The EOS main network is now live again, but the road to launch has been a rocky one.
This article does not contain investment advice or recommendations. Every investment and trading move involves risk, you should conduct your own research when making a decision.
Decentralized governance is hard, as we have already learned from the notorious Bitcoin community standoffs. But as the saga around the EOS mainnet’s onerous drag toward going live illustrates, it might be challenging to make things work collegially, even in a quasi-decentralized system where the number of stakeholders is limited.
Although the long-anticipated EOS blockchain technically took off on June 10, it was not operational, and its tokens were ‘frozen’ for almost five days. It was a stalled community-governance procedure that was holding the process back.
In order to kick off, the EOS community had to elect 21 entities responsible for the blockchain’s operations by a similarly designed vote, and it didn’t go smoothly. The delay in reaching the 15 percent threshold of voting tokens necessary to seal the election made investors nervous, who put in their money to make Block.one’s $4 billion ICO the biggest token sale in history.
Delegated proof of stake: a not-so-decentralized mechanics of EOS
EOS is heralded by its creators as a platform for the deployment of Decentralized Autonomous Organizations (DAOs) and Corporations (DACs) on an industrial scale. Its main selling point is meant to be the capacity to handle many more transactions per unit of time than the Ethereum blockchain, which EOS seeks to supersede. The technical solution behind this ambitious vision of increased speed and scalability is the consensus algorithm embedded into EOS’s design, called a delegated proof of stake (DPoS).
It relies on a restricted pool of Block Producers who work to validate transactions, which dramatically increases the network’s throughput — at the cost of decentralization, as many believe. One of the most outspoken critics of DPoS has been Ethereum’s Vitalik Buterin, who contended that it creates dynamics conducive for stakeholder collusion and cartelization of the token-holder community. Renowned blockchain researcher Emin Gün Sirer was also quite critical, suggesting that ‘EOS isn’t even trying’ to be a decentralized system.
The EOS mainnet relies on 21 groups called Block Producers, or supernodes, who are elected by the votes of token holders, cast in proportion to their token holdings. This immediately sounds like a whales’ game, doesn’t it? Factor in another consideration: according to a widely circulated Reddit user’s report, almost half of the EOS token supply is confined to just 10 wallets, while the top-100 sits on 75 percent of the overall wealth. The caveat is, however, that 10 percent are reserved for the founding entity Block.one, which pledged not to put them toward voting, while several other giants are likely major exchanges’ repositories. Yet, it would be a safe guess that many of the remaining money bags entered the competition to become block producers, as their holdings put them in a tremendous position to vote themselves in. Why, then, was it such an ordeal to surpass the 15 percent voting bar?
The whales in the backrooms: a struggle of the voting process
There are two mutually non-exclusive answers to this question. One is big-player politics. According to this view, the whales did not rush to cast their overweight ballots as soon as the election kicked off, waiting for others to vote and for the overall landscape to become clearer. While the main-street turnout was dragging percent by percent, the real talks were likely happening in backrooms, as coalitions were formed and agreements reached between those who had a real shot at the Block Producers’ spots. One convincing cue pointing to this dynamic is the fact that while the first 9 percent of the vote was painfully sluggish to come in, the stretch between 9 percent and 15 percent took less than a day — a likely sign that big chunks of votes rolled in as a consensus was finally reached. The covert struggle wasn’t all fair play, too: impostors surfaced here and there, and accusations of other candidates being fake were widespread.
Another probable reason for the lackluster vote during the first days was the how the procedure was complicated, and even risky for less-sophisticated investors. In order to cast a ballot, token holders had to use their private keys with a third-party’s software. While the only provably legitimate voting tool (CLEOS by Block.one) was command line-based, it was not always easy to tell where the tools with more user-friendly interfaces came from. Combined with the realization that one’s John Q. Public vote does not carry any real weight, it could have a chilling effect on potential lay voters. Ultimately, the pool of 21 elected Block Producers now looks like this, topped by EOS Cannon, Liquid EOS and EOS Beijing (as of press time).
The arduous election was far from the only speed bump on the way to the EOS mainnet launch. Several weeks prior to the event were riddled with controversies. On May 31, someone hacked Block.one’s Zendesk account and sent out fraudulent announcements of an ‘unsold tokens’ giveaway. Looking fairly authentic, these emails helped the impostors swindle investors out of several million dollars.
Just a few days prior to that, a Chinese cybersecurity firm found a series of massive flaws in the EOS system, pointing to loopholes that could allow attackers to take full control over any network node. In what looked more like a PR move, Block.one promptly responded by instituting a bug bounty program, which led to even more software-security professionals to start enthusiastically digging around in order to unearth further flaws and vulnerabilities. Cybersecurity researcher Guido Vranken gained fame when he discovered 12 vulnerabilities worth $10,000 each in a matter of a week, prompting a widespread murmur that the $4 billion enterprise could do somewhat better.
A messy path to launch
On top of all this, the draft of the EOS constitution that was unveiled in early June received a cold welcome from some influential voices in the crypto community as being poorly conceived. Blockchain titan Nick Szabo was especially harsh, calling the document ‘naively drafted’ and suggested that it would render EOS ‘labor-intensive, permissioned, jurisdictionally biased’ and having ‘poor social scalability.’ And then the transcript of the June 8 conference call of 200 EOS developers and Block Producer candidates went public. It had left the community perplexed for the overall organizational havoc that it showcased, as well as for some spicy particularities like the discussion on whether to ‘print’ more tokens (resolved with a yes):
“I am a little perturbed by these guys having the authority to print themselves more token,...What is this ---> [sic] A reserve banking system?”
However messy the path to the EOS launch has been, it finally happened on June 14, seeing the token price recover from recent blows. So, is it finally time for the developers to leave all the vitriol behind and get down to some real work? Not quite yet. Just two days after the launch, the EOS mainnet underwent a ‘freeze’ and effectively stopped processing transactions. It wasn’t until the next day that it was operational again. Ultimately, the number of really bad-looking debacles that accompanied the platform’s takeoff, weighted against its war chest, understandably left many influencers more and more critical of the whole enterprise.
The blockchain world, however, was never stingy with second chances. Some projects grow too hyped and too expensive to fail, retaining a massive vote of confidence despite a modest yield. Those who have stakes in EOS or are longing for a faster alternative to Ethereum will keep cheering Block.one and their confederates on, while the Ethereum loyals will anticipate new blunders with a degree of schadenfreude. All eyes are now on the performance of EOS-based applications and the ability of the network to withstand attacks. Whether EOS can survive and live up to its potential will shape the investors’ mood, as well as the fate of the DPoS consensus algorithm.