Dom Steil

Scalable Blockchain Protocols

Last week I attended the Wanxiang Global Blockchain Summit in Shanghai. There were a number of sessions but one that really stuck with me was about potential scaling solutions for distributed consensus protocols.

There is a concept of polyglot persistence which essentially means certain databases are meant for certain reasons. You have SQL database, noSQL databases, graph databases, blockchains, distributed fault-tolerant consensus ledgers, CRDTS; all of these different types of protocols. They have different consensus mechanisms, data models, query languages, state change and message ordering models such as validators or orderers; but in order to effectively scale they need to be to interoperate and need to be able to complement each other.

They will need to be able to share account state. They will need to be able to resolve any concurrency issues that arise. They will need to be able to resolve replication and eventually be consistent.

Most of my time is spent working at the application layer but an assumption that I have as that building solutions that will incorporate multiple protocols will be important in the future.

This type of approach is not necessarily found in the blockchain space specifically because of the economic/personal incentives one has if heavily invested in a specific protocol. This could be time learning Solidity versus Kotlin; this could be owning a bunch of Ether vs NEO; this could be working at a company that is heavily contributing to a particular open project.

Regardless of ones preference in blockchain/distributed ledger/consensus database; this past trip to Shanghai validated that taking a multi-blockchain approach and being open to learning the differences between them and how they could also complement each other is going to have a great effect on the way that solutions in the future will be built.

An example of this is the new Ethermint Token.

Ethereum and Tendermint In what is being called the Shanghai Accord; a new partnership between the two projects was announced in which they will work together to help scale blockchain technology. Tendermint a Proof of Stake consensus protocol and its new Internet of Blockchains Hub / project Cosmos is working closely with Ethereum project to essentially create derivate tokens on Tendermint infrastrucutre in what is being called a hard spoon. Essentially a new coin is minted on another chain (tendermint) that uses the account state or balances of an existing token on a chain (ethereum); and we have Ethermint. This could mean in an effort to provide a scalable cross-blockchain solution in which tokens now have shared state/accountbalances on the Ethereum Chain and the Tendermint Chain.

Content-Addressed Blockchain Resolvers This is but one approach to have inter-operable blockchains. Other approaches have been proposed historically such as pegged side chains but one to me that seems very interesting is leveraging IPFS at the network the content address and resolve through the different blockchain / databases (as long as there is a key value store) using multihash and the ipld thinwaist protocol. It is still much just concept but again it is aligned with the idea of having multiple chains interoperating to address some of the issues that blockchains are facing today.

Plasma Another approach is using Plasma; this allows transactions to occur offchain; only broadcasting them to the chain if there is a transaction dispute; you never sent me the money, vice versa). In this way the blockchain becomes an arbiter or a distributed court house; a way to have veridical computing if there is in fact a need for the chain based settlement.

Proof-of-Stake Another take away and interesting concept is the movement to proof-of-stake. Ethereum’s movement towards POS is in Vlad Zamfir’s Casper protocol. Other chains such as Lisk uses a dPOS system and Tendermint is on POS as well. A common theme between them is the concept of weight. The stake and the weight of stake has some very interesting implications on the security and threat models that could be introduced to the protocol. More on Proof-of-Stake/Casper in another post.

There have been a few recent cross chain atomic swaps recently aswell.

Litecoin + Decred

And cross chain transaction component verifications.

ZKSnarks + Ethereum The Byzantium hard fork went live on Ethereum’s Ropsten testnet at block 1,700,000; one of the features was support for ZKSnarks. On the Ethereum ropsten testnet a a zksnarks component of the Zcash transaction was verified.

Public Chains and Financial Services Lastly from Nick Szabo’s talk on Smart Contracts the idea of a smart contracts acting a digital distributed vending machine was very interesting. There are certain security guarantees; and the rules are deterministic and known (you put in money into the machine something comes out). I have heard the vending machine analogy from as well, the concepts share the same vision of taking the best of public blockchain technology and the best of existing financial services.