Near Protocol

Near is a blockchain and competes with other blockchains promising security, decentralization and scalability. It must be noted first, that this industry is highly competitive, but is attractive thanks to many infrastructural solutions proposing. The team demonstrates a good understanding and vision of the market. Its first step is to enable mobile devices to join the network - referring to an existing trend of mobile devices dominance. Their strategy aims to attract developers first, especially those, who have not touched blockchain yet. By the way, the key point is to join mobile devices that seems to be a right idea for the industry but still remaining complex. Current mobile devices' users are hardly to keep devices on - more battery usage and more internet traffic spent. Here the team understands the problem by assumes that these devices will be available to the network a few times a day with a low network connection. Focusing on mobile devices appears to be a good strategy initiative, but the telecom industry is not prepared good for a wide adoption of such ideas and we are likely to wait for about 5 years to see such networks performing effectively. The same issue is related to projects like Rightmesh, also ignoring that the telecom industry is not ready for maintaining such networks. The positive point is that the team provides a reader with the comparison among different blockchain and different consensus algorithms that helps to explain the need for developing their own blockchain with their own consensus.

Near aim to develop a blockchain that will be maintained by mobile devices arranged by its own consensus algorithm Thresholded Proof of Stake - shards both the state of the network and the processing of the transactions and scales capacity linearly. The consensus splits transaction verification (each verifier only executes and validates transactions in his shard) and block signing (here nodes create global blocks that encompass transactions from all the shards). Once a block is created, it is irreversible, so an attacker needs to corrupt all witnesses not just those assigned to a certain shard to corrupt a block. The consensus supports Direct Acyclic Graph (DAG) of messages built by verifiers. Also, Near will allow dApps that are also an integral feature among new blockchains. The prototype is scheduled to the August 2018, but we found no releases still. The roadmap is fast enough having the testnet ready in December 2018 and mainnet in March 2019.

It is still unclear how token will be implemented in the network, but the architecture designed makes us suggest, that token will appear as a means of payment and reward for nodes.

No financial information available

The team has experience in building sharding systems under MemSQL2, a distributed database that has large cluster deployments powering analytics at Uber, Pinterest, Goldman Sachs, Akamai and other large corporations. The team is experienced enough and includes 9 members.
Co-founder Alex Skidanov -  2 years in Microsoft, 5 years in MemSQL (Architect and Director of Engineering), 2x ICPC medalist (gold in 2008 and bronze in 2005) - respected collegiate programming competition with some distinguished winners (like Nikolay Durov).
Co-founder Illia Polosukhin - 6 years in Salford Systems, 3 years in Microsoft.
Senior Software Engineer Evgeny Kuzyakov - 3 years in Google, 3 years in Facebook.
Not to mention, other team members are also experienced on the same level with strong achievements as well.