Vote: Reduce delegation to 100k
Type: Snapshot
Although the direction of the EH project has shifted more prominently towards agentic governance as opposed to purely enabling individuals to band together and attain representation on a larger, more impactful scale, we are in support of seeing the agentic vision play out. Reduction of the total delegation amount to 100k enables further experimentation—plus we believe the EH team has attained strong enough goodwill on Arbitrum over the past year.
Vote: Wind down the MSS
Type: Snapshot
Allowing the MSS to be deprecated due to the existing issues is in our opinion a fair step. The AF is likely the best entity to take on this role for the time being, however, measures regarding transparency and proper disclosure should be top of mind for the AF. Since the AF is already involved in aspects like compliance checks for various payment programs, they should easily be able to adopt the MSS responsibilities. And in the future, a transition should be made for this system to run under the jurisdiction of the OpCo.
Vote: For
Type: Snapshot
Mountain Protocol unwinding makes the redemption of these funds obviously necessary. Since the capital that was allotted to USDM was already a part of the STEP program, we are in favor of rolling over the capital to STEP 2.
Vote: For
Type: Snapshot
This updated purview for the OpCo in our opinion is a fair extension—regardless of the potential overreach in the responsibilities, as mentioned by many delegates. Our perspective is that the OpCo will function with maximal efficiency if it is given a strong enough degree of flexibility while not undermining transparency. Discretionary spending and authorization to enter into agreements with vendors is well and good—as long as these matters are disclosed in a timely and justified manner. Constraints are also important to prevent excessive operational flexibility.
Vote: Abstain
Type: Snapshot
The initial post was asking for far too much for this integration, so, purely from a cost perspective, we were intending to vote Against this proposal. However, by the Snapshot phase, the cost was reduced to $60k. At this stage, because the costs aren’t too cumbersome, and the proposal seems to have a degree of merit in making the forums more robust, we are okay with the proposal passing. However, since the grant doesn’t present a surefire need for the DAO, we defer to an abstain vote.
Vote: For
Type: Onchain
Like in the Snapshot phase, we voted in favor of this proposal. There are too many incentive proposals that have failed to deliver long-term value to the DAO. Notably, STEP and LTIPP had some successes, but they could have been run more effectively. Namely, it’s very hard to determine whether or not a program has done well if the measures of success are difficult to surmise. Isolation of the different seasons to specific key areas of concern, like loop strategies or depth on xyz lending markets, makes assessing the efficacy of a program far easier. A large-scale incentive program introduced far too many confounding variables, which can be significantly controlled for if incentives are tranched over time and over disparate sectors.
Vote: For
Type: Snapshot
It seemed inevitable that a consolidation effort would eventually take place. Aggregation of the STEP, TMV, and GMC makes logistical sense. The OAT oversight aspect for the executing bodies is also an important inclusion. We are in support of reduced fragmentation and overhead—and are therefore in favor of this shift.
Vote: For
Type: Snapshot
Nova has solely become a cost sink for the AF. There was a spike in activity for Nova a couple of years ago when the Reddit team launched their rewards program on Nova since it was fast and cheap. But now, with the presence of new innovations like Orbit chains, the value proposition of a technology like Nova has since diminished. We are therefore in favor of removing the subsidies for Nova.
Vote: For
Type: Onchain
Since the Offchain Labs team is willing to collaborate with the Sky folks on implementation of this proposal, we are in favor of it. Ideally, proposals like this, and the previous Rari proposal, should not have to pass through an entire onchain vote—but we are sympathetic about the restrictive requirements around custom gateway contract inclusions. In a different forum post, we outlined potentially new approach that would make this process more seamless:
-
- Projects submit their custom gateway and L2 token contracts to OCL for a technical review to confirm compatibility with the Generic Custom Gateway Router, bridge security standards, etc. Community feedback on Sky’s proposal emphasized that delegates preferred that OCL was consulted as opposed to consulting the whole DAO, with most delegates having minimal to no key feedback on the proposal.
-
- Projects must attain an audit (like RARI was reviewed by OpenZeppelin and Sky by ChainSecurity) for their custom gateway and L2 token contracts, verifying security and compliance with Arbitrum’s bridge standards.
-
- After pre-validation by OCL and auditors, the project team can post a proposal with the stated stamps of approval so that delegates have an easier time conducting a review.
Vote: For
Type: Onchain
Ideally, quorum reductions do not occur simply because security is of paramount importance. A moving target when it comes to such an important parameter is unideal. However, after revising much of the ARDC data, it’s clear that issues around quorum margin are increasing especially with increases in supply. We are therefore—albeit reluctantly—in favor of this proposal.