A Brief Update on ETC and Grothendieck

This is Charles Hoskinson's first appearance on the Let's Talk ETC podcast, hosted by Carlos and Christian, to announce IOHK's Grothendieck team building a new Scala client for Ethereum Classic.

IOHK could not choose between an Argentine team and a Polish team during Grothendieck hiring, so it hired both, growing the group to seven people led by Alan McSherry.

The Grothendieck team has strong computer science backgrounds but no prior cryptocurrency experience, so months were spent on white papers and a smart contract Udemy course before writing any client code.

The team is roughly halfway through understanding the Ethereum protocol and has implemented 15 to 20 percent of it in Scala as a pedagogical prototype.

IOHK runs the Grothendieck team on daily reports, Gantt charts and scrum, and started broadcasting weekly live standups on February 6, with the roadmap due to go public within a week or two.

The team is now dissecting every EVM opcode to reach an understanding on par with Gavin Wood, a process expected to take about two more months; documentation has been solid apart from minor differences found between Geth and Parity, with a testable beta targeted for exiting April.

Turning that beta into secure, enterprise grade software should take another one to two months through a security audit, likely with the firm Grimm, while two Grothendieck developers co located with Rootstock in Buenos Aires give access to Sergio Demian Lerner's DECOR plus GHOST innovations and a five times faster EVM than EthereumJ that IOHK hopes to borrow.

The plan is no beta before June and no enterprise grade release for a few months after, with a fully independent client compatible with Geth and Parity targeted for the third quarter of 2017, bundled with an IDE and Solidity tooling such as Truffle; Daedalus, IOHK's Electron based multi currency wallet, will also be extended to support Ethereum Classic.

Monetary policy work is currently in the hands of Igor Artamonov, known as Splix, and the Geth and Parity core developers who have carried out both of Ethereum Classic's hard forks so far, since Grothendieck will not be ready in time to help set it.

The community is largely in favor of sticking with proof of work, and there is debate over whether to keep, upgrade or replace GHOST since it is tied to monetary policy, with a hope to finalize a proposal by March or April.

IOHK's Q4 2016 paper on a Dash style treasury system is being followed by new design documents due in a week or two, developed with a university in England over an arc of three to six months, with the choice still open between a final monetary policy like Bitcoin's or one with a hook to revisit and possibly add a treasury system a year later.

A gathering of the Ethereum Classic community running February 25 to 27 is a good venue to keep the monetary policy conversation going since key players will be there, and China in particular needs strong representation as one of the community's key drivers, even though IOHK cannot directly implement any change given Grothendieck's timeline.

Charles frames IOHK's approach to Ethereum Classic through Shuhari, the martial arts idea of learn, master, then innovate, before anyone earns the right to propose changes.

Being an Ethereum founder does not make him automatically credible on Ethereum Classic's direction since he left the project early, so IOHK must first prove mastery of the existing protocol.

Once Grothendieck has proven its mastery of the protocol through a delivered client, open code and public audits, IOHK plans to bring proposals for late 2017 and 2018 that differentiate Ethereum Classic from Ethereum with its own use cases and view of computation.

At the Shanghai winter school, Charles discussed state channels, new consensus concepts and blockchain partitioning with cryptographers, noting IOHK already has 10 people working on proof of stake every day even while it researches proof of work.

Aggelos Kiayias's proofs of sublinear complexity make a secure zero knowledge checkpointing system easier to build under proof of work, and IOHK currently has a better handle on proof of work side chains than proof of stake ones.

Proof of work needs better PR and honest acknowledgment of its downsides, since five mining pools control all of Bitcoin's hash power, but it remains a viable, smart contract compatible paradigm if mining can be decentralized further.

Cornell's IC3 group has been researching Intel SGX trusted hardware for concepts like sealed glass proofs and Town Crier, an Augur like oracle built on trusted hardware, and since ARM has an equivalent secure hardware zone, this kind of hardware could soon be democratized across every phone and computer.

Charles is excited about a Bitcoin-NG style two stage mining protocol, sometimes discussed as twin chains, and about treasury systems as a genuine alternative to centralized ICO style foundation funding, joking that a single funding decider turning out like Donald Trump is exactly the risk decentralized capital access avoids.

Formal verification is one of the most niche topics in development, rarely done outside academia, especially in programming language theory and functional programming; having already built a cryptocurrency in Haskell, IOHK wants formal verification too and is defining an ontology of what should be verified across the cryptography, the network stack, and consensus.

