in just hours. Whether the same will be repeated is yet to be seen, but codewise, both Parity and Geth, Ethereum’s most used clients, have already released a proposed hardfork code . While hardfork clients will be released as early as next week according to a public statement by Ethcore .
Some of the small miners we spoke to seem to be in favor or ambivalent with the arguments more practically focused on potential problems a lack of forking will cause when switching to POS, while some justified the decision on the basis that Ethereum is only months old, thus mistakes at such early stage are to be expected. No miner spoke against the fork, but we emphasize that the sample is very small, just five, and the time window was just two-three hours.
Unlike a Bitcoin hardfork, the proposed forks does not affect any other transaction or part of Ethereum. It only affects the DAO’s smart contract address while the rest of ethereum remains in an identical state. Lefteris Karapetsas, Lead Technical Engineer at Slock.it, described the fork changes in a simplified way as follows:
“Move all ETH from all contracts whose ultimate ancestor is 0xbb9bc244d798123fde783fcc1c72d3bb8c189413 and their extraBalance back to 0xbb9bc244d798123fde783fcc1c72d3bb8c189413.
Replace 0xbb9bc244d798123fde783fcc1c72d3bb8c189413 code with a simple refund contract (TBD). That contract would have a refund() function for DTH to call. Once called a refund proportional to the DAO tokens held in both the main DAO and in all (statically enumerated) child DAOs, except the dark DAOs could be claimed by the msg.sender.”
There are, however, more philosophical and political arguments to be made. The main argument is based on the premise that the blockchain should never be changed as code is law and there should be no exceptions. The argument is supported by some Ethereum developers, such as Sande, who was in favor of the softfork as it was temporary, but is not in favor of the hardfork.
As we were unable to quickly reach Ethereum developers and many of those we could contact failed to respond, perhaps due to the sensitivity of the matter as the decision is decentralized to the entire ecosystem, including businesses, node operators and miners, we asked Sirer whether he was in contact with Ethereum developers and, in his judgment, whether they support the hardfork:
“I’m in contact with some of them. It is a large team, with a diverse array of opinions. The vast majority of people I have spoken to (and some are eth devs, some are unrelated to the foundation), are in favor of a hard fork. As am I.”
Sirer emphasized that he’d be “perfectly happy with no fork as well,” before adding:
“I have no money in the dao, I just want the drama to end. I hold some cryptocurrencies, including btc and eth, small amounts. My investment is in terms of my own career direction and time. I don’t want the space to collapse.”
Stating that the situation had become a “high-stakes poker game,” Sirer compared it to “corewars”, a hacker (as in programmer rather than attacker) game from the 80s where:
“[T]wo people would write two programs, to execute on the same machine. The purpose of each program was to destroy the other. Every hacker played it and loved it. The DAO attack has foisted a similar game on us. There is the attack code, the whitehat response, the counter responses to come, and so forth. All played out on the Ethereum virtual machine.
This is at once exciting (Ethereum allows us to do amazing things, and this is a side effect of that), and at the same time, it wears everyone down.”
He stated that Ethereum could collapse:
“[I]f there isn’t a clear solution, and the current corewars saga continues to unfold. No Fork is a fine outcome as well. Any decisive outcome is great.”
To fork or not to fork has many financial, political, philosophical, reputational and emotional implications with each option having its advantages and disadvantages extensively discussed over the past two-three weeks. The current indications suggest that the majority is in favour of forking, but a conclusive decision won’t be known until pools open up their voting. Either way, any decision is better than no decision with Sirer urging Ethereum’s community to:
“[C]ome to a decision quickly and decisively. There are tons of exciting Dapps to come, let’s put the DAO story behind us and work to make the new Dapps the successes they deserve to be.”
We could not get a poll out in time for this article, but the comment section below could act as a poll. Please state whether you hold any Ethereum or Bitcoin and then state whether you support the hardfork or not.