What is EOS?
EOS is a blockchain that is designed specifically for facilitating “decentralized applications”, or DApps. This is different from Ethereum which has made its mark by focusing on “smart contracts”. A smart contract is a way to make a record of an agreement that is settled in one transaction. A DApp may continue to run indefinitely over many transactions, where each transaction is one step of an overall computation. EOS can, therefore, be thought of as a distributed computing resource, taking advantage of otherwise idle computing power and applying it to complex problems that require more power to solve than any one user may have in their possession.
There are two significant problems that must be overcome in order to make DApps worthwhile. One is that transactions must be inexpensive, because if a DApp becomes more expensive as the number of operations increase, then the cost of running the program might outweigh the benefits of its results.
The other problem is the speed of transactions per second. While every blockchain aspires to clear as many transactions as it can with every block, the ultimate goal of facilitating human beings making individual transactions, such as purchases, puts a rough upper boundary on how many transactions would be expected. For example, Bitcoin is often compared to whether or not it could handle a throughput of transactions similar to Visa, at 1600 transactions a second. However, any DApp could create an algorithm that expands to create nearly unlimited operations, limited only by what any programmer hopes to achieve.
EOS claims to solve both these problems. Its distributed proof-of-stake, or DPOS, a method of validating the blockchain requires no transaction fees from users, giving DApp developers access to all of the computing power of EOS at no cost. The same DPOS mechanism also shares computing power in such a way that EOS developers claim can process millions of transactions a second.
A DApp is a program that is run on a blockchain so that the work required to make it run is spread out over all the computing power supporting that blockchain. Examples of DApps running on EOS are Everipedia, a recreation of Wikipedia, Ono, a decentralized social network, and Oraclechain, an Oracle service provider.
One may note that all these DApps are versions of applications that exist in other forms elsewhere, so there does not yet seem to be a service that could not exist without the support of EOS. The current offering of DApps establish EOS as a viable way to run a program, but the killer app has yet to be found.
EOS uses a distributed proof of stake algorithm. Instead of blocks being more likely to be mined by people who have more computing power, blocks are verified by people designated as “witnesses” who are selected by an ongoing process of voting. In this way, bad actors can’t overtake the system as unwelcome actions will cause them to be voted out of their position. Also, it increases speed because blocks are not verified through a competition of computing power. A proof-of-work system puts all participants in competition with everyone else, driving an incentive for more and more computing power. In a proof-of-stake system, the designated witnesses are limited to a small pool, so there is reduced incentive to commit vast resources to compete with others. The savings in computing power by not competing can be put toward improving the throughput of calculations. According to EOS developers, they can handle millions of transactions a second, as compared to Ethereum’s 15 transactions per second.
Currently, EOS is 5th overall in market capitalization, with a little over 12.5 billion dollars invested. However, by having a clearly defined niche in distributed computing power, its success is not necessarily relevant to other top cryptocurrencies like Bitcoin or Ethereum, as they are going after different use cases. While a large market capitalization is an indicator of market confidence so far, EOS still has a ways to go to establish itself, in particular by demonstrating itself with a particularly popular DApp, which has yet to emerge. Until that time, competitors such as Golem, who are going after the same use case, could have a chance to catch up.