The Kolibri DAO Path To Mainnet
Thank you all for your patience as we’ve been working to roll out the Kolibri DAO & Governance process. Today, we’d like to communicate the specifics of the Kolibri Governance Platform, governance token (kDAO), as well as outline the final steps towards a mainnet release. We know it’s been a long journey but we wanted to ensure that this upgrade goes as smoothly as possible to ensure that Kolibri is around for many years to come — a few weeks seems like an easy cost to help solidify that.
We have spent the last several weeks doing extensive QA processes to allow us to create reproducible Kolibri deployments on testnet, test break-glass contract workflows, and generally help prove the changes we’re planning on making to the Kolibri protocol.
Today, we’re pleased to roll out the official testnet governance portal for general availability at https://testnet-governance.kolibri.finance! We highly encourage kDAO recipients to play around with the portal and let us know if you find anything amiss (or have suggestions for improvement!).
This governance portal will be the main interface for people to propose, vote, and execute governance proposals within the Kolibri protocol, and will very quickly be open sourced as part of the decentralization process.
Along with this testing and governance portal work, we’ve also airdropped kDAO to the recipients of the airdrop on the Florencenet testnet. The contract address is KT1QaYndtyAs12EdUWBb5vWUnY5JWD4CpHdE.
Please note that the original list has been revised after we found evidence of people using smurf accounts to game the airdrop. You can find the analysis here, and recipients should double check their airdrop allocations in the final list. If you notice any discrepancies, please contact Hover Labs as soon as possible at email@example.com or in our Discord. After the airdrop has occurred on mainnet, we cannot correct any issues so definitely double check to be sure!
Hover Labs would like to provide additional specifics about some parameters we’ve chosen for mainnet launch. All of these updates are configurable from within the DAO moving forward.
Initially, these will be the values used within the DAO — of course they’re all adjustable with a simple proposal, though it’s definitely worth collaborating off-chain to ensure there’s a good will to change them and ensure the proposal is well received.
Quorum Caps: 100,000 (lower) to 900,000 (upper)
This means that the lower quorum cap is only 10% of kDAO, and the highest it will ever get is 90% of kDAO. These represent conservative estimates that we suspect will be modified over time. A reminder that the quorum is floating (similar to the Tezos protocol’s quorum algorithm), so these caps are to prevent quorum from going to 0% or 100% of the protocol. Initial quorum (described below) will be 20% of the supply.
Escrow Amount: 3,000 kDAO
The escrow amount is the amount of kDAO required to make a proposal. If the proposal doesn’t achieve some threshold of ‘yay’ votes, it is confiscated to the community fund. This mechanism encourages off-chain coordination, and acts as a deterrent against spam proposals that could otherwise create denial of service attacks. Hover Labs feels that 3,000 kDAO represents an achievable, yet high, threshold.
Vote Length: 1 week @ 1 min blocks
Kolibri Governance is meant to be a slow moving process driven by collaboration and consensus. Allowing a full week per vote lets all users have a chance to consider and participate. It also allows larger entities (such as corporations) to have ample time to process any activities that may be required (for instance, some crypto custody companies will only do governance on one day a week due to cold storage operational constraints).
Minimum Yay Votes for Escrow Return: 20%
Hover Labs feels that if 20% of voters do not vote yes, the proposal is likely spam or could have benefitted from additional off chain coordination. If this threshold is not achieved by the time the vote ends, the escrow locked to submit the proposal is sent to the community fund instead of returned to the proposer.
We think that this helps decrease spam proposals early on in governance, as we may want to make more frequent adjustments.
Of course confiscated escrow tokens can be returned from the community fund via a separate successful governance proposal if there’s a strong community desire to do so.
Timelock for Execution: 3 days @ 1 min blocks
All successful proposals must sit in the timelock for at least this amount of time before they are able to be executed. Hover Labs feels 3 days represents ample time to allow protocol participants to adjust to changing economic conditions, and take action with their funds they deem appropriate. This acts as a powerful disincentive for malicious proposals to even be presented in the first place.
Timelock for Cancellation: 8 days @ 1 min blocks
If the submitting entity hasn’t executed a proposal after 8 days (from voting close), anyone may cancel it. Like voting, a conservative value allows ample time for the submitter to execute the proposal, and allows larger entities who may only participate 1x per week chance to execute their proposal.
Percentage for Super Majority: 65%
As with the core Tezos protocol, we feel that a reasonably high threshold makes sense to have to consider a proposal passed. In this case we want to start with a ⅔ supermajority to pass proposals.
Initial Quorum: 20%
Gauging governance interest is hard if not impossible before things are launched, and we think a relatively low value means the community can achieve a quorum on initial proposals. Quorum adjusts via a weighted moving average, which means as proposals are submitted the system can move towards a more appropriate value.
We’ve also finalized our parameters for farming allocations. 30% of the supply is earmarked for the farms (300,000 kDAO). The initial farming parameters are as follows:
- 75,000 kDAO allocated to the kUSD farm
- 75,000 kDAO allocated to the QLkUSD farm
- 150,000 kDAO allocated to the Quipuswap kUSD LP farm
These farms will be deployed with a lifetime of one year (1,048,320 blocks @ 30 second block times). Hover Labs suggests that after a 4 week “bonus” period, the DAO should reduce the rate of emissions such that the farms will run for two years.
As a reminder, the DAO can revoke or change the distribution schedule of any farm at any time. Additional updates may be needed as the block times on the Tezos network decrease.
Break Glass Multisig
It is important to understand that with the Break Glass (see “Break Glass” in our announcement), Hover Labs does have the ability to retake control of components within the system. We’ve written extensively about why this is important and how we plan to give away control of the system as part of our progressive decentralization process, but the TL;DR is that this break glass system serves as a counter-balance to any serious security vulnerabilities discovered as part of the security audit which should happen later this summer.
We will never use this functionality to override the will of the kDAO holders/as a veto mechanism, and will only use it to counter security vulnerabilities. The DAO can revoke these break glass contracts day 1 if there’s a strong willingness to do so, but we think this strikes a good balance between removing control as part of our progressive decentralization process, and ensuring the safety of the system.
We have not at this point, however, discussed the configuration of this multisig. We propose configuring the multisig as:
- 2 of 3 signers
- 8 hour time lock
Hover Labs Allocation Contracts
As part of the rollout process, we’ve developed time-lock contracts for each team member of Hover Labs. These contracts allow participation in the governance process, but constrain the ability to arbitrarily access the kDAO locked into them. This means that the team will be able to participate within the DAO (even just abstaining to help reach quorum), but can not move the kDAO out of the timelock contracts at a rate faster than (TOTAL_TOKENS / 1 year).
Since the Tezos network seems poised to uptake Granada, which will lower the block time to 30 seconds, and thus our contracts will make available to us over 1,051,200 blocks (This represents slightly more than one year, since 30second block times are not yet live).
Hover Labs has also advised its employees that liquid kDAO tokens from these timelock contracts will be subject to a one year lockup period. As such, no Hover Labs employees will sell kDAO tokens acquired from their timelock contracts before one year after the day the airdrop is executed.
Onward And Upward
In our testing we’ve encountered an issue we’re still working through in the QLkUSD farm, so we’ve disabled that on https://testnet.kolibri.finance for now. This issue is a show stopper for the launch to mainnet and we have to better understand what the issue is before we can give a concrete launch time for Mainnet.
As always, we’ll continue to communicate in our Discord if this plan changes. Rollout will require 9 individual operations through the current governance multisig, which Hover Labs will stagger injections of over the course of 12–24 hours. This will allow us to validate and test the system as it slowly reaches decentralization. Onward and upward!