The recursive call bug allowed the attacker to drain theDAO to the tune of 4 million Ether (or roughly $150M at the time). The community put forward the idea of refunding theDAO contributors via a hard fork — to the dismay of a minority within the community. From their point of view, Ethereum was initially sold as an immutable blockchain; forking to execute an irregular state change because of a badly written contract broke this property. However, there was a larger portion of the community which thought forking was justified and the correct course of action for the blockchain. Given the transition to proof-of-stake, leaving about 15% of the Ether supply in the hands of a malicious actor would put the network’s future at risk. There were also concerns that inaction from the community might lead to actions by regulatory agencies.

In order to gather the community’s opinion, a dapp named CarbonVote was created to let people voice their stance on whether a hard fork should occur or not occur, whereby 1 ETH equated to 1 vote. The pro-fork vote tallied 3.96 million votes, while the “code is law” camp gathered 0.57 million votes. In light of this vote, and of the general pro-fork stance by community members, the client teams implemented the hard fork changes and the network upgrade at block 1,920,000. A small community decided to not upgrade their nodes, to fork the code and to mine a new network called Ethereum Classic. Anyone with a balance on the Ethereum network would, at the fork block, automatically have their balance on the Ethereum Classic network too.

The fork added a new block processing rule: at the fork block, all balances in theDAO and child DAOs would be zeroed out and refunds would be available to contributors in a newly written refund contract.

Source: https://coincodex.com/article/50/the-dao-hack-what-happened-and-what-followed https://ethereum.stackexchange.com/questions/7832/give-a-summary-of-the-fork-state-changes-in-block-1920000 https://en.wikipedia.org/wiki/The_DAO_(organization)#History http://v1.carbonvote.com/