The Tally Newsletter, Issue 56

December 17, 2021

Dec 17, 2021

Welcome back for issue 56 of the Tally Newsletter, a publication focused on all things decentralized governance. We’ll keep you updated on key proposals, procedural changes, newly launched voting systems, shifting power dynamics, and anything else you need to know to be an informed citizen.

This week we cover:

Compound Voting on Competing Audit Proposals
Aave Community Selects Business License for Aave v3

Plus brief updates from the wider ecosystem.

Compound’s Competing Audit Proposals Are Up for Final Voting

TL;DR: Compound benefits from strong proposals from the three organizations, but the current governance process may not be well suited for competing proposals.

Following last month’s vote on OpenZeppelin’s auditing proposal, which was rejected at the last minute based on interest from competitors, Compound is again considering auditing service providers in the DAO ecosystem’s first competitive B2D (business to DAO) contracting process.

Robert Leshner @rleshner

This is a monumental step for DAOs and the future of work: Three companies are competing in the open for the chance to provide services to, and get paid millions of dollars by, a DeFi protocol. Whoever wins, the Compound protocol & community is in great hands.

Compound Labs @compoundfinance

Proposals 075, 076, & 077 are competing #B2DAO proposals submitted by @chain_security, @OpenZeppelin, and @trailofbits to provide audit & security services to the Compound protocol. Voting begins in 2 days. Discussion: https://t.co/ai0Krlk2uQ https://t.co/MUaLxqnbRH

While Compound has contracted with outside organizations before (for example Gauntlet’s risk management program), previous vendors have been unopposed which left governance with a relatively simple yes or no decision. The current batch of votes involve separate organizations with different offerings and cost structures, which creates a significantly more complex choice for voters.

Jeff Amico @_jamico

To conduct a competitive RFP for a service like auditing is challenging under any circumstances. To do it fully out in the open and subject to community input is another matter entirely.

As DAO’s begin to contract more with outside service providers, it could become very difficult for individual voters to make informed decisions. But the alternative option of leaving contracting to a centralized committee or core team risks creating a non-competitive environment with higher costs and possibility of self-dealing.

Competitive bidding also raises important issues about the Compound Governor framework; namely Compound style governance is not well suited for multiple choice proposals. Each compound proposal allows for yes, no, and abstain votes, and includes a single set of executable instructions (eg funds transfers). With no ability for proposals to directly count against each other, there’s a possibility that COMP voters could inadvertently approve multiple competing proposals with overlapping budgets and deliverables.

Source: Tally

In this case, it looks like OpenZeppelin’s resubmitted proposal is out to an early lead, having already received 900,000 votes (over twice the minimum 400,000 quorum of votes required to pass a proposal). Relatively close coordination between large Compound voters and investors may have helped avoid disorderly approval of multiple vendors. But this case still points to the need for more sophisticated DAO contract tendering processes.

Aave Votes for Business License for Aave v3 Codebase

TL;DR: Similar to Uniswap v3, a business license will restrict copied projects for several years before transitioning to open source.

Aave is gearing up to launch the v3 upgrade of their core protocol. Some of the top features of Aave v3 include:

Efficiency mode: allowing much better capital efficiency while borrowing and lending similar assets such as stablecoins
Isolation mode: limits borrowing capacity for risky collaterals, which allows for onboarding more long tail assets
Exposure ceilings: similar to MakerDAO’s debt ceilings, limits total market exposure to selected assets to reduce tail risks

Altogether, these upgrades help address one of the most glaring shortcomings of defi money markets - traditionally money markets are only as strong as the weakest asset accepted as collateral. We’ve seen this with the recent Cream Finance hack, here an issue within a single collateral asset led to marketwide insolvency. Considering this important advantage, we could expect an array of forked/copied projects to emerge using the Aave v3 codebase if it was immediately released under an open source license.

The Aave community had been voting on license options via a Snapshot poll, with a business license considered against open source options of MIT or AGPL license. The MIT license option took an early lead, but in the end the business license option prevailed with strong support from defiactivist.eth, one of the largest AAVE token holders.

marczeller.eth @lemiscate

- the very large voter that turned the fate of the vote represented less than half of total Business vote. His stack represents slightly more than 1% of the supply. A lot of people voted, but a whole lot didn't.

While some commentators were upset about the restrictive license choice, and outsize influence of a single large voter, participation took place in a gasless voting environment so any holders should have been free to participate if they favored a fully free and open source license.

Details are still being determined in a forum discussion, but it is expected that the business license would work similarly to Uniswap’s v3 license; reuse would be restricted for a period of 2-3 years, but governance would be able to grant exceptions to the license on a case by case basis. And the source code would be available for review from day 1, which allows for security review and responsible disclosures which are an important benefit of open source projects.