Recent events have reignited discussions surrounding the oDAO’s near-term future. First, the removal of bootstrap mode has transferred the role of gatekeeping contract upgrades to the oDAO. Second, and more importantly, two oDAO nodes, both operated by Rocket Pool Pty Ltd, were compromised and subsequently voted out of the oDAO. The remaining 12 oDAO members are planning a vote to re-add these nodes back soon.
I suggest that we take this opportunity to make changes to prepare the protocol for healthier long-term relationship with the oDAO, as it’s estimated that it will take at years of development before it becomes technically feasible to dissolve the oDAO and distribute its responsibilities. The below suggestions do not require smart contract changes and are purposefully informal so that ensuing discussion can influence a more formal RPIP eventually. Thanks in advance for your thoughts and feedback!
Community application and voting process.
Currently, oDAO members may, through a majority vote, admit new members. The oDAO is a powerful and critical piece of infrastructure for the protocol, however, and there are currently no mechanisms in place to check the power of this group. A formal governance requirement that any proposed member must also apply and be accepted by the community is both reasonable and relatively simple to implement.
Expand the oDAO
The recent security incident involved nearly 30% of the nodes required for a majority. If a majority of nodes were compromised, the protocol as a whole could be halted or otherwise dramatically harmed, and expansion reduces this threat dramatically. That said, an oDAO which is too large prevents effective and timely coordination between members. Therefore, I suggest we expand the oDAO to include at least eighteen members.
Tenure limits
I suggest a term limit of one year. If a member wishes to continue participating in the oDAO, they may reapply and be voted in again with the next round of applicants. While the community did not get a chance to express its opinion of the current oDAO members, introducing a tenure limit would allow for these members to be evaluated fairly for future inclusion if they desire.
More restrictions on the type of members
I believe oDAO membership should be restricted to a) individuals and b) no more than two members who principally represent the same organization. The reasoning is that individuals are more appropriately incentivized to take ownership of oDAO duties than organizations for which the compensation is just a small percentage of overall revenue. I also suggest excluding Rocket Pool Pty Ltd from this second restriction until inflation can be adjusted to pay development costs via the pDAO instead. In the future, we can make more extensive changes like oDAO inflation adjustments which will allow the development team to be paid through alternative means so as not to conflate the incentivizing of oDAO duties with other expenses. For reference, I’ve written on this topic more extensively here.
I would go further and would let every active veteran node operator to be able to become a odao member after 2 years of active staking and enough stake in rpl. These are the people who are committed to the protocol. This would also fit with additional measures like community application and voting process.
Agree on all counts, with the possible exception of term limits (I see the benefit but am concerned that would introduce campaigning and governance overhead).
Perhaps not “near-term”, but I’d also like to suggest we expand the scope of oDAO changes to also include economic adjustments, specifically:
We need to find another way to fund the dev team that doesn’t rely on oDAO inflation allocation, as @thomasg mentioned here. It’s not long-term sustainable for the team to rely on oDAO funds for operating expenses as we expect the oDAO to be expanded or eventually removed. pDAO Budget Definition - #19 by thomasg
Currently the oDAO has an incentive to keep the number of oDAO members small, to increase the size of their allocation. It would be great to align our design for a decentralized oDAO with their financial incentives. Arguably with the current design they carry a conflict of interest. An example would be switching from a percentage to a fixed payout.
Since the oDAO members are paid in RPL but have expenses denominated in ETH (gas), their profitability depends on the RPL/ETH ratio and gas prices. At a low RPL/ETH ratio or high gas fees, it may not be profitable to run an oDAO node which introduces a risk to the protocol. OTOH, at a high RPL/ETH ratio we’re likely overpaying for their services and not being capital efficient with our treasury funds. One could argue that we’re currently overpaying oDAO for their services today.
#1 and #2 definitely depend on smart contract changes. #3 could be managed manually if we’re willing to dynamically adjust the oDAO inflation percentage to whatever is reasonable given a criteria at current RPL/ETH ratio and gas prices.
Pulling the greater RP community into the oDAO admission process seems reasonable, is there an elegant way to achieve this off-chain until we have the bandwidth to implement any SC changes?
I support this, the current budget would easily support a number of new members.
I’d suggest we hold off on implementing term limits, and if we were to do so, have them be longer periods of time to reduce management overhead. If the total oDAO set expands to say 20 members in the medium term, it seems easier to focus on kicking out underperforming members and voting on replacements. My gut tells me that in 95% of cases the exact same oDAO members will just be reaffirmed every year in a “check the box” type process.
No strong opinion here, although I do agree that we are likely overpaying for oDAO duties right now. I’d be in favor of reducing the current oDAO inflation budget or distributing it among a larger population of oDAO participants.
Strongly agree. Voting could probably be done through something like Snapshot
Seems reasonable. Alternatively the kicked oDAO members could be replaced by entities other than Rocket Pool Pty Ltd. This would mitigate the risk of having such a big part of the voting under one roof.
