Ithaca and liquidity baking explained
Ithaca was the Tezos network's proposed ninth upgrade that, earlier this year, fell short of the necessary votes to reach a majority and move forward to implementation. At the heart of it was a planned shift to Tenderbake deterministic finality and a polarizing liquidity baking model.
To understand what happened, we need to go back to the beginning. Tezos, like most blockchains, uses inflation funding to pay bakers and delegators. Delegators go through bakers to get their stake. The principle is that each year, a new amount of currency is created, which is given to bakers. "Using inflation to fund other things is the baking part of liquidity baking," Alfour explains.
The liquidity part works like this. "If you have a device system that lets you trade between two tokens, A and B, people can use these tokens to help with the liquidity of this automatic trader. If there are only 10 of each token, the trading rate will change very quickly," he adds. "If I buy nine tokens, the price of the last token will be super high, so by buying those nine tokens, I have changed the price to something almost infinite. If there are 1000,000 of a token, you need to buy a lot more to significantly change the price."
Why does this matter? "You want price changes as a reaction to short-term buys and sells to be as low as possible," Alfour says. "With liquidity baking, part of the inflation is used to add more liquidity to a given pair, so people have an incentive to add even more liquidity. It's also known as liquidity subsidy. It's not meant to replace it, [it is] just there as an additional incentive."
Alfour believes the ‘no vote’ wasn't due solely to people rejecting Tenderbake and liquidity baking. On the day, he says, "two big bakers voted ‘no’ right at the end, and there wasn't enough time for other bakers to vote ‘yes’. Not all bakers participated, so the vote of a few was enough." While the ‘no’ vote caused a frenzy on blockchain social media, Alfour feels it was a storm in a teacup. "It was only a setback of three months or so," he says. "Most of the actors voted in good faith even if they didn't understand the technical details."
The future of Marigold and Tezos
Founded in November 2020, Marigold is focused on testing and developing upgrades to the Tezos protocol. Alfour and his team have two major workstreams in place for future improvements.
"The first is a protocol amendment that adds optimistic roll-ups to transactions like tickets," he explains. "The second is optimistic roll-ups in smart contracts. You benefit from better code, performance, and integration, and having the workstream in a smart contract de-risks the first workstream. If the tech lead of the first workstream dies tomorrow, you don't want the entire layer 2 effort of Tezos to stop. The other advantage is you don't need to wait for a vote or an amendment, so it can be deployed more quickly. If there is an issue, you can redeploy your contract, so it lets you be more iterative. Whereas with an amendment, you have to do a lot of reuse and tests before you can even think of merging it."
Alfour is also building a new, more accessible infrastructure for the Tezos network. "It's very sad to have an epic blockchain with epic consensus but institutions don't use it because they don't have this building block they need."
Alfour is also building Deku, Marigold's new sidechain software. There are two main components to this work. "The first is a canonical Tezos sidechain, which means if you care more about an application's speed and less about decentralization, you can move to the canonical side chain and interact with everyone else who has done so."
The other is enterprise. "A lot of institutions want to come to Tezos but don't want to start on a big blockchain. They want interactions with Tezos but don't want their main work on Tezos itself," Alfour explains.
"Our enterprise sidechain is a nice trade-off. A lot of projects are skittish about moving everything to a new technology. Being on the sidechain means you still have things you can control. It's a gateway solution between fully custodial cloud and fully decentralized blockchain."
Image by Kelly Sikkema via Unsplash