Hard forks and contract hacks – Tezos aims to solve blockchain's problems

If you follow the news in the crypto space – Bitcoin divided, Parity wallets hacked – it's not surprising the recent Tezos fundraiser attracted so much attention. Tezos is a smart contract blockchain platform created by Arthur and Kathleen Breitman, which recently raised $232m (£177.9m, €196.9m).

It's worth remembering two key problems Tezos sets out to solve. It presents a way to carry out decentralised upgrades to its protocol via an on-chain voting system, addressing the governance problem inherent to decentralised systems and the likelihood of contentious forking scenarios. It also proposes a more restrictive programming model for writing smart contracts, so their behaviour will be more predictable and more easily formally analysed.

The recent Parity bug was a reminder that even the most talented coders can make mistakes and smart contracts require formal methods of verification. "We had an eye from the beginning on formal verification and formal methods," said Arthur Breitman.

"Even if you're a great programmer bugs are going to happen, and the only way we can guard against this is by using formal methods: rigorous mathematical proofs that a program does what we want it to do."

"The most advanced proof assistant, Coq, is written in OCaml, the same language used to develop Tezos, so there are affinities between the two communities".

Tezos is also presenting a decentralised system that can deploy upgrades automatically, like in Chrome or Windows, "but instead of having the upgrades signed by your vendor – because there is no central vendor – the upgrade is, in some sense, signed by everyone who is using the protocol".

He made a clear distinction between the Tezos governance system and its proof of stake consensus. "To have a system like this you need to start with proof of stake [PoS] because proof of work creates a political class or miners whose interest may differ from token holders.

"But PoS, by itself, isn't sufficient for governance. There is a misconception that PoS consensus can be used as-is for governance.

"If you try to use consensus over chain branches as governance you end up with a kind of Keynesian beauty contest. You win if you follow the herd, and that doesn't have anything to do with merit. It doesn't have anything to do with what branch you prefer, it has to do with what branch you think is the Schelling point, the most natural one."

Arthur explained the Tezos governance system starts off with proposals, which anyone can make. They are then rated using approval voting. "Several proposals are made and people who own tokens indicate the ones they like and dislike.

"You then take the one with the highest approval rating and you submit it to another vote to pass it. Tezos requires something like a 70% majority in order to adopt a proposal, which then undergoes testing for a month. If the testing phase goes well, then a second confirmationvote is held before adoption."

"Crucially, there is no penalty if you support a version that doesn't end up winning."

He said this is only the very first layer, and the process will look to improve the governance model itself, as well as the ledger. For instance, some proposals and decisions that are not critical will not require a full referendum vote.

"You can have something like elected representatives making decisions over some narrow technical aspects of the ledger, for instance. So we can progressively build on that very simple governance layer, a more complex and better adapted governance structure."

Asked about the ICO, the team made the point that a high volume of tokens must be distributed in order to set up a wide proof of stake network. Co-founder Kathleen Breitman said: "Proof-of-stake lacks the elegant distribution that Bitcoin has. So the easiest way to bootstrap the network was this type of fundraiser, with the side benefit of endowing a non profit with funds that can help promote these technologies – in this case for quite some time."

She said it's important to distinguish this process from people who might be thinking about doing a Series B for their company and decide it's easier to just crowdfund it.

Arthur Breitman added that some people said it was irresponsible not to "cap" the funds raised. But to have done so would have meant the majority of tokens would have been snapped up by professional ICO scalpers looking only to make a quick arbitrage, rather than reaching a network of participants.

Arthur Breitman, co-founder and CTO of DLS, said: "When we started, a whole coterie of people in the space insisted there ought to be a cap on the fundraiser. Otherwise it's irresponsible they said, never exactly articulating a good reason why. They've made up this idea that some sort of self-imposed cap is a kind of duty, and attempt to back their argument with poorly understood business principles sprinkled with moralistic undertones. Not having a cap felt like crossing a picket line."

Breitman believes imposing a cap of, say, $5m when people want to contribute $50m turns the process into race whereby a professional industry of ICO scalpers grab all the tokens to book some 1,000% returns.

"The absurd practice of letting professional scalpers capture most of the value of these projects is basically killing analysis technical due diligence in the space. At best, it's often reduced to superficial indicators, like the number of stars on GitHub, instead of actually doing an in-depth analysis of the value proposition of a technology, or how well it's implemented."

Regarding the distribution of tokens across participants in the fundraiser, he said there were some 30,000 wallets created.

"The Gini coefficient, a statistical measure of disparity within a statistical sample, is about 0.87, similar to the Ethereum fundraiser, a bit higher than the distribution of wealth in the US – so it's not a surprising number."

In terms of next steps, Arthur said from technical standpoint there needs to be a few security audits, and the network needs to be polished ready for launch. "I think the path to doing that is very clear, whereas the path to growing Tezos into a much larger project is still a little new," he revealed.

Kathleen added: "We've been helping out the Foundation by reaching out to people who might be a good fit for the team and the ecosystem. The next three or four months you are probably going to hear a bit from the dev team, but the next month or so it will heads down and getting the right people at the table."