I agree with @NickS here. Although a good idea in principle, the overhead seems not worth it right now.
I would go further and would let every active veteran node operator to be able to become a odao member after 2 years of active staking and enough stake in rpl. These are the people who are committed to the protocol. This would also fit with additional measures like community application and voting process.
I’m a node operator who’s been with Rocketpool through all early betas. I agree that offering oDAO membership to those of us who are not commercial and not otherwise affiliated with Rocket Pool helps to balance the future direction of the protocol. 16 ETH is nothing to some of the larger groups but could represent a NO’s entire nest egg. NO’s have a high incentive to see Rocketpool succeed.
Agreed on all points, I think there should be a review system for current oDAO participants but I’m not sure annual reviews would be the best method. I wonder if the community or pDAO can initiate reviews after hitting a quorum?
Following the recent security issues with two oDAO members I think the best solution is a quick expansion of members, but I would like to see this on top of your excellent research on the oDAO tokenomics (Rocket Pool Tokenomics 2.0, Pt 4: - HackMD) and the idea from thomasg that devs are paid from inflation outside of the oDAO payments.
To @melbo :
Interesting point around smaller NOs committing a greater proportion of the wealth in an RPL bond than current oDAO members, and that they may be more economically aligned with the success of the protocol. I do think the team could make clearer the pathway to a more decentralised oDAO and how existing NOs could be included.
If considering NOs as oDAO members, I think on top of the bond they should be able to show they have competent equipment and location to run an Oracle Node. Perhaps this could be part of the review process.
For example: I have over spec’d bare metal equipment hosted in a Tier III Data Center. The DC has 24 hour security, active cooling, low latency gigabit internet and redundant power (both UPS and generators). The DC is fed by hydro power and claims 100% uptime over the past 10 years. I also have a secondary node ready and sync’d should I have a failure on my primary.
Security and resiliency should be critical for a node of this importance
Appreciate the replies and varied perspectives. It seems like the most controversial suggestion is term limits, so I’ll expand a bit on my thinking here.
Grant committee membership is comparatively inconsequential in terms of potential protocol harm vs the oDAO, but term limits for grant committee membership don’t seem to be controversial. Why is this? Is it lower overhead for grant committee elections? If so, what makes the overhead of electing new grant committee members lower than the oDAO? oDAO members are compensated more than even a fully corrupt grant committee would be able to achieve, and term limits introduce an additional check on oDAO member power.
I can see the benefits but feel like term limits lead to elections which can become popularity contests. Annual oDAO election brigading in forums, Discord and Reddit might not be the kind of PR Rocketpool is looking for. This is just my gut feeling as a non-member looking in.
I wanted to take plenty of time to think about this before sharing my perspective as an oDAO member. I do appreciate the RPL rewards paid to the oDAO, but I also want to prioritize the protocol above my personal gain. I want to say openly that being a part of the oDAO has given me more time and opportunity to champion Rocket Pool, and I believe it’s in all of our best interest to choose other oDAO members who will engage in similar activities. I’d like to thank RP and the community for giving me the opportunity to participate, and I want everyone to know that I take the responsibility seriously.
My two thoughts:
I do agree with a switch to static payments for oDAO members. I acknowledge that a conflict of interest exists for oDAO members who might onboard new members, with the current system discouraging the addition of new members because of payment dilution. My suggestion is that the payment for each oDAO member should be 500 RPL for each payment cycle, I think this leaves room for 15 members, while still making the payment an attractive incentive. I believe that lowering the payment below 500 RPL will result in a reduced quality of oDAO members, especially if the intention of the community is that these members are highly engaged.
Term limits make sense. They prevent stagnation and encourage new participants to get involved, but it needs a better design than simply “checking a box” and keeping the member because that will lead to stagnation. I think a better system would be a rotating system where one oDAO member is up for review every two months. This gives enough time to consider the contributions of each member, and it gives a tenure of about 30 months for each member (assuming 15 oDAO members). As an extension to this, I don’t know how I feel about a constantly changing oDAO set and the eventual campaigning and political drama it’s likely to lead to. Let’s work to design a system that limits the need for campaigning and the concept of “booting someone so I can make room for my own candidate [or myself]”. I feel like our community is better than that and we should work to keep it that way. I sort of like the idea of a ranked-choice vote every two months for an oDAO member spot, but I also have some reservations about it.
Given the responsibilities and power given to the oDAO members, I support the proposition to revise periodically their seat.
However before we review any of their performance, it should be clearly stated on which criteria they are being evaluated. There might be an oDAO charter already in place, but it would be best to be as transparent about what is expected of them by the protocol and what makes a good oDAO member. Without a clear benchmark we are running into a popularity contest like scenario.
There has been some campaigning for ENS and Optimism delegation. I think it could eventually look something like this for oDAO membership. Up to now that has been acceptable. But, long term, you’re right that this should be part of the discussion.