IOHK is retaining a specialized firm to teach it formal verification techniques, hoping to bring them into the Grothendieck team's practices by June to produce the first formally verified Ethereum Classic client.

Related work includes a Chalmers University paper on using dependent typing in smart contracts, and Yoichi Hirai's effort with the Ethereum Foundation to formally verify EVM bytecode.

The goal is to unify scattered formal verification research into real developer tooling within six months, though most everyday smart contracts will keep using Solidity, with formal verification reserved for library level work like OpenZeppelin.

The research roadmap covers proof of work efficiency and trusted hardware, decentralized funding, and formal verification of both the client and its smart contracts, with outputs expected in the second half of 2017 and production by early to mid 2018.

Charles admits he did not think Ethereum Classic would survive when he first got involved, since it began as a protest movement rather than a serious project.

He compares abandoning a defended philosophy for convenience to Ron Paul suddenly endorsing the Iraq war, arguing that the DAO fork broke a social contract that should not be quietly accepted.

The early Ethereum Classic community was badly fractured, with colorful figures like Michael Trout among them, forcing a decision on whether it was a real project worth building or just a protest.

Charles credits the podcast hosts and other community organizers for solidifying a passionate group of people, which is what convinced him to commit developers to the project.

He compares the early Ethereum Classic community's passion to the early days of Ethereum itself, when there was no clear roadmap but plenty of inspiration to work harder.

Ethereum Classic's market cap once dropped to about 60 million dollars, partly after the Ethereum Foundation sold 90 percent of its stake, before recovering and stabilizing at a plateau.

Charles is optimistic about Ethereum Classic's future, expecting some scams along the way just as Mt Gox did not kill Bitcoin, calling ETC the honey badger of smart contracts.

He apologizes for being too busy to properly discuss dApp pitches people send him, then thanks Elaine, Carlos, Cody Burns and Avtar Sehra, as well as Roy Zou and the rest of the Chinese community, and mentions plans for sushi with a cryptographer colleague in Japan.

A show with Alan McSherry, the Grothendieck team lead, is scheduled for this Friday to go deeper on the team's work.

Alan McSherry's backstory includes basing Grothendieck's technology on Scorex, built in Russia, and personally traveling to meet team members in Russia, Poland and Argentina, with Malta and a smart contract meetup in Athens planned next; amid joking about him being a great Irish guy, Charles genuinely admires how Alan sets realistic deadlines rather than bowing to pressure.

Once the Grothendieck client is finished it will not need seven full time people to maintain, just occasional sprints for forks, and the team is expected to surface many dApp ideas for Ethereum Classic later in the year.

Community management has been mostly positive for Charles despite occasional hate mail, and he takes particular pride in how organically Ethereum Classic's DAO hard fork was executed without any grand planning committee.

He credits Igor Artamonov, known as Splix, as a coding machine, and calls Matthew Mazur's spreadsheet driven ECIP-1017 monetary policy proposal a genuinely organic piece of community work that he actually likes.

Once a monetary policy is set it should stay fixed, the way Bitcoin's 21 million cap is fixed and accepted despite its imperfections, which Charles compares to Seinfeld's Soup Nazi: the soup is good, but there is no changing the order once it is placed.

Working with the Grothendieck team reconnected Charles with the technical magic of Ethereum after feeling bitter and divorced from it for about two years, and he is grateful for the chance to contribute even if the community's own ideas end up winning instead of IOHK's.

Charles addresses the viral photo of him with Vitalik Buterin, taken at the Shanghai winter school, offering to explain the backstory for anyone who took issue with it.

Charles had not spoken to Vitalik Buterin in person since June 2014, even though the two attended the same multi day Shanghai winter school alongside IOHK's own cryptographers.

Charles was invited onto a Shanghai conference panel alongside Vitalik Buterin and Jonathan Katz, who banned any Ethereum versus Ethereum Classic talk from the panel's opening moments.

After the panel, an awkward silent standoff with Vitalik Buterin ended with a hug and the peace sign photo that went viral.

Charles denies building Ethereum Classic out of revenge against Ethereum, calling it simply good business to train a smart contract team on a live network.

He frames Ethereum Classic as ultimately good for Ethereum too, offering a familiar proof of work home for any minority resisting the move to proof of stake, which he calls a major philosophical and security shift rather than a minor tweak.