Ethereum’s founder, Vitalik Buterin, addressed a crowd yesterday at the platform’s annual conference, Devcon, sharing his vision for the blockchain’s near-term future. The centerpiece of the speech was sharding, a process whereby data are partitioned into subsets; in the blockchain context, these smaller data packets can be stored on nodes that would otherwise have to store the entire history of all transactions across the network.
While the software that would enable this is still being developed, Buterin believes that this approach represents Ethereum’s most likely option for reducing transaction costs and processing times and enhancing scalability, which he identified as “probably problem number one” for the platform.
Furthermore, as he hinted during his speech, the Ethereum team may be close to completing a sharding proof of concept in the programming language Python, though the deployment of such a solution across the network is further off.
Ethereum’s researchers have been exploring several challenges relating to sharding systems, Buterin explained, including the problem of enabling trust between nodes that are storing only some of the data that comprise the network’s distributed ledger.
He also took the opportunity to propose a new type of sharding, in which full nodes storing the entire ledger exist alongside other nodes storing sharded data – an architectural scheme that he anticipates will enable developers to trial experimental changes on shards that won’t substantially impact operations and processing times platform-wide.
He alluded to other projects in Ethereum’s pipeline as well, including upgrades to the Ethereum Virtual Machine; a mechanism called Ethereum flavored WebAssembly, or eWASM, which would allow web browsers to act as Ethereum nodes; and a concept called “stateless clients,” which, if developed and implemented, could streamline the process of syncing clients with the network.