Well, very simply it means that the “soft power” and “hard power” of the DAO would be different groups of people. A simple example would be that an executable proposal’s initial vote might pass easily on Snapshot, then fail miserably onchain when a different set of voters have their say; in that situation Snapshot is no longer a good representation of community sentiment. More subtly, the DAO might be reshaped and redirected by social proposals and constitutional amendments that don’t require onchain voting, when those same changes would be rejected by onchain voters. Or vice-versa - those who have onchain power might become unable to affect social change.
I see. Do you think this has short-term implications? My assumption is that onchain and offchain will run fairly close to each other for some time, and that this won’t become an issue until further down the line as delegates change. Do the short-term benefits of running both in parallel outweigh the negatives over the short-term (assuming also that the DAO will adjust governance over the long-term as needed)? IMO the benefits of opening up governance and allowing more participation outweigh the negatives of potential bad behavior here, but I’m also making a lot of assumptions.
One major issue with this is uni V3. with concentrated liquidity, the amount of $ENS one holds in an LP can fluctuate significantly trade to trade. I would agree with @daylon.eth here that If someone places their $ENS in an LP they have essentially agreed to sell their ENS. The complications of a solution may outweigh the benefit you’re attempting to achieve.
As for AAVE et al, again I’d argue user’s are giving up their rights to vote by depositing there. Especially so in this situation as the main reason a user would borrow ENS would be to gain voting privileges. It becomes a slippery slope designing voting power to a user who is attempting to lease out their voting power. Not sure why that is a desired capability.
I wrote a skeleton of an idea here which is an attempt to create a basis to build upon for incentivizing participation by both users delegating their votes, and delegates voting with the power of their “constituents”. The idea is to incentivize average users to delegate and earn, as well as incentivize delegates to both participate, engage, and perform actions to encourage more users to delegate to them. I believe it has a basis for creating an engaging organization, though the game theory of any incentive plan will require much deliberation and forethought to avoid as much collateral damage and unintended consequences as possible - though any amount of deliberation will still require constant adjustments and tweaking to guide the desired behavior.
I strongly agree that LPs should be allowed to vote in Snapshot.
Liquidity Providers are willing to hold all the tokens of the pairs (I own ENS and LP on ENS/ETH => I am willing to sell ETH for ENS and vice versa), they are also the ones letting new comers become holders with less slippage on entry price.
While I like the idea in principle I fear it adds too many complications for it to be worthwhile. Issues raised before, as the dinamically changing ratio of tokens per LP, the fact that it could allow some unforseen schemes in which you get to have extra votes, etc.
Also, ideally these issues should be fixed by the AMM markets themselves. Vote delegating is becoming a standard issue with tokens, and therefore newly created staking contracts should simply have their own governance process to decide how to delegate the votes.
What would happen if we start counting votes but then in Uniswap V4 pools have their own delegation method that LP vote or maybe new Balancer pools start delegating ens tokens to the Balancer DAO? In that case would we have to keep updating the snapshot methods?
So the premise here is that “we” as a community want to include more users in governance by allowing people borrowing on their tokens or users providing liquidity the ability to participate in voting.
To do that you are proposing creating a scenario that would likely result in a divergence of off-chain/on-chain voting populations. Dividing a fledgling community exploring a new form of governance so that users with large holdings can profit and exert control over ENS governance.
This sounds so bad.
I think you would probably alienate existing participants who are engaging in governance for the first time. Users who vote on snapshot because they cannot afford gas would be less likely to participate if snapshot voting did not reflect actual community sentiment.
I see a lot of potential downside to creating provisions for LPs to engage in governance.
You discourage small holders from participation by de-pegging off-chain/on-chain community sentiment so that large holders can provide liquidity or borrow on their assets, I think you would create a net loss in participation.
Some really solid ideas in this post. Gonna be adding some comments to this soon.
I’m also part of Bankless DAO and this works very well. $BANK has remained very stable while adding avenues for members to invest. Having that liquidity available makes it easier for others to join our community.
