Whiteblock CTO EOS is a distributed database rather than a blockchain!
Recently blockchain testing and research firm Whiteblock Inc. released a review report that is extremely detrimental to EOS and refers to EOS as a “distributed homogeneous database” that is disguised as a blockchain. In this report entitled "EOS An Architectural Performance and Economic Analysis" Whiteblock analyzes several aspects of the EOS protocol and draws conclusions. EOS is considered to have serious flaws in security and network performance is much lower than what it claims to achieve.
Amazing discovery
According to a report by Brent Xu Dhruv Luthra Zak Cole and Nate Blakely of the Whiteblock research team the EOS network which is titled "Ethereum Terminator" has many shocks in security and agreement. Failures and fatal damage to a variety of applications proposed by the EOS network.
In the two months since EOS went live in September this year the Whiteblock team has continued to test whether the EOS network's trading throughput can achieve its stated performance. In addition the test included EOS response to adverse network conditions how to deal with variable transaction rates and scale average transaction completion time tolerance to partitions and fault tolerance. The test results are far from satisfactory.
Whiteblock bluntly stated in a press release on the EOS test on November 2
“EOS is not a blockchain but a distributed homogeneous database management system. The most significant difference is that EOS transactions are not encrypted. In essence the EOS token and RAM market is a kind of Cloud services the network provides users with a computing resource commitment in a black box and the user accesses through credit. Because EOS lacks transparency in the computing power of the block generator EOS does not provide an accountability mechanism."
Whiteblock measured the actual throughput of EOS under “practical” network conditions and the results were significantly lower than the claimed flux in the EOS marketing materials. In addition the Byzantine fault tolerance mechanism is missing from the EOS network and there are basic security issues such as the failure of repetitive consensus.
According to CCN's report in June this year EOS was controversial within a week after the release of its main network version as a result of an incident caused by block creators which led many to question the extent of decentralization of EOS networks ( Annotation This July EOS has security risks due to private key generation tools. The created private key was discovered by hackers and implemented a "rainbow" attack resulting in theft of digital assets in the account resulting in the loss of tens of millions of digital assets. . Whiteblock's test report confirms that people are concerned about the existence of the problem which will seriously affect the value of EOS.
Whiteblock gave concluding observations on the overall network of EOS
“Our findings indicate that the performance claimed by EOS is inaccurate. The EOS system is based on an incorrect model that is not truly decentralized.”
CCN's exclusive interview with Whiteblock CTO
After the test report was released CCN interviewed Whiteblock CTO Zak Cole to explain in detail the report on the EOS community and blockchain ecosystem.
Question 1 The research presented by your team concluded that EOS transactions are not encrypted and should be referred to as a distributed homogeneous database rather than a blockchain. What does this mean for the EOS ecosystem? Whether this will significantly change the goal that EOS promises to achieve (ie become the Ethereum killer). Should EOS investors and users be worried about this?
Cole I hope that our findings will provide a healthy basis for discussion in the community rather than provoke a political war between competitors. I believe that a clear roadmap to achieve the original intention of EOS requires an assessment of the long-term goals of the EOS ecosystem. Given the fundamental differences between Ethereum and EOS comparing these two systems is completely meaningless. The former is a decentralized P2P network supported by cryptographic authentication while the latter is an optimized distributed database which is functionally more similar to the IaaS service products provided by some general cloud computing platforms.
Whiteblock is a blockchain fan not a fan of a particular system such as EOS or Ethereum. The purpose of our research is not to prove that the systems are better or worse but to provide a reference objective and scientific analysis for the community in the process of building high performance and functional systems. The Whiteblock team will also provide guidance for the EOS hackathon (Hackathon) in San Francisco next week. Our only goal is to help build bridges that transform blockchain technology from a marginal science to a viable solution that delivers real-world applications and shapes the future of a decentralized world. That's why we launched the Whiteblock test framework.
The development tools required by the community should provide transparent and objective performance data. These performance data will identify facts and marketing terms to capture the capabilities of the systems under construction. The most fundamental problem with EOS is that it is currently unavailable and does not provide the claimed flux in the short term. EOS simply cannot support its multi-billion dollar vision. There is still a lot of work to be done and I hope that the EOS team can deliver on all the promises. In any case it is a very meaningful experimental exploration of distributed computing.
EOS investors and users don't have to worry too much unless they want to profit from emerging technology markets that are outside of regulation through speculative betting.
Question 2 Your research also points out that the actual flux of EOS is significantly lower than what it claims to achieve. What does this mean for users and DApp developers in the eyes of laymen?
Cole When choosing the platform that best fits your DApp developers should first evaluate their priorities. If you want to try to implement the functionality of decentralized P2P transaction logic you should ensure that the selected system actually provides the functionality needed to perform the operation. If you want to achieve high transaction throughput you should consider the problems with existing payment gateways such as Shopify or Stripe. Of course if the traditional client-server architecture is equally applicable it is understandable to stick to it.
Another important thing to note is that EOS is not really free of transaction costs. Instead transaction costs are transferred to DApp developers making the cost of running these DApps very high. We have seen in many software systems that this will create an app market similar to Apple's App Store and users may eventually pay far more than expected. I don't know if you still noticed that as the delay and the number of users increased the transactions that were successfully processed dropped significantly. This is not just because of flux but there are some more important factors at work.
Question 3 Does EOS pose a security risk to users in essence or is the above disadvantage of EOS repairable?
Cole I think there are some inherent security holes in the EOS system. EOS does not effectively implement game theory or adopt some additional algorithmic mechanisms to ensure that the current stored assets of the user are available for future use or access. All of the value of the EOS consensus model is based on the capabilities of the block creators who are voted by the token holders and perform functions on their behalf. Once it is impossible to prevent the block creator from voting for his own interests does this consensus still make sense? Even if it exists EOS does not provide functions encryption calculations or other means of governing the behavior of block generators. This problem is very obvious and does not need to be discovered through our three-month research project.
Having said that these shortcomings can be fixed. But if you fix the above problem EOS may become the same as many master systems (masternode) such as Dash or Syscoin.
Question 4 The study was funded by ConsenSys so are there any conflicts of interest?
Cole Our research has also received funding from more than a dozen institutions in addition to ConsenSys. For example Bo Shen the co-founder of Bitshares and former partner of Dan Larimer also provided funding but most of Bitshares' technology is based on EOS. ConsenSys also funded a portion of the research program which does not affect the scientific process although this should be considered a controversial issue. Similarly we conducted the same test on Ethereum and pointed out some shortcomings of Ethereum. The Ethereum community accepted our findings and supported our further research. Currently we have worked with dozens of blockchain systems. The purpose of our test is not to point out the merits of a system this is not a beauty contest. In order to build a more efficient and higher performance system we should be objective and transparent identify weaknesses in the system and help the system to consider these issues and optimize them in the further design process. If the EOS community chooses a conflicting attitude toward testing and observing such properties then the entire ecosystem will be doomed to fail and will certainly never achieve its claimed scale.
Our research report is publicly available which also points to a number of significant issues in the safety and performance of Ethereum.