EOS, the blockchain protocol designed to host industrial-scale decentralized applications (DApps) and launched this summer following a $4 billion initial coin offering (ICO), has created no shortage of buzz in the enterprise blockchain space. If fully realized, it could be an incredible tool for large corporations to begin reaping the benefits of decentralization. However, its delegated-proof-of-stake (DPoS) consensus model and some hiccups around how this functions have been controversial.
To get answers to some of the most pressing questions around EOS, we spoke with Thomas Cox, executive director of the EOS Alliance , and Myles Snider, CEO of Aurora EOS , an EOS block producer.
1. What new developments can we expect to see for EOS this month?
Thomas Cox: The big and obvious ones are the San Francisco Hackathon and the release into beta of the Referendum contract that will eventually allow for on-chain elections and voting on policy. The Referendum, once it passes out of beta testing, will be the cornerstone for a majority of EOS's governance activities.
Myles Snider: There are two major improvements to EOS in progress. One is the resource exchange (REX), which is a protocol-level liquidity pool for token holders to trustlessly rent out their bandwidth to other users who need it. This will lower costs for both users and developers and also allow token holders to earn a return by leasing out their unused capacity. Participation in the REX also requires that users vote, so it should increase voter participation, as well. It's a major improvement to EOS all around.
The other improvement being worked on is inter-blockchain communication (IBC). This is software that allows different blockchains built on the EOSIO software to communicate and interoperate with one another. This will allow for the creation of EOS sidechains and interoperable EOS code forks.
2. Do you see EOS and Ethereum working together, as interoperability protocols are now being developed for this?
Cox : Yes, very much so. Ethereum is amazing technology and has a robust community of brilliant contributors. EOS brings in speed and flexibility. The use cases and design envelopes of Ethereum and EOS are complementary. With the release recently of BancorX and other initiatives that allow for bidirectional integration, the EOS Alliance expects to see an explosion of interesting new use cases. We look forward to increasing collaboration across the two ecosystems.
Snider: I hope so! Recently we've seen the development of the EOS21 protocol by shEOS, and BancorX by Bancor/LiquidEOS, and both of those are huge steps toward cross-chain interoperability. I think we'll see many more teams that have issued tokens on one platform or the other offer users the ability to port between platforms. I'm really excited to see strong projects from Ethereum that could also be used on EOS. It would be really exciting if Maker, for example, allowed users to use Dai stablecoins on EOS.
3. What's your response to the claim that EOS is not actually a blockchain but simply a distributed database?
Cox : I released this official response on November 6. I'm pleased to report that the CTO of Whiteblock is in ongoing dialogue with many top EOS BPs (block producers) and core developers to better understand the gap between what Whiteblock found and what we know to be true about our technology. I expect that dialogue to be a prelude to a collaboration that supports Whiteblock's already announced public live-streaming of its EOS findings set for later in November. All parties in that discussion are being highly professional and respectful, and are focused on increasing mutual understanding.
One early result is a tentative plan to devise a shared, standard lexicon for blockchain phenomena and behaviors, ideally under the auspices of the IEEE Blockchain Initiative standards committee. I plan to broach this at the November meeting.
4. How would you describe the delegated-proof-of-stake model of governance of EOS to the average, nontechnical person?
Cox : You could think of DPoS as being a method whereby the members of the EOS community pick out their most able, technical members and temporarily give them the role of making blocks for the rest of us to validate. By requiring BPs to compete for election every two minutes, and to collaborate once elected, we get much faster block confirmation, a vast reduction in power consumption and very high reliability.
Snider: DPoS is a consensus algorithm that uses token-holder voting to decide on the set of network validators. DPoS comes in different variations, but generally speaking, it is the most performant consensus algorithm in existence.
In EOS DPoS, each token holder is allowed to cast votes for up to 30 BPs. Each block producer the user votes for receives an equal number of votes based on that person's total stake. For example, if a token holder owns 1,000 EOS, each of the 30 block producers they vote for receives 1,000 votes. Thus, any user's voting power is proportional to their total ownership stake in the network's tokens.
The top 21 block producer candidates by total number of votes received are then allowed to sign and produce blocks on the network. These 21 nodes form the core group of validators, and they are compensated by the network in the form of block rewards. While the core 21 BPs earn the majority share of rewards, the network also compensates a number of standby block producers who are ready to step in should any of the core 21 BPs be voted out.
DPoS is a form of liquid, representative democracy using a transparent on-chain voting system. Users can vote directly for block producers or they can trustlessly proxy their voting power to another user to vote on their behalf. DPoS is the first example of a global, 24/7, digital election. Block producers must not only be voted in by token holders, but they must also maintain token-holder approval, as they can be voted out at any time and replaced by a standby block producer.
Finally, DPoS is an incentive system that motivates block producer candidates to compete with one another to bring value to the EOS community. In proof-of-work systems, any miner with enough computing power can produce blocks on the network. With DPoS, block producer candidates not only have to secure the network by reliably producing blocks, but they must also earn the approval (and votes) of the community by offering additional value-add services to the network and the community. Once voted in, however, block producers work together to secure the network and produce blocks reliably and with as little latency as possible.
5. Which use case for EOS are you most excited about?
Cox: My personal favorites include the IRYO approach to using the EOS blockchain to protect access to medical records, decentralized credit scoring, smart-contract-based insurance, self-sovereign identity and logistics and supply chain.
Snider: I think that EOS stands to capture some of the most exciting use cases of blockchain technology that have not yet been possible on other platforms because of issues around throughput, latency and fees. EOS is extremely high-throughput and low-latency, and it has zero individual transaction fees. This opens up the design space for DApps substantially. I think that the first major use case we'll see take off on EOS is blockchain-based games. Game users need to make many transactions and don't want to pay to do so, and EOS is the only blockchain that can provide that experience. In the future, I think we'll see blockchain-based social networks, fully on-chain decentralized exchanges and tokenized securities on EOS, as well.