JP Morgan's two key developers who were working on the bank’s private blockchain project Juno have started their own blockchain project.

Will Martino and Stuart Popejoy have branched off to start Kadena.io as of June 2016. Will Martino was a lead engineer on the Haskell Desk at JP Morgan’s Emerging Technologies Group while Stuart Popejoy built the advanced technology team for JP Morgan's Emerging Technologies Group which focussed on solutions for banking.

Kadena the blockchain is powered by ScalableBFT, the first consensus-BFT capable of scaling into thousands of nodes. The move by Will and Stuart made quite a stir in the blockchain world.

Open Source and big corporations

Cointelegraph asks Will Martino about what he thinks about large corporations using open source technologies if key developers leave and set out on their own projects.

Martino says:

“The decision to open source Juno was a well-executed strategic decision by JPM for the specific purpose of presenting it to the HyperLedger Foundation as a candidate for their technology stack. The decision resulted in JPM garnering significant thought-leadership in the private blockchain space. JPM’s Emerging Technology group has other open-source efforts as well, which it sees as a competitive recruiting advantage to garner attention and hire, retain, and motivate top-notch developers.”

New darling of the financial world

Private blockchains are attracting the attention of many banks and financial institutions. However how soon these new technologies will become part of our everyday life is an interesting question.

Will Martino thinks that these technologies will be witnessed by consumers as soon as financial technology moves through the adoption cure.

He thinks features like straight-through-processing wire transfers would be an example of such a tech making an impact on daily lives of people.

However at the same time the biggest beneficiaries would be larger institutions as infrastructure costs for them would decline significantly and they would be able to offer their client’s access to new technology.

Kadena addresses key blockchain problems

Will says that Kadena is the first blockchain to address the ‘pressing issues’ of blockchains in a private setting. He says that Kadena does not compromise the core guarantees that a blockchain should provide.

According to him, private blockchains have fundamental problems with inadequate miners’ incentives. Unacceptable performance is also an important issue.

Will thinks that performance issues are unavoidable but the incentive breakdown can be interesting.

He says:

“In a private blockchain, all miners are also participants. Is there an incentive large enough to negate the potential business advantage of mining a fork? Legal repercussions could become the new, negative incentive, but if we need a court’s involvement to get mining in private contexts to function can we really still say that mining in this context works? Centralization, e.g. registering transactions in a central 3rd party before mining them, is another potential solution, however that it isn’t distributed and begs the question “why are we bothering with mining if we’re centralizing anyway?”

While Classic BFT consensus approaches can and are being used to construct private blockchain solutions they tend to discard too much of the mining security model.

He thinks that one potential solution is to abandon public-key signature verification for TLS-connection based authentication but this would only offer ephemeral security in his view as the ‘proofs’ of authentication are lost when the connection is closed. In his opinion, private blockchains need durable security and every transaction’s authorship should be verifiable and auditable through cryptographic means.

Kadena, according to him, is the first credible alternative because it achieves high performance at scale while still providing durable security in the Bitcoin model.

Kadena’s difference from Juno

The why and how of how Kadena and Juno differ is an interesting question and one that we posed to Will. He told us that Kadena was founded on the belief that Tangaroa (Juno started as a fork of this project) foreshadowed a potential breakthrough in blockchain technology.

He told Cointegraph that their work on Juno helped inform this belief but ultimately it lacks a full smart-contract solution and has fundamental issues preventing it from scaling.

Talking about Kadena Will says:

“Our new blockchain, also called Kadena, is powered by ScalableBFT, the first consensus-BFT capable of scaling into the thousands of nodes. Until now, BFT consensus protocols like PBFT and SmartBFT have faced fundamental problems that prevent them from scaling. ScalableBFT, a “Raft+BFT”-family protocol descending from Tangaroa (Copeland/Zhang), features fundamental innovations that allow forflat scaling. We’re excited to announce that with ScalableBFT, Kadena has achieved an industry first: the Kadena blockchain can process upwards of 7,000 transactions per second with latencies well below 100ms, supporting hundreds or even thousands of consensus nodes. This achievement is even more notable in that it sacrifices none of the cryptographic robustness expected from a blockchain, proving full durable security. Our consensus whitepaper has more details.”

Smart Contract Language

Apart from this there are other differences as well. Kadena features Pact, which is a new declarative Turing-incomplete smart-contract language, whereas Juno was based on Hopper, which according to Will, never reached maturity.

Talking about the need for a smart-contract language for private blockchains that would be suitable for enterprise level transaction processing, Will talks about Bitcoin, Ether and Dao: 

“Bitcoin contracts are far too limited to support diverse smart-contract use-cases, leading to tokenization-based solutions.

Until now, Ethereum’s Solidity has been “the only game in town,” but is a terrible choice for transaction processing: it lacks major safety features such as database-style transaction commit and rollback; it allows risky, Turing-complete general-purpose computation within the blockchain, resulting in unsafe solutions as amply illustrated by the $50m DAO hack; it is an error-prone and unverifiable development environment, with primitive tooling and opaque, illegible bytecode, with no way to verify what Solidity code is running on the blockchain.”

Infrastructure engine

The Kadena project and the Pact contract language is seen by both Will Martino and Stuart Popejoy as ‘infrastructure engines’ that power blockchain technologies up the private-use adoption curve.

They think that private blockchains solve a raft of issues for transaction processing that normally requires a lot of operational and engineering expertise.

Talking about blockchains in the future Will Martio says:

“Remarkably, blockchain solves all of these problems in a single stroke, letting IT focus efforts on solving business problems by coding transactional logic into smart-contracts. Indeed, the much-lauded benefit of supporting utility-like inter-organization use-cases like payment clearing is almost an “added bonus”, albeit an obviously attractive and game-changing one in its own right. Also, a smart-contract language like Pact encourages the development of lucid, user-verifiable business logic, similar to the trust embodied in a Bitcoin system based on the use of a Pay-To-PubkeyHash or a multisig contract. In the inter-organizational case, we can even see wide standardization of logic which could provide for terrific business acceleration.”