Last fall, research and testing firm Whiteblock published its report on EOS, casting doubt on its scalability credentials, and more damningly, arguing persuasively that it was not even really a blockchain. Put succinctly, their basis for this stunning claim was the fact that in EOS, the ordinary nodes do not actually validate the ledger and, instead, merely interact with the 21 block producers. Since then, EOS representatives have tried, but failed, to provide a convincing refutation of Whiteblock’s claims.

The DPoS approach to block validation used by EOS also appears to sell decentralization down the river, while paying lip service to it. There is nothing in it that prevents the 21 designated block validators from forming cartels, buying votes and even reverting past transactions at will.

Presumably, the real motivation behind this kind of behavior is to benefit from the hype around the blockchain space as fast as possible, bypassing the difficulties posed by attempts to actually implement genuine decentralization at scale.

Last week gave us an indication that at least some actors in the EOS community have also carried this dubious approach to DApp development. The pinned tweet at the time of writing proclaims Eva to be a “mass market decentralized [italics mine] application… building a bridge between the old and the new world.”

 

Eva Coop’s claims to have solved decentralized ridesharing

This author’s article on the difficulties faced by decentralized ridesharing caused active discussions on Reddit (here and here). It also attracted representatives of a project in the EOS ecosystem called Eva Coop. They claimed that the article was beside the point because Eva Coop already had a solution operational in Montreal. The solution is in the form of a ridesharing DApp powered by the EOS platform.

According to the project’s Twitter, their app, has, indeed, been available in Montreal since at least this month. I could not test whether it is functional, however, since I do not reside there. Nonetheless, even setting aside the question of its current functionalities, a much larger problem presents itself.

 

Why Eva Coop’s approach falls short of decentralization

Despite its community’s claims to the contrary, Eva Coop’s solution manifestly does not qualify as a decentralized ridesharing DApp. If we look at the project’s whitepaper, we will find that:

  • Eva is an application allowing P2P ridesharing interactions between members of territorially defined cooperatives
  • Those interactions are handled through smart contracts on the EOS platform (presumably, one smart contract per cooperative)
  • A server will be used for fiat payment and navigation services

 

This implies that not only will Eva use the EOS platform that is not really a decentralized blockchain, in the first place, but it will also involve cooperatives as a required intermediary layer and involve centralized servers.

Naturally, critics can argue that having data storage and messaging outside traditional client-server architectures is not essential for ridesharing to be considered decentralized. However, this response does not withstand scrutiny.

While cooperatives seem very different from large corporate intermediaries like Uber and Lyft, they still are legal entities from the standpoint of the existing legal systems. This implies that they can be ordered by governments to engage in, or refrain from, certain activities, as well as shut down completely. There will be no way to bypass the cooperative structure at least because centralized servers are involved in navigation.

Of course, it may be somewhat more difficult politically to shut down the activities of a cooperative compared to a multinational as multinationals have a negative reputation, while cooperatives have an air of being for the people. However, much of the opposition to ridesharing is driven by pressure groups of legacy taxi service providers. They will not hesitate to virulently attack cooperative-based ridesharing, and will still largely enjoy the political advantage of established interest groups.

Perhaps less importantly, it will be the cooperatives that will be responsible for setting prices for rides in the Eva ecosystem, or more precisely, the parameters listed on page 16 of the whitepaper. This means that the advantages of more bottom-up price formation compared to the Uber model will not be realized by Eva. If anything, given that cooperatives will lack the resources that companies like Uber have for fine-tuning their algorithms to better adapt to fluctuations in demand.     

 

***

To sum up, it appears that the approach consisting of having only semblance of decentralization while making grandiose claims may be a feature and not a bug of the EOS ecosystem. Why conduct serious research and development with uncertain prospects, if you can just choose a fast hack and add a lot of hype? Unfortunately for the blockchain space, this strategy is well-funded thanks in large part to the EOS’s ICO, so it is probably not going away anytime soon.

3 COMMENTS

  1. Disagree. This is hype anti-hype pice. EOS will evolve quite happily in spite of this article. When it comes to dapp platforms you need some centralisation. All sort of solutions coming, thanks to dPoS

LEAVE A REPLY

Please enter your comment!
Please enter your name here