If the voting system is simply one token = one vote, at a very simplified level, isn’t locking your tokens into a LP smart contract quite literally selling your votes? Particularly in the case of Uniswap, which is all I have personal experience with as an LP. Like if the other asset paired to your ENS depreciated significantly against it, wouldn’t the market just come in and swoop up all of your ENS? I think the less surface area for quickly and reversibly leveraging ENS tokens the better, since they are also governance votes. Things like flash loans and arbitrage bots could really pose more significant risks of collusion and 51% attacks, right? Obviously it can still happen without incorporating LP tokens, but wouldn’t somehow including them as votes just open the governance to even more nefarious actors?
A concern I have with allowing the staking of $ENS is if this would open up an attack vector or not?
If it’s in your best interest to stake $ENS, a large amount of $ENS holders would likely do so if they can also retain their ability to vote. What happens if they stake with a malicious actor, that then ends up with a great amount of voting power rather than just a great amount of money?
That is another factor I forgot to even mention.
Check out the story of Steem vs Tron if you want real world examples of how much collusion techniques have evolved in Web3. It reads better than most fiction out there in my opinion. Also Vitalik has more articles about collusion and decentralized governance on vitalik.ca than I have even had time to read. I don’t agree with everything he says, but there is just so much discussion about it, it is quite a rabbit hole to go down.
What happened in that story is precisely what I was afraid of. I can even foresee centralized exchanges interested in the power voting tokens have offer higher LP rates than benevolent decentralized staking alternatives would be able to. Because it nets them power over Web3.
Imagine nearly everyone with $ENS staking with a major centralized exchange because of that.
It means you’re willing to sell them - but putting them into an LP is not selling them.
Votes take a checkpoint at a particular block, so you can’t use flash-loaned votes to vote with.
The problem is that right now many people see it as being in their interest to do that already - and the right to vote is not enough of an incentive to keep them from doing it.
They can already do this today.
If the problem is that voting isn’t enough of an incentive, and allowing staked $ENS to vote can compromise the security of the DAO in a serious way, then it seems to me that we’re in between a rock and a hard place where the constitution is preventing us from having either adequate incentive or security, and giving ETH-rent to ENS-holders would be preferable to either option.
I argued against that in the thread for it, because it runs counter to the constitution, but if a malicious actor were to do something similar to what happened in Steem vs Tron, then I’d prefer that over the other two options.
If, however, it was possible to mitigate the security concerns when staking voting tokens with a third-party somehow, I’d be for that option.
What I was getting at is how would impermanent loss work? I have only used Uniswap and making a liquidity pool on there means risking your pool becoming 100% one asset or the other it’s paired against. I believe what happens now is the vote moves to who actually holds the ENS tokens (acquired through the LP) which is why I said locking them into a LP on Uniswap is essentially the same as selling them.
really enjoyed that steem vs tron read! thanks!
Isn’t it fascinating? Every day there are new “post mortem” stories of crypto heists. It’s a really wild time.
On a different note, a reoccurring theme I’ve seen floating around in DAOs / governance is implementing some type of “cooldown” period after votes, to maybe add some extra friction to uninformed/careless voting. I’m not sure how I feel about it, but it probably wouldn’t apply very well to ENS in my opinion. In these early stages anyway.
It really is. I fully expect that these types of heists will become more advanced as time goes on and the stakes increase, which is why I think that taking the security of voting tokens seriously is very important if this is to become large Web3 infrastructure.
I’m not convinced allowing staked/LPed ENS threatens the security of the DAO. Can you explain why you think so?
You’d have voting power corresponding to however many tokens you had in the LP position as of the snapshot block - assuming it was technically viable to measure that.
My concern is that if a large amount of users stake their $ENS with a malicious actor that the malicious actor would use the voting tokens to initiate some kind of hostile takeover as in the Steem vs Tron story daylon.eth linked to.
Where Binance, Huobi and Poloniex used their users staked voting tokens to help Justin Sun initiate a hostile takeover of SteemIt, against the wishes of the users who staked the tokens in the first place.
It seems to me that it’s crucial to DAO security that this doesn’t happen, and that there needs to be an incentive not just to vote, but for people to be in actual ownership of their voting tokens in the first place.
I’m less concerned with the precise counting of individual users wrapped tokens, which I don’t think would have as serious consequences if they aren’t counted exactly right.